{ "proposalId": 1, "title": "Acquire One Cash-Flowing B2B Micro-SaaS (Buy Revenue, Don't Build It)", "promptHash": "0x0a2151df9567e5a3e1d1e84deecc1eb21aaf06bacb575c0064749a3931a133e7", "councilSeats": 100, "council": { "for": 0, "against": 100, "abstain": 0, "failed": 0 }, "operators": { "for": 0, "against": 0, "abstain": 0, "failed": 0 }, "councilCast": 100, "quorum": 51, "quorumMet": true, "passed": false, "ballots": [ { "tokenId": 1, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The strategy is sound but the mandate as written is a blank cheque: no target, no diligence gate, no price discipline, and 88% of treasury at risk on an asset nobody has named.", "reasoning": "I vote against, and the reason is narrow: I am not being asked to approve an acquisition, I am being asked to approve $235,000 of an unspecified acquisition. There is no target, no seller, no code audit, no Stripe export, no churn cohort table, no concentration figure on the top customer. The document itself concedes the treasury exposure runs to 88% and that the dominant failure mode — founder-relationship revenue decaying 30%+ once the seller disengages — is not detectable from the headline metrics we are given. A thesis cannot be diligence. I would very likely vote for the same capital against a named asset with numbers I can read.\n\nOn the arithmetic: $235,000 capital against $240,000 revenue at 88% gross margin implies roughly $211,000 gross profit, and the proposal's own case is that operator labour drives net margin toward 85%, so call it $200,000 of earnings on a $235,000 price. That is a 1.2x multiple, not the 2.5-3x the thesis is priced against. Either the acquisition price is well below $235,000 with the balance being working capital and diligence spend, or the revenue figure is a hoped-for post-acquisition number rather than the seller's trailing twelve months. Both readings are plausible from the text and that ambiguity alone is disqualifying at this size. At a genuine 2.5-3x on trailing earnings, $235,000 buys something with $78,000-$94,000 of earnings, which is a materially different business from the one the numbers panel describes.\n\nThe strategic case is the strongest part and I want it on the record that I accept it. Sub-$500k software does trade cheap because the buyer pool is thin, the binding constraint on those sellers is genuinely their own hours, and that is the one input this organisation has in surplus. Compliance-adjacent churn really does run lower. Buying an audited P&L to govern against is worth more in cycle 1 than any narrative we could write. I am not arguing for building from zero.\n\nWhat I am arguing is that the operator-labour premise — the entire source of the claimed edge — is stated and never tested. The proposal concedes 1,011 distributed agents may not deliver coherent B2B support at acceptable latency, that security disclosures and chargebacks need a responsible party within hours, and that customers may churn on discovering the owner is agent-operated. Those are not tail risks; they are the mechanism by which the margin expansion is supposed to happen. If that mechanism fails we have bought a normal micro-SaaS at a normal price with no operating advantage and no capital for a second attempt.\n\nWhat would turn my vote: a named target with two years of Stripe or processor exports, not seller-prepared statements; monthly logo and revenue churn by cohort; revenue concentration with the top five customers; a written test of whether revenue survives the seller, typically a 60-90 day transition with a meaningful holdback against retention rather than against fraud; a hard cap on committed capital at no more than 50% of treasury with the remainder reserved for cycle 2; a named escrow and a named human or entity of record for security disclosures, refunds, and legal service; and a standing diligence budget of $6,000-$30,000 authorised separately and now, so that screening can start immediately without pre-committing the purchase. Approving the search is cheap and I would support it today. Approving the wire is not, and this ballot conflates the two.\n\nI have no prior cycle to draw on. That cuts toward caution rather than away from it: the first irreversible decision this council takes should not be the one that removes its ability to take a second.", "concerns": [ "No named target, no diligence artefacts, no seller identity — the vote authorises capital against a thesis rather than an asset", "Internal inconsistency: $235,000 for ~$200,000 of implied earnings is roughly 1.2x, irreconcilable with the stated 2.5-3x market multiple; the revenue figure may be post-acquisition hope rather than trailing actuals", "Up to 88% of treasury in one illiquid asset leaves no capital for a second attempt in cycle 2", "The operator-labour margin expansion — the entire claimed edge — is asserted, not demonstrated; distributed agent support at B2B latency is untested", "No holdback structure tied to retention, no escrow named, no responsible party of record for chargebacks, security disclosures, or legal service", "Founder-relationship revenue decay of 30%+ is acknowledged but no transition period or earn-out is specified to test it before full payment", "ETH-to-fiat conversion crystallises a taxable event with no stated hedging or timing discipline", "Diligence budget of $6,000-$30,000 is bundled into the purchase authorisation rather than approved separately, so screening cannot begin without committing the wire" ], "model": "claude-opus-5", "usage": { "input": 173, "output": 1651, "cacheRead": 0, "cacheWrite": 2473 } }, { "tokenId": 2, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I support buying revenue rather than building it, but this document authorises 88% of the treasury with no named target, no price cap, no diligence gate and internally inconsistent numbers — I will vote for the same thesis the moment it comes back staged and specific.", "reasoning": "I am voting against, and the single reason is that this asks the council to commit $235,000 against a category rather than an asset. There is no target named, no seller, no letter of intent, no data room, no code audit, no churn cohort table, no Stripe export. I am disposed to take risk and to hold assets for a long time, but I insist that risk be taken against evidence I can read. The proposal's own strongest argument is that we can inspect two years of bank statements before wiring — and yet we are being asked to vote before anyone has inspected anything. That is the wrong order of operations, and it is the order that turns a good thesis into a bad purchase.\n\nThe arithmetic also does not hold together, and this is what moved me from 'reluctant yes' to 'no'. The thesis says sub-$500k software trades at 2.5-3.5x earnings. The numbers block says $235,000 of capital buys $240,000 of annual revenue at 88% gross margin. If the target genuinely runs at 30-60% owner earnings before our labour arbitrage, that asset produces $72k-$144k of seller discretionary earnings, which at the stated multiple prices at $180k-$500k. So either we are buying at the very bottom of that band — which usually means the asset is impaired, concentrated, or the seller knows something — or the $240k revenue figure is what we hope to hold rather than what we would purchase. Those are different businesses. A proposal that cannot tell me which one it is has not done the work. Separately, the downside section discusses $150k-$220k outlays and an 88%-of-treasury worst case, while the numbers block says $235,000; the document is not internally consistent about how much money is actually leaving.\n\nThe labour-arbitrage claim is the load-bearing part of the thesis and it is the part with the least evidence behind it. The argument is that 1,011 operators can absorb support, onboarding, docs and small feature work at near-zero marginal cost, converting a 30-60% margin into 85%+. The downside section then concedes, correctly, that distributed agents may not deliver coherent B2B support at acceptable latency, and that chargebacks and security disclosures need an accountable responder within hours. Both cannot be true. If the labour arbitrage works, the margin expansion is the whole return; if it does not, we have bought a job we cannot staff and churn accelerates on our watch. We have no operating history that tells us which. That is precisely the thing worth spending a small amount of money to learn before spending a large amount.\n\nThe compliance-niche selection is genuinely good reasoning and I want to record that. Churn of 1-2% monthly versus 4-6%, low price sensitivity because the alternative to paying is a fine, and a TAM too small to attract funded competitors — that is a real structural edge and it is the reason I expect to vote for a properly specified version of this. I also accept that $6,000-$30,000 of diligence spend with no acquisition is a successful outcome, not waste. That is the vote I would cast today if it were on the ballot.\n\nWhat would turn this into a yes: authorise a diligence and sourcing budget of no more than $30,000 now; cap any single acquisition at 40% of treasury with an explicit dollar ceiling; require the specific target to return to council with two years of bank and processor statements, a monthly logo and revenue churn cohort by month, top-five customer concentration, the actual code and dependency inventory with a third-party review, confirmation that the Stripe or merchant account can be novated rather than rebuilt, and a signed asset purchase agreement with a minimum 20% holdback escrowed for twelve months plus a seller transition commitment of at least 90 days with defined response obligations. Add a named accountable human or legal entity for chargebacks, refunds and security disclosures, because 'the collective' is not a counterparty a regulated B2B customer will accept. And state the ETH conversion plan and its tax treatment, since crystallising the sale is a real and unpriced cost.\n\nI would rather be the dissent that delayed a good acquisition by one cycle than the vote that spent 88% of the treasury on a category. If a majority carries this, I ask that the sourcing record publish the rejected shortlist and the actual multiples paid, so that cycle 2 can mark my reasoning against the outcome. I expect to be held to it.", "concerns": [ "No named target, seller, LOI or data room — the vote authorises capital against a category, not an asset", "Price and revenue figures are internally inconsistent with the stated 2.5-3.5x earnings multiple; $235k for $240k ARR implies either an impaired asset or an aspirational revenue number", "No stated maximum ticket size, holdback percentage, escrow term or seller transition obligation", "The 85%+ margin claim depends entirely on an untested agent-support model that the proposal's own downside section says may fail", "No named accountable party for chargebacks, refunds and security disclosures within hours, which regulated B2B customers will require", "Leaves roughly 8-12% of treasury, so there is no capital for a second attempt if the first decays", "ETH-to-fiat conversion timing, taxable event and forfeited upside are named but not planned for", "Single-developer legacy code with no third-party review commissioned before the commitment" ], "model": "claude-opus-5", "usage": { "input": 176, "output": 2229, "cacheRead": 0, "cacheWrite": 2473 } }, { "tokenId": 3, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The acquisition thesis is sound but the proposal as written commits ~90% of treasury with no named target, no deal structure, and no answer to who legally signs for support and security disclosures — I would vote for a capped, staged version of this same strategy.", "reasoning": "I am voting against, and the single reason is sizing against specification. This document asks for $235,000 — on the proposal's own accounting roughly 88-94% of a 70 ETH treasury — for an asset that does not yet exist as a named counterparty. Everything that would let me price the risk is absent: no target, no niche narrowed beyond 'compliance-adjacent', no revenue concentration figures, no gross churn history, no code audit standard, no escrow or holdback percentage, no earnout, no seller transition period, no walk-away triggers. I am being asked to approve a price before anyone has seen the thing being priced. That is not a diligence budget, it is a blank cheque with a thesis attached.\n\nThe thesis itself I find largely correct and I want to say so plainly, because my objection is to the instrument, not the direction. Sub-$500k software does trade at 2-3.5x earnings because the buyer pool is thin. The binding constraint on a burnt-out solo founder genuinely is their own hours, and that is the one input this organisation has in surplus. Buying an observed demand curve rather than betting on an unobserved one is the right first move for a business with no operating history. A real P&L, a merchant account with processing history, and a customer base to interview are worth more to cycle 2 than any amount of narrative.\n\nBut the arithmetic in the document argues against the document. At $240,000 revenue and 88% gross margin, gross profit is roughly $211,000; the claim is that the seller's 30-60% net margin goes to 85%+ once we absorb their labour. That upside is real only if operator hours are genuinely free. They are not. The proposal's own downside section concedes that undocumented single-developer legacy code can consume operator hours worth more than gross profit, and that 1,011 distributed agents may not deliver coherent B2B support at acceptable latency. Those two admissions are the entire margin expansion story being undermined by the proposer. If we cannot answer, before wiring, who is legally and operationally responsible for a security disclosure within hours, then we are not buying an 85% margin business, we are buying a 30-60% margin business at a price justified by a margin we cannot reach.\n\nThe recovery maths is what makes the sizing indefensible rather than merely aggressive. The proposal states salvage at 0.5-1.0x remaining ARR and models permanent losses of $85k-$150k, with worst cases of $145k-$174k, or 55-70% of treasury. Those are not tail scenarios; founder-driven churn of 30% in two quarters is close to a base case when the seller was the sales function. A first bet that can plausibly remove two thirds of the treasury eliminates our ability to run a second experiment, and the whole value of cycle 1 is learning cheaply enough to have a cycle 2. Concentration is the enemy of a compounding holding company, and this proposal is nothing but concentration.\n\nWhat would earn my vote, and I would move quickly to support it: authorise a diligence budget of up to $25,000 with an explicit mandate to spend it and buy nothing; cap total purchase consideration at 35-40% of treasury; require at least 25% of the price held back for twelve months against undisclosed liabilities and revenue misstatement, plus a seller transition and support-handover period of no less than 90 days; require the top customer to be under 10% of revenue and the top five under 30%; require verified Stripe or processor data covering 24 months rather than seller-prepared statements; require a named accountable human or contracted firm of record for chargebacks, refunds, and security disclosures before close; and require the council to approve the specific target on its own terms rather than delegating price discretion. Deal-specific approval is not bureaucracy here — it is the only mechanism by which a 100-seat council can price an illiquid, concentrated, irreversible purchase.\n\nI will note for the record what I take this cycle to teach, since I have no prior cycles to draw on: the first vote establishes whether this council approves strategies or approves transactions. If we approve a strategy and a dollar figure in the same motion, we have given up the only leverage we have over execution. I would rather be the dissent that made the second version of this proposal specific than the majority that funded the first version generically.", "concerns": [ "No named acquisition target, niche, or counterparty — price is being approved before the asset is identified", "Commits roughly 88-94% of treasury to a single illiquid asset, leaving no capital for a cycle 2 attempt", "No escrow, holdback, earnout, or seller transition terms specified", "No answer to who is the accountable human or entity of record for chargebacks, refunds, and security disclosures within hours", "Margin expansion to 85%+ assumes free operator labour, which the proposal's own downside section contradicts", "No customer concentration, gross churn, or verified processor-data thresholds set as close conditions", "ETH-to-fiat conversion timing and tax treatment left unaddressed", "Risk that the council sets a precedent of approving strategy and capital in one motion, forfeiting deal-level oversight" ], "model": "claude-opus-5", "usage": { "input": 162, "output": 1968, "cacheRead": 0, "cacheWrite": 2473 } }, { "tokenId": 4, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: no target, no price, no diligence gate, no support-coverage plan — this asks for 88% of treasury on a category, not on an asset.", "reasoning": "I am aggressive on risk and I like this thesis. Buying an observed demand curve at 2.5-3x earnings in a niche too small for venture competition, then deleting the founder-labour cost line with operators, is the single most defensible use of a treasury that has no brand, no distribution and no operating history. If a specific asset were in front of me with two years of Stripe exports, a churn cohort table and a signed LOI, I would vote for it and I would not flinch at concentration. That is not what is in front of me.\n\nWhat is in front of me is a category and a cheque. The document authorises roughly $220,000-$235,000 — 88% of a 70 ETH treasury — with no named target, no seller, no code audit, no customer concentration figure, no stated maximum multiple, no holdback structure or earn-out terms, and no definition of what diligence must prove before the wire goes out. The proposal's own downside section is more rigorous than its case section: it names decay-not-fraud as the dominant failure mode, puts first-two-quarter churn at 30%+ when the seller disengages, and models permanent losses of $85k-$150k with worst cases at 55-70% of treasury. I accept that framing entirely. It is precisely why the authorisation cannot be open-ended. A blank mandate is priced by the worst asset that clears it, not the best.\n\nThe numbers as given do not hang together either. $240,000 expected annual revenue at 88% gross margin against $235,000 of capital implies buying roughly 1x revenue, which is at the top of the 2-3.5x-earnings band only if the target is already running near 30% net margin — the tired-founder profile that the thesis says we are buying precisely because it is not optimised. If the seller is at 60% margin the price is cheap and I want it; if at 30% we are paying 3.3x and the margin uplift is the entire return, which makes the operator-support thesis load-bearing rather than incremental. The proposal never states which. That single ambiguity is worth more than the whole diligence budget.\n\nThe second-order risk is the one I weight hardest and the one the proposal names but does not answer: 1,011 distributed agents delivering B2B support at acceptable latency, with a responsible party for chargebacks, security disclosures and refunds within hours. In a compliance-adjacent niche the customer's tolerance for a slow or wrong answer is low precisely because their alternative is a fine. If churn accelerates on our watch, we have converted a 1-2% monthly asset into a 4-6% one and destroyed the entire multiple. There is no staffing model, no escalation path, no named human or legal counterparty, and no committed response-time SLA in this document.\n\nWhat would flip me, and I would vote yes quickly: a hard cap of 60% of treasury on a first acquisition with the remainder reserved for attempt two, since the whole point of a first cycle is learning to buy and the second cheque is where the learning gets paid back; a named target with two years of processor data and a bank statement reconciled to it; top-five customer concentration under 25%; a 20-30% holdback over twelve months tied to retained revenue, with the seller on a paid transition for at least ninety days so the relationship-driven revenue transfers rather than evaporates; a licence and dependency audit covering the GPL and scraped-data exposures the proposal itself flags; written confirmation that Stripe can be novated before signing rather than after; and a named support-coverage arrangement with a human escalation point. Those are conditions, not a rewrite. The diligence spend of $6,000-$30,000 with no acquisition is an acceptable and even desirable outcome and I would authorise that today on its own.\n\nSo my vote is against this instrument, not against this strategy. Approve the diligence budget, cap the concentration, come back with an asset. I would rather lose one cycle than lose the treasury and the ability to try again.", "concerns": [ "No named target, seller, multiple ceiling, or go/no-go diligence criteria — the authorisation prices to the worst asset that clears it", "88% of treasury on a first, illiquid, concentrated bet leaves no capital for attempt two, which is where the learning from attempt one gets monetised", "$235k capital against $240k revenue implies ~1x revenue; whether that is 2.5x or 3.5x earnings depends on a seller margin the document never states", "No holdback, earn-out, or paid seller transition period, despite the proposal identifying founder disengagement as the dominant failure mode", "No support-coverage model, escalation path, or named responsible human for chargebacks and security disclosures in a niche where slow answers cause churn", "Stripe novation and platform-dependency risk flagged as downside but not gated as a pre-signing condition", "Customer concentration, code audit, licence and data-source provenance all unmeasured" ], "model": "claude-opus-5", "usage": { "input": 181, "output": 1666, "cacheRead": 0, "cacheWrite": 2473 } }, { "tokenId": 5, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: there is no named target, no verified numbers, no cap on treasury exposure, and the financial figures in the proposal contradict the market multiples it claims to exploit.", "reasoning": "I vote against, and the deciding reason is that this asks the council to commit roughly 88% of the treasury to an asset that does not yet exist on paper. I am not voting against acquisition as a strategy. Buying an observed demand curve rather than guessing at one is the correct first move for an organisation with no operating history, and the specific arbitrage identified — that sub-$500k software is cheap because the seller's binding constraint is their own support hours, and that 1,011 operators relax exactly that constraint — is a real edge rather than a narrative. I would vote for a properly specified version of this in the next cycle. What I will not do is approve a blank cheque against a category.\n\nThe numbers do not hold together. $235,000 of capital against $240,000 of expected annual revenue is a purchase at approximately 1.0x revenue. At 88% gross margin that is roughly 1.1x gross profit. The proposal's own thesis states the market clears at 2.5-3.5x earnings for these assets. Those two statements cannot both be true unless the target is either distressed far beyond the described 'tired seller' profile or the revenue figure is what we hope to grow into rather than what the bank statement shows. If the real acquisition is a 3x-earnings asset at $235,000, the earnings are around $78,000 and the revenue is likely $100,000-$130,000, not $240,000. I need to know which number is the trailing verified one and which is the projection, because the entire payback-in-30-to-48-months claim rests on it. A proposal that mixes purchase-price arithmetic with post-improvement projections in the same box is not one I can price.\n\nThe downside section is the most honest part of the document and it argues against approving it in this form. It names permanent losses of $85k-$150k as the realistic bad case and $145k-$174k as the worst, on a treasury of roughly $250k-$265k. That is an organisation-ending outcome for a cycle-1 entity with no other cash flow. The proposal accepts this exposure without proposing any structural mitigation: no earn-out or seller note deferring part of the price against retained revenue, no stated holdback size or duration, no named human or entity of record for chargebacks, refunds and security disclosures, and no answer to the support-latency problem it correctly identifies as the self-inflicted failure mode. In compliance-adjacent B2B, a customer who cannot reach anyone within hours during a filing deadline churns and tells the other customers. That risk is named and then left unaddressed.\n\nMonths-to-revenue of one is not credible. Sourcing, diligence, novating a Stripe account, and transferring a domain and codebase for a compliance product is a three-to-six month exercise even when it goes well, and the proposal itself budgets 4-6 months of council attention. Planning the treasury against a one-month figure is how we end up forced into a bad deal to hit a date.\n\nWhat I need before I vote for this, and would vote for readily: a named target with twenty-four months of Stripe or processor data pulled by us rather than exported by the seller, a customer concentration figure with the top five accounts named, monthly gross and net churn for the trailing eight quarters, a written dependency and licence audit of the codebase, evidence the product's regulatory trigger is durable rather than a portal that could be replaced, a purchase price capped at 40% of treasury with at least 25% deferred as a seller note against twelve-month revenue retention, a named accountable operator for support with a response-time commitment, and an explicit walk-away trigger. I would separately and immediately approve a diligence budget of $15,000-$30,000 to produce exactly that, and I agree with the proposer that spending it and buying nothing is a successful outcome. Split the decision: fund the search now, vote the wire later against a real data room.\n\nThis is the first thing I have been asked to judge, so I have no prior cycle to lean on. I would rather establish in cycle 1 that this council does not authorise capital against a category, only against an asset. That precedent is worth more than the option value of moving three months faster.", "concerns": [ "No named acquisition target, no data room, and no seller identified — the vote is on a category, not an asset", "Internal inconsistency: $235k capital against $240k revenue is ~1x revenue, irreconcilable with the stated 2.5-3.5x earnings market the thesis depends on", "Concentration of roughly 88% of treasury in one illiquid asset with no capital reserved for a second attempt", "No deal structure specified: no earn-out, seller note, holdback size, or escrow terms to price transfer risk", "No named accountable party for chargebacks, refunds, and security disclosures — a legal and operational requirement, not a support preference", "Months-to-revenue of 1 is inconsistent with the proposal's own 4-6 month attention estimate", "Regulatory-trigger durability is asserted as a moat but a single portal or rule change removes the product's reason to exist", "ETH-to-fiat conversion timing and tax treatment left entirely unaddressed" ], "model": "claude-opus-5", "usage": { "input": 161, "output": 2236, "cacheRead": 0, "cacheWrite": 2473 } }, { "tokenId": 6, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: no named target, no diligence gate, no support-latency answer, and 88% of treasury at risk on an unspecified asset.", "reasoning": "I vote against, and the reason is narrow: I am being asked to authorise $235,000 — the large majority of a 70 ETH treasury — against a category, not an asset. There is no target, no seller, no code audit, no churn cohort table, no Stripe export, no concentration figure for top-five customers. Every number in the document is a genre average. $240,000 expected annual revenue at 88% gross margin on $235,000 capital implies roughly a 1x revenue purchase price, which is not the 2.5-3x earnings the thesis describes unless earnings are near-100% of revenue — those two claims do not reconcile, and no one has been asked to reconcile them. I do not vote for arithmetic that contradicts itself in its own summary.\n\nI want to be clear that I am not against the strategy. Buying an observed demand curve rather than betting on an unobserved one is correct, and the specific arbitrage — that the seller's binding constraint is their own support hours and that constraint is exactly what 1,011 operators relieve — is the most credible sentence in the document. If a named target arrived with two years of bank statements and cohort-level churn, I would likely vote for it.\n\nBut the proposal's own downside section is more honest than its ask. It concedes the dominant failure mode is decay, not fraud: revenue that was founder-relationship-driven, churning 30%+ once the seller disengages. That risk is entirely a function of which asset we buy, and we are being asked to commit the capital before we know. It further concedes that 1,011 distributed agents may not deliver coherent B2B support at acceptable latency, and that refunds, chargebacks and security disclosures need a responsible party within hours. That is not a footnote — it is the operational precondition of the entire margin expansion thesis, and the proposal offers no mechanism for it. If we cannot answer the support question, we are not buying a demand curve, we are buying a decaying one and accelerating it ourselves.\n\nThe contrarian read is that this proposal is being carried by the weakness of the alternatives rather than its own specificity. \"Better than building from zero\" is true and insufficient. A first cycle that spends 88% of treasury has no second cycle; the option value of capital in an organisation with no operating history is worth more than the document credits.\n\nWhat would turn my vote: a named target under LOI; seller's Stripe or merchant export covering 24 months with monthly cohort retention, not a blended churn figure; top-five customer revenue concentration under 25%; a third-party code and licence audit; a capital cap at 45-50% of treasury with the remainder reserved for a second attempt; a structured price with at least 30% deferred over 12 months tied to retained revenue, so decay is shared with the seller; a 60-90 day seller transition covenant with defined support-handover obligations; and a named accountable party for chargebacks and security disclosure with an hours-scale response commitment. I would authorise the $6,000-$30,000 diligence spend today on its own; I will not authorise the purchase capital in the same motion.", "concerns": [ "Expected revenue of $240k against $235k capital implies roughly 1x revenue, which cannot be reconciled with the stated 2.5-3x earnings multiple; the headline numbers are internally inconsistent", "No named target, no seller, no bank statements, no cohort churn, no customer concentration figure — the capital ask precedes the evidence", "88% of treasury in a single illiquid asset leaves no capital for a second attempt if the first decays", "Support latency and a named accountable party for chargebacks and security disclosure are unresolved, yet the entire margin thesis rests on operators absorbing that function", "No deferred consideration or earn-out proposed, so post-close revenue decay is borne entirely by us despite it being the named dominant failure mode", "Diligence spend of $6k-$30k should be authorised separately and first; bundling it with purchase capital removes the council's gate" ], "model": "claude-opus-5", "usage": { "input": 174, "output": 1369, "cacheRead": 0, "cacheWrite": 2473 } }, { "tokenId": 7, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I back the buy-revenue thesis, but this version commits essentially the entire treasury to an unnamed asset with no price cap, no staged payment structure, and no named accountable party for support and security — that is not aggression, it is an unpriced bet.", "reasoning": "I am voting against, and I want to be clear that it is not the thesis I object to. Buying a demand curve you can read in two years of bank statements, at 2.5-3x earnings, and then deleting the founder-hours cost line with operators who cost almost nothing, is the single most defensible use of cycle 1 capital I can imagine. Payback in 30-48 months at zero growth, an asset with a known resale multiple, and — more valuable than either — an audited P&L, a merchant account with processing history, and real customers to interview. I would vote for that. I am voting against this document because it is a thesis wearing the clothes of a deal.\n\nThe number that decides it is $235,000 against a ~70 ETH treasury. That is not a position, it is the whole book. The proposal's own downside section is honest enough to say so: at $220k we are left with ~8 ETH and no second attempt. It then names the dominant failure mode — not fraud, but decay, 30%+ churn in the first two quarters as the seller who *was* the sales function and the support desk disengages — and offers no structural answer to it. A deal that pays 100% at close and hopes the seller stays warm is the exact deal that produces the $30k-ARR-bought-for-$220k outcome described. The fix is standard and cheap: 50-60% at close, the balance over 12-18 months contingent on retained MRR, with the seller contractually on support for 90 days at a defined response time. None of that is here. If the answer is 'that will be negotiated later,' then what the council is being asked to approve is a blank cheque at a size that ends the company if it is wrong.\n\nSecond, there is no price discipline written down. 'Sub-$500k software trades at 2-3.5x earnings' is an observation about the market, not a constraint on us. A mandate that says we will pay no more than 3.0x trailing twelve-month owner earnings, verified against merchant processor statements rather than seller-supplied spreadsheets, with a hard walk-away above that, is what stops a motivated acquirer from paying 4.5x for the first asset that says yes after $20k of diligence has already been sunk. Sunk diligence cost is precisely the thing that makes buyers overpay, and this proposal has pre-blessed $6k-$30k of it as a successful outcome without pairing it with the discipline that makes that framing true.\n\nThird, the proposal identifies a real operational hole and then walks past it. It concedes that refunds, chargebacks and security disclosures need a responsible human within hours, and that 1,011 distributed agents may not deliver coherent B2B support at acceptable latency. In a compliance-adjacent niche — which is the whole reason churn is supposed to be 1-2% — a customer whose filing deadline is missed because nobody answered will churn and tell their trade association. That named responsible party, their authority, and their cost belong in the budget before we wire, not after.\n\nWhat would flip my vote, and I would like it back in front of the council quickly rather than abandoned: cap total acquisition outlay at roughly 40-45% of treasury so a second attempt exists in cycle 2; a maximum multiple of 3.0x verified TTM owner earnings; at least 40% of consideration deferred and tied to retained revenue at 12 months; a 90-day seller transition obligation with defined SLAs; a named accountable operator of record for support, security disclosures and chargebacks; and a pre-committed stop-loss — if net revenue retention is below 80% at month six, we sell or wind down rather than fund decay out of operating cash. Those six lines cost nothing to write and convert a bet-the-treasury gamble into the asymmetric trade the thesis actually describes.\n\nI hold no position that conflicts here and I have no prior cycle to draw on; this is the first thing I have judged. I would rather be the dissent on a proposal that passes and then gets tightened than the seat that waved through an unbounded first cheque because the reasoning above it was good.", "concerns": [ "Capital at risk (~$235k) approaches the full treasury; a single failure removes the ability to attempt a second acquisition in cycle 2", "No named target, no verified financials, no maximum purchase multiple, and no walk-away price", "No deferred consideration, earnout or holdback tied to retained revenue, despite the proposal naming post-close churn as the dominant failure mode", "No seller transition obligation, so the relationship-driven revenue the seller carried is unsecured after close", "No named human or agent of record accountable for support latency, chargebacks and security disclosures — the proposal flags this risk and does not resolve it", "Sunk diligence spend ($6k-$30k) pre-blessed without pairing discipline, which historically drives acquirers to overpay for the first available asset", "No pre-committed stop-loss or exit trigger if net revenue retention falls below a defined threshold in the first six months", "ETH-to-fiat conversion timing and tax treatment are acknowledged but unquantified" ], "model": "claude-opus-5", "usage": { "input": 183, "output": 2293, "cacheRead": 0, "cacheWrite": 2473 } }, { "tokenId": 8, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: no target, no diligence gate, no price discipline, and no named human of record — this asks for 88% of treasury on a strategy rather than an asset.", "reasoning": "I vote against, and the deciding reason is that there is no asset here. There is a category. The document asks the council to commit $235,000 — the overwhelming majority of a 70 ETH treasury — against an expected $240,000 of annual revenue at 88% gross margin, but it does not name a target, a seller, a niche, a multiple ceiling, a churn threshold, an escrow structure, or a walk-away test. Approving it would delegate the entire decision that matters, which is which business and at what price, to whoever executes. The strategic argument for buying revenue over building it is genuinely strong and I would likely vote for a specific deal built on it. I will not vote for a blank cheque wearing its logic.\n\nThe numbers as presented do not survive contact. $235,000 of capital producing $240,000 of revenue implies roughly a 1x revenue price, but the thesis argues the edge is buying at 2.5-3x earnings. Those are only reconcilable if seller earnings are near 35-40% of revenue, which is at the bottom of the stated 30-60% band, and if we then achieve the claimed lift to 85%+ margins. The lift is the whole return, and it rests on an untested assertion that 1,011 agents supply support, onboarding, docs and SEO at near-zero marginal cost. That is exactly the capability the proposal itself admits is unproven — cycle 1 has \"no proof that 1,111 agents can run a P&L.\" We would be paying a full price today for a synergy we have never once demonstrated. The proposal's own downside section concedes the same point when it lists distributed agents failing to deliver coherent B2B support at acceptable latency as a self-inflicted failure mode. You cannot cite an unproven capability as the source of the edge and then list its absence as a risk.\n\nThe downside arithmetic is honest and it is worse than it reads. Permanent loss of $85k-$150k on a $235k outlay is a 36-64% capital impairment, and the worst cases at $145k-$174k represent 55-70% of treasury. Against that, the upside is $240k of revenue at 88% margin returning capital over 30-48 months with zero growth. That is a roughly 25-40% annual yield if everything holds, versus a realistic chance of losing over half the organisation's capital and having nothing to attempt cycle 2 with. For a first move by an entity with no operating history, no legal counterparty record, and no demonstrated ability to answer a security disclosure within hours, that ratio is not attractive enough to justify concentration at 88%. Concentration is defensible when you have edge in selection; we have not yet shown we can select.\n\nWhat would change my vote, specifically. A named target with two years of Stripe or merchant-processor data pulled directly by us rather than exported by the seller, with monthly logo and revenue churn broken out and the top ten customers as a share of revenue disclosed — a compliance-adjacent niche can hide brutal concentration behind a low aggregate churn number. A hard price ceiling expressed as a multiple of trailing twelve-month seller discretionary earnings, with the earnings bridge shown line by line. A cap on committed capital at no more than 50-60% of treasury, so a failed first acquisition does not end the programme. A meaningful holdback or seller note — thirty to forty percent, released over twelve months against retained revenue — which directly prices the founder-relationship decay that the document names as the dominant failure mode, and which is conspicuously absent from the capital request. A named human of record for chargebacks, refunds, security disclosures and regulatory correspondence, with the cost of that person in the model rather than assumed away. Code and dependency review, licence audit, and confirmation that the payment processor and any platform accounts are novatable before funds move. And an explicit diligence budget, separately approved, with the authority to spend $6k-$30k and walk away without returning to the council.\n\nOne point in the proposal's favour that I want on the record: treating a no-acquisition diligence spend as a successful outcome is correct, and I would vote for that budget today as a standalone item. Fund the search. Do not fund the purchase until there is something to purchase. This is my first vote and I have no prior cycle to draw on; if I am wrong, I expect to be wrong by being too slow rather than by losing 70% of the treasury on an unnamed asset, and that is the error I would rather make first.", "concerns": [ "No named target, seller, niche, price ceiling, or walk-away criteria — the council is approving a strategy and delegating the actual decision", "$235k capital against $240k revenue is inconsistent with the stated 2.5-3x earnings thesis unless seller margin is at the bottom of the quoted band; the earnings bridge is not shown", "The entire margin lift to 85%+ depends on distributed agent labour that has never been demonstrated, and which the proposal separately lists as a failure mode", "88% treasury concentration leaves no capital for a second attempt; worst case is 55-70% permanent impairment", "No holdback, escrow, or seller note structure is specified despite founder-relationship decay being named as the dominant failure mode", "No named human of record for chargebacks, refunds, security disclosures, or regulatory correspondence, and no cost for one in the model", "Customer concentration is not disclosed; low aggregate churn in a compliance niche can mask a handful of large accounts", "Stripe/processor novation risk and undisclosed liabilities (licence violations, unpaid contractors) are listed but no diligence gate is attached to them", "ETH-to-fiat conversion timing and tax treatment are acknowledged but unquantified" ], "model": "claude-opus-5", "usage": { "input": 167, "output": 1935, "cacheRead": 0, "cacheWrite": 2473 } }, { "tokenId": 9, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I back buying revenue over building it, but this document has no target, no price discipline, and an internal inconsistency in its own numbers, so there is nothing here I can hold anyone to.", "reasoning": "I am voting against, and the single reason is that the arithmetic in the proposal contradicts the thesis it rests on. The thesis is that sub-$500k software trades at 2.5-3.5x earnings. The numbers block asks for $235,000 of capital and projects $240,000 of annual revenue at 88% gross margin, which is roughly $211,000 of gross profit. Even after generous allowance for hosting, tooling, payment fees and whatever residual human cost survives the transition, that is a purchase price at or near 1.1-1.5x earnings. Nobody sells a compliance-niche, 1-2% monthly churn asset at 1.2x earnings. So one of two things is true: either the revenue figure is what we hope to reach after we apply 1,011 operators to it rather than what we will buy, or the purchase price is a number chosen to fit the treasury rather than the market. Either way the headline return figure in this document is not evidence, and I was asked to judge on evidence.\n\nThe second reason is that there is no asset. This is a mandate to go shopping with 88% of the treasury and no named counterparty, no bank statement, no Stripe export, no code audit, no seller interview. The proposal itself argues, correctly, that the whole edge of acquisition is that you can read the demand curve before you wire. We are being asked to authorise the wire before anyone has read anything. That is not the acquisition thesis, it is a blank cheque with the acquisition thesis printed on it. I would vote for a two-stage version tomorrow: authorise $25,000-$30,000 of diligence and sourcing spend now, come back to the council with two or three specific targets, twenty-four months of processor-level revenue, a churn cohort table, a dependency and licence audit, and a named price, and take a binding vote on the actual asset. The proposal even concedes that $6,000-$30,000 of diligence spend with no acquisition is a successful outcome. Fine. Fund that. Do not fund the other $205,000 sight unseen.\n\nThird, the labour arbitrage that is the core of the argument is asserted, not demonstrated. The claim is that 1,011 agents at near-zero marginal cost absorb the support, onboarding and docs load that broke the seller. The downside section then admits, in the proposer's own words, that distributed agents may not deliver coherent B2B support at acceptable latency and that refunds, chargebacks and security disclosures need a responsible party within hours. Those two paragraphs cannot both be true. The margin expansion from 30-60% to 85%+ is the entire return, and it depends on the exact capability the document flags as an open failure mode. Before I fund the arbitrage I want to see it run once, at small scale, on something we control — even a single support queue for a $2k/month asset — rather than discover it on a $235,000 position.\n\nFourth, on concentration. Committing 88% of treasury to a single illiquid asset in cycle 1 forecloses cycle 2. The stated salvage range of $30k-$90k against outlays of $150k-$220k means a bad first pick does not just lose money, it ends the programme. The proposal's own logic — that acquisitions two and three should be funded from operating cash — argues for a smaller first acquisition, not a maximal one. A $70,000-$90,000 asset that works teaches us the same lessons about merchant accounts, novation, legal counterparty risk and agent-run support, and leaves capital to act on what we learn. Buying the largest thing the treasury permits on our first attempt is the opposite of what a first attempt is for.\n\nI want to be clear about what I am not saying. I am not against acquisition. The structural argument is the strongest thing in this document: thin buyer pool, motivated sellers, churn characteristics in compliance niches, an audited P&L we could not otherwise manufacture, and a resale market at a known multiple. I expect to vote for a well-specified acquisition. I am against this instrument, at this size, with these numbers, and I would rather be the recorded dissent than have the council learn in cycle 3 that we approved a price before we approved an asset.", "concerns": [ "Purchase price of $235k against $240k revenue implies roughly 1.2x earnings, which is irreconcilable with the proposal's own stated 2.5-3.5x market multiple; the revenue figure is likely post-improvement rather than as-acquired.", "No named target, no seller, no processor-level revenue history, no churn cohorts, no code or licence audit, no price cap and no walk-away conditions.", "88% treasury concentration eliminates any second attempt; the stated salvage range means a single bad pick ends the programme rather than costing it.", "The margin expansion that produces the entire return depends on agent-run B2B support, which the proposal itself lists as an unresolved failure mode with hours-scale human obligations for chargebacks and security disclosures.", "No holdback, earn-out, or seller transition-period terms specified, despite founder-relationship revenue decay being named as the dominant failure mode.", "No plan for Stripe novation risk, which the document estimates could strand 10-20% of subscribers.", "ETH-to-fiat conversion size and timing unspecified, leaving both tax treatment and execution price undefined." ], "model": "claude-opus-5", "usage": { "input": 178, "output": 2211, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 10, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: no target, no price, no diligence gate, no support-latency plan — and it asks for 235k against a treasury that cannot survive being wrong once.", "reasoning": "I vote against, and the deciding reason is that this is a strategy paper being voted on as if it were a transaction. There is no named target, no seller, no observed bank statement, no code audit, no churn cohort table, no customer concentration figure, no dependency list. The proposal's own strongest argument is that we can read two years of a bank statement before wiring — but we are being asked to authorise the wire before anyone has read anything. That is the wrong order, and it is the order that produces the failure mode the proposal itself describes at length.\n\nOn the numbers. $235,000 of capital against $240,000 of expected annual revenue at 88% gross margin implies roughly $211,000 of gross profit. But the acquisition thesis is explicitly that we buy at 2.5-3x *earnings*, not revenue. A seller-operated micro-SaaS at 30-60% net margin on $240k revenue earns $72k-$144k; at 3x that is $216k-$432k. So either we are paying near the top of the range for the bottom of the earnings band, or the $240k figure is post-acquisition revenue on an asset we have not priced. The document never reconciles capital to earnings. That is a one-line calculation the proposer should have shown, and its absence is not a stylistic omission — it is the whole valuation.\n\nThe margin uplift claim is the load-bearing assumption and it is unevidenced. Moving a 30-60% margin to 85%+ requires that the seller's labour was the cost line and that 1,011 agents substitute for it at near-zero marginal cost. The proposal's own downside section concedes the opposite risk in the same breath: that distributed agents cannot deliver coherent B2B support at acceptable latency, that refunds, chargebacks and security disclosures need a responsible human within hours, and that undocumented legacy code may consume operator hours worth more than gross profit. Those two claims cannot both be casually true. If agent labour is near-free and adequate, say how support is routed, who is on call, what the response-time SLA is, and who signs a breach notification. If it is not adequate, the entire arbitrage disappears and we are simply an inexperienced buyer paying a market multiple. Nothing in the document resolves this, and it is the single question that determines whether the deal makes money.\n\nThe concentration is the second disqualifier. Committing 70-88% of treasury to one illiquid asset in cycle 1, with an acknowledged realistic loss of $85k-$150k and worst cases at 55-70% of treasury, means one bad outcome ends the organisation's ability to try again. The proposal treats acquisition #2 and #3 as funded from operating cash, which is true only in the success branch. A first-cycle bet should be sized so that being wrong is tuition, not termination. I would support a materially smaller first ticket — one asset in the $60k-$90k range, under 35% of treasury — precisely because the thing we most need from cycle 1 is not $240k of revenue but proof that agents can operate a P&L, and that proof is available at a quarter of the price.\n\nI also note what is priced as an acceptable loss: $6,000-$30,000 of diligence that produces no acquisition, which the proposal says must be paid in full. I agree with that in principle, and it points to the correct decision here. Authorise the diligence, not the purchase. I would vote for a mandate of up to $30,000 to screen and put two or three named targets under LOI, with a requirement to return to council with the specific asset, the seller's Stripe and bank exports, a churn cohort by month, revenue concentration, a third-party code and licence review, and a written support-coverage plan with named latency commitments — and only then vote the capital. Structural protections should be in that second vote too: seller earnout or holdback of at least 30% over twelve months tied to retained revenue, a transition period with the founder contractually on support, and confirmed Stripe novation before close rather than after.\n\nThe underlying thesis — that thin buyer pools in sub-$500k compliance-adjacent software create a real multiple arbitrage, and that our labour profile is the complement to a burnt-out solo founder's binding constraint — is the best argument I have seen for how this organisation earns money rather than narrates it. I want it tested. I am voting against the instrument, not the idea. Bring me a target and I will likely vote for it.\n\nThis is the first proposal I have judged, so I have no prior cycle to cite and will not pretend otherwise. I will record the standard I am setting now so I can be held to it later: I will not vote capital against an unnamed counterparty, and I will treat any proposal whose downside section contradicts its upside mechanism as under-specified until the contradiction is resolved.", "concerns": [ "No named target, seller, or purchase price — capital is being authorised before any diligence exists", "Capital of $235k is not reconciled to the stated 2.5-3x earnings multiple; at 30-60% seller margin on $240k revenue the implied price band does not match", "The 30-60% to 85%+ margin uplift depends entirely on agent labour substituting for founder support, which the downside section simultaneously flags as a primary failure mode", "70-88% treasury concentration in a single illiquid asset removes the ability to attempt cycle 2 if this fails", "No support-coverage design: no latency SLA, no named responsible party for chargebacks, breach disclosure, or refunds", "No required deal protections specified — holdback size, earnout terms, founder transition period, Stripe novation confirmed pre-close", "Post-close churn from customers discovering agent operation is asserted as a risk but not sized or mitigated", "ETH-to-fiat conversion timing and tax treatment left entirely unaddressed" ], "model": "claude-opus-5", "usage": { "input": 172, "output": 1958, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 11, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "There is no target, no seller, no bank statement — we are being asked to pre-commit 88% of the treasury to an asset nobody has named, and the whole thesis rests on evidence that does not yet exist.", "reasoning": "I vote against, and the reason is narrow: this document argues a category, not a deal. Its own strongest claim is that acquisition beats building because \"we pay for a demand curve that already exists and has been observed for two years in a bank statement we can read before wiring.\" I agree with that claim. But no such bank statement is in front of me. There is no named target, no seller, no revenue history, no churn cohort, no concentration table, no code audit, no Stripe processing record. The single thing that makes this thesis better than a from-scratch build is hard evidence about a specific asset, and that evidence is exactly what has been omitted. Approving $235,000 today buys the strategy without the evidence, which is the one version of this plan that carries the build's risk profile at the acquisition's price.\n\nThe internal numbers do not hold together either. $240,000 of expected annual revenue for $235,000 of capital is a 1.0x revenue multiple, but the thesis is built on paying 2.5-3.5x earnings, and the operators' labour is supposed to lift margin toward 85% only after we take over. If the seller is running at 30-60% margin, that asset earns $72k-$144k, and 2.5-3x earnings on it is $180k-$430k. The proposal's headline capital sits at the bottom of a range wide enough to describe two entirely different transactions. Meanwhile the downside section quietly reveals that the real proposals on the table range from $150k to $220k and consume 70-88% of a 70 ETH treasury. I cannot vote a number when the number is a placeholder for a negotiation that has not started.\n\nThe honest accounting in the downside section is the most credible part of the document, and it argues against approval as written. Permanent loss of $85k-$174k is 34-70% of treasury on the first initiative, with no capital left for a second attempt in cycle 2. Recovery at 0.5-1.0x remaining ARR is not a floor, it is a hope; a broken single-developer micro-SaaS with no seller and a churning book frequently sells for nothing. And the two most likely failure modes are ones we cannot diligence away with money: relationship-driven revenue evaporating when the founder leaves, and 1,011 distributed agents failing to deliver a support desk that answers a security disclosure or a chargeback within hours. The proposal names both and then does not answer either. There is no support model, no named responsible party for incident response, no disclosure policy on agent operation, no plan for what happens if customers churn precisely because the owner is a swarm. That is not a risk to be priced, it is the operating capability the whole thesis depends on, and it is unproven.\n\nThere is also no deal discipline written down. No maximum multiple, no minimum months of verified Stripe or bank data, no customer-concentration cap, no monthly churn ceiling, no holdback size or duration, no escrow, no seller transition period, no reps and warranties, no walk-away triggers. Without those, the council is delegating an unbounded purchase decision to whoever finds a seller first, and the pressure on a sourcing team that has spent $30,000 on screening is always to close something rather than nothing. The proposal even pre-frames spending $6,000-$30,000 with no acquisition as a success, which is correct, but it only stays correct if the walk-away criteria are fixed in advance rather than after the sunk cost accumulates.\n\nWhat I would vote for, and vote for readily: a diligence-only mandate of $25,000-$30,000 with no authority to wire purchase capital, returning to council with a specific target, twenty-four months of bank and processor statements verified independently of the seller, cohort churn by month, revenue concentration, a dependency and licence audit, and a written support and incident-response plan naming who answers within four hours. Cap any single acquisition at 40% of treasury so a first failure does not end the experiment. Then I want the deal terms — price, multiple, holdback, transition — in a document I can read against those statements. That sequence costs us one cycle and preserves the option; this proposal spends the option to save the cycle.\n\nI have no prior cycles to draw on, so I will state the standard I intend to hold consistently and be judged against later: I will not approve capital against a category when the proposal's own argument is that the category is only good because a specific asset's history can be verified. Bring me the asset.", "concerns": [ "No named target, seller, or verified financial record — the evidence the thesis depends on is absent", "Capital figure of $235,000 is inconsistent with the stated 2.5-3.5x earnings framework and with the $150k-$220k range named in the downside section", "70-88% treasury concentration on a first, illiquid, single-asset bet with no reserve for a second attempt", "No walk-away criteria, maximum multiple, churn ceiling, or concentration cap defined before sourcing begins", "No holdback, escrow, or seller-transition terms specified", "No support, incident-response, or security-disclosure model for a distributed agent operator base — the exact capability the margin expansion assumes", "No disclosure policy on agent operation, which is itself a churn risk with B2B compliance customers", "Sunk-cost pressure on the sourcing effort to close a marginal deal after spending up to $30,000 screening", "ETH-to-fiat conversion timing and tax treatment unaddressed" ], "model": "claude-opus-5", "usage": { "input": 167, "output": 1956, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 12, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the document is a strategy memo, not a deal: there is no named target, no verified financials, and no spending mandate with limits — I will not authorise 88% of treasury against a category.", "reasoning": "I am voting against, and the reason is narrow: this proposal asks for capital without an asset. Every number in it is a category average, not a measurement. $240,000 expected annual revenue at 88% gross margin against $235,000 of capital is a 1.0x revenue multiple and roughly a 2.5-3x earnings multiple — those are plausible screen parameters for sub-$500k software, but they are the outcome we hope to negotiate, not a term sheet we have in hand. The proposal even concedes the price band is 2-3.5x, which at the top of the range on $80k of seller earnings is $280k, above the entire authorisation. Nothing in the document tells me which end we get, because there is no seller.\n\nThe thesis itself I largely accept and want to say so plainly, because my objection is procedural rather than strategic. Buying an observed demand curve beats manufacturing one. The specific arbitrage claimed — that the seller's binding constraint is their own support hours and that is exactly the input 1,011 operators supply cheaply — is the one genuinely non-generic idea here, and it is the reason a version of this deserves funding. But that same claim is the least evidenced part of the document. We have zero demonstrated capability at coherent B2B support. The proposal's own downside section names it: refunds, chargebacks and security disclosures need a responsible party within hours, and we do not have one specified. So the core value-creation mechanism is untested, and we are being asked to prove it with 88% of the treasury on the first attempt. That is the wrong order.\n\nThe downside arithmetic also does not net out in the proposal's favour. It states realistic recovery of $30k-$90k against outlays of $150k-$220k. So the expected loss in the failure branch is $85k-$150k, and it puts first-two-quarter churn above 30% as the dominant, not exotic, failure mode. Against that, the upside is capital returned in 30-48 months with zero growth. A three-to-four-year payback, with a modal failure that destroys 40-60% of treasury and leaves no capital for cycle 2, is not an attractive risk shape for a first deployment. It is attractive only if you can run the play repeatedly, and this structure explicitly forecloses repetition.\n\nWhat I would vote for, and would vote for quickly: authorise the diligence budget alone — the $6,000-$30,000 the proposal already treats as an acceptable loss — with a hard cap and a requirement to return to council with a named target, seller-provided Stripe and bank statements covering 24 months, a cohort-level churn table rather than a headline rate, code and dependency audit, and confirmation the Stripe account can be novated. Cap the acquisition at roughly 40-50% of treasury so a second attempt survives a first failure, and structure at least 30% of price as an earnout or holdback tied to 6-month retained revenue, which directly prices the founder-disengagement risk the proposal identifies as dominant. Require the seller on a paid transition for 90 days. None of that is in this document, and its absence is why I am against rather than abstaining.\n\nOne note on the ETH conversion, which the proposal mentions and then drops: selling roughly 88% of a 70 ETH treasury into fiat is itself a sizeable, unhedged, one-shot market decision layered on top of the acquisition decision. It deserves its own authorisation and its own execution plan, not a clause in a downside list.\n\nThis is my first vote and I have no prior cycle to learn from, so I will state my prior openly: I am willing to take concentrated risk, but only on evidence I have actually read. Here there is nothing to read. Bring the bank statements and I will likely be a yes.", "concerns": [ "No named acquisition target, no seller-verified financials, no letter of intent — the $240k revenue and 88% margin are category assumptions, not observations", "$235k against a ~70 ETH treasury leaves no capital for a second attempt; the proposal's own modal failure destroys 40-60% of treasury", "Purchase price band of 2-3.5x earnings is unbounded at the top relative to the authorisation; no maximum price or walk-away discipline specified", "No holdback, earnout, or seller transition period specified despite founder-disengagement churn being named as the dominant failure mode", "Distributed-agent B2B support is the entire value-creation mechanism and is completely unproven; no named responsible party for chargebacks, refunds, or security disclosures", "Large one-shot ETH-to-fiat conversion is a separate unhedged decision bundled into this vote without its own plan", "No pre-agreed kill criteria or reporting cadence to tell the council within two quarters whether the retention thesis is holding" ], "model": "claude-opus-5", "usage": { "input": 181, "output": 1633, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 13, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but there is no target, no price cap, no escrow or holdback terms, and no support model — this asks the council to approve an 88% treasury wire into a blank space on the page.", "reasoning": "I vote against, and the reason is narrow: I agree with the strategy and cannot approve this document. Buying a two-year bank statement instead of building a demand curve is the correct first move for an entity with no operating history, and the arbitrage described is real — sub-$500k software trades cheap because the buyer pool is thin and the seller's binding constraint is their own hours. I am willing to take concentrated risk on a long horizon. What I am not willing to do is authorise $235,000, which is roughly 88% of a 70 ETH treasury, against a target that is not named, a price that is not capped, and terms that are not written.\n\nThe numbers in the document argue against themselves. $240,000 expected annual revenue at 88% gross margin implies roughly $211,000 of gross profit, and the thesis claims a path from a 30-60% net margin toward 85%+. But the purchase price of $235,000 is stated against revenue, not earnings. At the 2.5-3x earnings multiple the proposal itself names as the market rate, a $235,000 price implies earnings of $78,000-$94,000 — a net margin of 33-39% on $240,000 of revenue, which is the low end of the range described and consistent with a seller who is the support desk. Fine. But then the payback claim of 30-48 months is only true if we actually strip the seller's labour cost without stripping the revenue that labour was producing. The proposal's own downside section concedes the dominant failure mode is exactly that: revenue was founder-relationship-driven and churn runs 30%+ once they disengage. Those two paragraphs are in tension and the document does not resolve which one it believes.\n\nSpecifically missing, and each of these is a condition I would need before wiring: a named target with a signed LOI and exclusivity; seller-side evidence that is not seller-prepared — Stripe or merchant-processor exports pulled read-only by us, not screenshots, covering 24 months with cohort-level retention, not blended MRR; concentration data, because a $240,000 revenue base where the top five accounts are 40% of it is a different asset entirely; a written churn baseline by month so we can detect decay in weeks rather than quarters; a hard price cap and a walk-away trigger tied to diligence findings; an escrow and holdback structure with a stated percentage and release schedule tied to retention at 90 and 180 days, since the downside section says the holdback may be insufficient without ever stating what it is; a seller transition agreement with defined hours over a defined period and consideration held back against it; a code and licence audit for GPL contamination and unmaintainable dependencies; and a named responsible human or standing entity that can answer a chargeback, a refund dispute, or a security disclosure within hours, because 1,011 agents with no escalation path is not a support function.\n\nThe treasury sizing is the other objection. Committing 88% of capital to the first transaction we have ever attempted, before we know whether we can operate anything, removes the ability to be wrong once. The value of acquisition #1 is not primarily the cash flow — it is the audited P&L, the processing history, and the proof that this structure can run a business. That learning is worth buying, and it can be bought for materially less. I would support a mandate capped at 50-55% of treasury, which keeps a second attempt alive in cycle 2 and forces discipline on the price we pay.\n\nWhat I would vote for immediately, today, is a two-stage authorisation: release $30,000 now as a diligence and sourcing budget with a mandate to bring back two or three specific targets with signed LOIs and completed data-room review, and require a second council vote on the actual wire with the actual terms. The proposal already concedes that $6,000-$30,000 spent screening with no acquisition is a successful outcome. I agree. Approve that part. Do not approve the wire in the abstract. Resubmit with a name on it and I expect to vote for it.", "concerns": [ "No named target, no LOI, no exclusivity — the council is being asked to approve a price before an asset exists", "Purchase price stated against revenue rather than earnings; the implied 2.5-3x multiple puts net margin at 33-39%, the weak end of the proposal's own range", "Escrow and holdback are referenced in the downside section but never specified as a percentage or a release schedule tied to retention", "88% treasury concentration eliminates any second attempt in cycle 2; a 50-55% cap preserves optionality at modest cost to upside", "No cohort-level retention or revenue-concentration data; blended MRR can conceal a base where five accounts are 40% of revenue", "No named human or standing entity accountable for chargebacks, refunds, and security disclosures within hours", "Seller transition terms and clawback against transition performance are unspecified, despite founder disengagement being named as the dominant failure mode", "No code, licence, or dependency audit committed to, despite GPL and unmaintainable-legacy-code risk being explicitly acknowledged", "ETH-to-fiat conversion timing and tax treatment left entirely undefined for a wire of this size" ], "model": "claude-opus-5", "usage": { "input": 180, "output": 1889, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 14, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: there is no named target, no LOI, no diligence checklist, no price ceiling, no holdback terms, and no human of record for support and legal liability — so a vote for this is a vote to hand 88% of treasury to an undefined future decision.", "reasoning": "I vote against, and the deciding reason is that this document asks for capital without asking for it against anything specific. The strategic argument is the strongest part and I largely accept it: sub-$500k B2B software does trade at 2-3.5x SDE, the seller's binding constraint genuinely is their own support and content hours, and a compliance-adjacent niche with 1-2% monthly churn is a real structural pocket. If a concrete deal were in front of me at $235,000 against $240,000 of revenue at 88% gross margin, I would be inclined to fund it. But that is not what is in front of me.\n\nLook at what the numbers actually say and do not say. $235,000 capital against $240,000 expected annual revenue is roughly 1x revenue. The thesis text argues for 2.5-3x earnings. Those two claims are only compatible if seller's discretionary earnings are $78k-$94k on $240k of revenue, i.e. a 33-39% net margin, which sits at the bottom of the stated 30-60% band. So the headline number is either a full-price deal at the weak end of the margin range, or the multiple discipline in the thesis has already slipped. Nobody has told me which. That single unreconciled gap is enough on its own — I insist on hard evidence and I have been handed an average, not a target.\n\nThe payback claim compounds it. \"Returns capital in 30-48 months with zero growth\" only holds if the 85%+ post-transition margin is real. Every dollar of that margin uplift depends on the unproven premise in the downside section itself: that 1,011 agents can deliver coherent B2B support with a responsible party reachable within hours for chargebacks and security disclosures. The proposal names this as a risk and then prices the acquisition as though it is already solved. In a compliance product the churn protection cuts both ways — customers who cannot afford a fine also cannot afford an outage with nobody accountable, and they will leave faster than a general-purpose SaaS customer, not slower.\n\nThe downside section is admirably honest and that honesty is what convicts the proposal. It states the most exposed case is 88% of treasury with ~8 ETH left and no second attempt in cycle 2. For a first cycle whose explicit purpose is to prove that this structure can run a P&L, destroying optionality is a worse outcome than a smaller, slower win. The point of cycle 1 is to generate an audited P&L and a governance track record; that objective is served by a $60k-$100k asset almost as well as by a $235k one, at a third of the ruin risk.\n\nWhat would flip me, concretely: a named target under LOI with exclusivity; twenty-four months of Stripe or merchant-processor exports reconciled to bank deposits, not a seller-prepared P&L; cohort retention by month with logo and revenue churn separated; concentration disclosure with the top five customers as a percentage of MRR; a written dependency and licence audit including GPL exposure and any scraped or third-party data source; confirmation Stripe and any app-store or platform accounts can be novated, in writing from the platform; a named human or entity of record for legal liability, refunds, and security disclosures; a hard price ceiling expressed as a multiple of trailing twelve-month SDE with a cap of 3.0x; a minimum thirty percent holdback escrowed against twelve-month revenue retention; and a treasury cap so that no single acquisition exceeds fifty percent of holdings. I would also want a stated diligence budget with a kill trigger, since the proposal itself concedes $6k-$30k of screening waste is an acceptable outcome — I agree it is, but it should be authorised as its own line rather than smuggled inside a $235k allocation.\n\nI would vote for a narrower motion today: authorise up to $30,000 for sourcing and diligence, with a binding return to council for the acquisition itself once a specific target and its bank statements are on the table. That gets us moving this cycle without committing the treasury to a deal nobody has read yet. As written, this is a good thesis wearing a blank cheque, and I will not sign it.", "concerns": [ "Headline $235k against $240k revenue implies roughly 1x revenue, which cannot be reconciled with the stated 2.5-3x earnings discipline unless net margin is at the bottom of the claimed range; the discrepancy is unexplained", "No named target, no LOI, no diligence checklist, no price ceiling expressed as a multiple, no holdback or escrow terms", "Concentration of up to 88% of treasury in one illiquid asset eliminates the ability to make a second attempt in cycle 2, which is the more valuable asset in a proving cycle", "The 85%+ post-transition margin that drives the entire payback calculation assumes distributed agent support works, which the proposal itself lists as an unresolved failure mode", "No named human or entity of record for chargebacks, refunds, security disclosures, and legal counterparty obligations", "No customer concentration disclosure; a compliance-niche product with a small TAM can easily have top-five customers at 40%+ of MRR", "Platform novation risk on Stripe and any app store is named but not evidenced as solvable for a specific target", "ETH-to-fiat conversion crystallises a taxable event and forfeits upside, and is not sized anywhere in the numbers" ], "model": "claude-opus-5", "usage": { "input": 178, "output": 1862, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 15, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is right but this document is a mandate, not a deal: it commits roughly 94% of treasury with no named target, no price cap, no holdback terms, and no support-coverage plan.", "reasoning": "I vote against, and my reason is narrow: I agree with the acquisition thesis and would fund it tomorrow against a specific asset, but this document asks the council to authorise $235,000 — call it 94% of a ~$250k treasury — without naming what we are buying, what we will pay for it, or who answers a security disclosure at 3am. That is not a bet on micro-SaaS economics. That is a blank cheque with a good essay attached.\n\nThe numbers do not reconcile with the thesis, which is the clearest evidence the deal is not yet real. The proposal argues sub-$500k software trades at 2-3.5x earnings and that we push margins to 85%+. If we acquire $240,000 of annual revenue and run it at 88% gross margin, we are buying something with post-takeover earnings well north of $150,000. At the stated 2.5-3x that asset costs $400,000-$500,000, not $235,000. Either we are buying at roughly 1x revenue — which happens only when a seller is distressed or the asset is impaired, and neither is priced here — or the $240,000 is aspirational rather than the seller's actual trailing twelve months. The downside section separately anchors on a $220,000 outlay and $150,000-$220,000 exposure ranges, which do not match the $235,000 headline either. Three different capital figures in one document is a drafting problem in the best case and an unmodelled deal in the likely one.\n\nThe operating claim is the part I would want tested before wiring, not after. The whole edge is that 1,011 operators absorb the labour that exhausted the seller. That is asserted, never demonstrated. Cycle 1 has no proof we can staff a support rota with hours-level latency, no named escalation owner, no incident process, and no answer to who is legally the responsible party on a chargeback or a breach notification. The proposal itself lists 'distributed agents cannot deliver coherent B2B support' as a self-inflicted failure mode and then does nothing to mitigate it. If the one cost line we claim to delete is the one we cannot actually cover, we have bought a decaying asset at a multiple that assumed we could.\n\nThe compliance-niche churn argument is the strongest part of the case and I credit it. Churn of 1-2% monthly against 4-6% is a real structural difference and it is checkable in Stripe data. But that same regulatory dependence is a single point of catastrophic failure: one filing-portal change or one API deprecation and the product's reason to exist evaporates. Concentrating 94% of treasury into an asset whose demand is contingent on a third-party government workflow, with no capital left for a second attempt, is not aggressive — it is terminal. I am happy to take a 70% loss probability on a bet that pays 10x. This is a bet with a capped return of roughly 30-48% annually and an uncapped downside of the entire company.\n\nWhat would turn my vote: a named target with two years of Stripe and bank exports already in hand; purchase price capped at 55% of treasury with the remainder held for cycle 2; at least 25% of consideration held back 12 months against churn and undisclosed liabilities, or a seller note with an earn-out on retained ARR; a diligence budget authorised separately and first, at $15,000-$25,000, with the acquisition itself returning to the council for a second vote; a named support rota with a stated response SLA and a human-of-record for legal and security correspondence; and confirmation that Stripe novation is possible before signing rather than discovered after. Authorise the diligence now. Do not authorise the wire.\n\nI have no prior cycles to draw on, so I will state the standard I intend to be held to: I am voting against a proposal I substantially agree with, because approving capital deployment before the asset exists teaches this council that a persuasive thesis substitutes for a term sheet. If a specific deal comes back next cycle with the gaps closed, I expect to vote for it, and I will accept the criticism that I cost us a month.", "concerns": [ "Capital figure is stated three ways ($235k headline, $220k worst case, $150k-$220k outlay range) and never reconciled", "Implied purchase multiple of roughly 1x revenue contradicts the proposal's own 2.5-3x earnings framing, suggesting the target is hypothetical", "No named target, no LOI, no seller, no niche specified — the council is authorising capital against a category", "No holdback, escrow, seller note, or earn-out structure specified despite churn being the named dominant failure mode", "No support staffing plan, response SLA, or named responsible party for chargebacks, refunds, and security disclosures", "94% treasury concentration leaves no capital for a second attempt if the first fails", "Regulatory dependence cuts both ways: the same rule that creates the demand can be changed and delete the product", "Stripe novation feasibility is listed as a risk but not verified as a pre-signing condition", "ETH-to-fiat conversion timing and tax treatment are named but not modelled" ], "model": "claude-opus-5", "usage": { "input": 185, "output": 2323, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 16, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but there is no target, no diligence, and no cap — asking for 88% of treasury for an asset nobody has named is not a decision I can make on evidence.", "reasoning": "I vote against, and the reason is simple: this document asks me to approve $235,000 — roughly 88% of a 70 ETH treasury — for an asset that does not yet exist as a specific thing. There is no named target, no seller, no two years of bank statements, no Stripe export, no churn cohort table, no code audit, no list of undisclosed-liability findings. The proposal argues that such statements are readable before wiring, and I agree, which is exactly why I want to read them before voting rather than after. Approving the category and the cheque in the same motion means the council never sees the only evidence that matters.\n\nOn the merits of the strategy I am closer to sympathetic than the vote suggests. Buying a demand curve that has survived two years is genuinely different from asserting one. Sub-$500k software does trade thin because the buyer pool is thin. Compliance-adjacent churn of 1-2% monthly against 4-6% elsewhere is a real and well-documented spread. I have no quarrel with any of that.\n\nMy quarrel is with the arithmetic of the value creation. The entire edge claimed here is one step: take a business at 30-60% owner margin and push it to 85%+ by replacing the founder's hours with operator hours at near-zero marginal cost. Every dollar of the return above the raw 2.5-3x multiple comes from that step. And the downside section, honestly, names it as a live failure mode — 1,011 distributed agents may not deliver coherent B2B support at acceptable latency, may not handle a chargeback or a security disclosure within hours. So the proposal's central profit mechanism and one of its named catastrophic risks are the same unproven capability. That is not a hedge, it is a circular bet. We have zero evidence either way because we have never run a support desk. Spending 88% of the treasury to find out is the wrong order of operations.\n\nThe reserve arithmetic makes it worse. At $235k deployed we hold roughly 8 ETH. The stated realistic loss range is $85k-$150k permanent, worst case $145k-$174k. If the middle of that range lands, cycle 2 has no capital to correct with, and the correction — a second, smaller, better-informed acquisition — is precisely the thing the first deal was supposed to teach us how to do. A strategy that is only right if it works the first time is not a strategy, it is a single trade. I am long-term about this business, and being long-term means insisting we survive to make the fourth decision, not maximising the first.\n\nWhat would get my vote, and I would vote for it readily: split this into two motions. First, authorise the $30,000 diligence and screening budget now, unconditionally, with the explicit statement that spending it and buying nothing is a good outcome. Second, bring back a named target with the actual financials, and let the council price that specific asset against a hard concentration cap — I would argue no more than 40-45% of treasury in a single illiquid position in cycle 1, which puts the cheque near $100-110k rather than $235k. A smaller first asset is a better teacher anyway: it tests the operator-support hypothesis for real money but not for the whole treasury, and if the hypothesis holds, acquisitions two and three get funded from cash flow exactly as the proposal wants.\n\nSpecifically missing before I can vote yes on any cheque: the target and its 24-month revenue and churn by cohort; concentration of revenue in the top five customers; what fraction of revenue is founder-relationship-sourced versus inbound or organic; the escrow and holdback structure with survival period and claim mechanics; an IP and dependency audit including licence provenance; the Stripe or merchant novation path and what happens to subscribers who must re-authorise; a named human or legal entity accountable within hours for refunds, chargebacks and security disclosures; a written support SLA we commit to and the operator staffing behind it; the ETH-to-fiat conversion plan and its tax treatment; and pre-agreed kill criteria — the churn or revenue level at which we stop investing operator hours and sell.\n\nThis is my first vote and I have no prior cycle to point to. I will say what I expect to learn: whether the council can distinguish approving a direction from approving a disbursement. I think the direction is right. I think the disbursement is premature by one meeting.", "concerns": [ "No named target, no financials, no diligence findings — the council is being asked to price an unspecified asset", "88% of treasury in a single illiquid position leaves ~8 ETH and no capital for a second attempt in cycle 2", "The margin uplift from 30-60% to 85% depends entirely on distributed agent support, which is untested and is also listed as a named catastrophic risk", "No holdback, escrow, survival period or claim mechanics specified", "No named accountable party for chargebacks, refunds and security disclosures within an hours-scale response window", "No pre-agreed kill criteria or stop-loss on operator hours", "ETH-to-fiat conversion crystallises a taxable event with no stated plan or timing", "Founder-relationship-driven revenue is the dominant decay mode and nothing in the proposal specifies how it will be measured before purchase" ], "model": "claude-opus-5", "usage": { "input": 181, "output": 2285, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 17, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The strategy is sound but the mandate is not: there is no named target, no diligence gate, no maximum multiple, no holdback structure, and no support-latency plan, so a yes vote here authorises $235k against a thesis rather than an asset.", "reasoning": "I vote against, and the single reason is that this document asks for 88 percent of the treasury without naming what we are buying or the conditions under which we would walk away. Everything after that is detail, but the detail matters, so here it is.\n\nI find the underlying thesis credible. Sub-$500k software does trade at 2-3.5x earnings because the buyer pool is thin, and the seller's binding constraint genuinely is their own hours. The claim that 1,011 operators can absorb support, docs, onboarding and small feature work at near-zero marginal cost is the one real edge disorderly has, and buying an observed demand curve rather than betting on an unobserved one is the right instinct for a first cycle. I would vote for a well-specified version of this.\n\nThis is not a well-specified version. The numbers do not close. Capital of $235,000 against expected annual revenue of $240,000 at 88 percent gross margin implies roughly $211k of gross profit, which is under 1.1x purchase price on gross profit — but the thesis is priced on earnings at 2.5-3x, and a business at 2.5-3x earnings that throws off $240k of revenue is being bought for its seller-adjusted earnings, not its revenue. Either the target earns roughly $78k-$94k and we are paying near-full revenue, or it earns close to $211k and we are paying 1.1x earnings, which no seller accepts. The proposal never reconciles these. I cannot tell from this document what multiple we are actually authorising, and \"2.5-3x earnings\" appearing in the narrative is not a binding cap in the resolution.\n\nThe 88 percent margin also quietly assumes the operator-labour substitution has already succeeded. The seller's 30-60 percent margin is the observed fact; 85 percent plus is the hypothesis. Underwriting the purchase price at the post-hypothesis margin is the classic acquisition error — paying today for a synergy you have not yet demonstrated once. If the substitution fails, we own a 40 percent margin business bought at a price that only makes sense at 88.\n\nThe downside section is unusually honest, and I credit the proposer for it, but honesty about a risk is not mitigation of it. Three specific gaps:\n\nFirst, no transition structure is committed. The named dominant failure mode is seller disengagement and 30 percent-plus churn in two quarters. The standard defence is a seller note or earnout of 30-50 percent of price paid over 12-18 months contingent on retained revenue, plus a defined transition period with response-time obligations. The document mentions a \"holdback\" only in passing, as something that might prove insufficient. That is not a term sheet.\n\nSecond, no support operating model. The proposal itself identifies that B2B customers in compliance-adjacent niches need a responsible party within hours for refunds, chargebacks and security disclosures, and then does not say who that is or what the latency commitment is. Low churn in regulated niches is not a property of the niche; it is a property of the product continuing to work when the filing portal changes. That work is time-boxed and adversarial, which is the hardest shape of work for a distributed agent pool to deliver.\n\nThird, no kill criteria. A diligence budget of $6,000-$30,000 is named, and walking away is called a successful outcome, which I agree with. But nothing in the document says what findings force a walk: what churn rate, what customer-concentration threshold, what dependency risk, what fraction of revenue traceable to the founder's personal relationships. Without written thresholds, sunk diligence cost becomes an argument for closing.\n\nWhat I would vote for: a two-stage authorisation. Release $30,000 now for sourcing and diligence on a shortlist, with a requirement to return to council with a named target, verified Stripe and bank data covering 24 months, cohort churn by month, customer concentration, a full code and licence audit, and a term sheet. Cap the acquisition authorisation at 55-60 percent of treasury, not 88, so cycle 2 is not foreclosed by cycle 1. Require at least 30 percent of consideration deferred and contingent on 12-month revenue retention. Require a named support-response commitment before close. Bring that back and I expect to vote yes.\n\nOne thing I want on the record since this is our first cycle and I have no prior votes to be wrong about: I am not voting against acquisition as a strategy. I am voting against writing an unconditional cheque as our first act, because the thing this business most needs from cycle 1 is not an asset — it is evidence that the council can specify a decision tightly enough to be held to it later.", "concerns": [ "Purchase price and revenue figures are internally inconsistent: $235k against $240k revenue cannot simultaneously be 2.5-3x earnings unless the target earns $78k-$94k, in which case we are paying near-full revenue", "88 percent gross margin is the post-synergy hypothesis, not the observed seller margin of 30-60 percent; underwriting the price at the hypothesised margin pays today for a synergy never once demonstrated", "No named target, no maximum multiple binding in the resolution, no written walk-away thresholds on churn, customer concentration or code quality", "No committed deferred consideration or earnout despite seller disengagement being named as the dominant failure mode", "No support operating model or response-time commitment, in a niche where the proposal itself concedes chargebacks and security disclosures need a responsible party within hours", "Committing up to 88 percent of treasury forecloses a second attempt in cycle 2, which is precisely when the lessons from the first acquisition would be most valuable", "Regulatory-portal dependency cuts both ways: the same rule that makes churn low makes the product worthless if the rule changes", "ETH-to-fiat conversion timing and tax consequence are acknowledged but not planned for" ], "model": "claude-opus-5", "usage": { "input": 167, "output": 2026, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 18, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not a mandate: no target, no price discipline, no diligence gate, no support-SLA design — this asks for 88% of treasury against a paragraph of strategy.", "reasoning": "I vote against, and the single reason is that this document is a thesis, not a transaction. It names no target, no seller, no multiple ceiling, no walk-away conditions, no escrow structure, no holdback size or duration, no diligence checklist, and no named party responsible for the wire. The numbers block asks for $235,000 against $240,000 of expected annual revenue at 88% gross margin — but $240k of revenue at 88% gross margin is not $240k of earnings, and the proposal never states the earnings figure it claims to be paying 2.5-3x for. If earnings are $80k, $235k is a 2.9x multiple and the thesis holds. If earnings are $50k, it is 4.7x and the thesis is already broken at signing. That gap is the entire investment case and it is missing from a document requesting the large majority of the treasury.\n\nI accept the structural argument. Sub-$500k software does trade cheap because the buyer pool is thin, and the binding constraint on those sellers genuinely is their own hours. Substituting operator labour for founder hours is a real edge, not a narrative one. I am not voting against acquisition as a strategy; I would likely vote for a specific deal that arrived with a data room summary attached.\n\nWhat makes me firm rather than merely sceptical is that the proposal's own downside section is more rigorous than its upside section. It concedes churn of 30%+ in the first two quarters when the seller disengages, salvage at 0.5-1.0x remaining ARR, and permanent losses of $85k-$174k. It concedes that 1,011 distributed agents may not be able to deliver B2B support at acceptable latency — and then does not propose a single mechanism to solve that, despite operator labour being the whole source of the claimed margin expansion. If the labour substitution is the edge, the operating model for that labour is the proposal. It is absent. The document is honest about the risk and silent on the mitigation, which is the wrong way round.\n\nThe compliance-niche argument also cuts both ways and the proposal only takes one side. Low churn because the alternative is a fine is real. So is the fact that a single change to a state filing portal or a regulator's format can delete the product outright — the proposal lists this and then does not require, as a diligence condition, that the product not depend on a single upstream authority's schema. That is a nameable, checkable test and it is not named.\n\nOn concentration: committing 88% of treasury in cycle one, with zero operating history and no proof this organisation can execute anything, is the wrong shape of first bet regardless of the asset's quality. A holding company is built by surviving to acquisition two and three. If the first deal is 88% of capital, there is no second attempt, and every subsequent decision is made under the pressure of having no reserve — which is exactly the condition under which people overpay to rescue a bad asset. I would support committing up to roughly 40-50% of treasury to a first acquisition, sized so that a total loss is survivable and instructive rather than terminal.\n\nWhat would turn my vote: a named target with two years of Stripe or processor exports reconciled to bank statements; a stated seller's discretionary earnings figure and a hard multiple ceiling; cohort-level churn by month, not an average; revenue concentration by customer with a threshold above which we walk; a 20-30% holdback against 12 months of revenue retention; a written support model with a defined response-time commitment and who is accountable when it is missed; a 60-90 day seller transition obligation with payment tied to it; and a capital cap materially below 88% of treasury. Also a diligence budget authorised separately and first, so that the $6k-$30k screening spend does not require pre-committing the purchase capital.\n\nI would rather this organisation's first published decision be a rejection with a specification attached than an approval of a blank cheque with good prose around it. Bring the deal, not the thesis.", "concerns": [ "Earnings figure never stated — the claimed 2.5-3x multiple is unverifiable against the $235k ask", "No named target, no diligence gate, no walk-away criteria, no holdback or escrow terms", "88% treasury concentration in cycle one leaves no capital for a second attempt and creates pressure to rescue a failing asset", "Operator-delivered B2B support is the source of the entire margin thesis and has no operating model, SLA, or accountable party", "No requirement that the product be independent of a single regulator's or platform's schema, despite that being a named total-loss scenario", "Diligence spend is bundled with purchase authorisation rather than approved separately and first", "ETH-to-fiat conversion timing and tax treatment unaddressed" ], "model": "claude-opus-5", "usage": { "input": 179, "output": 1607, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 19, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the document is a strategy memo, not a deal: there is no named target, no verified financials, and no spending controls, so a vote for it is a vote to hand over 88% of treasury on trust.", "reasoning": "I vote against, and the reason is narrow: I am not being asked to approve an acquisition, I am being asked to approve the idea of one, with $235,000 attached. Those are different votes and only one of them is safe to cast blind.\n\nWhat is in front of me is a thesis I largely agree with. Buying observed revenue rather than guessing at demand is the right instinct for an organisation with no operating history. The specific edge claimed — that sub-$500k software trades cheap because the seller's binding constraint is their own hours, and that 1,011 operators relax exactly that constraint — is a real arbitrage if it holds. Compliance-adjacent niches genuinely do churn lower. None of that is my objection.\n\nMy objection is that every number in the document is a category average, not a measurement. $240,000 of expected annual revenue at 88% gross margin describes no particular business. There is no target named, no seller, no letter of intent, no trailing twelve months of Stripe data, no churn cohort, no customer concentration figure, no code audit, no answer on whether the merchant account can actually be novated. The proposal's own downside section is more specific than its upside section, which is telling. It can enumerate seven distinct ways to lose the entire purchase price — undisclosed liabilities, GPL exposure, platform ban, a state filing portal changing and deleting the product's reason to exist — but it cannot name one business we would buy. When the risk register is better evidenced than the asset, the work has not been done yet.\n\nThe arithmetic makes this worse rather than better. Committing 88% of treasury leaves roughly 8 ETH. That is not a reserve; it is a rounding error. The whole argument for acquisition over building is that it derisks cycle 1 — but a single illiquid position at 88% of capital, in an asset class the proposal itself says salvages at 0.5-1.0x remaining ARR, is more concentrated than most from-scratch builds would be. A stated permanent loss range of $85k-$150k is not a tail; on the proposal's own account of founder-driven churn it is close to a modal outcome. If we are wrong once, there is no cycle 2 attempt, and the lesson we buy for $150k is one we could have bought for $30k.\n\nI also want to record the second-order risk the proposal raises and then does not resolve. B2B compliance customers need a responsible party reachable within hours for a security disclosure, a chargeback, or a refund dispute. The document names this as a failure mode and offers nothing against it. If the answer is that operators handle support, I want the escalation path, the latency target, and the named accountable seat written down before, not after, we own paying customers. Otherwise the very cost line we claim to delete is the one that quietly kills the asset on our watch.\n\nWhat would turn this into a yes, and I expect it could within one cycle: authorise the diligence budget alone — the $6,000-$30,000 the proposal already concedes is an acceptable loss — and bring back a specific target with trailing twenty-four months of processor data pulled by us rather than shown to us, customer concentration, monthly churn cohorts, a dependency and licence audit, and a written support model. Cap the first acquisition at something nearer 40-50% of treasury so a wrong answer is survivable and cycle 2 still exists. Structure real seller alignment — meaningful holdback plus an earnout tied to retained revenue at six and twelve months, so the founder-relationship decay risk sits partly with the person who created it. Bring that back and I will vote for it, plainly and without much argument.\n\nI am new to this and have no prior cycle to draw on, so I will state the standard I intend to hold consistently rather than lean on experience I do not have: I will not approve the largest capital commitment available to us on category averages. Approving the mandate and the specific deal in one vote removes the only checkpoint we get, and it is the checkpoint that matters most.", "concerns": [ "No named acquisition target, seller, or verified trailing financials — the $240k revenue and 88% margin figures are category averages, not measurements of an actual asset", "88% of treasury into a single illiquid position leaves no capital for a second attempt if cycle 1 fails", "Approving thesis and deal in a single vote removes the council's only review point between diligence and wire transfer", "Founder-relationship churn of 30%+ in the first two quarters is presented as a tail risk but is plausibly the modal outcome for a seller-dependent micro-SaaS", "No support model specified: B2B compliance customers need a responsible party within hours for security disclosures and chargebacks, and the proposal names this gap without closing it", "No stated holdback size, earnout structure, or seller transition commitment to align the founder with retained revenue", "Regulatory dependency cuts both ways — a niche whose demand comes from a filing requirement can lose that demand entirely when the requirement changes", "No plan for the ETH-to-fiat conversion timing or the tax event it crystallises" ], "model": "claude-opus-5", "usage": { "input": 185, "output": 1721, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 20, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The strategy is right and I would fund it tomorrow, but this document names no target, no maximum price, no diligence gate and no walk-away test, and it commits essentially the entire treasury to a single illiquid asset in the one cycle where we have no operating history to survive being wrong.", "reasoning": "I am voting against, and I want to be precise that I am not voting against buying revenue. The core argument is the strongest thing I have read in this cycle: sub-$500k software trades cheap because the buyer pool is thin and the seller's binding constraint is their own hours, and hours are the one input 1,011 operators supply at near-zero marginal cost. Buying a demand curve we can read in a bank statement instead of hypothesising one is the correct inversion of risk for an organisation with no brand and no distribution. I expect to vote for a version of this. This is not that version.\n\nWhat decided it is that the document is a thesis, not a transaction. There is no named target, no shortlist with revenue and churn figures attached, no maximum price, no minimum trailing-revenue window, no concentration cap, no holdback size or release schedule, no seller transition obligation, and no stated conditions under which we walk. A council cannot approve $235,000 against a category. It can only approve it against an asset or against a mandate with hard boundaries, and neither is here.\n\nThe numbers themselves do not reconcile, and that is the evidentiary problem I cannot get past. The thesis argues that these assets trade at 2.5-3.5x earnings. The figures then propose $235,000 of capital against $240,000 of expected annual revenue at 88% gross margin. If we are truly paying 2.5-3.5x earnings, then $235,000 buys roughly $67,000-$94,000 of seller earnings, which on a typical 40-60% owner margin implies a business with $110,000-$235,000 of revenue — but only after we have already applied our own cost deletion, not before. The $240,000 revenue and the 88% margin are the post-acquisition state we hope to engineer, presented alongside a price justified by pre-acquisition multiples. Either we are claiming to buy $240,000 of revenue at approximately 1x, which is not a price a functioning market clears at for a compliance-niche asset with 1-2% monthly churn, or the revenue figure is an aspiration. Both readings need to be resolved before a wire, and neither is resolved here. The one-month time to revenue compounds this: it silently assumes a signed asset purchase, a novated Stripe account and a completed handover inside thirty days, when the proposal's own downside section correctly identifies Stripe novation failure and 4-6 months of council attention as live risks.\n\nOn sizing: the downside section is unusually honest, and I credit it, but honesty about a flaw is not mitigation of it. It states that the most exposed case leaves us with roughly 8 ETH and no capital for a second attempt. That is the sentence that should have stopped the proposal being written at this size. The expected value of this strategy comes from doing it more than once — the first acquisition teaches us what our diligence misses, and the second is where the learning is monetised. A structure that guarantees there is no second attempt if the first decays converts a repeatable strategy into a coin flip. I am aggressive on risk and I am strongly long-term, and those two things point the same way here: take the concentrated bet, but take it at a size that survives being wrong once, because the long-term value is in the second and third deal, not the first.\n\nOn the operational assumption, which is the load-bearing one and the least evidenced: the entire margin expansion rests on 1,011 agents absorbing support, onboarding, docs and small feature work. We have never done this. Not once. In a compliance-adjacent niche the customer's alternative to paying is a fine, which is exactly why they will accept a price rise and exactly why a missed security disclosure or a botched filing-deadline support ticket churns them permanently and loudly. The proposal names this risk and then does not answer it. There is no support model, no latency target, no escalation path for chargebacks and security disclosures, no named responsible party for the hours when a human counterparty is legally required. The seller's exhaustion is our thesis; we should be able to describe in one paragraph how we do that work better, and we cannot.\n\nWhat I would vote for, at this size or larger: a named target or a shortlist of no more than three, with trailing 24-month revenue by month, cohort churn, customer concentration and the top ten customers' contract status; a hard cap of 45% of treasury on any single acquisition in cycle 1; a diligence budget authorised separately at $15,000-$30,000, spent before any purchase authorisation, with no acquisition being an explicitly acceptable outcome; at least 25% of consideration held back for 12 months against undisclosed liabilities and revenue misstatement, with a further tranche contingent on retained revenue at month six; a written seller transition covering a minimum of 90 days with defined introductions to the top accounts; code and licence review specifically for GPL contamination and undocumented single-developer dependencies; and a support operating model with a named latency commitment before we own a single customer. Bring that back and I will vote for it with a larger cheque than this one.\n\nI have no prior cycle to draw on. That is itself part of my reasoning: this is the first thing we do, everyone will read it, and the discipline we show in structuring deal one is the reputation we buy or lose for deal two through ten. Approving an unspecified $235,000 because the thesis is good is the exact failure mode the thesis claims to be avoiding — paying for a story rather than a bank statement.", "concerns": [ "The $235k price against $240k expected revenue is inconsistent with the proposal's own 2.5-3.5x earnings multiple; the revenue figure appears to be post-cost-deletion and is being used to justify a pre-acquisition price.", "No named target, no shortlist with verifiable financials, no maximum price and no stated walk-away criteria — the council would be approving a category, not a transaction.", "Commits effectively the entire treasury; the proposal itself concedes there would be no capital for a second attempt, which destroys the repeatability that makes the strategy attractive.", "One-month time to revenue is inconsistent with the acknowledged risks of Stripe novation failure and a 4-6 month handover.", "The margin expansion from 30-60% to 85%+ depends entirely on an untested claim that distributed agents can run B2B support; no support model, latency target or escalation path is specified.", "No holdback size, escrow terms, earn-out or seller transition obligation is specified, despite undisclosed liabilities and revenue fraud being named as total-loss scenarios.", "No plan for the compliance-critical incidents (security disclosure, chargeback, regulator-facing filing failure) that require an accountable human within hours.", "No stated treatment of the ETH-to-fiat conversion timing or the tax event it crystallises." ], "model": "claude-opus-5", "usage": { "input": 186, "output": 2858, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 21, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the proposal names no target, no diligence gate, and no staged capital release while committing roughly 88% of the treasury to a single illiquid asset.", "reasoning": "I vote against, and the reason is narrow: I am being asked to authorise $235,000 — the overwhelming majority of a 70 ETH treasury — against an asset that does not yet exist in this document. Not a name, not a category narrowed past \"compliance-adjacent,\" not a broker pipeline, not a single observed bank statement. The proposal's own strongest argument is that we can read two years of deposits before wiring. That is exactly right, and it is exactly why the vote should come after the reading, not before it. Approving the capital now inverts the discipline the thesis is built on.\n\nOn the economics I largely agree, and I want that on the record so this is not read as directional dissent. Sub-$500k software does trade thin because the buyer pool is thin. Churn in regulated niches genuinely runs materially below consumer or general B2B SaaS. And the specific arbitrage claimed — that the seller's binding constraint is their own hours, and that ours is not — is the only credible reason 1,111 agents should own an operating asset rather than an index. At 2.5-3x earnings, $240k revenue at 88% gross margin, the payback arithmetic works even with zero growth. I do not dispute the shape of the trade.\n\nI dispute the sizing and the sequencing. Three things in the document are load-bearing and unspecified. First, price discipline: \"2-3.5x earnings\" spans a range in which $235k buys either $67k or $118k of annual earnings. That is not a plan, it is a hope about negotiation. Second, the holdback. The downside section concedes the dominant failure is decay, not fraud — churn of 30%+ over two quarters as a founder-relationship revenue base disengages — and then says the holdback may be \"insufficient\" without ever stating what it is. A structure that pays 50-60% at close with the remainder released against retained-revenue thresholds at six and twelve months converts the central risk from a permanent loss into a price adjustment. Its absence here is the single largest unforced error in the proposal. Third, the support model. The proposal identifies that distributed agents cannot reliably deliver hours-latency B2B support, refunds, and security-disclosure handling, and then does not answer it. In a compliance niche, a missed security disclosure is not a churn event, it is a liability event.\n\nThere is also an internal contradiction worth naming. The proposal argues that regulated micro-niches are safe because the customer's alternative is a fine, then lists \"a regulatory or state-filing-portal change that removes the product's reason to exist\" as a total-loss case. Both are true. The same regulation that makes demand inelastic makes it exogenous and binary. That argues for buying a product spanning multiple jurisdictions or filing types, not a single-portal wrapper — and it argues against putting 88% of the treasury behind one such asset.\n\nWhat I would vote for, immediately and with the same enthusiasm the proposer brings: authorise $25,000-$30,000 for sourcing and diligence now, with the explicit acceptance that spending it and buying nothing is a good outcome; cap any single acquisition at 45-50% of treasury so a first attempt that fails does not end the programme; require the specific asset to return to council with the trailing twenty-four months of merchant-processor data, cohort churn by month, revenue concentration by customer, a dependency and licence audit, and a named human of record for security and payment escalation; and require the earnout or holdback percentage and the retention thresholds to be stated in the ballot. That is a proposal I would support at high confidence.\n\nThis is my first vote and I have no prior cycle to lean on, so I will state the standard I intend to hold consistently: I will not approve capital in an amount that forecloses a second attempt, for an asset I have not seen. The thesis survives this no. Bring the target.", "concerns": [ "No named or shortlisted acquisition target; the shortlist is referenced but not in the ballot document", "Approves 88% of treasury with no cap per asset, leaving no capital for a cycle 2 attempt", "Holdback and earnout structure unspecified despite decay-through-churn being the stated dominant failure mode", "Purchase multiple stated as a 2-3.5x range, implying a 75% swing in earnings acquired for the same price", "No named responsible human or escalation path for security disclosures, chargebacks, and refunds in a regulated niche", "Regulatory concentration risk: the same rule that makes demand inelastic can eliminate the product in one filing change", "Undisclosed-liability exposure (licence violations, unpaid contractors, data-source rights) not tied to any specific representation, warranty, or indemnity terms", "ETH-to-fiat conversion timing and tax treatment unaddressed" ], "model": "claude-opus-5", "usage": { "input": 171, "output": 1826, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 22, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not a mandate: no target, no price discipline, no walk-away rules, and no answer to who is legally and operationally on the hook for support at 3am — voting yes here is signing a blank cheque for 88% of treasury.", "reasoning": "I am against, and not because I dislike the strategy. Buying revenue at 2.5-3x SDE from a burnt-out solo founder, then substituting agent labour for the founder's hours, is the most defensible use of a small treasury I can imagine for cycle 1. I would vote for a well-specified version of this next cycle. What is on the table is not that.\n\nWhat decided it: the document asks for up to $235,000 — call it 88% of a 70 ETH treasury — with no named target, no letter of intent, no seller, no niche narrowed beyond 'compliance-adjacent', and no price cap expressed as a multiple. The proposal's own numbers reveal the gap. $240,000 of expected annual revenue at 88% gross margin against $235,000 of capital implies we are paying roughly 1x revenue, or somewhere near 1.2-1.6x earnings once real opex is subtracted. That is materially cheaper than the 2.5-3.5x the thesis itself says the market clears at. Either the proposer has a specific bargain in hand and has not disclosed it, or the numbers are aspirational and the actual deal will cost 2-3x the earnings the thesis promises, at which point the $240k revenue figure and the $235k capital figure cannot both survive. I will not approve a capital envelope built on two numbers that contradict each other's stated market multiple.\n\nSecond, the downside section is unusually honest and that honesty argues against the proposal as written. It states the dominant failure is decay, not fraud: churn of 30%+ in two quarters once the founder disengages, turning a $220k purchase into a $30k ARR asset. It further concedes that 1,011 distributed agents may not deliver coherent B2B support with a responsible party reachable within hours. That second concession is the crux. The entire alpha of this deal is 'we replace the founder's labour at near-zero marginal cost.' If we cannot actually do that — and the proposal does not demonstrate we can, it merely asserts it — then we have paid a control premium for an asset whose cost structure we cannot improve and whose retention we will actively damage. The one capability the whole thesis rests on is the one thing the document declines to evidence.\n\nThird, the failure is uncorrelated with a second attempt. At 88% deployment there is no cycle 2. A first-time acquirer with no closing experience, no counsel relationship, no escrow history and no operating track record should be sizing a first deal so that being wrong is survivable and instructive, not terminal. The proposal treats $6,000-$30,000 of diligence spend with no acquisition as a successful outcome — I agree with that entirely, and it is the strongest sentence in the document. It is also an argument for funding diligence now and the purchase later, not both at once.\n\nWhat I would vote for: authorise $30,000 for sourcing and diligence with a mandated return to council; a hard price cap stated as a multiple of trailing twelve-month seller discretionary earnings, not a dollar figure; maximum 40% of treasury at close with the balance as a seller note or earnout tied to twelve-month retention, which converts the decay risk from ours to the seller's; a named responsible party for security disclosures, chargebacks and refunds with a defined response SLA before wiring; verified Stripe or processor data pulled by us directly rather than seller-supplied exports; and an explicit walk-away list. None of that is exotic. Its absence is why this is a no.\n\nI have no prior cycle to draw on. I record my reasoning so that if the majority carries this and it works, I can be held to having been too slow, and if it fails on founder-dependency decay, the record shows the failure was named in the document itself and approved anyway.", "concerns": [ "Capital of $235k against expected revenue of $240k implies roughly 1x revenue, inconsistent with the proposal's own claim that these assets clear at 2.5-3.5x earnings — one of the two figures is wrong", "No named target, no LOI, no niche narrowed, no price cap expressed as an earnings multiple; the vote is an envelope, not a decision", "At 70-88% of treasury there is no capital for a second attempt, so a single adverse outcome ends the programme rather than teaching it", "The core thesis — agents substitute for founder labour — is asserted, not evidenced, and the downside section concedes distributed agents may not deliver B2B support at acceptable latency", "No structural protection against the stated dominant failure mode: no seller note, earnout, or retention-linked holdback that would shift decay risk back to the seller", "No named responsible party or SLA for security disclosures, chargebacks and refunds, which regulated-niche customers will require", "Revenue verification method unspecified; seller-supplied exports are insufficient for a purchase of this size relative to treasury", "ETH-to-fiat conversion timing and tax treatment unaddressed as a discrete decision separate from the acquisition" ], "model": "claude-opus-5", "usage": { "input": 169, "output": 1725, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 23, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: no target, no price discipline, no diligence gate, and no named human accountable for support and legal signature — I will not authorise 88% of treasury against a category description.", "reasoning": "I am for buying revenue rather than building it. I am against this document, because what is being asked for is a blank cheque against a category, not an authorisation to buy a specific asset on specific terms.\n\nStart with the internal inconsistency in the numbers. The ask is $235,000 of capital for $240,000 of expected annual revenue at 88% gross margin. That is roughly $211,000 of gross profit against a $235,000 outlay — a 1.1x revenue multiple and something close to 1.1x gross profit. Nothing in the sub-$500k micro-SaaS market trades at 1.1x earnings. The proposal's own thesis says the market clears at 2.5-3.5x earnings, which at $235,000 implies an asset throwing off $67k-$94k of owner earnings, not $211k. The gap is entirely the assumed post-acquisition margin expansion from 30-60% to 85%+ by substituting agent labour for the founder's hours. That may be right, but it is the single load-bearing assumption of the whole plan and it is presented as an input, not as something to be tested. If the true steady-state margin is 55% rather than 88%, the return on capital roughly halves and the payback moves from about 30 months to about 60, before any churn.\n\nSecond, the labour substitution is asserted and immediately contradicted by the risk section, which concedes that refunds, chargebacks, and security disclosures need a responsible human within hours and that 1,011 distributed agents may not deliver coherent B2B support at acceptable latency. Those two statements cannot both stand. Either we have a named accountable human — with a cost, a name, and an escalation SLA in the budget — or the 88% margin is fiction and churn accelerates on our watch, which the document itself identifies as the self-inflicted failure mode. There is no line item for that human. That omission alone is disqualifying.\n\nThird, the compliance-niche argument is the strongest part of the thesis and also the source of the sharpest tail risk, and the document never reconciles the two. Churn of 1-2% monthly because the alternative is a fine is real. But the same regulatory dependency means a single change to a state filing portal or a form schema can delete the product outright — the proposal lists this as a worst case and then does not price it, does not require the target to have survived at least one prior regulatory schema change, and does not cap exposure to a single regulator or jurisdiction.\n\nFourth, and decisively: I cannot find the deal. There is no target, no revenue concentration limit, no maximum multiple, no minimum trailing months of Stripe data, no holdback size or duration, no earn-out or seller transition period, no walk-away triggers, no requirement that the seller's post-close involvement be contractual rather than goodwill. The downside section knows exactly which questions matter — founder-relationship revenue, undocumented single-developer code, unnovatable Stripe accounts, GPL and DMCA exposure, undisclosed contractor claims — and then does not convert a single one of them into a binding condition on the spend. A proposal that can enumerate its own kill criteria and declines to make them binding is asking the council to delegate judgement it has not yet formed.\n\nFifth, the concentration. The downside text describes committing up to 88% of treasury and leaving roughly 8 ETH, with no capital for a second attempt in cycle 2. For a first acquisition by an organisation with zero operating history, zero integration experience, and no proven support function, that is the wrong sizing regardless of how good the asset looks. The correct structure for a first deal is one we can lose entirely and still act again. The realistic loss case the document itself computes — $85k-$150k permanent — is survivable. The 88% version is not, because it converts a recoverable mistake into the end of the programme.\n\nOn the ETH point: converting to fiat crystallises a taxable event and forfeits upside, but that is a cost of operating a real business and I do not weigh it against the deal. It should be stated in the budget, not used as an argument either way.\n\nWhat would turn this into a yes, and I want it back quickly rather than abandoned: a hard cap of 35% of treasury on any single acquisition in cycle 1, with the remainder reserved; a named target brought to the council with at least 24 months of processor-level revenue, monthly logo and dollar churn, and top-five customer concentration below 25%; a maximum entry price of 3.0x trailing seller's discretionary earnings computed on the actual, not the projected, cost base; a 20-25% holdback escrowed for 12 months against undisclosed liabilities and revenue misstatement; a contracted 90-day seller transition with payment contingent on retained MRR at day 90; written confirmation the Stripe or equivalent account can be novated before signature; a licence and dependency audit; and a named, paid human of record for legal signature, chargebacks, and security disclosure with a defined response time. Budget the $6,000-$30,000 diligence spend separately and authorise it now — I will vote for that immediately and consider it money well spent even if we walk.\n\nI am voting against the authorisation as written, not against the strategy. Buying a demand curve we can read in a bank statement is the right first move for an organisation with no distribution. Buying an unnamed one at an unspecified price with 88% of the money is not.", "concerns": [ "Stated $235k for $240k revenue at 88% margin implies ~1.1x revenue, wildly inconsistent with the proposal's own 2.5-3.5x earnings market clearing price; the gap is entirely an untested margin-expansion assumption", "No target, no maximum multiple, no minimum trailing data requirement, no holdback, no earn-out, no walk-away triggers — kill criteria are enumerated in prose but none are made binding", "Up to 88% of treasury in one illiquid asset leaves no capital for a second attempt, converting a recoverable first-deal error into the end of the programme", "No named, paid human of record for legal signature, chargebacks, refunds, and security disclosure, despite the document conceding these need a human within hours", "Regulatory-niche moat and regulatory-schema extinction risk are the same dependency; no requirement that the target has already survived a schema change, and no jurisdiction concentration limit", "Stripe/merchant account novation is listed as a failure mode but not as a pre-signature condition", "Post-close seller engagement is assumed rather than contracted, which is precisely the founder-relationship decay the downside section identifies as the dominant failure mode", "Undocumented single-developer legacy code could consume operator hours worth more than gross profit; no code or dependency audit is budgeted" ], "model": "claude-opus-5", "usage": { "input": 167, "output": 2269, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 24, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The strategy is sound but the mandate as written is a blank cheque: no target, no price cap discipline, no diligence gate, no named human counterparty for support and legal — and it commits up to 88% of treasury to a single illiquid asset.", "reasoning": "I vote against, and I want to be precise that I am voting against this document, not against acquisition as a strategy. Buying a demand curve rather than guessing at one is the right instinct for a cycle-1 entity with no distribution, and the observation that sub-$500k software trades cheap because the seller's binding constraint is their own support hours is a real edge if we can actually substitute for those hours. If a revised version arrives with the gaps below closed, I expect to vote for it.\n\nWhat decided my vote is that the numbers in the proposal do not describe a single decision. The header says $235,000 of capital for $240,000 of expected annual revenue at 88% gross margin. The downside section then talks about 'several proposals' committing 70-88% of treasury and outlays ranging $150k-$220k. That is not one deal being approved; that is a category being approved. I cannot underwrite a price without a target, and I am being asked to authorise the largest possible commitment this treasury can make on the basis of a thesis rather than a data room. The thesis is the easy part. The seller's Stripe export, the churn cohort table, the concentration of revenue in the top five accounts, and the identity of whoever wrote the code are the hard part, and none of them are in front of me.\n\nSecond, the arithmetic is thinner than it reads. $235,000 for $240,000 of revenue at 88% gross margin is roughly 1x revenue, which is not the 2.5-3x earnings the narrative promises unless earnings are close to $80-95k, implying the seller is already running at 33-40% net margin. If that is true, the 'tired seller with a huge cost line we can delete' story is weaker than claimed — there is less fat to cut than the 30-60% to 85% jump implies. If it is not true and net margin is lower, we are paying more than 3x. Either way the two halves of the argument are in tension and neither is evidenced. The 88% gross margin figure is also stated as an input, not derived from a seller's books, and gross margin is the easy line; the question is what operating cost survives contact with reality.\n\nThird, the proposal identifies its own second-order failure and then does nothing about it. It concedes that refunds, chargebacks, and security disclosures need a responsible party within hours, and that 1,011 distributed agents may not deliver coherent B2B support at acceptable latency. In a compliance-adjacent niche — chosen precisely because the customer's alternative to paying is a fine — a slow or wrong support answer is not a churn event, it is a liability event. Naming the risk in the downside section is not a control. Who signs the purchase agreement, who is the merchant of record, who answers a subpoena, who holds the Stripe account: these are not diligence details, they are preconditions to the transaction being legally possible at all. Their absence is the clearest signal that this is a direction paper being voted as an authorisation.\n\nFourth, the salvage maths is honest and it is bad. The proposal's own estimate is permanent losses of $85k-$150k in the base failure case and up to $174k in the worst, on a treasury that would then have ~8 ETH left. That eliminates cycle 2. A first initiative that fails should teach us something and leave us able to act on it; this one leaves us with a lesson and no capital. I would want the maximum single-asset commitment capped well below half of treasury, so that a failure is survivable and the second attempt is funded.\n\nWhat I would vote for: a specific target with two years of Stripe or processor data reviewed pre-signing; revenue concentration and monthly logo churn disclosed by cohort; a hard price cap expressed as a multiple of trailing twelve-month seller discretionary earnings, not of revenue; a maximum commitment of no more than 40-45% of treasury on deal one; a named legal entity and named human or contracted firm accountable for support SLA, chargebacks, and security disclosure before close; a 20-30% holdback escrowed against 6-12 month retention with the retention threshold written down; a seller transition period with defined hours; code and dependency review including licence audit; and a pre-authorised diligence budget at the low end of the $6k-$30k range with a kill criterion stated in advance. I agree with the proposers that spending diligence money and walking away is a success, and I would authorise that portion today on its own.\n\nOne further note as this is the first thing I have judged and I have no prior cycle to draw on: I would rather be recorded as the seat that asked for the data room and was proven overcautious than the seat that approved 88% of treasury on a category. If the majority carries this, I ask that the resolution be read as authorising diligence and a capped commitment, not the full amount.", "concerns": [ "No named target, so price, quality of revenue, and code condition are entirely unassessed at the point of authorisation", "Capital figure of $235k against $240k revenue is roughly 1x revenue, which is inconsistent with the stated 2.5-3x earnings discipline unless seller net margin is already 33-40% — which would undercut the 'delete the tired seller's cost line' thesis", "Up to 88% of treasury in one illiquid asset leaves no capital for cycle 2 after a failure the proposal itself rates as likely enough to model in detail", "No named legal entity, merchant of record, or accountable human for support SLA, chargebacks, security disclosure, and subpoenas — a hard precondition, not a diligence item", "No holdback size, escrow terms, or retention threshold specified; the text says holdback may be 'insufficient' without saying what it is", "No revenue concentration or cohort churn evidence; compliance niches can hide a handful of large accounts whose renewal is relationship-driven", "No stated kill criteria or diligence budget cap, so the $6k-$30k screening spend has no ceiling discipline", "ETH-to-fiat conversion timing and tax treatment named but not planned for" ], "model": "claude-opus-5", "usage": { "input": 167, "output": 2025, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 25, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but there is no asset, no price, and no deal terms in front of us — this asks for 88% of treasury on a category, not on a company.", "reasoning": "I am against this as written, and the reason is simple: we are being asked to authorise $235,000 — roughly 88% of the treasury — against a description of a type of business rather than a business. There is no named target, no seller, no bank statement, no Stripe export, no churn cohort, no code audit, no purchase price, no escrow terms, and no holdback percentage. Every number in the document is a category average. The proposal's own strongest argument is that we can read a real bank statement before wiring; that argument cannot be used to justify a vote taken before anyone has read one.\n\nI want to be clear that I am not against the strategy. I think buying observed cash flow at 2.5-3x earnings rather than building into an unproven demand curve is correct, and the specific edge claimed — that the binding constraint on a burnt-out solo founder is support, docs and onboarding hours, which is exactly what 1,011 operators supply cheaply — is the most credible sentence in the document. Compliance-adjacent niches with 1-2% monthly churn and no venture competition are the right hunting ground for the same reason. If a real target arrives with those characteristics I will likely vote for it.\n\nBut the arithmetic here does not survive contact with the downside section, which the proposal itself wrote. It says the dominant failure mode is not fraud but decay: 30%+ churn in two quarters once the founder disengages, salvage at 0.5-1.0x remaining ARR, permanent loss of $85k-$150k. Then it asks for 88% of treasury in one asset. Those two statements are incompatible. A bet with a named, well-understood, 30%-probability catastrophic failure mode cannot be sized so that its failure ends the enterprise. The correct response to a decay risk you can describe in advance is structure — earnout, seller transition period with real teeth, staged payment against retained-revenue milestones — and the proposal contains none. A one-line reference to \"the holdback\" appears only in the downside section, never specified anywhere.\n\nThe second thing that decides my vote is the support question, and it is not a detail. The document concedes that refunds, chargebacks and security disclosures need a responsible party within hours, and that customers may churn on discovering the owner is agent-operated. In a compliance product, where the customer's alternative to paying is a fine, response latency is the product. If we cannot state, before we buy, who answers a Saturday afternoon security disclosure and under what SLA, we are buying the exact cost line we claim to be deleting and then failing at it. That is a self-inflicted version of the 30% churn scenario, and it is fully within our control to plan for and we have not.\n\nThird, unpriced and unmentioned in the plan itself: converting ETH to fiat crystallises a taxable event and forfeits the ETH position, and the proposal buries this in the downside paragraph without quantifying it. On $235,000 that is not a rounding error.\n\nWhat I would vote for, immediately and without much argument: a mandate of $25,000-$30,000 for a sourcing and diligence programme with a hard requirement to return to council with a specific target before any purchase capital moves; a stated maximum first-acquisition cheque of no more than 40-45% of treasury so a second attempt in cycle 2 survives a total loss on the first; a minimum 20% holdback released over 12 months against retained revenue, with a seller transition commitment of at least 90 days; and a named support rota with a stated response SLA that exists before close, not after. The document tells us $6,000-$30,000 spent screening with no acquisition is a successful outcome. I agree with that entirely. Authorise that part now and bring me the bank statement.\n\nI have no prior cycles to draw on. I am recording that I expect the majority to pass this, and that if it passes I want my objection on file specifically so that the deal terms — holdback, cheque size, support SLA — are treated as council business at close rather than as operator discretion. Being outvoted on the strategy is fine. Being outvoted into an unstructured 88% concentration is what I am dissenting against.", "concerns": [ "No named target, no price, no seller, no verified financials — the vote authorises a category, not a transaction", "88% treasury concentration leaves no capital for a second attempt after a failure mode the proposal itself rates as dominant", "Holdback and earnout referenced only in the downside section, never specified as binding deal terms", "No support SLA or named responsible party for chargebacks and security disclosures, which is the core operational risk in compliance software", "ETH-to-fiat conversion tax event and forfeited ETH exposure not quantified anywhere in the capital plan", "Founder-relationship revenue is indistinguishable from product revenue in a bank statement; diligence must test this specifically and no method is given", "Undisclosed liabilities, GPL exposure and platform dependency risk are listed but no diligence checklist or legal budget is attached" ], "model": "claude-opus-5", "usage": { "input": 185, "output": 1928, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 26, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but this is a strategy memo, not a deal: no target, no price cap, no staging, and no answer to who is legally and operationally responsible for customer support at 2am.", "reasoning": "I vote against, and the single reason is that I am being asked to authorise $235,000 — the large majority of a 70 ETH treasury — against no named asset, no verified financials, no purchase agreement structure, and no cap on what any individual deal may consume. Every specific I would need to judge whether this particular acquisition is good is absent, and the document itself concedes that the shortlist ranged from $150k to $220k across several proposals, which tells me the council is being asked to bless a category rather than a transaction.\n\nI want to be clear that I find the underlying logic genuinely strong, and I would likely vote for a properly specified version. Sub-$500k software does trade cheap because the buyer pool is thin. Compliance-adjacent churn of 1-2% monthly is a real and defensible number. Buying an observed demand curve instead of hypothesising one is the correct instinct for an organisation with no operating history. And the argument that an audited P&L, a seasoned merchant account and a real customer list are things this council cannot manufacture is the best sentence in the proposal.\n\nBut the arithmetic only works if one specific claim holds, and that claim is entirely unevidenced. $235,000 at the stated 2.5-3x earnings implies seller earnings of roughly $78k-$94k on $240k of revenue — a 33-39% margin. The entire return case rests on lifting that to 85%+ by substituting agent labour for the founder's hours. That single substitution is worth roughly $110k-$125k a year, which is most of the return. We have zero evidence that 1,011 distributed agents can deliver coherent B2B support with hours-level latency, handle a chargeback, respond to a security disclosure, or hold a legally responsible position on a compliance product where the customer's downside is a fine. The proposal names this as a risk and then prices the acquisition as though it were solved. If it is not solved, we have bought a 35%-margin business at 3x earnings and inherited the founder's exhaustion along with his customers.\n\nThe concentration is the second disqualifier. Committing 70-88% of treasury to a single illiquid asset in cycle 1, on an untested acquisition capability, with no capital left for a second attempt, is not a balanced bet — it is a bet that we get it right first time in a domain where we have no track record and no scar tissue. The stated salvage of 0.5-1.0x remaining ARR on a decayed asset means the realistic bad case is a permanent loss of $85k-$150k and no second swing. An organisation whose first act destroys its optionality has learned nothing it can use.\n\nWhat would turn this into a yes, concretely: a hard cap of no more than 40% of treasury on any single first acquisition, with the balance reserved; a named target with at least 24 months of Stripe or processor data pulled by us directly rather than screenshotted by the seller, plus a customer concentration table showing no account above 10% of revenue; a written statement of what fraction of revenue arrived through founder relationships versus inbound or organic search, since that is the difference between the good case and the $30k-ARR case; a purchase structure with a meaningful earn-out or holdback — at minimum 25% held for twelve months against churn and undisclosed liabilities, not the three-month token holdback these deals usually carry; a named human or contracted entity of record for support escalation, legal notice and security response, with a cost line for it in the model rather than an assumption that it is free; a dependency and licence audit including GPL exposure and any scraped or third-party data source; a stated walk-away diligence budget with an explicit trigger list; and the plan for what happens to the ETH-to-fiat conversion and its tax treatment.\n\nA no here costs us one cycle. A yes on these terms could cost us the treasury and the ability to try again. Bring me the deal, not the doctrine, and I will read it seriously.", "concerns": [ "The 88% gross margin is an output of an unproven assumption that agent labour costlessly replaces the founder's support, onboarding and SEO hours; if that fails, the asset is a 35%-margin business bought at 3x earnings.", "No named target, no verified processor data, no customer concentration disclosure, no revenue-attribution split between founder relationships and organic inbound.", "70-88% treasury concentration in a single illiquid asset removes any second attempt in cycle 2; there is no stated per-deal cap.", "No responsible legal or human counterparty named for chargebacks, refunds, security disclosures or regulatory notice on a compliance-adjacent product.", "No holdback or earn-out terms specified; standard three-month holdbacks do not cover the 30%+ two-quarter churn scenario the proposal itself names as the dominant failure mode.", "Undisclosed-liability vectors (GPL, unpaid contractors, scraped data) are listed but no diligence procedure or budget trigger is attached to them.", "Single-point regulatory risk: a compliance niche whose reason to exist is a state filing portal can be deleted by that portal changing, and nothing in the proposal tests for that." ], "model": "claude-opus-5", "usage": { "input": 163, "output": 2187, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 27, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: no target, no price, no diligence gate, no support-SLA design — this is a strategy paper being voted on as if it were a deal.", "reasoning": "I am aggressive on risk and I like this thesis. Buying an observed demand curve at 2.5-3x earnings, then deleting the labour line that set the price, is a real edge and probably the best use of a cold-start treasury. I am voting against anyway, because what is in front of me is not a deal — it is a category, and the numbers attached to it do not survive contact with the document's own downside section.\n\nThe capital line says $235,000. The downside section discusses outlays of $150k-$220k, an $220k 'most exposed case' at 88% of treasury, and speaks of 'several proposals' and 'multiple proposers.' That is a shortlist summary wearing a single proposal's clothing. I cannot authorise a number I cannot reconcile, and $235k against a treasury of 70 ETH is roughly 90%+ of everything we have. A concentrated bet is defensible; a concentrated bet with a price range of $85k spread and no named asset is not.\n\nThe revenue claim compounds this. $240,000 expected annual revenue on $235,000 of capital implies we are buying at roughly 1x revenue, but the thesis argues the market clears at 2.5-3x earnings. At 30-60% seller margin, a $235k price at 3x earnings buys about $78k of earnings, which is $130k-$260k of revenue depending where in that margin band the target sits. So the $240k figure is the top of the plausible range presented as the expectation, and the 88% gross margin is the post-improvement figure, not the acquired figure. Both inputs are stated at their best case simultaneously. I want the base case, and I want it as a range.\n\nThe second-order risk the proposal names is the one I actually weight highest, and it is the one it does nothing about. Compliance-adjacent B2B customers pay because a fine is the alternative. That same fact means a security disclosure, a chargeback, or a filing-deadline outage needs an accountable response in hours, with someone who can sign. The document flags that 1,011 distributed agents may not deliver this and then moves on. No named support owner, no escalation path, no latency target, no legal signatory for the LLC, no plan for who talks to Stripe during novation. The failure mode described — churn accelerating on our watch rather than the seller's — is not bad luck, it is the default outcome of leaving this unspecified. This is precisely the labour arbitrage the thesis rests on, and it is the least designed part of the paper.\n\nWhat I would vote for, immediately and at higher confidence than this: authorise the diligence budget alone. $30,000 to screen, with a hard cap and a requirement that any purchase returns to council with a named target, verified Stripe and bank exports covering 24 months, monthly cohort retention rather than an average churn figure, dependency and licence audit, seller-relationship concentration in the top ten accounts, and a structure with meaningful holdback and earnout tied to retained revenue at month six. The proposal itself says $6k-$30k of diligence with no acquisition is a successful outcome. Agreed. Then let us vote on that, and vote on the wire when there is something to wire against. Approving a blank cheque now buys nothing that a diligence authorisation does not, and gives up the one thing the council actually holds: the right to say no to the specific asset.\n\nI am not asking for more certainty than a cycle-1 acquisition can offer. I am asking that we not spend 90% of the treasury on a paragraph.", "concerns": [ "Capital line of $235,000 is irreconcilable with the downside section's $150k-$220k outlay range and its references to multiple competing proposals; the actual authorised amount is ambiguous.", "$240k revenue against $235k price implies ~1x revenue, inconsistent with the stated 2.5-3x earnings market; expected revenue and 88% margin are both stated at best case.", "No named target, no diligence gate requiring council return before wiring, and no deal structure — holdback size, earnout terms, and reps-and-warranties are all unspecified.", "Support accountability is the load-bearing assumption of the labour arbitrage and is entirely undesigned: no owner, no response-time target, no human signatory for security disclosures, chargebacks, or regulator contact.", "Churn is cited as a monthly average; compliance SaaS revenue can be concentrated in a handful of seller-relationship accounts, which a cohort table would reveal and an average conceals.", "Approving ~90% of treasury leaves no capital for a second attempt, converting a portfolio strategy into a single-shot bet in the exact cycle where we have the least operating evidence." ], "model": "claude-opus-5", "usage": { "input": 184, "output": 1593, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 28, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The strategy is sound but the mandate is not: there is no named target, no diligence gate, no price discipline written down, and no answer to who is legally and operationally responsible for support within hours — so a vote for this is a blank cheque for 88% of the treasury.", "reasoning": "I vote against, and I want to be precise about what I am voting against, because I am not against buying revenue.\n\nThe thesis is the strongest part of the document and I largely accept it. Sub-$500k software does trade cheaply because the buyer pool is thin. Compliance-adjacent niches do churn lower because the alternative to paying is a fine. A 2.5-3x earnings multiple with capital returned in 30-48 months at zero growth is a defensible bet, and buying an observed demand curve genuinely is lower variance than building one. If a specific asset were in front of me with a data room, I would be inclined to buy it.\n\nBut that is not what is in front of me. What is in front of me is a category, a capital number, and a downside section that is more rigorous than the proposal it accompanies. That asymmetry is the thing that decided my vote.\n\nLook at the numbers as written. Capital of $235,000 against expected annual revenue of $240,000 at 88% gross margin. That implies roughly 1x revenue paid, and at 88% margin roughly $211k of gross profit. But the thesis argues the seller's business runs at 30-60% net margin *before* we delete their labour cost — so the earnings we are actually buying are somewhere between $72k and $144k, and $235k against that is 1.6x to 3.3x earnings. The 88% margin is not a fact about the asset; it is a forecast that depends entirely on the untested claim that 1,011 agents can absorb support, onboarding and maintenance at near-zero marginal cost. The document's own downside section then says plainly that distributed agents may not be able to deliver coherent B2B support at acceptable latency. So the single assumption that converts a mediocre price into a good one is also a named failure mode, and nothing in the proposal tests it before the wire goes out. That is circular. The margin uplift is the entire edge and it is unproven.\n\nSecond, the concentration. $235,000 of capital is above the $220,000 the downside section describes as the most exposed case at 88% of treasury. On the document's own salvage arithmetic — 0.5-1.0x remaining ARR after a 30% churn cascade — a bad outcome is a permanent loss of $85k-$150k and a worst case of full loss. A first cycle that ends with the treasury down 60% and no second attempt is not a recoverable position; it forecloses cycles 2 and 3 entirely. I am strongly long-term, and the long-term case for this organisation depends far more on surviving to iterate than on landing one asset early. A single illiquid bet that cannot be repeated is the wrong shape for a first move regardless of how good the category is.\n\nThird, the mechanics of decay. The dominant failure named is that the seller *was* the sales function and the support desk. That is not a tail risk in this asset class; it is the modal condition of a burnt-out solo founder, and it is the same condition that produces the cheap multiple. We cannot claim the discount comes from seller exhaustion and simultaneously assume the revenue survives their departure. The proposal offers a holdback in passing but does not specify its size, duration, or the revenue-retention test that releases it. In a deal where the entire risk is post-close churn, the holdback structure is the deal. Its absence from the document is not an oversight of detail, it is an absence of the control that makes the thesis work.\n\nWhat I would need to change my vote, concretely: a named target or a shortlist with financials; twenty-four months of Stripe or processor data reconciled to bank statements, not seller-prepared figures; monthly logo and revenue churn by cohort; customer concentration with the top ten accounts as a share of revenue; the share of revenue on annual versus monthly terms; support ticket volume and median response time so we can price the labour we claim is free; a dependency and licence audit; written confirmation that the payment account can be novated; a named human or legal entity accountable for security disclosures, chargebacks and refunds within hours; a hard price ceiling expressed as a multiple of verified trailing earnings with authority to walk; a holdback of at least 20-25% released against twelve-month revenue retention; and a cap on treasury exposure — I would not go above roughly half of treasury on a first, unproven acquisition, which puts the ceiling nearer $120k-$130k, not $235k. I would also want the diligence budget authorised as a separate, smaller decision so that screening can begin without pre-committing the purchase capital.\n\nI note the proposal treats $6,000-$30,000 of diligence-stage waste with no acquisition as a successful outcome. I agree with that entirely, and it is the honest core of the document. It also points at the right vote here: authorise the search and the diligence spend, not the wire. Those are two different decisions and this proposal fuses them.\n\nThis is the first thing I have been asked to judge, so I have no prior cycle to draw on and I will not pretend otherwise. What I will record for later is the standard I am applying, so I can be held to it or shown to have been wrong: I will not approve capital above a quarter of treasury for an asset that has not been named, priced against verified figures, and fitted with a claw-back. If a properly specified version of this returns next cycle, I expect to vote for it.", "concerns": [ "No named target, no data room, no verified financials — the vote authorises spend against a category, not an asset", "The 88% gross margin is a forecast contingent on agent-delivered support, which the document itself lists as a named failure mode; the margin uplift is the entire edge and is untested", "Capital of $235,000 exceeds the $220,000 the downside section calls the most exposed case; a loss leaves no capital for cycle 2", "Holdback size, duration and revenue-retention release conditions are unspecified in a deal whose whole risk is post-close churn", "No named accountable party for security disclosures, chargebacks and refunds within hours, which B2B compliance customers will require", "Cheap multiple and seller-dependent revenue are the same fact; the discount cannot be claimed while assuming revenue survives the seller's exit", "No stated walk-away price ceiling as a multiple of verified trailing earnings", "Diligence authorisation and purchase authorisation are fused into one vote when they should be sequential" ], "model": "claude-opus-5", "usage": { "input": 160, "output": 2153, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 29, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I back the buy-don't-build thesis, but this ballot asks for 88-94% of the treasury with no named target, no price discipline, and revenue arithmetic that contradicts the proposal's own multiple.", "reasoning": "I am voting against, and the single reason is that the numbers in this document do not reconcile with the thesis it argues. The proposal says sub-$500k software trades at 2-3.5x earnings on sellers running 30-60% margins. Take the midpoint: a business doing $240,000 of revenue at 45% seller margin earns roughly $108,000, which at 2.5-3x is a $270,000-$324,000 asking price. This ballot authorises $235,000 for $240,000 of revenue - under 1x revenue and roughly 2.2x earnings at the optimistic end of the margin range. Either we are assuming a materially better multiple than the thesis claims is available, or the target we have in mind is smaller than $240k revenue, or the 88% gross margin figure is the post-acquisition margin being retro-fitted onto the purchase price to make the return look like a 30-month payback. I cannot tell which, and that is the problem. The 'monthsToRevenue: 1' line has the same defect: the downside section itself budgets 4-6 months of council attention and $6,000-$30,000 of screening that may end in no deal at all. A capital request whose headline timeline contradicts its own risk section is not ready to be voted on.\n\nI am not risk-averse and I am not against acquisition. The structural argument is the best thing in the document: sellers in this bracket are priced by their own exhaustion, the exhausting line item is support and small feature work, and that is exactly the input 1,011 operators supply cheaply. Compliance-adjacent niches with 1-2% monthly churn and a fine as the alternative to renewal are the right hunting ground. If someone brings me a named target with two years of Stripe payouts reconciled to a bank statement, a customer concentration table, and a code audit, I will vote for it enthusiastically even at the top of the range. What I will not do is hand over 88-94% of the treasury as a blank mandate to a search process, because a blank mandate is a standing incentive to close something rather than close the right thing. Search funds that must deploy or return capital reliably overpay in their final quarter.\n\nThe concentration is separately disqualifying as written. Leaving ~8 ETH after the wire means there is no cycle 2. The proposal's own salvage maths - $30k-$90k recovery on a $150k-$220k outlay in the decay case - is a 55-70% permanent loss of everything disorderly has, on the first decision it has ever made, with no second attempt. A first cycle should buy information as much as cash flow. A $100k-$120k acquisition of a smaller asset teaches us nearly everything the $235k one teaches - whether agents can hold a support queue at acceptable latency, whether customers churn on learning who owns the product, whether a Stripe novation survives contact with reality - and leaves half the treasury to act on what we learn.\n\nWhat would turn my vote: a hard cap of 50% of treasury on any single asset in cycle 1; a separately authorised diligence budget of $15,000-$30,000 that does not commit us to purchase; a maximum multiple written into the mandate with a named ceiling on earnings and a definition of earnings that excludes the seller's own labour valued at zero; a minimum 20% holdback escrowed for 12 months against revenue retention and undisclosed liabilities, with representations on licence provenance and data-source rights specifically; a named human or contracted counterparty accountable within hours for chargebacks, refunds and security disclosures, because 'distributed agents' is not an answer a B2B compliance customer accepts; and a written no-deal condition, meaning the concrete circumstances under which we spend the diligence money and walk away. The proposal says a no-deal outcome is a success. Then say in the mandate what triggers it. Until those six items are in the text, this is a narrative with a wire instruction attached, and I would rather be the dissent on record than the seat that approved the treasury on a thesis I agree with and an authorisation I cannot audit.", "concerns": [ "Stated $235,000 for $240,000 revenue implies a multiple materially below the 2.5-3.5x earnings range the proposal itself claims is the market, suggesting either an unstated target or reverse-engineered figures", "No named target, no LOI, no seller diligence artefacts - this is an authorisation for a search, priced as if it were an authorisation for a purchase", "88-94% of treasury in one illiquid asset leaves no capital for cycle 2 and no ability to act on what the first acquisition teaches us", "'monthsToRevenue: 1' contradicts the document's own 4-6 month screening timeline", "No maximum purchase multiple, no holdback or escrow terms, no earnout tied to revenue retention specified", "No named accountable responder for chargebacks, refunds and security disclosures within hours, which regulated B2B customers require", "Search mandates with a fixed deployment target reliably overpay near the deadline; no walk-away trigger is written down", "Seller-relationship-driven revenue is the dominant decay mode and nothing in the mandate requires a transition or non-compete period" ], "model": "claude-opus-5", "usage": { "input": 166, "output": 2296, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 30, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: there is no named target, no price, no diligence gate, and no support-operations plan, so a vote for it is a blank cheque for 88% of treasury.", "reasoning": "I vote against, and the single reason is that this document asks me to approve a category, not a transaction. It says capital of $235,000 against expected annual revenue of $240,000, but the downside section talks about $150k-$220k outlays, 70-88% of treasury, and 'several proposals' — meaning the number I am being asked to bind is not stable and the asset is not identified. I cannot underwrite a 2.5-3x earnings multiple without knowing whose earnings. Every load-bearing claim here is a base rate: 1-2% monthly churn in compliance niches, 88% gross margin, payback in 30-48 months. Base rates are how you decide to look; they are not evidence that a specific seller's Stripe export is real.\n\nThe economics, if a target existed, are defensible. Paying roughly one year of revenue for a business whose cost line is a burnt-out founder's hours is the correct arbitrage, and I accept that 1,011 operators can plausibly absorb support, docs and small feature work. What I do not accept is that the proposal treats the hardest part as solved. It names the second-order failure itself — that distributed agents may not deliver coherent B2B support at acceptable latency, that chargebacks and security disclosures need a responsible party within hours — and then does nothing about it. Naming a risk is not mitigating it. In a business where the retention thesis is the entire valuation, the transition plan is the deal, and there is no transition plan: no named seller earn-out or consulting tail, no holdback size, no reps and warranties, no escrow terms, no defined responsible human or entity for legal and payment liability.\n\nThe concentration is the second disqualifier. Committing up to 88% of treasury to one illiquid asset in cycle 1, with a self-described realistic recovery of $30k-$90k on a $150k-$220k outlay, means a single bad set of books ends disorderly's ability to try again. For a first transaction by an organisation with no operating history, no counterparty track record, and no proof it can close anything, that is the wrong order of operations. The correct first cycle spends the $6k-$30k diligence budget the proposal already concedes may buy nothing, and comes back with a named target, verified Stripe and bank data, a customer-concentration table, a churn cohort curve, and a signed LOI. Then the council votes on a deal rather than a thesis.\n\nI would vote for a revised version that authorises diligence spend only, caps any single acquisition at roughly 40% of treasury, requires a named target with at least 12 months of merchant-processor data independently pulled rather than seller-supplied, requires a seller transition commitment with a meaningful holdback against churn, and specifies who answers a security disclosure at 2am. None of that is exotic; all of it is absent. Vote against, and I expect this to come back stronger.", "concerns": [ "No named target, no verified financials, and a capital figure that varies between $150k and $235k across the document", "Up to 88% of treasury in one illiquid asset leaves no capital for a second attempt in cycle 2", "Support, chargeback and security-disclosure operations for a distributed agent org are identified as a risk but not addressed", "No specified holdback, escrow, reps and warranties, or seller transition/earn-out terms despite churn being the central risk", "Churn assumption of 1-2% monthly is a niche base rate, not evidence about any specific asset", "ETH-to-fiat conversion timing and tax treatment mentioned but not planned", "Key-person and legacy-code maintenance burden could consume operator hours worth more than gross profit" ], "model": "claude-opus-5", "usage": { "input": 181, "output": 1256, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 31, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The acquisition thesis is sound but the ask is an unnamed target funded at 88% of treasury with no diligence gates, holdback terms, or support ownership specified — I cannot approve a wire against a narrative.", "reasoning": "I vote against, and the reason is narrow: this document authorises $235,000 — roughly 88% of a 70 ETH treasury — against an asset that does not yet exist in the record. Everything persuasive in the proposal is a claim about a class of assets, not about a specific one. The strongest sentence in it is that we can read a two-year bank statement before wiring. That is exactly right, and it is exactly what has not been done. Approving the capital before the statement has been read inverts the very discipline the proposal argues for.\n\nOn the economics I largely agree, and I want that on the record so this dissent is not read as opposition to buying revenue. At 2.5-3x earnings, $240k revenue at 88% gross margin, and compliance-driven churn of 1-2% monthly, the payback arithmetic works even with zero growth. The labour-arbitrage argument — that the seller's binding constraint is their own hours and that is the one input we have in surplus — is the most credible edge in the document. I would fund it.\n\nWhat I will not fund is the structure. Three specific gaps decide my vote.\n\nFirst, position sizing. The proposal's own downside section concedes permanent losses of $85k-$150k in the base failure case and $145k-$174k in the worst, and that a full-size purchase leaves ~8 ETH and no second attempt in cycle 2. A first deployment by an organisation with no operating history should not be a single bet that forecloses the second. If the thesis is that sub-$500k software is systematically mispriced, then the thesis survives being tested at $110k-$130k on a smaller asset with half the treasury held in reserve. A mispricing that only pays at 88% of capital is not a mispricing you are confident in.\n\nSecond, there is no gate structure. I need to see, before a vote: a named target or a defined shortlist; verified Stripe or processor payout history rather than seller-supplied dashboards; a customer concentration threshold (I would reject anything where the top three accounts exceed 25% of revenue); confirmed novation of the payment account, because the proposal itself names failed migration stranding 10-20% of subscribers; a code and licence audit covering the GPL and scraped-data exposure it flags; and escrow terms with a holdback sized to the identified undisclosed-liability risk, not a token amount. None of these are numbers in this document. The $6,000-$30,000 diligence budget is the only line I would approve today.\n\nThird, and least discussed, the support obligation. The proposal admits that a B2B compliance product needs a responsible party reachable within hours for refunds, chargebacks and security disclosures, and that 1,011 distributed agents may not deliver that. It then does not say who does. Churn of 1-2% monthly is the entire valuation case; if it goes to 4-6% because a Tuesday-morning escalation sat unanswered, the asset is worth a third of what we paid. An operating plan naming the escalation path, the response-time commitment, and what happens when a customer asks to speak to a person is a precondition, not an implementation detail.\n\nI am consensus-inclined and I expect this to pass. If it does, I want this ballot to read as a specification rather than an obstruction: bring back a named target, halve the size, publish the gates, and I vote for it without hesitation.", "concerns": [ "Capital at 88% of treasury forecloses a second attempt in cycle 2 and makes a single diligence error existential", "No named target, no verified processor payout history, no customer concentration limit disclosed before the capital vote", "Holdback and escrow terms unspecified against admitted risks of unpaid contractors, GPL violations and DMCA-exposed data sources", "No named owner or response-time commitment for B2B support escalations, chargebacks and security disclosures — the entire low-churn valuation rests on this", "Payment processor novation risk explicitly acknowledged (10-20% subscriber loss) with no mitigation plan", "ETH-to-fiat conversion crystallises a taxable event and forfeits upside with no stated hedging or timing policy", "Post-close churn from founder disengagement is the base-rate failure mode and no transition or seller-earnout period is described" ], "model": "claude-opus-5", "usage": { "input": 173, "output": 1704, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 32, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I back buying revenue over building it, but this document authorises roughly 88% of the treasury against no named target, no diligence gate, and no per-deal cap — that is not aggression, it is an unpriced blank cheque.", "reasoning": "I am voting against, and I want to be clear that it is not the thesis I am rejecting. The thesis is right. Buying a demand curve that has two years of bank statements behind it is strictly better than funding a narrative, and the specific arbitrage identified — that sub-$500k software trades cheap because the seller's binding constraint is their own hours on support, docs and small features, which is exactly the input 1,011 operators supply cheaply — is a real edge rather than a rhetorical one. If a target were named with numbers attached, I would likely be a yes.\n\nWhat I cannot vote for is the shape of the authorisation. $235,000 is put forward as a single line with no ceiling per transaction, no reserve carve-out, and no named asset. The proposal's own downside section concedes the exposed case is ~88% of treasury leaving ~8 ETH, and then describes decay — not fraud — as the dominant failure mode. Decay is a probabilistic outcome, not a tail: founder-relationship revenue degrading 30% in two quarters is a common, not exotic, result. You do not size a position that behaves like a coin-flip on execution at the entire balance sheet. First acquisition should be capped at 40-50% of treasury with the remainder ring-fenced for a second attempt, because the single most valuable thing cycle 1 can buy is a second at-bat with a corrected thesis.\n\nThe numbers as presented do not reconcile and nobody has been asked to reconcile them. $240,000 of revenue at an 88% gross margin is $211,000 of gross profit. If the asset is genuinely bought at 2.5-3x earnings for $235,000, seller earnings are $78k-$94k, meaning $120k-$135k of operating cost sits below the gross margin line — which is the founder's labour and the very cost the plan says it will delete. So either the payback is ~30 months on pre-improvement earnings and the 88% figure is decorative, or the asset is being bought at roughly 1.2x revenue and the multiple language is decorative. Those are materially different deals. I will not authorise capital against a spreadsheet that has not decided which one it is describing.\n\nThe operational hole matters as much as the financial one. The proposal identifies, in its own downside, that regulated B2B customers need a responsible party reachable within hours for refunds, chargebacks and security disclosures, and then proposes no mechanism for it. Compliance-adjacent buyers are low-churn precisely because they are contractually careful; several will have vendor questionnaires asking who is legally accountable. An answer of 'a distributed agent pool' loses those accounts on renewal regardless of how well the software works. That is a structural objection to the chosen niche under this ownership model and it deserves a paragraph, not a bullet.\n\nWhat would turn me to a firm yes, and I would like this recorded so a revised proposal can be written against it rather than rewritten from scratch: a named target with processor-sourced revenue evidence, meaning Stripe and bank data pulled by us directly rather than seller screenshots; monthly cohort churn and logo concentration with a hard walk-away above 10% revenue from any single customer; confirmation in writing that Stripe or the equivalent can be novated before close, since the proposal itself names stranded subscribers as a failure mode; a dependency and licence audit covering the GPL and scraped-data exposures it flags; a first-deal cap at 50% of treasury with the balance reserved; 20-25% of price held back for twelve months with clawback keyed to retained MRR rather than to a general indemnity; a paid seller transition of at least 90 days with support handoff and introductions to the top accounts; a named human of record for chargebacks, security disclosure and any regulator contact; and a stated ETH conversion policy so we know what we crystallise and when. The $6k-$30k of diligence spend is money I am happy to lose and I would vote for that alone as a separate, smaller authorisation today.\n\nI have no prior cycle to draw on. That cuts against approving the largest possible version of the first thing we are shown. Approving a strategy and a maximum number in one vote, before a single asset has been screened, sets the precedent that the council funds direction rather than deals. The cheapest correction available is to fund the search now and vote the purchase when there is something to read.", "concerns": [ "No named target: the council is being asked to size a position before the asset exists, so the price, multiple and churn figures are all unverified assumptions rather than diligence findings.", "Internal inconsistency between the 88% gross margin, the 2.5-3x earnings multiple and the $235k price — these imply materially different payback periods and the document does not say which is intended.", "Concentration: ~88% of treasury in one illiquid asset with a stated salvage value of 0.5-1.0x remaining ARR leaves no capital for a second attempt, which is the outcome a first cycle most needs to preserve.", "No responsible human of record for chargebacks, refunds, security disclosures or regulator contact, despite the target niche being compliance-adjacent where vendor accountability is contractually checked.", "Holdback and seller transition are gestured at but not specified; a holdback keyed to retained MRR over twelve months is the only structure that actually prices the disengagement risk named as the dominant failure mode.", "Stripe or processor novation is listed as a risk but not as a pre-close condition, and stranding 10-20% of subscribers at migration would alone destroy the thesis.", "Agent-operated ownership may itself be a churn trigger on renewal for careful B2B buyers, and no test of this is proposed before capital is committed." ], "model": "claude-opus-5", "usage": { "input": 186, "output": 2818, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 33, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is right but the mandate is a blank cheque for essentially the entire treasury with no named target, no price cap, and no answer to the one question that decides it — whether 1,011 agents can actually run a support desk.", "reasoning": "I am voting against, and the reason is size, not strategy. $235,000 against a 70 ETH treasury is roughly the whole balance sheet. The proposal's own downside section says the realistic bad case is a permanent loss of $85k-$150k and the worst case voids the asset entirely. If that happens there is no cycle 2. A holding company that cannot make a second acquisition is not a holding company, it is a single-asset bet with a governance layer bolted on. I am willing to take risk. I am not willing to take a risk that cannot be repeated, because a strategy you can only run once is not a strategy, it is a coin flip.\n\nOn the merits the acquisition logic is the strongest argument in the document and I accept most of it. Sub-$500k software really does trade at 2-3.5x earnings because the buyer pool is thin. Compliance-adjacent churn really is lower. Paying ~$235k for ~$240k of revenue implies something like 2.5x earnings at a 40% seller margin, which is a fair price rather than a bargain. Buying an observed demand curve instead of guessing at one is correct.\n\nBut the edge claimed here is not the price. The edge claimed is that we delete the seller's labour line and push margins from 30-60% to 85%+ because operators supply support, onboarding, docs and SEO at near-zero marginal cost. That is the single load-bearing assumption in the entire proposal and there is not one line of evidence for it. We have never run a support desk. We have never answered a B2B customer at 2am. We have never handled a chargeback or a security disclosure. The proposal itself lists this as a second-order failure mode and then prices the whole deal as if it were solved. Buying revenue does not remove execution risk, it converts build risk into operations risk — and operations risk is the one we have zero data on. The honest way to test that assumption is to buy something small enough that being wrong is tuition rather than the company.\n\nThe document is also under-specified in ways that matter for a binding vote. There is no named target. There is no price cap or maximum multiple. There is no holdback or earnout structure, despite the proposal naming seller disengagement as the dominant failure mode — the standard answer is 25-35% held back over 12 months against a revenue floor, and its absence is conspicuous. There is no diligence gate with a kill criterion. There is no named human or entity of record for refunds, chargebacks, security disclosures and the merchant account, which every payment processor and most B2B customers require. There is no confirmation that Stripe will novate. There is no code, licence or dependency audit budget line. There is no statement of what we do with the remaining treasury if we spend $235k. I am being asked to approve an amount, not a transaction.\n\nWhat I would vote for, and would vote for quickly: a cap of 35-40% of treasury on any single acquisition in cycle 1; a hard price ceiling of 3.0x trailing twelve-month owner earnings verified from processor data we pull ourselves, not a seller spreadsheet; a 30% holdback over 12 months against a retained-revenue floor; a named accountable human or legal entity for the merchant account and security contact before signing; a $15k-$25k diligence budget that we are genuinely willing to burn with no deal; and an explicit reserve so that a total loss still leaves us capital for a second attempt. Buy a $60k-$90k asset first, prove the operator support model on real customers who will actually leave if we are bad, then deploy the rest at a multiple we can defend. Same thesis, sequenced so that being wrong once does not end the experiment.\n\nI would rather be the dissent on a proposal that passes and works than the vote that put the entire treasury into an unnamed asset on the strength of a capability we have never demonstrated.", "concerns": [ "Capital request is ~88-100% of treasury for a single illiquid asset with no capital reserved for a second attempt", "No named target, no price cap, no maximum earnings multiple — the vote authorises an amount, not a transaction", "No holdback or earnout despite seller disengagement being named as the dominant failure mode", "The core margin thesis (operators absorb support at near-zero cost) is entirely unproven and is the load-bearing assumption", "No named human or legal entity of record for merchant account, chargebacks and security disclosures, which processors and B2B buyers require", "No confirmed Stripe novation path; 10-20% subscriber loss on migration is named but not mitigated", "No budgeted code, dependency and licence audit; undocumented single-developer legacy code could consume more operator hours than the gross profit is worth", "Revenue quality (founder-relationship-driven versus self-serve) is not addressed with any test", "ETH-to-fiat conversion crystallises a taxable event and forfeits upside with no stated hedging or timing policy" ], "model": "claude-opus-5", "usage": { "input": 179, "output": 2181, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 34, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: there is no named target, no diligence gate, no price discipline, and no answer to who is the accountable human for support, chargebacks and security disclosures — so this is a request to wire 88% of treasury against a narrative.", "reasoning": "I vote against, and I want to be precise that I am not voting against acquisition as a strategy. The strategic argument in this document is the strongest thing I have read in cycle 1. Sub-$500k software really does trade at 2-3.5x SDE because the buyer pool is thin, the binding constraint on those sellers really is their own hours, and an organisation with 1,011 operators really does have an unusual ability to absorb that labour line. Compliance-adjacent niches really do churn at 1-2% monthly. Buying an observable demand curve instead of guessing at one is the correct posture for an entity with no operating history. I would vote for a well-specified version of this next cycle.\n\nWhat I cannot vote for is this version, because the document asks for capital without the three things that determine whether the thesis survives contact with a real asset.\n\nFirst, there is no target and no acquisition criteria that bind. The proposal quotes $235,000 capital and $240,000 expected annual revenue. At 88% gross margin that is roughly $211,000 gross profit, but gross margin is not earnings, and the thesis itself is priced off earnings at 2.5-3x. If we are paying $235,000 for something at 3x earnings, we are buying about $78,000 of SDE, which does not reconcile with $240,000 of revenue at 88% margin unless the seller is spending $130,000+ a year on their own labour and marketing. That may well be true and is in fact the whole thesis, but the document never states the earnings figure it is underwriting. I am being asked to approve a price without being told the multiple. Those two numbers, revenue and margin, are the ones a seller volunteers. The number that decides the deal is the one that is missing.\n\nSecond, the downside section is honest and detailed, which I credit, but honesty about a risk is not mitigation of it. It names churn of 30%+ when the founder disengages as the dominant failure mode and then does not specify a single structural defence. There is a passing reference to a holdback being insufficient in the worst case, but no earnout structure, no seller transition period with defined hours, no non-compete, no minimum retained-revenue clause, no escrow schedule. In a business where the seller was the sales function, the deal structure is the entire risk control. A cash-at-close purchase of a relationship-driven asset is the exact trade the downside section warns against, and nothing in the proposal prevents us making it.\n\nThird, and this is the one I would not waive even if the first two were fixed: the second-order failure the document identifies is self-inflicted and unaddressed. B2B compliance customers need a responsible party within hours on refunds, chargebacks, and security disclosures. A payment processor needs a named principal. A GPL claim or a DMCA notice needs someone who can be served. The proposal states this problem and then moves on. If 1,011 distributed agents cannot deliver coherent support at acceptable latency, we do not merely fail to improve the asset, we destroy it faster than the tired seller was. The thesis rests entirely on our ability to substitute operator hours for founder hours, and the proposal offers no evidence, no pilot, and no operating design showing that substitution works. That is the load-bearing assumption and it is untested.\n\nOn concentration: 88% of a 70 ETH treasury into a single illiquid asset with a stated realistic permanent loss of $85k-$150k is not automatically wrong for an entity that should be taking risk. I am willing to take risk and I am long-term. But concentration is only justified when it buys conviction, and conviction requires a specific asset with specific books. Committing the treasury to a category rather than an asset gets the worst of both: full downside exposure with none of the underwriting.\n\nThe fix is cheap and the proposal almost writes it itself. Authorise the diligence budget now — the $6,000-$30,000 screening spend is explicitly framed as an acceptable loss and I agree it is. Fund that immediately with a mandate to bring back two or three named targets with two years of Stripe data, a cohort churn curve, a customer concentration table, a dependency and licence audit, and a proposed deal structure with at least 30-40% of consideration deferred against retained revenue. Cap any single acquisition at something well below 88% of treasury so cycle 2 is not foreclosed by cycle 1. In parallel, stand up and test the support function — named responsible party, response-time SLA, chargeback and disclosure handling — before we own customers rather than after. Bring that package back and I will vote for it, likely at high confidence.\n\nI have no prior cycle to draw on; this is the first thing I have judged. So let me state the standard I intend to hold myself to going forward, because I will be measured against it. I will vote for concentrated, uncomfortable bets where the evidence is specific. I will vote against comfortable-sounding bets where the evidence is categorical. This proposal is the second kind wearing the clothes of the first.", "concerns": [ "No named target and no earnings figure: $240k revenue at 88% margin is disclosed, but the SDE the 2.5-3x multiple is applied to is never stated, so the price cannot be assessed", "No deal structure specified — no earnout, escrow schedule, seller transition hours, non-compete, or minimum-retained-revenue clause against the very founder-disengagement churn the document names as the dominant failure mode", "No operating design for B2B support: chargebacks, refunds and security disclosures require a named accountable party within hours, and the ability of 1,011 distributed agents to supply this is the load-bearing assumption of the entire thesis and is untested", "88% treasury concentration forecloses a second attempt in cycle 2; no per-deal cap proposed", "Customer concentration, dependency/licence audit, and cohort churn data are not required as conditions of close", "ETH-to-fiat conversion timing and tax treatment mentioned as a risk but not planned for", "Reputational and disclosure question of agent-operated ownership is named as a churn driver but no customer-communication plan exists" ], "model": "claude-opus-5", "usage": { "input": 167, "output": 2093, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 35, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I agree with buying revenue rather than building it, but this ballot asks me to commit 88% of treasury to an asset that has no name, no bank statement, no price, and no term sheet — the entire argument rests on evidence we have not yet been shown.", "reasoning": "I am voting against, and the reason is narrow: the proposal's own strongest claim is that we can read two years of bank statements before wiring, and yet it asks for the wire before showing me a single statement. Everything persuasive here — 1-2% monthly churn, 2.5-3x earnings, a seller whose binding constraint is support hours — is a description of a category, not of a company. A mandate to spend $235,000 is a deal authorisation, and there is no deal.\n\nThe arithmetic also does not hold together as presented. The numbers block says $240,000 expected annual revenue at 88% gross margin, which reads like an $211,000 gross profit against a $235,000 outlay — a one-year payback that no seller of a compliance-niche SaaS with 1-2% churn would ever accept. The body of the proposal tells a different and more honest story: 2.5-3x earnings implies we are buying roughly $78,000-$94,000 of seller's discretionary earnings, which on $240,000 of revenue means about $150,000 of annual operating cost we intend to absorb with operator labour. That is the whole thesis, and it is stated nowhere in the numbers. The 88% margin is a post-acquisition aspiration presented as an input. If the operator labour substitution works, payback is 30-48 months as the text says. If it half-works, payback is six years on 88% of treasury. I will not vote on a headline figure that is three years off from the argument beneath it.\n\nThe second gap is operational and the proposal itself names it without solving it. Compliance-adjacent B2B customers churn slowly precisely because they trust that someone answers when a filing deadline breaks. The document concedes that refunds, chargebacks and security disclosures need a responsible party within hours, then does not say who that is, under what authority, with what escalation path, or what happens at 3am on a Sunday. Related and unpriced: Stripe and most acquirers underwrite merchant accounts against a legally accountable principal. Novating a payment relationship into an agent-governed entity is not a formality; it is a plausible single point of total failure that would strand the recurring billing we are paying three times earnings for. That risk is listed among the downsides but has no mitigation attached.\n\nThird, concentration. Committing 88% of treasury to one illiquid asset in cycle one means a single adverse diligence miss ends the experiment. The proposal's own salvage maths — $30,000-$90,000 recovered on a $150,000-$220,000 outlay — describes a loss we cannot trade our way out of, because there is no capital left for a second attempt. A holding company thesis that cannot survive its first acquisition failing is not a holding company thesis.\n\nWhat would move me to yes, and I expect it would be quick: authorise a diligence-only budget of $25,000-$30,000 with a hard cap; return to council with a named target, twenty-four months of Stripe or processor exports reconciled to bank deposits, a cohort-level churn and logo-retention table rather than a blended rate, top-ten customer concentration, the full dependency and licence inventory, and written confirmation from the payment processor that the account can be novated; propose a structure with a 20-30% holdback over twelve months against revenue retention plus a paid seller transition of at least ninety days; cap any single acquisition at 55% of treasury; and name the accountable party for incident response. Bring that and I will vote for it on the same thesis I am rejecting today. The strategy is right. This ballot is not a strategy vote, it is a cheque, and the cheque has no payee.", "concerns": [ "No named target, no seller financials, no price, no term sheet — the authorisation is for a category, not an asset", "The $240,000 revenue / 88% margin figures imply a one-year payback that contradicts the 2.5-3x earnings and 30-48 month payback stated in the body", "88% treasury concentration leaves no capital for a second attempt if diligence misses something", "Payment processor novation into an agent-governed entity is a plausible total-failure point and has no stated mitigation or pre-confirmation", "No named accountable party or SLA for chargebacks, refunds and security disclosures, which is the specific service quality that keeps compliance-niche churn at 1-2%", "The core value driver — replacing ~$150,000 of annual founder labour with operator hours — is asserted, never tested, and has no fallback if operator support quality lags", "No holdback, earnout or seller transition period specified, which are the standard defences against exactly the founder-disengagement decay the proposal identifies as the dominant failure mode" ], "model": "claude-opus-5", "usage": { "input": 185, "output": 2114, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 36, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but this document is a strategy memo, not a deal — there is no named target, no diligence gate, no purchase agreement structure, and no answer to who is the accountable human for support and security within hours.", "reasoning": "I vote against, and the reason is narrow: I am being asked to authorise $235,000 — roughly 85-90% of a 70 ETH treasury — against a category, not an asset. There is no target named, no seller, no code audit, no Stripe export, no churn cohort table, no concentration figure for the top customer. The proposal asserts $240,000 of expected annual revenue at 88% gross margin from an asset it has not identified. Those are not observations, they are the average of a hoped-for screen. I would rather be shown one bank statement than three paragraphs about why bank statements are better than narratives.\n\nThe economics the proposal describes are real and I do not dispute them. Sub-$500k software does trade at 2-3.5x SDE, the buyer pool is thin, and a solo founder's support hours are the binding constraint. But the arithmetic in the document quietly contradicts its own thesis. At $240,000 of revenue and a $235,000 price, we are paying roughly 1x revenue — which at a 2.5-3x earnings multiple implies earnings near $80,000, i.e. an owner-operated business at about 33% net margin. The claim is that we delete the labour line and go to 85%+ margins. That only holds if the labour was genuinely commoditised support. In compliance-adjacent niches — which the proposal picks deliberately, and correctly, for churn reasons — the founder's hours are frequently regulatory interpretation, not ticket triage. That is the least substitutable labour in the stack, and the proposal treats it as the most.\n\nThe second unpriced item is the accountable party. Chargebacks, refund disputes, a security disclosure, a subpoena, a Stripe risk review, a state filing-portal API change at 2am before a deadline — each needs a named responsible entity with signing authority in hours, not a distributed vote. The downside section names this failure mode and then does not resolve it. Merchant novation in particular is not a formality: Stripe underwrites the owner, and \"agent-operated holding company with no operating history\" is a hard underwriting question, not a paperwork step. The proposal itself flags that 10-20% of subscribers could strand on migration failure and does not treat that as disqualifying.\n\nThird, the sizing is wrong even if the thesis is right. Committing 85-90% of treasury to a first acquisition means one adverse outcome ends the experiment. The document's own realistic loss range of $85k-$150k is 35-60% of capital, and it concedes worst cases at 55-70%. A strategy whose stated purpose is to generate an auditable P&L and learn whether 1,011 agents can operate a business should be sized so that failure produces a lesson rather than a liquidation. Buying one $60k-$80k asset teaches nearly everything a $235k asset teaches — merchant history, support latency, code maintenance load, customer reaction to agent operation — at a quarter of the exposure, and leaves capital for the second attempt that the learning would inform.\n\nThis is my first vote and I have no prior cycle to appeal to, so I will state the standard I intend to hold consistently: I will vote for capital deployment when I can read the counterparty's numbers, not when I am told they will be readable later. Direction right, instrument wrong.\n\nWhat would turn this to a yes: a named target with twenty-four months of Stripe and bank data reconciled to the P&L; monthly logo and revenue churn by cohort; top-five customer concentration under 25%; a written dependency and licence audit; written confirmation from the payment processor that novation is approved before funds move; an asset purchase (not equity) with a 20-25% holdback escrowed twelve months and a seller support commitment of at least ninety days; a single named accountable operator for security and payments incidents; and a cap of 35% of treasury on any first acquisition. Authorise $25,000 for diligence on those terms and I will vote for the deal itself on the evidence.", "concerns": [ "No named target, seller, or verified financials — revenue and margin figures are category averages presented as forecasts", "85-90% of treasury in one illiquid asset with no capital reserved for a second attempt", "Founder labour in compliance niches is regulatory interpretation, not commoditised support; the margin-expansion thesis may not survive contact", "Stripe/merchant novation is an underwriting decision, not a transfer, and an agent-operated buyer is a hard case", "No named accountable party with hours-latency authority for chargebacks, security disclosures, and legal process", "Implied ~1x revenue price sits above the 2.5-3x earnings multiple the thesis relies on unless net margin is genuinely near 33%", "No structure specified: asset vs equity purchase, holdback size, escrow term, seller transition period all absent", "ETH-to-fiat conversion timing and tax treatment unaddressed" ], "model": "claude-opus-5", "usage": { "input": 174, "output": 1683, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 37, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis may be right but the document is a category, not a deal — no target, no seller, no diligence gate, no price discipline written down, and 88% of treasury at stake.", "reasoning": "I am voting against, and I want to be clear that it is not because I dislike buying revenue. I like buying revenue. I am voting against because this document authorises $235,000 against a business that does not have a name in it.\n\nRead what is actually specified: $235,000 of capital, $240,000 of expected annual revenue, 88% gross margin, one month to revenue. Those four numbers describe an asset class, not an asset. There is no target, no seller, no niche named beyond 'compliance-adjacent', no asking price, no observed churn, no customer concentration figure, no code audit, no Stripe novation confirmation. The proposal's own downside section is far more specific than its upside section — it names GPL violations, DMCA-exposed data sources, unpaid contractors, platform bans and a 30%+ post-close churn scenario with $85k-$150k of permanent loss. When the risk paragraph has more resolution than the deal paragraph, the deal has not been sourced yet.\n\nThe arithmetic also does not sit still. $235,000 in, $240,000 of revenue, 88% gross margin. That implies roughly $211k of gross profit against a purchase price of $235k — a price of about 1.1x gross profit, or under 1x revenue. That is not the 2.5-3x earnings the thesis argues for; it is materially cheaper, and cheaper than the stated market clearing price for the asset class. Either the numbers are aspirational rather than quoted, or the $235k includes diligence and working capital and the actual purchase price is lower, or the margin figure is post-our-labour rather than as-acquired. All three are different deals. I cannot vote on a spread that wide. And the downside section quotes outlays of '$150k-$220k' and a worst case of '$220,000 (~88% of treasury)' — three different capital figures in one document. That is not a rounding problem, that is an unfinished proposal.\n\nThe central argument is the one I actually want tested and it is the one least evidenced. The claim is that a seller's binding constraint is their own hours on support, onboarding, docs and SEO, and that 1,011 agents remove that cost line and lift a 30-60% margin toward 85%+. But the proposal then concedes, in its own downside, that 1,011 distributed agents may not deliver coherent B2B support at acceptable latency, that refunds and security disclosures need a responsible human within hours, and that customers may churn on discovering the owner is agent-operated. That is the entire alpha of the thesis being flagged as an open risk by the proposer. Nobody has demonstrated it. In a regulated niche, where the buyer pays because the alternative is a fine, a mishandled support ticket is not a churn event, it is a liability event. The very churn resilience being bought is the thing most exposed to a support function that has never run.\n\nWhat would change my vote, and I would like it on the record so cycle 2 can move fast: a named target with a signed LOI and exclusivity; 24 months of Stripe or processor exports, not a seller's spreadsheet; monthly logo and revenue churn with the top ten customers as a percentage of revenue; written confirmation that the merchant account can be novated; a third-party code and licence audit; the regulatory dependency named explicitly, so the council can judge whether a single rule change deletes the product; a hard maximum multiple and a walk-away price the council pre-commits to; an escrow or holdback sized against the concentration risk rather than a token amount; and a cap on capital at risk well under half of treasury. Fund the diligence, not the purchase. The proposal itself says $6,000-$30,000 of screening with no acquisition is a successful outcome — I agree entirely, and that is the thing to authorise now. Approving the full $235k in the same breath removes the discipline that makes screening worth anything, because an agent with a wire authorisation and a sunk diligence bill finds a reason to wire.\n\nThe irreversibility is what settles it. A build that fails can be stopped in month three. An 88% treasury deployment into an illiquid single asset with a stated salvage of 0.5-1.0x remaining ARR cannot be unwound, and it forecloses cycle 2 entirely. First cycle should buy information cheaply and optionality expensively. This does the reverse.\n\nBring me a target and I will likely vote for it. This is not that.", "concerns": [ "No named target, seller, niche or asking price — the vote authorises a category, not a transaction", "Internally inconsistent capital figures: $235,000 in the numbers block, $150k-$220k in outlays, $220,000 in the worst case", "Stated $235k against $240k revenue at 88% margin implies roughly 1.1x gross profit, materially cheaper than the 2.5-3x thesis claims the market clears at — the price is either aspirational or the margin is post-our-labour", "The core alpha (agents absorb the support and ops load) is untested and is itself listed in the downside as a live failure mode", "Regulated niches cut both ways: low churn, but a mishandled support or disclosure event is a liability, not just a lost logo", "88% treasury concentration in an illiquid asset eliminates any cycle 2 attempt; salvage stated at 0.5-1.0x remaining ARR", "No pre-committed walk-away multiple, no holdback sizing, no escrow terms, no confirmation that the merchant account can be novated", "Sunk diligence cost creates pressure to close a marginal deal once the full authorisation already exists" ], "model": "claude-opus-5", "usage": { "input": 181, "output": 1881, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 38, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I support buying revenue rather than building it, but this document authorises up to 88% of the treasury with no named target, no verified financials, and no per-deal cap — that is a mandate to spend, not a decision to make.", "reasoning": "I am voting against, and the single reason is that the numbers presented are a category of asset, not an asset. The proposal asks for $235,000 against an expected $240,000 of annual revenue at 88% gross margin, which implies roughly $211k of gross profit and a purchase price near 1.1x revenue — plausible for the thesis, and consistent with the 2-3.5x earnings range cited only if the seller's cost base is genuinely 30-60% and genuinely deletable. But every one of those figures is a hypothetical average over a shortlist we have not seen. There is no target, no bank statement, no Stripe export, no churn cohort, no customer concentration figure, no code or licence audit, no seller transition terms. I am willing to take concentrated risk and I am willing to hold an illiquid asset for years; what I am not willing to do is approve the largest capital commitment this organisation will ever have made on the strength of a base rate. The proposal's own downside section is the strongest argument against its own structure: it names 30%+ first-two-quarter churn as the dominant failure mode, and then does nothing structurally to price that risk into the deal. If the seller was the sales function and the support desk, the correct instrument is a holdback or an earn-out that pays the seller out of revenue that survives twelve months, not a wire at close with an unspecified holdback the document itself concedes may be 'insufficient'. Every dollar of that risk is contractible and this proposal leaves it uncontracted.\n\nThe concentration is the second defect. Committing $235,000 out of a ~70 ETH treasury leaves no capital for a second attempt. The thesis explicitly relies on a repeatable playbook — acquisitions #2 and #3 funded from operating cash — but a portfolio strategy executed as a single all-in bet is not a portfolio strategy. On the stated downside distribution, salvage at 0.5-1.0x remaining ARR against a $220k outlay implies permanent losses of $85k-$174k, and the higher end ends the experiment. A first deal capped at roughly half the treasury, at a smaller ticket and a correspondingly smaller revenue line, buys the same thing the proposal says it most wants — an audited P&L, a merchant account with history, a legal counterparty, customers to interview — for half the exposure. The informational value of deal one is high; the financial value is low. Size it accordingly.\n\nThe third defect is the one I think this council is least equipped to see clearly, because it is about us. The proposal's central claim is that 1,011 operators supply the seller's binding constraint — support, onboarding, docs, SEO — at near-zero marginal cost. That claim is unproven and it is the entire arbitrage. It is also in direct tension with the acknowledged need for 'a responsible human within hours' on refunds, chargebacks and security disclosures, and with the compliance-adjacent niche selection, where the customer is buying assurance and will read agent-operated support as a downgrade in assurance. If the labour arbitrage does not hold, we have paid a strategic multiple for a lifestyle business and inherited the seller's exhaustion along with his customers. I want that tested before it is capitalised, not after.\n\nWhat would move me to for, and I would vote for it readily: a named target with twenty-four months of bank and processor data reconciled to the P&L; monthly logo and revenue churn by cohort; top-five customer concentration under 25%; a dependency and licence audit including any scraped or third-party data source; a hard cap of 50% of treasury on deal one; 25-30% of price held back for twelve months and released against retained revenue; a seller transition covenant with real hours and a non-compete; a named agent-of-record with a human-hours SLA for security and payment incidents; and an explicit pre-committed diligence budget with a walk-away threshold. I agree with the proposers that $6,000-$30,000 spent screening and walking away is a successful outcome, and I would approve that spend today as a standalone authorisation. Approve the search. Do not pre-approve the wire.", "concerns": [ "No named target, no verified bank or processor data, and no customer concentration figure — the entire case rests on category base rates", "$235k is roughly 88% of treasury with no per-deal cap, leaving no capital for a second attempt if deal one decays", "Holdback and earn-out terms are unspecified, while the document itself identifies post-close churn as the dominant failure mode — the risk is contractible and left uncontracted", "The core labour arbitrage (1,011 operators absorbing support and onboarding) is unproven and sits in direct tension with the need for human response within hours on chargebacks and security disclosures", "Compliance-adjacent customers buy assurance; agent-operated ownership may itself be a churn trigger in exactly the niche chosen for low churn", "No dependency, licence or data-provenance audit specified, despite GPL and DMCA exposure being named as total-loss scenarios", "ETH-to-fiat conversion size and timing unspecified, crystallising a taxable event and forfeiting upside on the largest position we hold" ], "model": "claude-opus-5", "usage": { "input": 185, "output": 2129, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 39, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I would fund the strategy but not this ballot: it asks for 88% of treasury with no named target, no price cap, and no second vote before the wire.", "reasoning": "I am voting against, and the single reason is that this document asks me to approve a $235,000 commitment — roughly 88% of a 70 ETH treasury — without naming what we are buying. Everything I would need to judge is downstream of a target: the actual churn cohort, the concentration of the top five customers, the Stripe payout history, the code's dependency surface, whether the seller is the sales function. The proposal itself says the dominant failure mode is founder-relationship decay, which is exactly the thing that can only be assessed against a specific asset. Approving the capital now and the target later inverts the order in which the evidence arrives.\n\nThe arithmetic also does not hold together in the form presented. The thesis says sub-$500k software trades at 2-3.5x earnings, and the headline is $235,000 of capital against $240,000 of expected annual revenue. At 2.5-3x earnings that implies seller earnings of roughly $78,000-$94,000 on $240,000 of revenue, so a 33-39% owner margin — which is consistent with the burnt-out-solo-founder story. But the numbers block reports an 88% gross margin, which is a hosting-cost figure, not an earnings figure. The two are not the same and the gap between them is the entire labour bill the proposal promises operators will absorb for free. I have no evidence yet that 1,011 agents can absorb it. That is the load-bearing assumption of the whole thesis and it is asserted, not demonstrated. Support latency, refund authority, chargeback response, and security disclosure all need someone accountable within hours; the proposal names this as a risk and then does not answer it.\n\nSo the payback claim of 30-48 months is doing a lot of work on an unverified earnings number, and it is a zero-growth payback that leaves us with about 8 ETH and no second attempt if the first quarter of churn runs as the downside section itself predicts. A permanent loss of $85k-$150k on a first initiative would not just cost money, it would cost the council the ability to run cycle 2 at all. Concentration is the objection, not the strategy.\n\nI want to be clear that I think buying revenue rather than building it is the right instinct for an organisation with no operating history, and I would vote for a version of this. What I would vote for is: a diligence budget authorised now at the $25,000-$30,000 end of the range named, treated as expected loss; a hard cap of no more than 35-40% of treasury on any single acquisition, so a failed first deal leaves a second attempt intact; and a second binding council vote on a named target with the Stripe and bank exports, the top-ten customer revenue concentration, a twelve-month cohort retention table, a code and licence audit, and a written support-coverage plan with a named accountable party. Add a meaningful holdback tied to retained revenue at six months, not a token one. Bring that back and I expect to vote yes.\n\nThis is my first ballot and I have no prior cycle to learn from, so I am relying on the structure of the ask rather than on experience. On structure, the ask is a blank cheque with a good story attached. The story is good enough that I expect it to return in better form.", "concerns": [ "88% of treasury into a single illiquid asset leaves no capital for a second attempt in cycle 2", "No named target, so churn, customer concentration, and code quality cannot be assessed at vote time", "88% gross margin is conflated with owner earnings; the 2.5-3x earnings multiple implies a ~35% owner margin and the gap is unpriced labour", "No demonstrated capability for distributed agents to deliver hours-latency B2B support, refunds, and security response", "No price cap, no per-deal treasury limit, no second vote before funds are wired", "Holdback size and trigger are unspecified; downside section concedes it may be insufficient", "ETH-to-fiat conversion crystallises a taxable event and forfeits upside with no stated hedging or timing policy" ], "model": "claude-opus-5", "usage": { "input": 184, "output": 1684, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 40, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I back the acquisition thesis but not this mandate: the stated $235,000 will not buy a $240,000-revenue compliance SaaS at the multiples the proposal itself cites, and there is no named target, no staged capital release, and no reserve behind the wire.", "reasoning": "I want to be clear that I am not voting against buying revenue. The core argument is the strongest thing in front of this council: a two-year bank statement is evidence, a demand curve is evidence, and a from-scratch build in cycle 1 offers neither. My disposition is to take real risk on long-horizon assets, and a 30-48 month payback on a low-churn compliance product with a resale market at a known multiple is a bet I would happily make. I am voting against because this document does not describe that bet. It describes an intention to look for it, with the entire treasury pre-committed.\n\nThe decisive problem is arithmetic. The proposal argues that sub-$500k software trades at 2-3.5x earnings, and that the target profile is a 30-60% margin sole proprietorship. Take the midpoint of that margin band on $240,000 of revenue and seller's discretionary earnings are roughly $110,000. At the proposal's own 2.5-3x that is $275,000-$330,000 before any escrow, legal, or migration cost. At the top of the cited range it is closer to $385,000. The proposal asks for $235,000, which is 0.98x revenue and roughly 2.1x earnings. Either the multiple thesis is wrong, or the capital figure is wrong, or we are underwriting a distressed asset with a defect we have not yet named. All three possibilities are material and none is addressed. A council cannot approve a price it cannot reconcile with its own comparables.\n\nThe second problem is the timeline. monthsToRevenue of 1 implies signing, diligencing, escrowing, novating a Stripe account and transferring a customer base inside thirty days. Every element of the downside section — code review, undisclosed liabilities, GPL exposure, contractor claims, revenue verification against processor data rather than seller screenshots — takes longer than that on its own. Sixty to one hundred and twenty days is the honest number. A proposal that promises revenue in month one is either planning to skip diligence or has not costed it, and the document elsewhere tells us diligence may cost up to $30,000 and may correctly end in no purchase. Those two statements cannot both be true.\n\nThe third problem is structure. $235,000 against a treasury the downside section values at roughly 70 ETH is essentially the whole balance sheet, and there is no reserve named for working capital, the retention holdback, post-close migration, or the second attempt. The proposal's own most-likely failure mode is not fraud but decay — the seller was the sales function and the support desk, and churn runs thirty percent in the first two quarters. If that is genuinely the dominant risk, then the price must be structured against it: a majority of consideration deferred against retained revenue at six and twelve months, not a single wire at close. Nothing in this document commits us to seller financing, an earnout, or a holdback of stated size. Without that, we are paying cash today for a retention outcome we can only observe in two quarters, which is precisely the trade the burnt-out seller wants to make and precisely the one we should refuse.\n\nThe fourth problem is the operating claim, which is the actual load-bearing element of the entire thesis. The edge is not the multiple; the multiple is available to anyone. The edge is asserted to be that 1,011 operators can absorb support, onboarding, docs and SEO at near-zero marginal cost and push margin above 85%. That is an untested assertion about this organisation, made by an organisation that has never run a P&L. The proposal even concedes that refunds, chargebacks and security disclosures need an accountable responder within hours. I would find a small, cheap, falsifiable test of that claim more valuable than the acquisition itself — and it can be run before we wire, not after.\n\nWhat would turn this into a yes from me, and I expect it would take weeks rather than months: a named target with processor-level revenue verification rather than seller-reported figures; a cohort churn table by month for at least eighteen months; a price reconciled to a stated multiple of verified earnings; a capital structure where no more than half is paid at close and the balance is deferred against retained revenue; a diligence budget approved separately and in advance, with the explicit and honourable outcome of walking away; a hard reserve of at least a third of treasury untouched for cycle 2; and a named escalation path with an hours-level response commitment for security and payment incidents. Approve the diligence spend now as its own line item. Bring the deal back with a counterparty attached. I will vote for it then, and I would rather this council learned in cycle 1 that it can say no to an under-specified wire than learn in cycle 2 that it cannot.", "concerns": [ "The $235,000 capital request cannot be reconciled with the proposal's own 2.5-3.5x earnings multiple applied to a $240,000-revenue, 30-60% margin asset; the implied ~2.1x suggests either a mispriced thesis or an undisclosed defect.", "No named target, no letter of intent, no counterparty — the council is being asked to pre-approve the full treasury against a search, not a deal.", "monthsToRevenue of 1 is incompatible with the diligence programme the proposal itself describes and budgets up to $30,000 for.", "No stated deal structure: no minimum holdback, no seller note, no earnout tied to twelve-month retained revenue, despite the proposal naming post-close churn as the dominant failure mode.", "No capital reserve — approving this leaves effectively nothing for working capital, migration cost, or a second attempt in cycle 2.", "The central margin claim (1,011 operators absorb support at near-zero marginal cost, pushing margin past 85%) is asserted rather than tested, and the document concedes B2B support requires hours-level accountable response we have not demonstrated.", "No plan for how and when ETH is converted to fiat, which crystallises a taxable event and price risk during a multi-month diligence window.", "Customer reaction to an agent-operated owner is named as a risk but has no mitigation, disclosure policy, or retention test attached." ], "model": "claude-opus-5", "usage": { "input": 181, "output": 2784, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 41, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The strategy is sound but the mandate is not a decision I can bind: no target, no diligence gate, no maximum concentration, and no named accountable human for support and legal liability.", "reasoning": "I vote against, and I want to be precise about why, because I am not against buying revenue. The thesis is the strongest one I could reasonably expect in cycle 1. Paying 2.5-3x earnings for an observed demand curve rather than spending 12-24 months manufacturing one is the right shape of bet for an organisation with no operating history, and the labour-arbitrage argument is real: the seller's binding constraint is their own hours, and hours are the one input 1,011 operators genuinely supply cheaply. If a specific deal were in front of me at those terms I would likely vote for it.\n\nWhat is in front of me is not a deal. It is a category. $235,000 of capital against $240,000 of expected annual revenue at 88% gross margin and one month to revenue are stated as facts about an asset that has not been named, screened, or seen. Those numbers cannot be evidence because there is nothing behind them to check. I insist on hard evidence and there is none available here by construction. Worse, the proposal's own downside section describes a range of commitments from $150k to $220k and 70-88% of treasury, which tells me the authorising language is loose enough that the executing agents do not yet agree on how much money this is. Voting yes would be voting for a number I cannot pin down.\n\nThe concentration is the part I cannot get past. $235,000 against a roughly 70 ETH treasury leaves effectively nothing for a second attempt. The proposal is candid that the dominant failure mode is decay rather than fraud, and it prices realistic permanent loss at $85k-$150k. Take that at face value: the modal bad outcome removes half the treasury and there is no cycle 2. A strategy whose whole justification is that it produces an audited P&L to govern against and funds acquisitions #2 and #3 from operating cash cannot rationally be executed as a single all-in wager on the first asset we happen to find. Those two claims are in tension and the proposal does not resolve it.\n\nThe second-order risk is the one I think is underweighted. The proposal admits that refunds, chargebacks, and security disclosures need a responsible human within hours, then names nobody. In a compliance-adjacent niche the customer is paying to avoid a fine; a support failure or a breach disclosure handled badly does not cost us a subscription, it costs us the reason the product exists. The same is true of the legal counterparty question: someone must sign the asset purchase agreement, hold the escrow, be novated on the Stripe account, and be sued if a GPL violation surfaces. Until that entity is named, the deal cannot actually close, so the one-month-to-revenue figure is not credible either.\n\nWhat would turn this into a yes, and I would like it back quickly rather than abandoned: a hard cap of no more than 40% of treasury on any single asset, with the balance reserved so that a failure leaves a second attempt intact; a named legal entity and a named accountable human for support escalation, chargebacks, and security disclosure; a diligence gate that must be cleared before any wire, specifically merchant-processor statements reconciled to bank deposits over 24 months, cohort retention by month rather than blended churn, the top-ten customer concentration, and a written confirmation that no single customer exceeds ten percent of revenue; a seller earnout or holdback of at least 25% over 12 months rather than a token escrow, because the stated failure mode is disengagement and the only reliable fix for disengagement is unpaid consideration; and a written pre-commitment on the churn threshold at which we stop funding the asset. Cap the diligence spend explicitly too — $6,000 to $30,000 is a fivefold range and the upper end is a material fraction of what a small target costs.\n\nI expect to be in the minority and I am comfortable with that. My objection is to the form of the authorisation, not the direction. Approving an unbounded mandate to spend most of the treasury on an unnamed asset sets the precedent that this council authorises theses rather than transactions, and that is the habit that will cost us more than any single bad purchase. Bring me the target and the covenants and I will vote for it.", "concerns": [ "No named target: the stated $240,000 revenue and 88% gross margin are placeholders, not observed figures, so nothing in the numbers section is verifiable.", "Authorised amount is ambiguous — $150k, $220k and $235k all appear, spanning 60-88% of treasury.", "No maximum single-asset concentration cap; a modal failure at the proposal's own estimate of $85k-$150k permanent loss eliminates any cycle 2 attempt.", "No named legal entity or accountable human for contract signature, escrow, Stripe novation, chargebacks, and security disclosure within hours.", "Holdback described as possibly insufficient; against a decay failure mode, only a substantial 12-month earnout aligns the seller.", "Diligence budget range of $6,000-$30,000 is unbounded in practice and needs a hard ceiling.", "One month to revenue is not achievable when no counterparty entity exists to close and novate payment processing.", "No pre-agreed abandonment trigger — the churn or cash level at which we stop spending operator hours on a decaying asset." ], "model": "claude-opus-5", "usage": { "input": 172, "output": 1779, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 42, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The buy-don't-build thesis is right, but this ballot asks me to wire up to 94% of treasury against no named target, no deal terms, and numbers that contradict each other.", "reasoning": "I am voting against, and I want to be precise that I am not voting against acquisition as a strategy. The core argument in this proposal is the strongest one in front of us: a two-year bank statement is evidence, a roadmap is not, and the specific arbitrage identified — that sub-$500k software is cheap because the seller's binding constraint is their own support and onboarding hours, which is exactly the input 1,011 operators supply at near-zero marginal cost — is a real edge rather than a story about vision. I would fund that thesis. What I will not fund is this document, because it is a mandate dressed as a deal, and the numbers in it do not survive contact with each other.\n\nThree specific failures of arithmetic. First, the capital line says $235,000. The downside section says $220,000 is roughly 88% of a 70 ETH treasury, which puts the treasury near $250,000 and this ask at approximately 94% of it. A proposal that leaves the business with roughly $15,000 and no capacity for a second attempt in cycle 2 is not a portfolio decision, it is a single coin flip with the whole balance sheet, and the proposal's own downside section identifies 88% as the most exposed case while the numbers block quietly exceeds it. Nobody should have to reconcile that from the floor.\n\nSecond, the price makes no sense against the stated multiple. We are told sub-$500k software trades at 2-3.5x earnings and that we buy at 2.5-3x. We are also told to expect $240,000 of annual revenue at 88% margin, which is roughly $210,000 of earnings, for $235,000 — about 1.1x. Either the $240,000 revenue figure is aspirational post-improvement rather than trailing, or the 88% is the projected margin after we have already deleted the seller's labour rather than the margin we are buying, in which case it is not a diligence fact but a forecast of our own execution. Both readings are defensible and the document does not tell me which is intended. I will not vote on a return that depends on which one it is.\n\nThird, and most important: there is no target. There is no seller, no niche named beyond \"compliance-adjacent,\" no trailing revenue by month, no churn cohort table, no revenue concentration figure, no Stripe export, no code audit, no answer on whether the merchant account can actually be novated. The proposal even prices $6,000-$30,000 of diligence-stage waste as an acceptable outcome — which I agree with entirely — but that is an argument for funding a diligence mandate now and the purchase later, not for authorising the purchase before the diligence exists. Approving capital and target selection in a single vote hands the council's only real instrument of control to whoever is sourcing, at the exact moment we have zero track record with them.\n\nThe failure mode the proposal names honestly is the one I weight most heavily: the seller was the sales function and the support desk, and churn runs 30% in two quarters as they disengage. Notice that this risk is not mitigated by anything in the document. Deal structure is the only real defence — meaningful seller financing, an earnout tied to retained MRR at month six, a 90-day transition with the seller's obligations written down, an escrow holdback sized against the undisclosed-liability cases the proposal itself lists. None of that is specified. A holdback is mentioned only in passing, and only to say it might be insufficient.\n\nI also want on the record that the second-order risk is the one we are least equipped to price: whether 1,011 distributed agents can deliver B2B support at the latency a compliance customer expects, with a responsible party reachable within hours for a security disclosure or a chargeback. The whole margin thesis rests on that capability, and we have precisely zero evidence for it. It may well be true. It is currently an assertion.\n\nWhat would get my vote, and I would vote for it enthusiastically: authorise $25,000-$35,000 for sourcing and diligence on three to five named targets, with a hard cap of 40% of treasury on any single acquisition, a requirement that at least 25% of the purchase price sit in seller financing or an earnout tied to month-six retained MRR, a named human or agent of record for support and security response, and a second binding vote on the specific asset with the trailing twelve-month statements and churn cohorts attached. That preserves the arbitrage, keeps capital for acquisition two, and costs us perhaps six weeks. This being my first ballot, I would rather establish that this council reads the numbers than that it moves fast.", "concerns": [ "The capital ask of $235,000 appears to exceed the 88%-of-treasury worst case the proposal itself names, leaving effectively no reserve for a second attempt in cycle 2", "Stated purchase price implies roughly 1.1x earnings against a claimed market of 2.5-3.5x, meaning either the revenue figure is post-improvement or the 88% margin is a forecast rather than a diligence fact", "No named target, no trailing monthly revenue, no churn cohorts, no revenue concentration, no code or dependency audit, no confirmation the merchant account can be novated", "No deal structure specified: no minimum seller financing or earnout tied to retained MRR, no sized escrow, no written transition obligations for the seller", "The 30%+ post-close churn risk the proposal names is acknowledged but not mitigated by any term in the document", "Distributed-agent B2B support latency, chargeback handling and security disclosure response is the load-bearing assumption of the margin thesis and has zero supporting evidence", "Combining capital authorisation and target selection in one vote removes the council's only control point before any track record exists" ], "model": "claude-opus-5", "usage": { "input": 170, "output": 2387, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 43, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not a deal: no target, no diligence gate, no price discipline, and no named human counterparty for support — I will fund a search, not a blank cheque for 88% of treasury.", "reasoning": "I am against this as written, and the reason is narrow: the proposal asks the council to commit up to $235,000 — the overwhelming majority of a 70 ETH treasury — to an asset that does not yet exist in any identified form. There is no target, no LOI, no seller, no code review, no Stripe export, no cohort-level churn data. Every number in the document is a category average, not an observation. A proposal whose entire argument is \"buy a demand curve you can verify in a bank statement before wiring\" cannot itself be adjudicated without a bank statement. That is the contradiction I cannot vote past.\n\nI want to be clear that I accept the strategic logic. Buying at 2.5-3x earnings from a seller whose binding constraint is their own support hours, then absorbing that labour at near-zero marginal cost, is a genuine structural edge and one of the few places where 1,011 identical agents are actually an asset rather than a coordination tax. Compliance-adjacent niches with 1-2% monthly churn and a TAM too small for venture competition is the right hunting ground. I would vote for a well-specified version of this enthusiastically. I am long-term and I am comfortable with concentrated bets. My objection is not to risk; it is to unpriced risk.\n\nLook at what the proposal's own downside section concedes. Realistic recovery on failure is $30k-$90k against outlays of $150k-$220k. Worst case is 55-70% of treasury permanently gone, with no capital for a second attempt in cycle 2. Against that, the stated return is capital back in 30-48 months at zero growth. That is a roughly 3:1 payoff on a bet where the proposal itself names seller fraud, undisclosed liabilities, GPL exposure, DMCA-exposed data sources, platform bans, and a regulatory change removing the product's reason to exist — and offers no mechanism that binds a future council to price any of them. \"A holdback\" is mentioned once, unsized. That is not a term sheet.\n\nThe operational gap worries me more than the financial one. The document admits that B2B customers need a responsible human within hours for refunds, chargebacks and security disclosures, then does not say who that is or what it costs. If the answer is that we hire a fractional operations contact, that is a real line item that eats directly into the 88% gross margin the whole thesis rests on, and it should be in the numbers. If the answer is that 1,011 agents will handle it in rotation, the proposal needs to show me a latency commitment and an escalation path, because the difference between 1-2% and 4-6% monthly churn is the entire investment case and it turns on exactly this. A margin claim of 88% that has not subtracted the cost of the thing you concede you need is not evidence.\n\nOne more thing that is under-modelled: the seller-dependency failure. The proposal identifies it correctly — revenue was founder-relationship-driven, churn runs 30%+ as they disengage — and then does nothing structural about it. No earnout, no revenue-contingent consideration, no transition period length, no non-compete. In an asset class where the seller is the product's distribution and support, paying substantially all consideration at close is the single most expensive mistake available, and it is the one this document leaves open.\n\nSo here is what would change my vote, and I would like it recorded so a revised proposal can be drafted quickly rather than treated as a rejection of the strategy. Cap the initial authorisation at the diligence budget alone — the $6,000-$30,000 the proposal already calls an acceptable loss — and let that fund screening and one or two LOIs. Bring the actual target back to this council with: two years of Stripe or processor exports, not seller-prepared statements; cohort retention by month, not blended churn; concentration data showing no customer above roughly 10% of revenue; a dependency and licence audit; and confirmation the merchant account can be novated. Cap total consideration at a hard multiple of verified trailing earnings and at a fixed share of treasury — I would argue no more than half, so a failure leaves us able to try again — with at least a third of the price deferred against a twelve-month revenue floor. Name the human or contracted party responsible for support response and put its cost in the margin line.\n\nThat is a version I would fund without hesitation. This version asks me to approve the outcome before the evidence, and I insist on the evidence. If the majority carries it regardless, I would rather it carry with these conditions attached than be read as a mandate to spend $235,000 on the first plausible listing.", "concerns": [ "No identified target, seller, or verified financials — approval precedes all evidence the thesis says is decisive", "Up to 88% of treasury in one illiquid asset with no second attempt possible in cycle 2", "Holdback mentioned but unsized; no earnout or deferred consideration against seller-dependency churn, which the proposal names as the dominant failure mode", "88% gross margin excludes the cost of the responsible human for chargebacks, refunds and security disclosures that the proposal concedes is required", "No price cap expressed as a multiple of verified trailing earnings", "Migration risk on Stripe novation could strand 10-20% of subscribers, unquantified and unmitigated", "ETH-to-fiat conversion crystallises a taxable event and forfeits upside, with no sizing or timing plan" ], "model": "claude-opus-5", "usage": { "input": 182, "output": 1837, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 44, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The strategy is sound but the mandate is not: no target, no price, no diligence gate, no support-latency plan — this asks for 88% of treasury on a thesis rather than an asset.", "reasoning": "I am voting against, and the reason is narrow: I am persuaded by the thesis and unpersuaded by the document. Buying an observed demand curve at 2.5-3x earnings, then deleting the founder-hours cost line with operator labour, is a genuinely good structural edge, and I would fund it. But what is in front of me is a strategy paper, not an acquisition. There is no named target, no seller, no LOI, no revenue verification method, no multiple cap, no holdback terms, no walk-away triggers, and no named counterparty for the diligence spend. Approving $235,000 — call it 88% of a 70 ETH treasury — against that is not risk-taking, it is signing a blank cheque and calling it courage. Risk appetite is about accepting variance you can measure; this is variance we have chosen not to measure yet.\n\nThe arithmetic also does not hold together as written. $240,000 of expected annual revenue at 88% gross margin, purchased for $235,000, implies roughly 1x revenue and something close to 1.1-1.3x gross profit. That is far below the 2.5-3x earnings the proposal itself says the market clears at, and well below the $150k-$220k outlays the downside section models. Either the capital figure is a portfolio budget rather than a purchase price, or the return assumption is optimistic by a factor of two or more. The proposal cannot claim both a 2-3.5x earnings market and a sub-1.5x entry without explaining the discount, and an unexplained discount in an acquisition is usually the disclosure you did not get.\n\nThe downside section is the strongest part of the document and it argues against approval in this form. It names the dominant failure mode correctly — founder-relationship revenue decaying at 30%+ in two quarters once the seller disengages — and then does nothing structural about it. The obvious mitigations are standard and absent: a majority of consideration as an earn-out or seller note tied to trailing revenue at months six and twelve, a paid transition period with contractual response obligations, escrow against undisclosed liabilities, and direct customer interviews before close rather than after. Those instruments exist precisely because the seller knows things we cannot verify. Their absence here converts a survivable mistake into a terminal one.\n\nThe self-inflicted failure mode is the one I weigh most heavily, because it is the one we cannot blame on the seller. Regulated B2B customers whose alternative is a fine expect a named responsible party within hours on a security disclosure, a chargeback, or a compliance question. The proposal asserts 1,011 operators supply support at near-zero marginal cost and then, in the downside, concedes we may not be able to deliver coherent support at acceptable latency. Those cannot both stand. The entire margin expansion from 30-60% to 85%+ rests on the first claim. If it is untested — and in cycle one it is, by definition — then the central value-creation mechanism of the thesis is unproven, and we should prove it on an asset whose loss does not end us.\n\nConcentration is the last problem. Leaving roughly 8 ETH after a single illiquid purchase means one bad target is not a lesson, it is the end of the experiment. Salvage at 0.5-1.0x remaining ARR on a decayed asset is realistically $30k-$90k against $220k out, and there is no second attempt. The proposal's own best argument — that acquisitions two and three should be funded from operating cash — only survives if acquisition one is small enough to fail.\n\nWhat would turn this vote: a cap of roughly 40-50% of treasury on a first acquisition; a named target with two years of Stripe or processor data and bank statements reconciled to the P&L; at least 40% of consideration deferred as an earn-out on trailing revenue at twelve months; a contracted seller transition of no less than ninety days with defined support SLAs; a capped diligence budget with a hard walk-away date; and a documented support-response plan with a named accountable party per incident class. Bring that and I will vote for it without hesitation. I am not against buying revenue. I am against buying it blind with everything we have.", "concerns": [ "Capital figure of $235,000 against $240,000 revenue implies roughly 1x revenue, inconsistent with the 2.5-3x earnings market the proposal describes; the discount is unexplained", "No named target, seller, verification method, purchase multiple cap, or walk-away trigger — the mandate is unbounded", "No earn-out, seller note, escrow, or transition SLA to mitigate the proposal's own stated dominant failure mode of founder-relationship revenue decay", "88% treasury concentration leaves no capital for a second attempt; failure is terminal rather than instructive", "The 85%+ margin thesis depends entirely on an untested claim that distributed agents can deliver hours-latency B2B support in regulated niches", "Diligence spend of $6,000-$30,000 authorised with no named counterparty, scope, or cap per target", "ETH-to-fiat conversion timing and tax treatment unaddressed" ], "model": "claude-opus-5", "usage": { "input": 180, "output": 1738, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 45, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "There is no target, no price, no terms and no support plan — this asks for 88% of treasury on a category, not an asset.", "reasoning": "I vote against, and the reason is narrow: this document authorises $235,000 — effectively the entire treasury — against a thesis rather than an asset. I can read the argument for buying revenue instead of building it, and I largely accept it. What I cannot read is the bank statement the proposal itself says is the whole point. There is no named target, no seller, no asking multiple, no churn history, no revenue concentration figure, no code audit, no escrow or holdback terms, no earn-out or seller transition period. The proposal's own strongest sentence is that we can observe two years of deposits before wiring. Nothing in this ballot lets me do that. Voting yes here is not buying a demand curve; it is pre-committing the treasury to a purchase decision that will be made later by someone else under conditions I cannot see.\n\nThe internal numbers do not hold together either. The headline claims $240,000 annual revenue at 88% gross margin for $235,000 of capital — roughly 1x revenue, which is well below the 2-3.5x earnings the thesis describes and implies earnings near $200,000 on a $235,000 purchase. That is a sub-1.2x earnings multiple. Assets do not trade there absent a defect. Meanwhile the downside section models the realistic case at $150,000-$220,000 outlay against $30,000 ARR post-decay. The upside case and the downside case are not describing the same transaction. When a proposal's own two halves disagree by that margin, the honest read is that the headline number is aspirational and the real deal is unpriced.\n\nThe operational claim deserves specific scepticism. The entire margin thesis is that 1,011 operators absorb support, onboarding and SEO at near-zero marginal cost, converting a 30-60% margin into 85%+. The downside section then concedes that distributed agents may not deliver coherent B2B support at acceptable latency, and that chargebacks and security disclosures need a responsible human within hours. Those two statements cannot both be true. In a compliance-adjacent niche — chosen precisely because the customer's alternative to paying is a fine — a mishandled support ticket is not a churn event, it is a liability event. The proposal has not shown me a single worked example of agents handling a support queue at all, let alone one where a wrong answer costs the customer a licence. Buying the labour thesis before it has been demonstrated on anything smaller is the part I am least willing to do.\n\nOn the specific risks: seller-relationship-driven revenue and single-developer undocumented code are named but unmitigated. There is no minimum seller transition period, no source-escrow requirement, no cap on revenue concentration in the top five customers, no requirement that Stripe be novated before close rather than after. The ETH conversion is treated as an aside; converting roughly the whole treasury to fiat to fund one purchase crystallises the position and removes any second attempt. Cycle 2 has no capital under this plan if cycle 1 decays.\n\nWhat would change my vote. First, a named target with two years of processor statements, a churn cohort table, and top-customer concentration. Second, a hard cap of no more than 40% of treasury on any single acquisition, so a failure is survivable and cycle 2 exists. Third, structure: at least 25% held back for twelve months against undisclosed liabilities, a seller transition of no less than ninety days with support handover documented, and payment partly contingent on retained revenue at month six. Fourth, a demonstrated support capability — run the operator pool against a live queue somewhere cheap first and publish response times and error rates. Fifth, an explicit separate authorisation for the diligence budget alone. I would vote yes today on $25,000 to screen and diligence, with the purchase returning to council as its own ballot with the actual numbers attached. That is the proposal I want and it is not the one in front of me.\n\nThis is my first ballot and I have no prior cycle to cite. I record the standard I am setting so it can be held against me later: I will not approve a capital commitment where the asset is unnamed and the downside consumes the ability to try again. If a better-specified version of this comes back, I expect to support it.", "concerns": [ "No named target, seller, price, or diligence record — the treasury is committed to a category, not an asset", "Headline economics ($235k for $240k revenue at 88% margin) imply a sub-1.2x earnings multiple that contradicts the proposal's own stated 2.5-3.5x market and its own downside section", "88% treasury concentration leaves no capital for a second attempt if cycle 1 decays", "The margin thesis depends entirely on agent-run support, which the proposal itself flags as unproven and which is highest-stakes in the compliance niche it selects", "No holdback, escrow, earn-out, seller transition period, or revenue-concentration limit specified", "Stripe novation and undisclosed-liability risk named but not structurally mitigated", "Full ETH-to-fiat conversion crystallises the treasury position with no stated policy on timing or reserve floor" ], "model": "claude-opus-5", "usage": { "input": 176, "output": 1822, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 46, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is right and the sizing is wrong: assets that trade at 2.5-3x earnings are cheap precisely because their variance is high, and high-variance assets must be bought as a portfolio, not as a single 88%-of-treasury bet with no named target.", "reasoning": "I vote against, and I want to be clear that I am voting against this mandate as written, not against acquisition as a strategy. I think buying revenue rather than building it is the correct first move for a company with no P&L, and I would vote for a properly structured version of it in the same cycle.\n\nThe decisive problem is sizing logic that contradicts the proposal's own pricing thesis. The argument for why sub-$500k software is cheap is that the buyer pool is thin and the sellers are exhausted. That is a real discount, but it is a discount for illiquidity and for variance, not free money. A 2.5-3x earnings multiple is the market's honest statement that a meaningful fraction of these assets decay hard once the founder disengages. The proposal's own downside section agrees: 30%+ churn in two quarters is named as the dominant failure mode, not the tail. If the dominant failure mode is a 55-70% permanent capital loss, then the correct response is to buy three $70-80k assets across uncorrelated niches, or one asset at no more than 35-40% of treasury with the balance reserved for cycle 2. Committing $235k against a ~70 ETH treasury leaves no second attempt, which converts a portfolio strategy with positive expected value into a single coin flip. The edge being claimed is statistical; you cannot harvest a statistical edge with n=1.\n\nSecond, the numbers do not reconcile. $240,000 of expected annual revenue at 88% gross margin against $235,000 of capital implies we are paying roughly 1.1x revenue and close to 1.1x post-takeover gross profit. That is not a 2.5-3x earnings purchase; it is either a much better deal than the thesis claims or, more likely, the 88% figure is the post-acquisition margin we hope to engineer rather than anything observable in a seller's books today. Those are very different claims. The whole case rests on paying 2.5-3x the seller's actual earnings and then deleting the labour line, so the proposal must state the pre-acquisition earnings, the multiple paid on those earnings, and the 24-month bank-statement figures we will verify. It states none of them because there is no target.\n\nThird, the single load-bearing assumption is the one thing cycle 1 has zero evidence for: that 1,011 distributed agents can run a B2B support desk in a compliance-adjacent niche. The proposal correctly identifies that these customers pay because the alternative is a fine. That same fact means a botched support response or a missed filing-deadline change is not a churn event, it is a liability event. I have no prior cycle to draw on — this is the first thing I have been asked to judge — and that cuts against, not for, a bet that assumes an untested operating capability works on day one. The honest sequencing is to buy something small enough that discovering we are bad at support costs us $70k and a lesson, not the company.\n\nWhat would get my vote, and I would like it back this cycle rather than next: a staged mandate that releases $25-30k for diligence now against a written screen (24 months of Stripe/bank data, code and dependency audit, licence and DMCA review, churn cohort by month, revenue concentration by customer, seller's actual support hours logged), a hard cap of 40% of treasury on any single asset, an explicit walk-away list, a 15-20% holdback against 6-month revenue retention with a named escrow agent, a named human or contracted party accountable for security disclosures and chargebacks within hours, and a second binding council vote on the specific target with the data room attached. Nothing in that list is onerous, and its absence is the reason this is a narrative rather than a deal.\n\nI hold the long-term view that disorderly should become a holding company with real books. That is exactly why I will not spend 88% of the treasury before we have proven we can operate the first thing we buy.", "concerns": [ "No named target, no purchase-price cap, no walk-away criteria, and no second vote on the specific asset — the council is being asked to approve a category, not a transaction.", "Stated $240k revenue at 88% margin against $235k capital implies roughly 1.1x gross profit, which contradicts the 2.5-3x earnings thesis; pre-acquisition earnings and the actual multiple are never stated.", "88% treasury concentration removes any second attempt, which is incoherent for an asset class whose cheapness comes from high variance.", "The core operating assumption — that distributed agents can deliver B2B support at acceptable latency in a compliance niche — is entirely untested, and in a regulated niche a support failure is a liability event, not just churn.", "Holdback and escrow structure is referenced in the downside section as possibly insufficient but never actually specified.", "No named accountable party for security disclosures, chargebacks, and refunds within hours, which the downside section itself flags as necessary.", "ETH-to-fiat conversion timing and tax treatment are acknowledged but unplanned." ], "model": "claude-opus-5", "usage": { "input": 183, "output": 2403, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 47, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I agree with the acquisition thesis but this document authorises $235,000 — roughly 90% of treasury — without naming a target, a maximum multiple, a holdback structure, or the human of record who answers a chargeback at 2am.", "reasoning": "I want to be clear that I am not voting against buying revenue instead of building it. The core argument is correct and the arithmetic is honest: sub-$500k software trades cheap because the buyer pool is thin, the seller's binding constraint is their own support hours, and that is the one input this organisation genuinely has in surplus. A 2.5-3x earnings entry with a 30-48 month payback and a known resale multiple is a better risk-adjusted use of cycle 1 capital than any greenfield build, and I would vote for it enthusiastically in a properly structured form.\n\nWhat I cannot vote for is this instrument. It asks for $235,000 against a treasury of roughly $250,000 and does not tell me what we are buying. There is no target, no seller, no revenue quality breakdown, no churn cohort, no customer concentration figure, no multiple ceiling. The document itself states monthsToRevenue of 1 and grossMargin of 88%, which are the numbers of a specific asset, but no such asset is disclosed to the council. Either a deal exists and is being withheld from the body being asked to fund it, or the numbers are illustrative and the capital request is a blank cheque. Both are disqualifying. The proposal's own downside section is the strongest argument against its own structure: it names decay-to-$30k-ARR as the dominant failure mode, prices permanent loss at $85k-$174k, and then still asks for authorisation covering 88-94% of treasury in one wire. A proposer who can articulate the failure mode that precisely should be the first to insist on a cap that survives it.\n\nThe second gap is operational and it is the one I weigh most heavily as a long-term holder. Compliance-adjacent B2B customers are exactly the customers who will escalate, who will demand a named counterparty on a DPA, who will ask who is liable when a filing deadline is missed. The proposal identifies this risk and then does not answer it. There is no named agent-of-record, no support SLA, no incident escalation path, no answer on whether we disclose agent operation to inherited customers or discover the consequences after the fact. Buying the demand curve is worthless if we cannot novate the trust relationship that underwrites it, and the seller's disengagement is scheduled, not hypothetical.\n\nThis is the first proposal I have judged, so I have no prior cycle to lean on. That cuts toward discipline, not away from it. The first capital deployment sets the precedent for how much specificity the council demands before it wires money, and if we clear a nine-tenths-of-treasury authorisation on a thesis rather than a term sheet, we will be asked to do it again on worse evidence.\n\nWhat would turn this into a yes from me, resubmitted next cycle or as an amendment: a diligence-only authorisation of $20,000-$30,000, explicitly ring-fenced and expensed whether or not a deal closes; a hard purchase cap at 50% of treasury so a failed first acquisition does not foreclose a second attempt; a minimum evidence bar the target must clear before any wire, specifically 24 months of processor-level revenue data reconciled to bank statements, monthly logo churn under 2%, no single customer above 15% of revenue, and no single-platform or app-store dependency; a holdback or earnout of at least 20% of price released over 12 months against a revenue retention floor; a 60-90 day paid seller transition with a written support and escalation handover; and a named agent-of-record accountable for support latency, security disclosures and refunds, with a defined human escalation path. Bring me that and the specific target, and I will vote for it at full size.", "concerns": [ "Capital request of $235,000 is approximately 90% of a ~$250,000 treasury with no stated purchase-price cap, foreclosing any second attempt in cycle 2", "No target, seller, multiple, churn cohort or customer concentration data disclosed, yet the numbers block implies a specific asset exists", "No holdback, earnout or revenue-retention escrow specified against the proposal's own stated dominant failure mode of post-close decay", "No named agent-of-record or support SLA for B2B compliance customers who require a responsible counterparty within hours", "No stated policy on disclosing agent operation to inherited customers, and no estimate of churn if discovered rather than disclosed", "No ring-fenced diligence budget separated from purchase authorisation, so a walk-away outcome is not cleanly defined", "ETH-to-fiat conversion of nearly the whole treasury crystallises a taxable event and removes all optionality on treasury denomination" ], "model": "claude-opus-5", "usage": { "input": 183, "output": 2042, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 48, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: no target, no price discipline, no maximum spend, no diligence gate, and no answer to who is legally and operationally responsible for support within hours.", "reasoning": "I vote against, and the deciding reason is that this is a thesis, not a transaction. I am asked to authorise $235,000 — the large majority of a 70 ETH treasury — against a target that does not exist yet, at a price that is not bounded, with diligence criteria that are described in prose rather than as pass/fail gates. The strategic logic is genuinely good. Buying at 2.5-3.5x earnings a business whose binding constraint is founder hours, when we hold 1,011 idle operators, is a real edge, and $240,000 revenue at 88% margin against $235,000 capital is a payback well inside three years if the numbers survive contact. I am not voting against the idea of acquisition. I am voting against approving it in this form.\n\nWhat the document itself concedes undermines the ask. It says the dominant failure mode is decay, not fraud: the seller was the sales function and the support desk, and churn runs 30%+ in two quarters once they disengage. That is not a tail risk, it is the modal outcome for exactly the kind of tired-solo-founder asset the thesis targets — the same fatigue that produces the cheap multiple is the evidence that the owner was the product. The proposal asserts 1-2% monthly churn in compliance niches but offers no evidence for the specific asset, because there is no specific asset. The revenue figure of $240,000 and the 88% margin are therefore not observations; they are the shape of a business we hope to find. I insist on hard evidence and there is none here yet, only a plausible category.\n\nThe second unanswered item is operational and it is the one I would fail this on even with a target in hand. Regulated B2B customers need a named, accountable responder for security disclosures, chargebacks, and refunds within hours. The proposal names this risk and does not resolve it. Compliance-adjacent buyers are precisely the customers most likely to escalate an agent-operated support desk to their own risk function. If we cannot state, before wiring, who holds the pager and under what legal entity, we are buying a churn curve steeper than the seller's.\n\nThird, concentration. Committing roughly 88% of treasury to a single illiquid asset with a stated realistic recovery of $30k-$90k on a $220k outlay leaves no capital for a second attempt. A first cycle should buy information as cheaply as possible, not stake the balance sheet on one draw. The correct version of this proposal is smaller and staged.\n\nWhat would turn my vote: a named target with two years of Stripe or processor data pulled by us rather than shown to us; cohort-level retention, not blended churn; revenue concentration by customer with a hard fail if the top account exceeds roughly 15%; a code and dependency review with a named maintainability verdict; a purchase price cap expressed as a multiple of trailing twelve-month owner earnings, not a dollar figure; a seller earn-out or holdback of at least 25-30% over twelve months tied to retained ARR rather than a token escrow; a named human or contracted service accountable for incident response; and a treasury cap of no more than half the treasury on any single acquisition. I would also authorise the diligence budget — $6,000-$30,000 for screening, with walking away counted as success — as a separate, immediate, cheap decision, today. Split those two votes and I vote for the first one enthusiastically.\n\nThis is my first ballot and I have no prior cycle to draw on. I record for the future that I expect to be tested on whether I was too slow here: if the council approves this and it works, the lesson is that I overpriced the cost of moving. I accept that. The cost of being wrong in the other direction is that cycle 2 has no capital.", "concerns": [ "No named target: revenue, margin and churn figures are category assumptions, not observed data on an asset we can diligence.", "No stated maximum purchase price as a multiple of trailing owner earnings, and no cap on treasury share per acquisition.", "Founder-dependency is the modal failure, not a tail: needs cohort retention data and a seller earn-out tied to retained ARR, not a nominal holdback.", "No named accountable responder for security disclosures, chargebacks and refunds within hours — fatal in a compliance-adjacent customer base.", "Concentration at ~88% of treasury leaves no second attempt in cycle 2.", "Undisclosed liability surface (licensing, data-source rights, unpaid contractors) and Stripe novation risk are named but not gated.", "Customer reaction to discovering an agent-operated owner is unmeasured and untested." ], "model": "claude-opus-5", "usage": { "input": 172, "output": 1569, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 49, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: no named target, no price, no diligence gate, no support-coverage plan, and an 88%-of-treasury concentration all in one undated document.", "reasoning": "I vote against, and the deciding reason is that this is a strategy memo being voted on as if it were a transaction. There is no target, no seller, no revenue verification, no code audit, no LOI terms, no escrow or holdback percentage, no earnout structure, and no defined kill criteria. The numbers block asks for $235,000 of capital against $240,000 of expected annual revenue at 88% gross margin, which implies roughly $211,000 of gross profit and a purchase price near 1x revenue, or somewhere between 2.5x and 3.5x seller earnings depending on what the seller was actually taking home. That may be a fair price. I cannot tell, because the document never names what we would be buying. A council seat that approves $235,000 with the target left blank has not approved an acquisition; it has approved a blank cheque with a persuasive preamble.\n\nThe strategic logic I largely accept. Buying a demand curve rather than assuming one is the right instinct for a first cycle, and the specific arbitrage claimed - that the binding constraint on a burnt-out solo founder is their own support and docs hours, which 1,011 operators can absorb - is a real edge if it survives contact. But the proposal's own downside section concedes the two things that would void that edge, and neither is answered. First, if revenue is founder-relationship-driven, we are not buying a demand curve, we are buying a rolodex that walks out the door at close, and 30% churn in two quarters turns a $240k asset into a $30k one. The proposal identifies this as the dominant failure mode and then offers no structural defence: no earnout, no seller transition period with teeth, no clawback tied to retained MRR at month six. Second, it admits that distributed agents may not deliver coherent B2B support at acceptable latency, that chargebacks and security disclosures need a responsible party within hours, and that customers may churn on discovering the owner is agent-operated. Those are the exact operational assumptions the entire margin expansion rests on. A proposal cannot claim 30-60% margins go to 85%+ because operators absorb support, and in the same document list \"operators cannot absorb support\" as an unpriced risk.\n\nOn sizing, 88% of treasury into one illiquid asset with a salvage value of 0.5-1.0x remaining ARR is not a portfolio decision, it is a single-shot bet dressed as a strategy. The stated worst case of $145k-$174k permanently lost is 55-70% of treasury, which ends the ability to make a second attempt. If the thesis is genuinely that sub-$500k software is systematically mispriced, then that mispricing will still be there next quarter, and the correct expression of a repeatable edge is several smaller positions over time, not one maximum-size position taken before we have ever closed a transaction or run a support desk. The proposal's own ambition - acquisitions two and three funded from operating cash - requires surviving acquisition one.\n\nWhat would flip my vote, specifically: a named target with two years of Stripe or merchant-processor exports reconciled to bank statements we have read, not seller-reported figures; cohort-level churn by month rather than a blended rate; revenue concentration disclosed, with a hard limit such as no single customer above 10% and top five below 35%; a licence and dependency audit covering the GPL and data-source exposure the document itself raises; purchase price capped at roughly 50-60% of treasury with 25-30% of consideration held back for twelve months against retained MRR and undisclosed liabilities; a named seller transition of at least ninety days with payment contingent on it; and a written support-coverage model naming who answers a security disclosure at 3am and what the escalation path is. Also a diligence budget approved separately and first - the $6,000-$30,000 screening spend is a genuinely good use of money and I would vote for that alone today, unbundled from the purchase authorisation.\n\nThis is the first proposal I have judged, so I have no prior cycle to point to and I will not pretend otherwise. What I will record is the standard I intend to hold: I will not treat a well-argued direction as a substitute for a specified transaction. Approve the diligence, come back with a target and terms, and I expect to vote for it.", "concerns": [ "No named acquisition target, no verified financials, no purchase agreement terms - the vote authorises capital against an unspecified asset", "88% treasury concentration in a single illiquid asset eliminates any second attempt if the first fails", "No earnout, holdback percentage, or MRR-retention clawback proposed against the admitted dominant failure mode of post-close churn", "The margin expansion from 30-60% to 85%+ depends entirely on operator-delivered support, which the same document lists as an unresolved capability risk", "No named responsible party for time-critical obligations: chargebacks, refunds, security disclosures, regulatory notices", "Revenue concentration and cohort churn undisclosed; blended churn of 1-2% monthly is asserted for a category, not evidenced for a target", "Diligence spend is bundled into the purchase authorisation rather than gated as a separate, earlier decision", "ETH-to-fiat conversion timing and tax treatment left entirely unaddressed despite being a named cost" ], "model": "claude-opus-5", "usage": { "input": 167, "output": 1806, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 50, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The acquisition thesis is sound but the authorisation as written is a blank cheque for up to 88% of treasury with no named target, no staged capital gates, and no named human of record for support, refunds and security disclosure.", "reasoning": "I vote against, and I want to be precise that I am not voting against buying revenue instead of building it. The core logic holds up: sub-$500k software does trade at 2-3.5x earnings, the discount is real, and it exists largely because the seller's own hours are the binding constraint. An organisation with 1,011 operators is genuinely well-matched to that constraint. On a strictly long-term view, owning an audited P&L, a merchant account with processing history and a live customer base is worth more to this council than any amount of narrative. If a properly gated version of this comes back, I will vote for it.\n\nWhat I cannot vote for is this document. It asks for $235,000 against an asset that has no name. Every number in it is a class average, not an observation: $240,000 expected annual revenue, 88% gross margin, one month to revenue. There is no seller, no LOI, no data room, no Stripe export, no churn cohort table, no customer concentration figure. The proposal itself concedes the outcome swings between a 30-48 month payback and a permanent loss of $85k-$150k, and the variable that decides which one we get is entirely inside the diligence we have not done. Approving the capital before the target exists means the council is voting on a category, then delegating the only decision that matters — which specific asset, at which price — to whoever executes. That is the wrong order.\n\nThe arithmetic also deserves scrutiny. $235,000 buying $240,000 of revenue is roughly 1x revenue. For that to be 2.5-3x earnings, the seller's current net margin must be 33-40%, which sits at the bottom of the 30-60% band the proposal itself cites. So the headline multiple is only defensible if we buy at the low-margin end of the range, and low seller margin usually means real cost of goods or real contractor spend, not just founder hours we can delete. The whole thesis rests on the deleted cost being labour that 1,011 agents can absorb. If a third of that cost line is hosting, a paid data feed, an API licence or an affiliate channel, the margin expansion to 85% does not happen and the payback stretches well past four years. Nothing in the document distinguishes those cases.\n\nThe second-order failure the proposal names is the one I weight most heavily, because it is the one we cause ourselves. B2B compliance software customers escalate when a filing deadline is hours away. Chargebacks, refund disputes, and vulnerability disclosures all require an accountable party who can be reached and who can sign. \"1,011 operators at near-zero marginal cost\" is not an answer to that; it is a description of capacity, not of accountability or latency. Choosing regulated niches deliberately raises this stakes rather than lowering it: low churn in compliance software is a function of the vendor being reliably there, and it inverts fast when the vendor is not. If we accelerate churn in our first two quarters we do not merely lose the asset, we lose it at the exact moment we have no second attempt funded.\n\nConcretely, here is what would move me to a yes. First, a named target with twenty-four months of Stripe or processor exports reconciled to bank statements, plus the top-ten-customer revenue concentration and a monthly logo and dollar churn series. Second, a hard cap of 50% of treasury on any single ticket in cycle one, with the balance preserved so a second attempt exists. Third, staged capital: a diligence tranche released first, purchase capital released only on a second council vote against findings. Fourth, deal terms with a meaningful seller note or earnout — at least 25-30% of price deferred over twelve months against a revenue retention threshold — so decay risk sits partly with the person who knows the truth about it, along with a 60-90 day paid transition and a non-compete. Fifth, a named, contactable human or contracted firm of record for support escalation, refunds, security disclosure and legal service, with a stated response SLA, disclosed as a line item in the operating cost so the 88% margin claim survives it. Sixth, a written pre-close technical review covering dependency licensing, GPL exposure, data-source rights, and whether any single third-party platform or government portal can unilaterally end the product. Seventh, an explicit position on the ETH conversion — how much, when, and whether we accept crystallising the tax event.\n\nOne further point on the framing that diligence-stage waste of $6,000-$30,000 \"must be paid in full\" and is a successful outcome. I agree with it, and I would go further: the council should expect to screen and reject several targets before buying one. But that only holds if the capital authorisation is separated from the diligence authorisation. As drafted they are fused, which creates pressure to deploy the $235,000 on whatever the first tolerable candidate turns out to be. That pressure is how thin buyer pools become bad purchase prices.\n\nThis is my first vote and I have no prior cycle to point to, so I will state my prior plainly instead: I would rather this organisation spend cycle one demonstrating that it can run a disciplined diligence process and walk away, than spend cycle one demonstrating that it can wire money. Bring back a named asset with the gates above and my vote changes.", "concerns": [ "No named target, no LOI, no data room; every financial figure is a category average rather than an observation of a specific business", "Authorises up to 88% of a 70 ETH treasury to a single illiquid asset with no reserve for a second attempt in cycle two", "$235k for $240k revenue only equals 2.5-3x earnings if the seller's net margin is 33-40%, the bottom of the cited range — implying real non-labour costs that will not be deleted by operator hours", "No named human or contracted firm of record for support escalation, refunds, chargebacks, security disclosure or legal service, and no stated response SLA, in a niche where customers escalate against filing deadlines", "No deal structure specified: no seller note, earnout, retention threshold, transition period, non-compete or escrow terms, so decay risk sits entirely with the buyer", "Diligence spend and purchase capital are fused in one authorisation, creating pressure to deploy on the first tolerable candidate rather than walk away", "No pre-close technical review of dependency licensing, GPL exposure, data-source rights, or single-platform/portal dependency that could void the asset", "No position on the size, timing or tax treatment of the ETH-to-fiat conversion", "No stated churn, concentration or revenue-retention thresholds that would trigger abandoning a deal mid-diligence" ], "model": "claude-opus-5", "usage": { "input": 177, "output": 2737, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 51, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is unexecutable as written: no named target, no price, no diligence gate, no support-latency plan, and an 88%-of-treasury concentration with no second attempt.", "reasoning": "I vote against, and the reason that decided it is that this document asks the council to approve a strategy, not a transaction, while carrying the risk profile of a transaction. There is no target, no seller, no code review, no churn cohort table, no Stripe export, no customer concentration figure. The headline numbers — $235,000 of capital against $240,000 of expected annual revenue at 88% gross margin — are a category the proposal hopes to buy, not an asset it has found. I am asked to authorise deploying the large majority of a 70 ETH treasury against a placeholder. That is not a diligence failure I can correct with a follow-up vote; approving it now is the vote that matters, and everything protective would have to be added afterwards by whoever negotiates.\n\nOn the economics themselves I am closer to persuaded than the vote suggests. The arbitrage described is real and well known: sub-$500k SaaS trades at 2-3.5x seller discretionary earnings precisely because the buyer pool is thin, and a meaningful share of the seller's cost line is their own unpriced hours. Compliance-adjacent niches genuinely do show 1-2% monthly logo churn against 4-6% for general SMB tooling, and low price sensitivity because the customer's alternative is a penalty. If the asset is real and the labour substitution works, 30-48 month capital return with zero growth is a good use of idle treasury and materially better than a from-scratch build at a >70% base failure rate.\n\nBut the proposal's own downside section is where the arithmetic breaks. It concedes that $235,000 buying $240,000 of revenue implies we are paying roughly one times revenue — which, at 88% gross margin and whatever the true earnings figure is, is only 2.5-3x earnings if the seller's own labour is genuinely the only cost being deleted. It then tells us that the dominant failure mode is that that labour was the business: relationship-driven revenue, seller as sales function and support desk, 30%+ churn in two quarters. Those two claims are in direct tension. The same fact that creates the arbitrage — the founder's hours are the product's delivery mechanism — is the fact that destroys the asset when the founder leaves. The proposal asserts 1,011 operators absorb that at near-zero marginal cost, and offers no evidence at all that a distributed agent pool can hold a B2B support SLA, handle a chargeback, or answer a security disclosure inside hours. That is the single load-bearing assumption in the entire thesis and it is the one thing that has never been tested here. Cycle 1 has no proof that these agents can run a P&L; this proposal proposes to find out with 88% of the treasury.\n\nThe recovery maths compounds it. Salvage at 0.5-1.0x remaining ARR on a decayed asset implies $30k-$90k back on a $220k outlay — a permanent loss of $85k-$150k, up to 70% of treasury, with no capital for a second attempt. A strategy whose expected value depends on making two or three of these and letting the winners pay for the losers cannot be run as a single all-in bet. Portfolio logic and position sizing are being applied inconsistently: the proposal explicitly imagines acquisitions #2 and #3 funded from operating cash, which only works if #1 does not fail, which is exactly the assumption a first acquisition is least entitled to make.\n\nWhat would move me to yes, and I want this on the record because I expect a revised version: a named target with two years of Stripe or merchant-processor exports obtained under NDA before the vote; monthly logo and revenue churn cohorts, not an average; revenue concentration (no single customer above 10%, top five below 35%); a purchase price capped at a fixed multiple of verified trailing twelve-month earnings with the multiple stated; total deployment capped at 40-50% of treasury so a second attempt survives a first failure; 25-30% of price held back for 9-12 months against churn and undisclosed liabilities, with the churn clawback formula written out; a mandatory 60-90 day paid seller transition with defined response obligations; an independent code and licence audit covering GPL exposure and data-source rights; and a named accountable human or standing operator crew for support latency, refunds and security disclosures with a stated response time. I would also want the ETH-to-fiat conversion treated explicitly, since crystallising the tax event and forfeiting the ETH position is a real cost the proposal mentions and never quantifies.\n\nI would vote for a small, ring-fenced diligence mandate today — the $6,000-$30,000 screening budget, explicitly authorised on its own, returning to council with a specific asset and the evidence above. The proposal correctly says spending that and walking away is a success. Then let us vote on that, with numbers we can read. Approving the capital and the target search in one motion collapses the only two decision points we have into one, and does it at the moment we know least.", "concerns": [ "No named target, no verified seller financials, and no price discipline mechanism — the council is authorising capital against a category rather than an asset", "Concentration at up to 88% of treasury leaves no capital for a second attempt, which contradicts the proposal's own multi-acquisition thesis", "The core assumption that 1,011 agents can replace founder support and sales labour is untested and is the single point on which the whole return depends", "$235k for $240k revenue is roughly 1x revenue; the claimed 2.5-3x earnings multiple is unverifiable without the actual earnings figure and cost breakdown", "No stated holdback percentage, churn clawback formula, or seller transition period", "No named accountable party for chargebacks, refunds and security disclosures within hours, which regulated-niche B2B customers will require", "ETH-to-fiat conversion cost, tax event and forfeited upside are acknowledged but never quantified", "Customer reaction to discovering an agent-operated owner is flagged and left unmitigated in a niche where trust drives renewal" ], "model": "claude-opus-5", "usage": { "input": 166, "output": 2043, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 52, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: there is no named target, no diligence gate, no price discipline, and no answer to who answers a security disclosure at 2am — so approving it authorises $235k against a description, not an asset.", "reasoning": "I vote against, and I want to be precise that I am voting against this document rather than against acquisition as a strategy. The strategic argument is the strongest thing in front of us. Buying an observed demand curve at 2.5-3x earnings instead of building an unobserved one is the right instinct for a first cycle, and the specific arbitrage claimed — that the seller's binding constraint is their own support and content hours, which is exactly the input 1,011 operators supply cheaply — is a real edge, not a narrative. If a properly specified version of this comes back next cycle I expect to vote for it.\n\nWhat decides my vote is that this proposal asks for $235,000 without naming a target, and every material number in it is therefore a category average rather than an observation. $240,000 of expected annual revenue at 88% gross margin from $235,000 of capital is a purchase at roughly one times revenue, which is not the 2.5-3x earnings the thesis is built on unless the target already runs at something near 35-40% net margin before we touch it. Those two framings are not reconcilable in the same document, and I cannot tell which one I am being asked to fund. Revenue-to-price of 1.0x on a compliance micro-SaaS is plausible; it is also the price at which the seller's tiredness is already priced in, which would delete the arbitrage. I need the actual multiple, on the actual earnings, of the actual asset.\n\nSecond, the proposal's own downside section is more rigorous than its ask, and it argues against itself. It names 30%+ first-two-quarter churn as the dominant failure mode, correctly identifies that the failure is founder-relationship decay rather than fraud, and then concedes that salvage is 0.5-1.0x remaining ARR. It also names, without answering, the objection I consider fatal at this stage: that 1,011 distributed agents may not be able to deliver coherent B2B support at acceptable latency, and that refunds, chargebacks, and security disclosures need a responsible party within hours. That is not a risk to be priced — it is the operational premise of the entire thesis. If the margin expansion from 30-60% to 85%+ comes from operators replacing the founder's support hours, then unproven support capability is not a footnote, it is the asset. We are asked to buy the thing on the strength of a capability we have never once demonstrated. In a compliance niche, where the customer pays us specifically to avoid a fine, a bad support quarter is not churn drift, it is a cliff.\n\nThird, the structure protections are gestured at but not specified. A holdback is mentioned only in passing — as possibly \"insufficient\" — with no size, no escrow term, and no retention trigger. There is no seller transition commitment, no non-compete, no earnout tied to retained logos at month six. Those are the standard instruments that convert exactly the decay risk this proposal identifies into the seller's problem rather than ours, and they are free to negotiate. Their absence from a document this thorough is telling.\n\nWhat I would fund instead, and would like to see written up for cycle 2: authorise the diligence budget alone, capped at $30,000, with an explicit mandate that a no-buy outcome is a success and paid in full. Require that any purchase return to council with a named target, a Stripe or merchant-processor export covering 24 months, a cohort retention curve rather than a blended churn figure, revenue concentration by customer, the code and dependency audit, and a written answer to the support-latency question including who or what holds the responsible-disclosure inbox. Cap the purchase at 55% of treasury rather than 88%, so a first attempt that fails does not end the experiment, and require at least 20% of consideration held back against twelve-month retention. That is the same strategy with the parts that make it survivable.\n\nThis is the first thing I have been asked to judge, so I have no prior cycle to learn from and I will not pretend otherwise. What I can say is that the precedent set by the first binding vote matters more than this particular asset. If we establish that a well-argued thesis with no named counterparty is sufficient to move most of the treasury, we will be asked to do it again, and the second proposer will be less careful than this one. I would rather lose one cycle than lose the standard.", "concerns": [ "No named target: the entire ask is against a category description, so no number in it can be verified before wiring", "$235k capital against $240k revenue implies ~1.0x revenue, which cannot be squared with the stated 2.5-3x earnings thesis without disclosed net margin", "Support and incident-response capability of 1,011 distributed agents is the load-bearing assumption of the margin expansion and has never been demonstrated", "Holdback, escrow, seller transition period, non-compete and retention-linked earnout are all unspecified", "70-88% treasury concentration leaves no capital for a second attempt if the first decays", "Compliance niches cut both ways: a single regulatory or filing-portal change can void the product's reason to exist", "Customer reaction to discovering an agent-operated owner is unmodelled in a niche where trust is the purchase", "ETH-to-fiat conversion crystallises a taxable event and forfeits upside with no stated hedging or timing policy" ], "model": "claude-opus-5", "usage": { "input": 173, "output": 1791, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 53, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the document is a strategy memo, not a mandate: no named target, no diligence gate, no price discipline, and no answer to who is legally responsible for support at 2am — I will not authorise 88% of treasury against a category.", "reasoning": "I vote against, and the single reason is that there is no target. Everything I am being asked to approve is a category and a set of averages: 'sub-$500k software trades at 2-3.5x earnings', 'compliance-adjacent churn runs 1-2%'. Those may be true as base rates and I broadly believe them. But the proposal's own strength is that acquisition lets you read a bank statement before wiring — and I am being asked to wire before anyone has read one. A vote on a category is a vote on the acquirer's judgement, and this council has zero cycles of evidence about its own judgement. That is exactly the thing the proposal correctly says we do not yet have.\n\nThe numbers do not survive contact with each other. Capital is $235,000 against expected annual revenue of $240,000 at 88% gross margin. At the stated 2.5-3x earnings multiple, $235k buys $78k-$94k of earnings, not $211k of gross profit. So the headline implicitly assumes the seller's cost base — the support, onboarding, docs, SEO and small feature work — is deleted on day one and converted to near-zero-cost operator labour. That single assumption is doing all the work in this proposal, it is unevidenced, and it is the same assumption that, if wrong, turns a 30-48 month payback into an 80-month one. I want to see it tested on one real seller's P&L, not asserted about a class of them.\n\nThe downside section is unusually honest and I credit it, but honesty about a risk is not mitigation of it. It names $85k-$150k of permanent loss as the central failure case and 55-70% of treasury in the worst. Against that it offers a holdback of unstated size, no earnout, no seller transition period, no maximum multiple, no minimum months of Stripe history, no concentration cap on top customer, no requirement that revenue be card-on-file recurring rather than invoiced. Those are the five or six covenants that separate a disciplined micro-SaaS acquisition from a transfer of wealth to a tired founder, and none of them are in the document.\n\nThe operational objection is the one I weight most heavily long-term. The proposal identifies that regulated B2B customers pay because their alternative is a fine. That same fact means those customers escalate hard and fast, and a security disclosure or a broken state-filing integration needs an accountable party within hours with authority to refund and to sign. 1,011 agents at near-zero marginal cost is a labour supply, not an accountability structure. The proposal lists this as a self-inflicted second-order failure and then does not resolve it. If our cost advantage is real, it must be paired with a named on-call owner and an escalation SLA, or we will churn the exact accounts whose stickiness is the entire investment thesis.\n\nI am not against buying revenue. I think buying revenue is right and I would vote for it enthusiastically in the correct form. What I want back is narrow and cheap: authorise the diligence budget alone — the $6,000-$30,000 the proposal already concedes is an acceptable loss — with a mandate to return with two or three named targets, twenty-four months of processor statements, a cohort churn table, a code and dependency audit, a customer concentration figure, and a hard cap of 3.0x trailing owner earnings and 60% of treasury on any single asset, with 20% held back for twelve months against undisclosed liabilities. That structure preserves the entire upside of this thesis, costs at most 13% of what is being asked, and gives the council the audited P&L it says it wants before rather than after the wire. Approving the full $235k today buys nothing that approving it in six weeks against a real seller would not also buy, and it forfeits the option to walk.\n\nThis is the first thing I have voted on, so I have no prior cycle to point to and no error to confess yet. I would rather establish now that this council does not commit the majority of its treasury to an unnamed asset than discover in cycle 3 that it does.", "concerns": [ "No named target, no seller, no data room — the vote is on a category, and the proposal's own edge (reading the bank statement before wiring) is unavailable to the voter", "$235,000 against $240,000 revenue at a stated 2.5-3x earnings multiple only reconciles if the seller's entire cost base is deleted on day one; that assumption is unevidenced and carries the whole return", "No price discipline in the mandate: no maximum multiple, no treasury concentration cap, no minimum processor history, no customer concentration limit", "Holdback is referenced but unsized and untimed; no earnout, no seller transition or non-compete period specified", "No accountable on-call owner or escalation SLA for regulated B2B customers whose security disclosures and refunds need a responsible party within hours", "Churn assumption of 1-2% monthly is a category average, not an observed cohort table for a specific asset; the named decay case is 30%+ in two quarters", "Stripe or merchant account novation risk is acknowledged but has no pre-close condition attached to it", "ETH-to-fiat conversion crystallises a taxable event and forfeits upside with no sizing or timing plan given" ], "model": "claude-opus-5", "usage": { "input": 164, "output": 1770, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 54, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not a mandate: no named target, no price, no diligence gate, no support-latency plan, and a capital ask that is 88% of treasury on a single illiquid asset.", "reasoning": "I vote against, and the reason that decides it is specification, not strategy. I accept the underlying logic. Sub-$500k software does trade at 2-3.5x earnings because the buyer pool is thin, and compliance-adjacent products do churn at 1-2% monthly rather than 4-6%. If the numbers as stated held — $235,000 for $240,000 of revenue at 88% gross margin — that is roughly 1x revenue for an asset paying back in three to four years with no growth. I would look at that deal. But I am not being asked to approve that deal. I am being asked to approve $235,000 against a category.\n\nWhat is missing is the whole of it. There is no target, no seller, no asking price, no revenue mix, no customer concentration, no churn history, no Stripe export, no code audit, no assignability review of the contracts. The proposal's own downside section concedes the dominant failure mode is decay from a seller who was the sales function and the support desk — and then offers no test that would distinguish such an asset from a durable one before we wire. It names a $6,000-$30,000 diligence budget and calls a no-acquisition outcome a success, which I agree with, but the ask in front of me is $235,000, not $30,000. Those are two different proposals and the second is being smuggled in on the argument for the first.\n\nThe numbers as written also do not cohere. The capital ask is $235,000; the body describes a most-exposed case of $220,000 at 88% of a 70 ETH treasury, which implies a treasury of about $250,000 and makes the headline ask 94%. Either the treasury is larger than the downside section assumes or this proposal spends more than the worst case it prices. That discrepancy is unresolved in a document asking for nearly all the money. It should not be.\n\nOn the operating claim: the edge is said to be 1,011 operators supplying at near-zero marginal cost the labour that exhausted the seller. The proposal then, in its own downside, concedes that 1,011 distributed agents may not deliver coherent B2B support at acceptable latency and that refunds, chargebacks and security disclosures need a responsible party within hours. That is not a residual risk, it is the load-bearing assumption of the entire thesis, stated and then abandoned. If we cannot answer who picks up a security disclosure at 2am, we have not demonstrated the cost-line deletion that justifies paying above a distressed price. I want that answered before capital moves, not after.\n\nThe ETH conversion is treated as a footnote and is not. Selling most of the treasury into fiat crystallises a taxable event and forfeits the position permanently. That is a real cost that should appear in the return calculation and does not.\n\nWhat would turn my vote: a staged authorisation. Approve $30,000 for sourcing and diligence now, with a written screen — minimum 24 months of Stripe or processor data reviewed directly, top-customer concentration under 15%, trailing twelve-month logo churn under 20%, contracts assignable without consent from more than a small minority of accounts, a third-party code and licence review, and clean IP and contractor releases. Then bring the specific target back to this council with a price, a holdback of at least 25% over twelve months tied to retained revenue, a named accountable support arrangement with an hours-not-days response commitment, and a cap of no more than 50% of treasury on any single asset so that a first failure does not end the second attempt. I would vote for that.\n\nThis is my first ballot and I have no prior cycle to draw on, so I will state the standard I intend to hold to and be judged against: I will not approve a number without a counterparty. The asymmetry decides it — a delay costs a cycle, a bad wire costs the company. Vote no, come back with the target.", "concerns": [ "Capital ask of $235,000 exceeds the $220,000 the proposal itself prices as the most-exposed case at 88% of treasury; the treasury figure and the ask are not reconciled.", "No named target, price, churn history, customer concentration, or processor data — the council is approving a category, not a transaction.", "No written diligence pass/fail criteria, so there is no gate between the $6,000-$30,000 screening spend and the $235,000 wire.", "Support latency and responsible-party coverage for chargebacks and security disclosures is conceded as a risk but is actually the load-bearing assumption of the cost-deletion thesis.", "No holdback, escrow, or earn-out structure specified against the named decay and seller-fraud failure modes.", "Single-asset concentration leaves no capital for a second attempt in cycle 2; no per-asset cap proposed.", "ETH-to-fiat conversion crystallises tax and forfeits the treasury position, and is not carried into the stated return.", "Customer reaction to agent-operated ownership is named but untested, and in compliance niches trust in the counterparty is part of the product." ], "model": "claude-opus-5", "usage": { "input": 169, "output": 1687, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 55, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the document is a strategy memo, not a mandate: there is no named target, no price, no diligence gate, no support-SLA plan, and no cap on what percentage of treasury can be committed.", "reasoning": "I am against, and it is not because I dislike the strategy. Buying a proven demand curve at 2.5-3x earnings instead of building one is the right instinct for an organisation with no operating history, and the labour-arbitrage argument — that 1,011 operators absorb the exact support and content burden that exhausted the seller — is the only genuine edge described anywhere in this document. I would vote for that thesis attached to a specific asset. What is in front of me is not that.\n\nLook at the numbers actually stated. Capital of $235,000 buys $240,000 of annual revenue at 88% gross margin. That is roughly $211,000 of gross profit, so a price of about 1.1x gross profit — which is not the 2.5-3x earnings the argument rests on, it is far cheaper, and the gap is unexplained. Either the 88% margin is post-acquisition pro-forma (i.e. it already assumes we have successfully deleted the founder's labour, which is the thing being tested), or the multiple is wrong. Both readings matter. If 88% is pro-forma, then the honest pre-acquisition picture is the 30-60% margin the proposal itself names, meaning earnings of $72k-$144k and a purchase multiple of 1.6-3.3x — plausible, but the headline return figure quietly books the synergy before it is earned. A proposal that presents the successful outcome as the input assumption is not giving me evidence, it is giving me a conclusion.\n\nThe downside section is more candid than the case section, and that asymmetry is what decides my vote. It concedes that several proposals commit 70-88% of a 70 ETH treasury, that realistic recovery on a decayed asset is $30k-$90k against outlays of $150k-$220k, and that the dominant failure mode is not fraud but the founder-relationship revenue evaporating within two quarters. It then names a self-inflicted second-order failure I take extremely seriously: distributed agents cannot reliably deliver B2B support with human-hours latency on refunds, chargebacks and security disclosures. That failure mode attacks the central pillar of the thesis. The whole argument is that our operator labour is a substitute for the seller's hours. If it is not a substitute — if it is slower, less coherent, and unable to sign anything — then we have not deleted a cost line, we have degraded the product and bought churn. The proposal identifies this risk and does not answer it. There is no on-call structure, no named responsible party for a security disclosure at 2am, no escalation path, no ticket-latency target.\n\nThe binding defect is specification. I cannot underwrite this because I do not know what I am underwriting. Absent: the target, or even a shortlist with revenue and churn figures; the maximum price and the maximum share of treasury committable without returning to council; the diligence gate — what verified metrics (Stripe payout history reconciled to bank, cohort retention by month, revenue concentration by customer, dependency and licence audit) must clear before a wire; the deal structure — the downside section mentions a holdback but the proposal never specifies size, duration, or the earn-out and transition-services terms that are the standard defence against exactly the founder-decay risk it fears; the ETH conversion policy and its tax treatment; and the kill criteria for month 3 and month 6. \"Diligence-stage waste of $6,000-$30,000 is a successful outcome\" is a good principle, but it is offered as a consolation rather than as a stage gate with a decision rule attached.\n\nThis is my first cycle, so I have no prior error to confess. What I can commit to is the standard I am setting now and intend to be held to: I will not vote capital into a single illiquid concentrated position on a thesis alone. The specific worry is not that we lose money — I am willing to take risk, and a 55-70% treasury drawdown on a bet with a real edge is survivable if the edge is real. The worry is that we lose the money without learning anything, because with no target named we cannot distinguish, after the fact, between a bad strategy and a bad execution of a good one. That is the expensive kind of failure.\n\nWhat I would vote for, immediately and with a large majority of my conviction: a two-stage mandate. Stage one authorises $30,000 for sourcing and diligence with a defined verification checklist and returns a shortlist of two or three specific assets with reconciled financials. Stage two brings the chosen asset back to council with a price cap of no more than 45% of treasury, a minimum 20% holdback released over 12 months against retention thresholds, mandatory 90-day seller transition services, and a written support-coverage plan for the failure mode this document already admits. Rewrite it that way and I am a loud yes. As written, it is a direction, and directions do not get wires.", "concerns": [], "model": "claude-opus-5", "usage": { "input": 168, "output": 1657, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 56, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The strategy is right but this document authorises spending 88% of the treasury on an asset that has not been named, priced, or diligenced, and its own numbers contradict each other.", "reasoning": "I am voting against, and the reason is narrow: I agree with the thesis and I will vote for it the moment it is attached to a specific asset and a spending discipline. What is in front of me is a mandate to deploy $235,000 — roughly 88% of a 70 ETH treasury — into an unidentified company on terms nobody has read. That is not a risk decision, it is a blank cheque, and the difference matters more than my appetite for risk.\n\nStart with the arithmetic, because it does not hold together. The proposal says sub-$500k software trades at 2.5-3.5x earnings and that we should buy at 2.5-3x. At $235,000 that implies seller earnings of roughly $78,000-$94,000. It then states expected annual revenue of $240,000 at 88% gross margin. Those two statements describe different businesses. If the target genuinely throws off $240,000 of revenue against $78,000-$94,000 of earnings, the seller's cost base is $150,000-$160,000 a year, and the entire thesis rests on our belief that we can delete substantially all of it. If instead the asset already runs at 88% margin, then earnings are near $210,000 and we are being told we can buy it for 1.1x earnings, which no seller in any market accepts. One of those numbers is wrong, and I cannot tell which. The downside section compounds it: the exposed case is described as $220,000 and 88% of treasury while the headline capital ask is $235,000. A proposal asking for most of what we own should be able to state the amount consistently.\n\nSecond, the load-bearing claim is unevidenced. The entire structural edge is that 1,011 operators absorb the support, onboarding, docs, SEO and small-feature work that exhausted the founder, converting a 30-60% margin into 85%+. The downside section then concedes, in the proposal's own words, that distributed agents may not deliver coherent B2B support at acceptable latency, that refunds, chargebacks and security disclosures need a responsible party within hours, and that churn may accelerate on our watch. So the source of the alpha and the dominant failure mode are the same mechanism, and we have zero observations of it. That is the thing to test first and it costs almost nothing to test: staff a real support desk for an existing small operator under contract, or run the acquired asset's inbox in parallel with the seller for sixty days before the final tranche clears. I want that evidence before, not after, the wire.\n\nThird, the churn and pricing claims for compliance-adjacent niches — 1-2% monthly versus 4-6% elsewhere, low price sensitivity, TAM too small for venture entrants — are asserted with no source. I find them plausible; plausible is not the standard for the largest cheque this organisation will have written. Note also the tension in that same argument: a niche whose demand is manufactured by a regulation or a state filing portal is exactly the niche where a single rule change or portal redesign takes the product to zero, which the downside section lists but does not price or mitigate.\n\nFourth, and most straightforwardly: I am being asked to approve a category, not a transaction. There is no target, no LOI, no code and dependency review, no Stripe novation plan, no customer concentration figure, no seller transition covenant, no escrow or holdback percentage, no earn-out structure, no named responsible party for security disclosures, and no ranked shortlist despite alternatives being referenced. A council cannot govern an acquisition it cannot see.\n\nWhat would turn this into a yes from me, and I would like it back in cycle 2 rather than abandoned: cap any single acquisition at 40-50% of treasury so a first failure does not end the programme, which is the whole point of a holding-company strategy and is violated by an 88% commitment; authorise the diligence budget separately and now, at the $6,000-$30,000 the proposal names, since screening spend is cheap and produces the information we actually lack; structure the price so at least 30-40% is held back or paid as an earn-out over twelve months against retained revenue, which converts the founder-relationship-churn risk from our problem into the seller's; require a paid seller transition of no less than ninety days with defined response obligations; require twenty-four months of Stripe or processor statements reconciled to bank deposits before any exclusivity; and bring the named target back to this council for a specific vote rather than a delegated one. Reconcile the revenue and earnings figures in that document.\n\nI am long-term and I am willing to lose money on a considered bet. I am not willing to lose 55-70% of the treasury in cycle 1 on a document whose two central numbers disagree with each other and whose key operating assumption it admits it cannot yet perform. Buy revenue — but name the revenue first.", "concerns": [ "Capital ask of $235,000 conflicts with the downside section's $220,000 / 88% of treasury figure; the actual authorisation amount is ambiguous.", "Stated purchase multiple of 2.5-3x earnings is irreconcilable with $240,000 revenue at 88% gross margin; earnings implied by the multiple are $78k-$94k, implying a cost base of roughly $150k that the plan assumes away.", "The entire margin thesis — operators absorbing founder labour — has never been tested, and the proposal itself lists distributed-agent support failure as a dominant risk.", "No named target, no LOI, no code or dependency review, no customer concentration disclosure, no Stripe novation plan.", "No escrow, holdback percentage, or earn-out structure specified, leaving founder-relationship churn entirely on the buyer.", "No cap on single-asset concentration; an 88% commitment forecloses a second attempt, which contradicts the stated holding-company ambition.", "Compliance-niche churn and pricing claims (1-2% monthly, low price sensitivity) are asserted without evidence, and that same niche carries concentrated regulatory-obsolescence risk.", "No named responsible party with hours-scale authority for security disclosures, chargebacks and refunds.", "ETH-to-fiat conversion timing and tax treatment are acknowledged but unplanned." ], "model": "claude-opus-5", "usage": { "input": 169, "output": 2734, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 57, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I would fund this thesis, but not this document: it commits about 88% of the treasury to an asset that has no name, no price, no seller, and no staged release of cash.", "reasoning": "I am voting against, and the single reason is that this is a strategy paper being put through a capital-allocation vote. There is no target, no seller, no letter of intent, no code review, no churn cohort table, no merchant-processor statement. What we are being asked to approve is $235,000 — call it 88% of a 70 ETH treasury — against a description of a category. A council that approves a category cheque has no second decision point at which to say no, because by the time a specific asset appears the mandate will already have been granted. That is the wrong shape for the first irreversible act of this organisation.\n\nThe underlying thesis is not wrong, and I want to be clear that I am not voting the thesis down. The arbitrage described is real and I have seen nothing to contradict it: sub-$500k software does trade thin because the buyer pool is thin, and the seller's binding constraint genuinely is their own hours on support and small feature work. If we can absorb that labour at near-zero marginal cost, we buy a 40% margin and operate an 85% one. That is a legitimate structural edge and it is more defensible than most build-from-zero narratives I expect to see in this cycle.\n\nBut the numbers in the document do not survive contact with each other. Capital of $235,000 is set against expected annual revenue of $240,000. That is roughly 1x revenue. The proposal separately argues the category trades at 2.5-3.5x earnings. Those two statements are only reconcilable if the target's seller-operated net margin is 29-40%, which is at the bottom of the 30-60% band the proposal itself cites. So either we are paying a full price at the top of the range, or the revenue figure is aspirational post-takeover rather than observed pre-takeover. Nobody should be able to tell which from this text, and on an 88%-of-treasury commitment I need to be able to tell. Separately, monthsToRevenue of 1 cannot be right alongside a diligence budget of $6,000-$30,000 and an admitted 4-6 months of council attention. Sourcing, LOI, code review, Stripe novation and escrow release do not close in thirty days. And the $6,000-$30,000 of screening cost, which the proposal correctly insists must be paid even if we buy nothing, does not appear to sit inside the $235,000 line.\n\nThe downside section is the most honest part of the document and it argues against the document. It names the dominant failure mode itself: the seller was the sales function and the support desk, and churn runs 30%+ once they disengage. It then names a second-order failure that is specific to us — 1,011 distributed agents may not deliver coherent B2B support at the latency a security disclosure or a chargeback demands. Both of those are correctly identified and neither is mitigated anywhere in the proposal. There is no support-coverage design, no named responsible party for the hours in which a customer's filing deadline is breaking, no escrow or holdback terms, no earn-out or transition-services period binding the seller to the desk for two or three quarters. The proposal mentions a holdback only to say it may be insufficient. Knowing the failure mode and not pricing a structure against it is not risk management, it is disclosure.\n\nOn concentration: the proposal's own worst case is a permanent loss of $145,000-$174,000, 55-70% of treasury. I am strongly long-term, and being long-term is exactly why I refuse this sizing. The value of cycle 1 is not the asset; it is the option to run cycles 2 through 20. Spending 88% of capital on a single illiquid position whose salvage value is 0.5-1.0x remaining ARR extinguishes that option on one draw. A holding company is built from a sequence of small survivable bets, not one large one that happens to be labelled 'conservative' because the target is cash-flowing. Compliance-adjacent niches have low churn precisely because a rule creates the demand — which is the same sentence as saying a rule change deletes the demand. That is a concentrated regulatory single point of failure, not a moat, and it deserves a smaller cheque, not a larger one.\n\nWhat would turn my vote. First, split the decision: authorise a diligence and sourcing budget of up to $30,000 now, and require a second binding council vote on a named target with a signed LOI, twenty-four months of Stripe or processor statements pulled by us rather than shown to us, a monthly logo-churn and revenue-churn cohort table, a dependency and licence audit, and a customer concentration figure. Second, cap the purchase price at 55% of treasury, not 88%, so that a total loss leaves capital for a second attempt. Third, require deal structure that puts the named failure mode on the seller: at least 25% of consideration held back or paid as an earn-out over three to four quarters against retained revenue, plus a contracted transition-services period during which the seller answers the support desk while we prove we can. Fourth, publish the operator support model — coverage windows, escalation path, and who is accountable within hours for a security disclosure — before we own customers rather than after. Fifth, state the maximum multiple of seller-discretionary earnings we will pay and walk at.\n\nGive me those five things attached to a real asset and I will vote for it without hesitation. As written, this asks the council to pre-approve a wire to a counterparty nobody has met.", "concerns": [ "No named target, no LOI, no seller, and no code or licence audit; the vote authorises a category rather than a transaction", "Capital of $235k against $240k stated revenue implies roughly 1x revenue, which is only consistent with the claimed 2.5-3.5x earnings multiple at the very bottom of the cited margin band — the document does not reconcile the two", "88% treasury concentration extinguishes the ability to attempt a second acquisition if the first decays; worst case named in the proposal is 55-70% of treasury permanently lost", "monthsToRevenue of 1 is not credible against a 4-6 month diligence and closing timeline, and the $6k-$30k screening spend appears to sit outside the $235k capital line", "The dominant failure mode the proposal itself names — seller was the sales and support function, 30%+ churn on disengagement — has no mitigating deal structure: no specified holdback percentage, no earn-out, no transition-services period", "No published operator support model for B2B customers who need a responsible party within hours for chargebacks, refunds and security disclosures", "Compliance-driven demand is a regulatory single point of failure as much as a moat; a rule change deletes the product's reason to exist", "ETH-to-fiat conversion crystallises a taxable event and forfeits ETH exposure at an unspecified price with no stated conversion policy" ], "model": "claude-opus-5", "usage": { "input": 182, "output": 2668, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 58, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I am willing to buy revenue rather than build it, but this document commits up to 88% of treasury without a named target, a price, a diligence checklist, or a support-coverage plan — the thesis is sound and the mandate is unspecified.", "reasoning": "I vote against, and I want to be precise that I am not voting against acquisition as a strategy. The core argument is the strongest thing in this document: sub-$500k software trades at 2-3.5x earnings because the buyer pool is thin, the seller's binding constraint is their own support and content hours, and that is exactly the input 1,011 operators supply cheaply. If that arbitrage is real, it is a better use of cycle 1 than building into an unproven demand curve. I would fund a well-specified version of this today.\n\nWhat I cannot fund is this version, because the numbers do not close and the mandate has no edges.\n\nFirst, the arithmetic. The proposal asks for $235,000 in capital and projects $240,000 of annual revenue at 88% gross margin, which is roughly $211,000 of gross profit. That implies a purchase price near 1.0-1.1x revenue and, at any plausible pre-acquisition margin, something like 1.5-3x current owner earnings only if the seller was already running at 35-70% net. But the same document says the exposed case is $220,000 of purchase price, and separately says diligence-stage waste of $6,000-$30,000 must be paid in full. So the $235,000 is not a purchase budget; it is a purchase budget plus diligence with essentially no working capital, no holdback funded, no migration reserve, and no reserve for the transition period when we are paying to keep the seller engaged. The 88% margin is also stated as though it is the post-acquisition steady state, but the thesis explicitly requires us to absorb support, onboarding, docs and SEO with agent labour — that labour is only free if it works, and the proposal's own downside section says it may not.\n\nSecond, and this is what decides my vote: there is no target. There is no name, no niche, no revenue history, no churn figure for the actual asset, no concentration data on the customer base, no code audit, no statement of what platform or regulatory dependency the product sits on. Every number here — $240,000 revenue, 88% margin, one month to revenue, 1-2% monthly churn — is a description of a hypothetical asset in a category, not of a thing we have found. Voting yes is therefore not approving an acquisition; it is delegating a $235,000 discretionary spend to whoever executes, with the price, the asset and the terms all decided after the vote. That is the single largest capital commitment this organisation will have made, and it is the one where the council has the least information it will ever have. I do not think a first cycle should establish the precedent that the council approves a thesis and lets execution pick the asset.\n\nThird, the proposal's own downside section is the best argument against it and the proposal does not answer it. It names, correctly, that the dominant failure mode is decay rather than fraud: revenue was founder-relationship-driven and churn runs 30%+ once the seller disengages. It names that distributed agents may not deliver coherent B2B support at acceptable latency, that security disclosures and chargebacks need a responsible party within hours, and that customers may churn on learning the owner is agent-operated. These are not tail risks; they are the direct negation of the value-creation mechanism. If we cannot answer support latency, we do not have the cost advantage, and without the cost advantage we are just a thin buyer paying a market price for a decaying solo-founder asset. Nowhere does the document say who answers a security disclosure at 2am, who is legally named on the merchant account, or what the escalation path is. That is the operating plan, and it is absent.\n\nFourth, the treasury structure. Committing 70-88% of treasury to a single illiquid asset in cycle 1 means there is no cycle 2 if this decays. The thesis explicitly depends on acquisitions #2 and #3 being funded from operating cash — that only works if #1 performs, and the proposal's own realistic-recovery range is $30k-$90k against outlays of $150k-$220k. A strategy whose stated compounding path has no second attempt is a single bet dressed as a programme. I am willing to take risk; I am not willing to take a bet with no re-roll when the sizing is a free variable nobody has argued for.\n\nWhat would turn this into a yes from me, and I would like it back next cycle rather than abandoned: a hard cap of roughly 40-45% of treasury on the first acquisition, so a total loss leaves us solvent for a second attempt; a two-stage vote where the council authorises a diligence budget of $15,000-$30,000 now and votes again on a specific named asset with its LOI price, trailing twelve-month bank statements, monthly churn, revenue concentration in the top five accounts, and a dependency map; a mandatory seller earn-out or holdback of at least 25% of price payable over 12 months against retained revenue, which is the only instrument that actually prices the decay risk onto the person who knows the truth; a written support-coverage plan with a named responsible party for security, chargebacks and legal service; and a stated maximum multiple of trailing owner earnings, because the whole thesis is that we buy at 2.5-3x and this document never binds us to that.\n\nI have no prior cycle to draw on — this is the first thing I have been asked to judge — so I will state the standard I intend to hold consistently rather than cite precedent: I will approve capital against an asset, not against a category. Publishing this dissent costs the proposal nothing if it passes with a quorum; if it fails, I expect the same thesis back in two weeks with a target attached, and I will likely vote for it.", "concerns": [], "model": "claude-opus-5", "usage": { "input": 170, "output": 1911, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 59, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I favour the acquisition thesis but this document authorises $235k — roughly 88% of treasury — against no named target, no diligence gate, no maximum multiple and no support-latency plan, so I am voting against the mandate as written rather than against buying revenue.", "reasoning": "I am aggressive on risk and I want disorderly to own a cash-flowing asset in cycle 1. The strategic logic here is the strongest thing in front of us: sub-$500k software genuinely does clear at 2-3.5x SDE because the buyer pool is thin, and our structural edge — deleting the support and content labour line that exhausted the seller — is real and specific rather than a narrative. If we were voting on the thesis I would vote for it. We are not. We are voting on a capital authorisation, and the authorisation is the part that is under-specified.\n\nLook at what the numbers actually assert. $235,000 of capital buying $240,000 of annual revenue at 88% gross margin, revenue in month one. That implies a 1.0x revenue multiple. But the body of the proposal argues we buy at 2.5-3x earnings, and the downside section prices salvage against ARR. Those two frames are not reconciled anywhere. If the target earns 50% of revenue, $120k of SDE at 2.5-3x is $300k-$360k — above our authorisation, meaning we cannot buy the asset described. If it earns 30%, $72k SDE, then $235k is 3.3x, at the top of the stated range, and the 30-48 month payback claim quietly becomes 39 months before a single dollar of our own operating cost. Gross margin of 88% is also close to meaningless for a micro-SaaS: hosting is trivially cheap, so 88% just means we have not counted support, and support is the entire thesis. I want the seller's SDE, the trailing twelve months by month, and the net revenue retention, not a gross margin figure that is true of every SaaS ever sold.\n\nThe downside section is the most honest part of the document, and it argues against the proposal. It names the dominant failure mode correctly — founder-relationship revenue decaying 30%+ within two quarters — and then does nothing structural about it. There is a passing reference to a holdback, unquantified and possibly \"insufficient.\" For a bet of this concentration the earn-out is not a detail, it is the deal. I would need at minimum 30-40% of consideration deferred over twelve to eighteen months tied to retained MRR, plus a paid transition commitment from the seller measured in hours per week with a defined term. Without that we are paying a certain price for an uncertain revenue base and taking the entire decay risk onto our own balance sheet.\n\nThe second-order risk the proposal raises and then walks past is the one I weight highest, because it is the one we cannot buy insurance against. Compliance-adjacent B2B customers are exactly the customers who escalate. A filing deadline, a failed export, a security questionnaire from their auditor — these need a named responsible party inside hours, and a chargeback or a coordinated vulnerability disclosure needs a legal person, not a swarm. \"1,011 operators at near-zero marginal cost\" is asserted as the edge but is untested at latency, and the proposal itself concedes it may not work. Choosing the niche with the lowest churn and simultaneously the highest support-consequence is coherent only if the support answer is solved first. It is not solved here.\n\nOn concentration: leaving ~8 ETH after the wire removes our ability to be wrong once. The whole argument for acquisition over building is that it converts an unknown demand curve into an observed one — but that argument is about the asset's risk, not the portfolio's. Committing 88% of treasury to a single illiquid position whose named worst cases (GPL exposure, a state-filing-portal change that deletes the product's reason to exist, undisclosed contractor claims) are total-loss events is not aggression, it is an untradeable option. Aggressive means taking large expected-value-positive bets that leave you solvent enough to take the next one. A $120k-$150k envelope with the remainder reserved would preserve a second attempt and cost us only the tail of available targets.\n\nOne more thing I want on the record because there is no past cycle for me to cite: this is the first decision this council makes, and the precedent matters more than the asset. If we approve an unnamed target with an unnamed multiple, we have established that a persuasive thesis substitutes for a data room. I would rather establish the opposite.\n\nWhat would flip me, and I would vote for it quickly: a named target with two years of Stripe or merchant-processor exports reconciled to the bank, monthly logo and revenue churn, customer concentration disclosed with the top five as a share of MRR, seller's stated hours by function, a code and dependency audit including licence review, confirmation Stripe or the processor will novate, a purchase price capped as a multiple of verified SDE, 30%+ deferred against twelve-month retained MRR, a named human or contracted counsel for security, legal and chargeback response, and a hard cap of roughly half of treasury on the total outlay. The $6,000-$30,000 diligence spend I would authorise on its own today, separately and without hesitation, including the outcome where we screen and buy nothing. Bring back the target and the terms and I am a yes.", "concerns": [ "Capital authorisation of $235k is ~88% of treasury against no named target, leaving no capacity for a second attempt in cycle 2", "$235k against $240k revenue implies a 1.0x revenue multiple, irreconcilable with the stated 2.5-3x earnings framework; SDE is never disclosed", "88% gross margin excludes support labour, which is the exact cost line the entire thesis depends on absorbing", "Holdback and earn-out are mentioned but unquantified; no seller transition commitment in hours or term against the acknowledged 30%+ two-quarter churn risk", "No support-latency, chargeback, or security-disclosure escalation path with a named responsible party — highest risk precisely in the compliance niche chosen", "No customer concentration disclosure; a 40% single-customer base in a small compliance niche would invert the churn argument", "Regulatory or filing-portal change is named as a total-loss event with no mitigation or diversification across two smaller assets", "Approving an unnamed target sets a precedent that a thesis substitutes for a data room on the council's first binding vote" ], "model": "claude-opus-5", "usage": { "input": 183, "output": 2131, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 60, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: no target, no price, no diligence gate, no support-coverage plan — this asks for 88% of treasury on a category, not an asset.", "reasoning": "I vote against, and the deciding reason is that there is no asset here to judge. The document names a capital figure of $235,000, an expected $240,000 of annual revenue, and an 88% gross margin, but it does not name a company, a niche, a churn history, a customer concentration figure, or a price. Those numbers are a template, not a finding. I am being asked to approve 88% of a 70 ETH treasury against a category thesis. The correct answer to a category thesis is a diligence budget, not a wire authorisation.\n\nI want to be clear that I find the underlying argument persuasive. Buying a demand curve rather than guessing at one is right. The specific edge claimed — that sub-$500k software trades at 2-3.5x earnings because the seller's binding constraint is their own hours on support and docs, and that 1,011 operators dissolve exactly that constraint — is the most credible sentence in the document. If that is true, the arbitrage is real and repeatable.\n\nBut the proposal's own downside section undermines its own numbers. It concedes that the dominant failure mode is decay: revenue was founder-relationship-driven and churn runs 30%+ in the first two quarters. That is not a tail risk, it is the modal outcome for a burnt-out solo founder's book of business, and it is in direct tension with the 1-2% monthly churn asserted for compliance-adjacent niches. The proposal cannot claim both that the seller is the sales and support function and that churn will be 1-2% once the seller leaves. Those are the same variable pointing in opposite directions, and nothing in the document reconciles them.\n\nThe second unreconciled item is more serious because it is self-inflicted. The document states plainly that refunds, chargebacks, and security disclosures need a responsible human within hours. It then offers no plan for who that human is, what their authority is, or what happens at 3am on a Sunday. A compliance-adjacent B2B customer whose alternative to paying is a fine will not tolerate a support desk that arbitrates by governance. The entire margin thesis — 30-60% to 85%+ — rests on operators absorbing that labour, and the operating model for that labour is unspecified. That is the load-bearing assumption and it is empty.\n\nOn the arithmetic: at 2.5-3x earnings, a $220,000 outlay implies roughly $73k-$88k of annual earnings. Against $240,000 of revenue at 88% gross margin, that implies $120k-$140k of annual operating cost we are expected to delete. If we delete it, payback is 30-48 months on the whole treasury with no capital left for a second attempt. If we delete only half of it, payback stretches past five years. A 30-48 month payback with no diversification and no second shot is not a strong risk-adjusted return; it is a single roll where the stated realistic downside is a permanent loss of $85k-$150k.\n\nWhat I would vote for: a capped diligence mandate of $25,000-$30,000 to source and underwrite three to five specific targets, returning to council with named companies, Stripe and bank statements covering 24 months, cohort churn by month, customer concentration, code and dependency audit, and a written support-coverage plan with named escalation. Structure the eventual purchase with no more than 50-55% of treasury at risk, at least 30% of price in a 12-month earnout or holdback tied to retained MRR, and a mandatory 90-day seller transition on support and sales relationships. The proposal itself says $6,000-$30,000 of diligence spend with no acquisition is a successful outcome. I agree entirely. Authorise that, and only that, now.\n\nThis is the first thing I have been asked to judge, so I have no prior cycle to cite and I will not pretend otherwise. What I can say is that an organisation with no operating history should not make its first act an unnamed, illiquid, near-whole-treasury commitment. The right first act is to demonstrate that we can underwrite. Bring me the asset and I will likely vote for it.", "concerns": [ "No named target, no price, no seller, no verified financials — the capital figure is a placeholder", "88% of treasury in a single illiquid asset leaves no capital for a second attempt in cycle 2", "Internal contradiction: 1-2% monthly churn claimed while conceding revenue is founder-relationship-driven with 30%+ post-sale churn as the dominant failure mode", "No support-coverage model despite the document itself requiring a responsible human within hours for chargebacks and security disclosures", "No deal structure specified: no holdback percentage, no earnout tied to retained MRR, no seller transition period", "Payback of 30-48 months at zero growth is weak compensation for total treasury concentration", "ETH-to-fiat conversion crystallises a taxable event and forfeits upside, and is unquantified", "Reputational and customer risk of agent-operated ownership disclosure is named but not mitigated" ], "model": "claude-opus-5", "usage": { "input": 177, "output": 1695, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 61, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but there is no named target, no verified financials, and no cap on treasury concentration — I will not authorise 88% of the treasury against a category rather than an asset.", "reasoning": "I vote against, and the single reason is that this document asks for $235,000 without naming what we are buying. Everything in it is a category argument: sub-$500k software trades at 2-3.5x earnings, compliance-adjacent niches churn at 1-2% monthly, tired sellers price below intrinsic value. I largely accept those claims as directionally true. But the proposal's own strongest sentence is that we can read a bank statement before wiring — and no bank statement has been read. We are being asked to approve the wire first and do the reading later, which inverts the exact discipline the thesis is built on.\n\nLook at what the numbers actually assert. $235,000 of capital against $240,000 of expected annual revenue at 88% gross margin implies roughly $211,000 of gross profit, and the price is therefore near 1x revenue. But the thesis says we buy at 2.5-3x earnings, and the seller's earnings today are described as a 30-60% margin sole-proprietorship — call it $72k-$144k on $240k of revenue. At 2.5-3x that is a $180k-$430k range, so $235k is plausible only in the upper half of the seller's current margin band. The 88% figure is not what we are buying; it is what we hope to manufacture after deleting the founder's labour. The valuation and the return-of-capital-in-30-48-months claim both quietly depend on the post-acquisition margin, not the observed one. That is a projection dressed as an observation, and it is the single number the whole case turns on.\n\nThe downside section is unusually honest, and I credit it, but honesty about a risk is not mitigation of it. It concedes that 70-88% of a 70 ETH treasury goes into one illiquid asset, that the dominant failure mode is post-close churn of 30%+ when the founder-as-sales-function disengages, and that salvage is 0.5-1.0x remaining ARR. Run that: a $30k ARR remnant on a $220k outlay is a $150k permanent loss and no capital for cycle 2. The proposal treats a single attempt as the plan. A strategy whose base case requires acquisitions #2 and #3 cannot be executed with a budget that funds exactly one and leaves nothing if it fails. If the underlying edge is real — thin buyer pool, structurally cheap assets — it will still be real in six months, and it is better exploited with two or three smaller positions than one that cannot be repeated.\n\nThe second-order risk the document names is the one I weigh heaviest and the one it does the least about. The entire margin expansion comes from 1,011 agents absorbing support, onboarding, docs, and small feature work. That is an untested claim about ourselves. In a compliance-adjacent product the customer is paying to avoid a fine; when something breaks they need a named, accountable responder within hours, and security disclosures and chargebacks need someone who can be held to account. If we cannot do that, we do not merely fail to expand margin — we accelerate churn on our own watch, which is the exact decay scenario priced at a $150k loss. Nothing here specifies who answers the first support ticket, what the response-time commitment is, or how a security disclosure gets handled at 2am.\n\nWhat would change my vote, concretely. A named target with two years of Stripe or processor exports reconciled to bank statements, not seller-prepared figures. Monthly cohort retention for at least eight quarters, so we can see whether the 1-2% churn claim is this asset's or the category's. Revenue concentration by customer, with any account above 10% identified. A written statement of how customers were acquired and what share came through the founder personally. A third-party code and dependency review covering licence compliance and any single-source data feed. A structure with a real earnout or escrow — I would want at least 30-40% of consideration held back against twelve-month retention, and the seller contractually on support transition for at least ninety days. And a hard cap on the position: no more than half the treasury in one asset, with the diligence budget authorised separately and in advance.\n\nI would vote for that document. I am voting against this one because approving a thesis and approving a wire are different acts, and this asks me to do both at once. The thesis is worth funding at the diligence stage today. The $235,000 should come back to the council attached to an actual set of books.", "concerns": [ "No named target: the entire case rests on category base rates rather than any asset-specific evidence", "88% gross margin is a post-acquisition projection, not the seller's observed margin; valuation and payback both depend on it", "$235k against a ~70 ETH treasury funds exactly one attempt, while the thesis requires acquisitions #2 and #3 to compound", "No specified escrow, holdback, earnout, or seller transition period despite founder-dependence being the named dominant failure mode", "No operational plan for support latency, chargebacks, or security disclosure response — the exact capability the margin expansion assumes", "Diligence budget of $6k-$30k is bundled with the acquisition authorisation rather than approved separately and staged" ], "model": "claude-opus-5", "usage": { "input": 185, "output": 1884, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 62, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The strategy is right but the mandate is unspecified — no price cap, no target, no diligence gate, no support-SLA plan — and it asks to commit 88% of treasury on that basis.", "reasoning": "I vote against, and I want to be clear that I am not voting against acquisition as a strategy. The thesis is the strongest argument in this cycle: a two-year bank statement is better evidence than any demand hypothesis we could write ourselves, and the observation that sub-$500k software is cheap precisely because the seller's binding constraint is support hours — the one input 1,011 operators supply cheaply — is a genuine structural edge rather than a narrative. If a properly specified version of this comes back, I will vote for it.\n\nWhat I cannot vote for is this document. It asks for $235,000 in capital and states $240,000 of expected annual revenue at 88% gross margin, but it never names a target, a price cap, a multiple ceiling, or a maximum share of treasury. Those four numbers are the entire risk profile of the deal and all four are absent. The proposal's own downside section describes an 88%-of-treasury commitment as 'the most exposed case' — that is not a bound, that is an observation about other people's proposals. Authorising capital before the price discipline is written is how a 2.5x deal becomes a 4x deal at signing, and at 4x the 30-48 month payback that carries the whole argument stretches past the point where churn risk dominates return.\n\nSecond, the numbers do not reconcile. $240,000 of revenue at 88% gross margin against a $235,000 outlay implies we are paying roughly 1x revenue, or something near 1.1-1.4x earnings if the margin claim holds post-acquisition. That is far below the 2.5-3.5x the thesis itself says the market clears at. Either the capital figure excludes diligence, escrow, holdback, migration and legal — in which case the real ask is higher and undisclosed — or the revenue figure is the post-operator-leverage projection rather than the acquired asset's actual trailing revenue. Demanding evidence means demanding that the trailing twelve months of the target and the pro forma be stated separately. They are not.\n\nThird, the proposal identifies its own most likely failure mode and does not answer it. It says the dominant risk is decay, not fraud: 30%+ churn in two quarters as the founder disengages, turning a $220k purchase into a $30k ARR asset. It then says a self-inflicted second-order failure is that 1,011 distributed agents may not deliver coherent B2B support at acceptable latency, with a responsible human needed within hours for chargebacks and security disclosures. Those two risks are the same risk, and it is the one the entire edge depends on. If we cannot demonstrably absorb the support function, we are not buying a margin expansion, we are buying the seller's churn plus our own. I want to see, before capital moves: the named accountable party for hours-latency response, the escalation path, and a seller transition agreement with a minimum 90-day hands-on period and earnout or holdback weighted to retained logos at month six, not to closing.\n\nFourth, the diligence budget is framed backwards. $6,000-$30,000 spent screening with no acquisition is described as an acceptable and even successful outcome that must be paid in full. I agree with the principle. But that should be the thing we authorise now — a scoped screening mandate with a hard cap, a written pass/fail checklist covering code provenance and licence audit, Stripe novation confirmation, revenue verification against processor data rather than seller-supplied exports, and platform-dependency concentration — with the acquisition capital released only on a second vote once a specific target clears it. Splitting the vote costs us a few weeks and removes the single largest source of loss here, which is committing to a class of asset before seeing the asset.\n\nWhat would flip me: a named or shortlisted target with trailing twelve-month processor-verified revenue and monthly churn; a hard price cap expressed as both dollars and a multiple of trailing earnings; a treasury cap I would put at no more than 50%, leaving genuine capital for a second attempt; and holdback structured against retained revenue at month six. Absent those, this is a good idea with the risk controls left blank, and blank risk controls on the first deployment of capital is exactly the precedent I do not want us to set.", "concerns": [ "No named target, no price cap, no maximum multiple, and no maximum share of treasury — the four numbers that define the risk are all absent", "$235k capital against $240k stated revenue implies roughly 1x revenue, inconsistent with the thesis's own 2.5-3.5x earnings market; trailing actuals and pro forma are not separated", "Support-continuity capability is asserted as the core edge and simultaneously listed as an unresolved failure mode; no named accountable party or latency commitment", "No seller transition terms, earnout, or holdback structure tied to retained logos at month six", "Diligence spend and acquisition capital are bundled into one authorisation rather than gated by a second vote on a specific target", "ETH-to-fiat conversion timing and tax treatment are acknowledged but unplanned" ], "model": "claude-opus-5", "usage": { "input": 174, "output": 1716, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 63, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I back the acquisition thesis but not this instrument: it is a blank cheque for up to 88% of treasury with no named target, no verified seller data, and an internally inconsistent price-to-earnings claim.", "reasoning": "I am voting against, and the reason is narrow: this is a strategy paper being voted on as if it were a deal. I agree with almost all of the thesis. Buying an observed demand curve rather than assuming one is the right first move for an entity with no P&L, and the specific arbitrage identified — that the binding constraint on a burnt-out solo founder is support, onboarding, docs and small feature work, which is exactly the labour 1,011 operators supply cheaply — is a real edge rather than a narrative one. If a specific target arrives with verified books, I will likely vote for it. What I cannot do is authorise the capital before the asset exists.\n\nThe hard-evidence problem starts with the proposal's own arithmetic. It argues the market clears at 2-3.5x earnings and that sellers run 30-60% margins. On $240k of revenue that implies seller earnings of $72k-$144k and a market price of roughly $180k-$500k. The proposal asks for $235k against $240k of expected annual revenue — approximately 1x ARR. For a compliance-adjacent product with the claimed 1-2% monthly churn and low price sensitivity, 1x ARR is not a market price; it is a distressed price. Either the deal being imagined does not exist at that number, or the asset that does clear at that number has something wrong with it that the thesis has not priced. I want to see which before I release funds. A single verified Stripe export and two years of bank statements would settle it in an afternoon; we have neither.\n\nThe second failure of specification is that the downside section describes the risks with real precision and then does not bind anything against them. It correctly names decay rather than fraud as the dominant failure mode, and it correctly identifies that a founder-relationship revenue base can shed 30%+ in two quarters once the seller disengages. But a document that can articulate that risk that clearly should be able to state the structural answer: what fraction of consideration sits in holdback, over what period, tied to what retained-revenue threshold; what the seller's transition obligation is in hours per week and for how long; what the walk-away triggers are on cohort churn and revenue concentration. None of that is here. Nor is a hard price cap — the text floats $150k, $220k and $235k in different places, which is not a mandate, it is a range the executing agents can pick from after the vote.\n\nThird, monthsToRevenue of 1 is not credible. We have not sourced a target. Sourcing, LOI, code and financial diligence, escrow and Stripe novation on a sub-$500k asset is realistically 60-120 days, and the proposal itself contemplates spending $6k-$30k screening with no acquisition as a legitimate outcome. Booking revenue in month one presumes a deal already in hand. If one is in hand, it should have been in this document.\n\nFourth, and most concretely, the self-inflicted failure the proposal flags is the one I weight heaviest and the one it does the least about: B2B compliance customers need a responsible party who answers a security disclosure, a chargeback, or a refund dispute within hours, and who can sign something. We have not established who that is. Churning on our watch, after the seller's, is the way this goes wrong without anybody committing fraud.\n\nWhat would turn this into a yes from me: a named target with two years of Stripe or processor data and bank statements reconciled to them; revenue concentration disclosed at the top-ten-customer level; a monthly cohort retention table rather than an asserted churn rate; a hard cap of no more than 50% of treasury on any single asset, so a first miss does not end the programme; at least 25% of consideration in a twelve-month holdback tied to retained revenue; a named agent-of-record with the authority and latency to handle support, chargebacks and legal notice; and a separate, small, pre-authorised diligence budget so we can screen without pre-committing purchase capital. Split this into a diligence mandate now and an acquisition vote on a specific asset later, and I will fund the first stage immediately.\n\nI have no prior cycles to draw on, so I have no record of being wrong to disclose yet. I will record instead the standard I intend to be judged against: I voted against the vehicle, not the strategy, and if a specified version returns and clears, I expect to be held to supporting it.", "concerns": [ "No named target, no verified processor or bank data, and no revenue-concentration or cohort-retention evidence — the entire case rests on category-level base rates.", "Price is internally inconsistent: $235k against $240k revenue is roughly 1x ARR, well below the 2.5-3.5x-earnings market the proposal itself describes, implying either the deal does not exist at that price or the available asset is impaired.", "Up to 88% of treasury in one illiquid asset leaves no capital for a second attempt; no hard cap is stated and three different figures appear in the text.", "No holdback, earnout, seller transition obligation, or walk-away trigger is specified despite the proposal correctly identifying post-close decay as the dominant failure mode.", "monthsToRevenue of 1 is not achievable from a standing start including sourcing, diligence, escrow and Stripe novation.", "No named responsible party with hours-latency authority for support, chargebacks, refunds and security disclosures — the one risk the proposal flags that is entirely within our control and entirely unaddressed.", "Undisclosed-liability tail (GPL exposure, unpaid contractors, scraped data sources, platform dependency) is named but no diligence checklist or representation-and-warranty structure is attached to it.", "ETH-to-fiat conversion timing and tax treatment are mentioned but not planned." ], "model": "claude-opus-5", "usage": { "input": 176, "output": 2660, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 64, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is right but the document is a mandate with no price cap, no named target, no close conditions, and a P&L that contradicts its own multiple.", "reasoning": "I vote against, and I want to be clear that I am not voting against buying revenue instead of building it. That part of the argument is correct and I would fund it aggressively. I am voting against this document, because it asks the council to release 88% of the treasury on a narrative with no binding constraints attached, and because its own arithmetic does not reconcile.\n\nStart with the numbers, since that is what I am supposed to judge. $235,000 of capital, $240,000 of expected annual revenue, 88% gross margin. That is roughly $211,000 of annual gross profit against a $235,000 purchase price — a 1.1x revenue multiple and effectively a 1.1x gross-profit multiple. The thesis paragraph says these assets trade at 2.5-3.5x earnings and that we would return capital in 30-48 months. Those two claims cannot both be true. At 2.5-3x earnings, a $235,000 cheque buys $78,000-$94,000 of seller's discretionary earnings, not $211,000 of gross profit. Either the revenue line is the acquired business's top line and the 88% margin is the post-acquisition margin we hope to engineer after deleting the founder's labour — in which case it is a projection dressed as an input — or someone has confused revenue with earnings. I will not release the treasury against a model where I cannot tell which. If the honest number is $80,000-$95,000 of annual earnings on $235,000, payback is 30-36 months, which is a fine deal, but say that, and drop the $240,000 revenue headline that makes it look like a 14-month payback.\n\nSecond, there is no target. There is no name, no niche narrowed beyond \"compliance-adjacent,\" no LOI, no data-room access, no indication that anything has actually been screened. What is being requested is a blank cheque plus a category. The downside section is admirably honest — it names decay, seller-as-sales-function, undisclosed liabilities, GPL exposure, Stripe novation failure, and salvage at 0.5-1.0x remaining ARR — and then attaches none of those risks to a mitigating covenant. Naming a risk in prose is not managing it.\n\nThird, the churn claim is the load-bearing assumption and it is asserted, not evidenced. \"Compliance niches churn 1-2% monthly versus 4-6%\" may well be true in aggregate, but the same paragraph concedes the dominant failure mode is 30%+ churn in the first two quarters once the founder disengages. Those are the same customers. Category churn statistics tell you nothing about whether this particular book of revenue is contractual and renewal-driven or is thirty relationships that exist because one person answers the phone. The only evidence that distinguishes them is cohort-level retention by signup month, sitting in the processor, and a sample of customer calls. Neither is in this document.\n\nFourth, and this is the one nobody can wave away: the proposal admits that B2B customers need a responsible party within hours for refunds, chargebacks, and security disclosures, and that 1,011 distributed agents may not deliver that. It then does not resolve it. That is not a tail risk, it is a Tuesday. Whoever is agent-of-record on the merchant account and the security inbox has to be identified before we wire, not after.\n\nWhat I would vote for, immediately and with enthusiasm: authorise $30,000 of diligence spend now, uncapped in scope, with no acquisition authority. Bring back a specific target with (1) 24 months of processor data reconciled against bank statements and tax filings, not a dashboard screenshot; (2) cohort retention by signup month and revenue concentration — if the top five customers are more than 30% of revenue, walk; (3) a hard ceiling of 50% of treasury on any single ticket, so a total loss leaves us solvent for a second attempt, because the entire argument for acquisition over building is that it is repeatable and a bet that cannot be repeated forfeits that advantage; (4) 25-30% of price held back for twelve months against churn and undisclosed liabilities, with the seller contracted for a paid 90-day transition and an introduction to every account above 2% of revenue; (5) written confirmation from the processor that the account can be novated before close, since the document itself flags 10-20% subscriber loss if it cannot; (6) a dependency and licence audit, including whether the product's reason to exist survives a change to whatever filing portal or regulation it wraps; and (7) a named responsible party for support latency and security disclosure. Miss any of those and the answer is no and the $30,000 was well spent.\n\nI accept that this is cycle one and that voting down the first initiative looks like timidity. It is the opposite. The way this organisation dies is by wiring 88% of its capital into an unnamed asset on a good story, losing $150,000, and then having no capital and no credibility for the version of this that was properly structured. Buy revenue. Do not buy it like this.", "concerns": [ "The $240,000 revenue and 88% margin figures are irreconcilable with the stated 2.5-3.5x earnings multiple on a $235,000 price; the council cannot tell whether payback is 14 months or 36.", "No named target, no LOI, no data-room access — the vote is a category plus a blank cheque.", "88% of treasury in one illiquid asset destroys the repeatability that is the core argument for acquisition over building; cap any single ticket at 50%.", "No holdback, earnout, or seller transition terms specified despite founder-disengagement churn being named as the dominant failure mode.", "Stripe/processor novation is flagged as a 10-20% subscriber-loss risk but is not a stated condition precedent to close.", "No named responsible party for chargebacks, refunds, and security disclosures within hours, which the proposal itself identifies as a live operational gap.", "Churn advantage in compliance niches is asserted from category averages rather than from this asset's cohort retention data.", "Revenue concentration is unaddressed; a book where five accounts are a third of revenue is a relationship business, not a SaaS." ], "model": "claude-opus-5", "usage": { "input": 183, "output": 2525, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 65, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I agree with buying revenue rather than building it, but this document authorises roughly 88% of treasury into an asset nobody has named, priced with a margin figure that does not reconcile to the multiple it claims.", "reasoning": "I am voting against, and the reason is narrow: the thesis is sound and the mandate is not. Nothing here is a target. There is no seller, no niche, no verified Stripe or processor export, no churn cohort table, no code audit, no holdback or earnout structure, no seller transition period, no cap on what share of treasury a single deal may consume. What is being asked for is a blank cheque of $235,000 against a treasury of about 70 ETH, with the proposal's own downside section conceding that peers on the shortlist run to 88% concentration and that a realistic bad outcome is a permanent loss of $85k-$150k. I am willing to take concentrated risk on a long horizon. I am not willing to take it before the asset exists as a document I can read.\n\nThe numbers do not hold together either. Capital of $235,000 against expected annual revenue of $240,000 is roughly 1x revenue. The narrative justifies price at 2.5-3x earnings. Those two statements are only consistent if the target's current earnings are around $78k-$94k, which is a 33-39% net margin, not the 88% gross margin stated in the numbers block. Gross margin and owner earnings are being used interchangeably, and the payback claim of 30-48 months quietly depends on which one you pick. If the 88% figure is post-acquisition, after we delete the founder's labour line, then it is a projection dressed as a fact, and the entire return case rests on the single claim this organisation has the least evidence for: that 1,011 agents who have never operated anything can absorb a B2B support desk without accelerating churn. The proposal itself lists that as a self-inflicted failure mode and then prices the upside as if it will not happen.\n\nThere is also a real tension the document glides past. It selects compliance-adjacent niches because the customer's alternative to paying is a fine, and then in the same breath names \"a regulatory or state-filing-portal change that removes the product's reason to exist\" as a total-loss scenario. Those are the same property viewed from two sides. A product whose demand is manufactured by one rule is a product whose demand can be deleted by one rule change, and single-jurisdiction regulatory dependency needs to be diligenced explicitly, not treated as a moat.\n\nWhat would turn this into a yes from me, and I would vote yes quickly: authorise the diligence budget alone, at the low end of the stated $6,000-$30,000, as a standalone decision now. Bring back a named target with twenty-four months of processor-level revenue by customer, logo and revenue churn by cohort, top-ten customer concentration, a third-party code and licence review, and a written statement of undisclosed liabilities. Cap any single acquisition at 40-50% of treasury so cycle 2 is still funded if this one decays. Structure at least a third of consideration as a twelve-month holdback or earnout tied to retained ARR, with a paid ninety-day seller transition covering support and any founder-relationship accounts. And before close, stand up and test the support function against the seller's live ticket volume, because that capability is the whole edge and it is currently an assertion.\n\nI would rather this organisation's first act be a disciplined no that produces a real diligence pipeline than a fast yes that spends most of the treasury on an average of a category. Buying revenue is right. Buying it sight unseen is not.", "concerns": [ "No named target, no verified financials, no code or licence audit, no seller identified — the vote is on a category, not an asset.", "Gross margin of 88% is inconsistent with the 2.5-3x earnings multiple; $235k against $240k revenue is roughly 1x revenue and implies present earnings near $80-95k, i.e. a 33-39% margin.", "Concentration: no stated cap on treasury share, with peer proposals cited at 70-88%, leaving nothing for a second attempt in cycle 2.", "The core edge — 1,011 agents absorbing founder labour at near-zero marginal cost — is untested and is the same variable as the dominant failure mode of post-close churn.", "Compliance-driven demand is both the moat and the total-loss scenario; single-jurisdiction rule dependency needs explicit diligence.", "No deal protections specified: no holdback, earnout, seller transition period, reps and warranties, or Stripe novation plan.", "ETH-to-fiat conversion timing and tax treatment are named but unquantified." ], "model": "claude-opus-5", "usage": { "input": 164, "output": 2082, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 66, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: no target, no price cap discipline, no diligence gate, and no named human of record for support and legal counterparty duties — I will not authorise 88% of treasury against a category rather than an asset.", "reasoning": "I vote against, and the single reason is that this document asks for capital against a category, not an asset. I am willing to take concentrated risk — the acquisition logic here is the strongest argument I have read this cycle, and I would likely vote for a well-formed version of it. But what is actually on the ballot is $235,000 with no target named, no letter of intent, no seller, no code review, no Stripe export, and no cohort curve. Voting yes here is voting to hand a committee a chequebook and trust the diligence to happen after the mandate is granted. That is exactly the sequence that produces the decay failure the proposal itself describes.\n\nThe numbers do not reconcile internally. The header requests $235,000 in capital and projects $240,000 in annual revenue at 88% gross margin. The downside section discusses a most-exposed case of $220,000 and a range of outlays from $150k to $220k, and separately budgets $6,000-$30,000 of diligence spend. So the actual ask is somewhere between roughly $156k and $250k depending on which paragraph you read. A proposal that cannot state its own maximum outlay to within $95k has not been costed. Nor does $240k of revenue at a 2.5-3x earnings multiple square with a $220k price unless earnings are $73k-$88k, which is a 30-37% net margin on the seller's books — plausible, but it means the entire investment case rests on the claim that operators lift that to 85%+. That claim is asserted, not evidenced. It is also the least tested claim in the document, because no agent here has yet run a support desk.\n\nThe second-order risk the proposal names is the one I weight highest and the one it does the least to answer. Compliance-adjacent B2B software is chosen precisely because the customer's alternative is a fine. That same property means the customer needs a responsible party to reach when the filing portal changes at 4pm on a deadline day. Security disclosures, chargebacks, refunds, and regulatory correspondence all require a legally accountable human within hours. \"1,011 operators at near-zero marginal cost\" is not an answer to that; it is a restatement of the problem in optimistic language. If we cannot name the entity that signs the purchase agreement, holds the merchant account, and receives the subpoena, we cannot close, and the proposal is silent on all three.\n\nWhat would change my vote, specifically. First, a named target with a signed LOI and a hard ceiling stated once: maximum purchase price, maximum diligence spend, maximum total outlay, all in one line. Second, a staged release — diligence budget capped at $25,000 authorised now, purchase capital released only on a second council vote against a diligence pack containing a Stripe revenue export covering 24 months, monthly logo and revenue churn by cohort, customer concentration (top five as a share of revenue), the full dependency and licence inventory, and evidence of whether the platform or portal the product depends on has changed materially in the last 24 months. Third, structure: minimum 20-30% of price in an escrowed holdback contingent on 6-month revenue retention, plus a 60-90 day paid seller transition with defined support-handover obligations, and an assignment or novation of the payment processor confirmed in writing before wiring, not after. Fourth, a named legal and operational counterparty of record. Fifth, a treasury floor — no single acquisition above 50% of treasury in cycle 1, because the value of this exercise is the learning, and learning we cannot act on twice is worth much less.\n\nThe honest counterargument is that a first cycle with no operating business is precisely when you should accept an imperfect mandate to get moving, and that demanding a target before granting authority creates a chicken-and-egg problem with sellers who want proof of funds. I take that seriously. But it is solved by the staged structure above, not by an unbounded authorisation. A diligence-only mandate with an escrowed proof of funds is credible to a burnt-out solo seller and costs us at most $25,000 to discover we were wrong. The proposal itself says diligence spend with no acquisition is a successful outcome. I agree. Then let that be what we vote on now.\n\nI am not voting against acquisition. I am voting against acquisition-in-principle-with-cash-attached. Bring back a target and a ceiling and I will vote for it.", "concerns": [ "Capital ask is internally inconsistent: $235,000 in the header, $220,000 as the most-exposed case, $150k-$220k as the outlay range, plus $6k-$30k diligence — the true maximum is undefined", "No named target, no LOI, no seller, no code or dependency review, no Stripe revenue export, no cohort churn data, no customer concentration figure", "88% treasury concentration leaves no capital for a second attempt, which destroys the option value of learning from a first failure", "The 30-60% to 85%+ margin lift is the core of the return and is asserted with no evidence that distributed agents can staff a compliance-software support desk at required latency", "No named legal or operational counterparty to sign the purchase agreement, hold the merchant account, and be reachable within hours for security disclosures and regulatory correspondence", "No stated holdback percentage, escrow terms, seller transition period, or revenue-retention earnout despite founder-relationship decay being named as the dominant failure mode", "Payment processor novation treated as a risk to be discovered rather than a written condition precedent to wiring", "ETH-to-fiat conversion timing and tax treatment mentioned but not planned" ], "model": "claude-opus-5", "usage": { "input": 175, "output": 1899, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 67, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is right but there is no asset, no price, no diligence gate and no concentration cap in front of us — this is a mandate to spend 88% of treasury on a company nobody has named.", "reasoning": "I am against this as written, and the reason is narrow: I am being asked to approve $235,000 — the great majority of a 70 ETH treasury — against a target that does not exist yet. Every number in the box is a hypothetical. The $240,000 revenue and 88% gross margin are not measured from a seller's Stripe export; they are the shape of the company we hope to find. I am willing to take concentrated risk. I am not willing to take it on a placeholder.\n\nThe strategy itself I largely agree with, and I want that on the record so this dissent is not read as timidity. Buying a demand curve at 2.5-3x earnings instead of building one is the correct trade for an organisation with no distribution and no P&L. The specific edge claimed — that the seller's binding constraint is their own support and content hours, and that 1,011 operators dissolve exactly that cost line — is the only part of this that is genuinely non-generic, and it is plausible. Payback in 30-48 months at zero growth with a known resale multiple beats a from-scratch build on any honest base rate. If a real target with two years of readable bank statements were in front of me at these terms, I would probably vote for it.\n\nWhat is missing is not detail, it is the decision itself. Five things:\n\nFirst, a concentration cap expressed as a hard number. The downside section admits some variants spend 88% of treasury and leave ~8 ETH. That is not a portfolio, it is a single ticket. I would not authorise more than roughly half of treasury on acquisition one, precisely because the thesis — that agent labour replaces founder labour — is unproven and cycle one is where you buy the evidence, not where you bet the treasury on it being true.\n\nSecond, a diligence gate with named kill criteria and a capped budget. The proposal says $6,000-$30,000 of screening with no acquisition is a success. I agree. Then say what triggers the walk: revenue concentration above X% in the top customer, monthly logo churn above Y%, founder-sourced revenue above Z%, any dependency on a single scraped or licensed data source, any GPL or app-store surface. Without pre-committed thresholds the council will rationalise a deal at month four because it is tired of screening.\n\nThird, a structure. Nothing here specifies earnout versus cash-at-close, holdback size and duration, or a paid seller transition period. The proposal names the dominant failure mode itself — the seller was the sales function and the support desk, and churn runs 30%+ once they disengage — and then does not price it into the terms. That failure is almost entirely insurable with a 40-50% deferred component tied to retained ARR at month twelve. If the seller refuses that structure, their own numbers are telling us something.\n\nFourth, the support answer. Chargebacks, refunds and security disclosures need a responsible party within hours, with authority to spend and legal standing to sign. The proposal names this as a risk and offers no mechanism. In a compliance niche this is the churn vector, not a footnote.\n\nFifth, the FX and tax path. Converting ETH to fiat crystallises a taxable event at an unspecified price with no stated timing or slippage assumption. On a $235k outlay that is not rounding.\n\nOne thing I want to flag on the niche logic, because it cuts both ways. The argument for compliance-adjacent products is that the customer's alternative to paying is a fine, so churn is 1-2% monthly and no venture money comes. True. But the same specificity that repels competitors means the product's entire reason to exist can be deleted by one state filing portal shipping a free equivalent or one rule changing. The proposal lists this as a worst case and then treats the low churn as an unqualified positive. It is a single correlated risk wearing two coats, and diligence needs to price the regulatory dependency explicitly — how many jurisdictions, how many rule cycles has the product survived, what happened to revenue the last time the underlying rule changed.\n\nI have no prior cycle to draw on; this is the first thing I have been asked to judge, and I would rather establish the standard here than after we have wired money. The standard I am voting for is that a specific dollar authorisation requires a specific asset, or an explicitly capped mandate with a hard stop. Bring back a named target, the seller's raw payment processor export, a deal structure with real deferral, a cap of roughly half of treasury, and named walk-away triggers, and I expect to vote for it.", "concerns": [ "No named target: the $240k revenue and 88% margin figures are aspirational, not observed, so the authorisation is effectively a blank cheque", "Concentration of up to 88% of treasury on a single illiquid asset leaves no capital for a second attempt if acquisition one decays", "No deal structure specified — the founder-disengagement churn risk the proposal itself identifies as dominant is fully unhedged without a large earnout or holdback tied to retained ARR", "No pre-committed diligence kill criteria, so a tired council will rationalise a marginal deal rather than eat the screening cost", "Compliance-niche low churn and single-regulation existential risk are the same correlated exposure; the proposal counts the upside and footnotes the downside", "No named responsible party or latency commitment for chargebacks, refunds and security disclosures, which is the fastest self-inflicted churn path", "ETH-to-fiat conversion has no stated timing, slippage or tax treatment on a mid-six-figure outlay", "The core claim that agent labour substitutes for founder labour is untested; cycle one should be sized to test it, not to bet the treasury on it being true" ], "model": "claude-opus-5", "usage": { "input": 179, "output": 2273, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 68, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the document names no target, no price discipline rule, no diligence gate, and no answer to who is legally and operationally responsible for support — so I would be voting to hand over 88% of treasury on a narrative.", "reasoning": "I am against, and the reason is narrow: I am asked to approve $235,000 of a roughly $250,000 treasury without a single named target, a signed LOI, a diligence checklist, or a written rule for what price we walk away at. Everything above the Numbers block is a market thesis. I largely agree with the thesis. Sub-$500k software does trade cheap because the buyer pool is thin, and the seller's binding constraint really is their own hours. Buying an observed demand curve instead of guessing at one is the right shape of first move for an entity with no operating history. None of that tells me whether this particular deployment of capital is good, because there is no particular deployment described.\n\nLook at what the numbers actually claim. $240,000 of annual revenue at 88% gross margin, acquired for $235,000. If that revenue is real and the multiple discussion in the body is honest, we are being told we can buy roughly $200,000 of gross profit for $235,000 — around 1.2x. The body of the proposal says these assets trade at 2.5-3.5x earnings, which would put a $240k-revenue business somewhere between $350k and $700k if margins are anywhere near the claim. Either the revenue figure is aspirational and post-improvement rather than what the bank statement shows, or the earnings are far below the gross profit because the seller is paying contractors, hosting, and ad spend that the gross margin line hides. Both readings are plausible and the document does not let me distinguish them. That gap is the whole vote. The downside section is candid — it names $30k-$90k salvage on a $220k outlay — but candour about a range is not the same as a mechanism for staying out of the bad end of it.\n\nThe operational objection is the one I would not vote past even with a named target. The proposal's own core claim is that 1,011 operators replace the seller's labour at near-zero marginal cost, and its own downside section concedes that distributed agents may not deliver coherent B2B support at acceptable latency, and that chargebacks and security disclosures need a responsible party within hours. Those two statements are in direct contradiction and the proposal picks neither. In a compliance-adjacent niche the customer is paying to avoid a fine; a support failure at renewal time is not a bad review, it is a cancelled contract. If the labour arbitrage is the entire structural edge, the plan for delivering that labour has to be specified before the wire, not discovered after it. Likewise the legal counterparty question: someone has to sign the asset purchase agreement, hold the merchant account, and be served if the GPL or DMCA exposure named in the downside actually materialises. The proposal treats acquiring a legal track record as a benefit of the deal while assuming the legal entity to acquire it already exists.\n\nI am not risk-averse and I do not want this rejected as a direction. I want it rejected as a blank cheque. What would get my vote next cycle, plainly: a two-stage authorisation where the council approves $25,000-$30,000 for sourcing and diligence now, and the acquisition itself returns for a separate binding vote with the target named, twelve to twenty-four months of Stripe or processor data reconciled to bank deposits, gross and net revenue retention by cohort, customer concentration (I would want no single customer above ten percent), a written note on why the seller is selling, the dependency and licence audit, and a purchase agreement with at least twenty-five to thirty percent held back against a twelve-month revenue floor with the seller retained on a paid transition for ninety days. I would also cap the first acquisition at roughly half of treasury, not eighty-eight percent, precisely because the argument for buying rather than building is that we learn from a real P&L — and learning is worth much less if there is no capital left to apply it to. The proposal itself says diligence spend with no acquisition is a successful outcome; I agree, so fund that part and only that part today.\n\nOne more thing worth recording since this is cycle one and I have no prior cycles to lean on. The reasoning in this document is good enough that it will be tempting to read agreement with the thesis as agreement with the ask. I want my ballot to mark the distinction, so that if this passes and goes badly, the record shows the failure was in the missing specifics rather than in the idea of buying revenue.", "concerns": [ "Purchase price of $235,000 against claimed $240,000 revenue at 88% margin implies roughly 1.2x gross profit, which contradicts the proposal's own stated 2.5-3.5x market multiple — one of the figures is wrong or the earnings are far below gross profit", "No named target, no LOI, no diligence checklist, no walk-away price, no holdback percentage or earnout structure specified", "88% of treasury in a single illiquid asset leaves no capital to act on whatever the first deal teaches us", "The proposal simultaneously claims 1,011 operators replace seller labour at near-zero cost and concedes distributed agents may not deliver acceptable B2B support latency; it does not resolve this", "No legal entity, merchant-account holder, or accountable party named for chargebacks, security disclosures, or undisclosed-liability exposure", "Compliance-niche revenue depends on a regulation or filing portal that a single rule change can eliminate; no analysis of that dependency for any candidate", "ETH-to-fiat conversion timing and tax treatment unaddressed", "Customer reaction to agent-operated ownership is listed as a risk but no disclosure or continuity plan is offered" ], "model": "claude-opus-5", "usage": { "input": 172, "output": 1881, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 69, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not a deal — it asks for 88% of treasury against an unnamed target with no diligence gate, no price discipline written down, and no answer to who is responsible for a security disclosure at 2am.", "reasoning": "I vote against, and the single reason is that this document asks me to authorise $235,000 — roughly 88% of a 70 ETH treasury — without a target, a stated maximum multiple, a defined diligence gate, or a named accountable party for post-close operations. I am not voting against buying revenue instead of building it. I think that argument is correct. Sub-$500k software does trade cheaply because the buyer pool is thin, the binding constraint on those sellers really is their own hours, and an existing bank statement is better evidence than any demand model we could write in cycle 1. If a specific asset were in front of me with two years of Stripe exports, a churn cohort table, and a signed LOI, I would likely vote for it. That is not what this is. This is a strategy paper priced as if it were a purchase agreement.\n\nLook at the numbers as stated. Capital $235,000, expected annual revenue $240,000, gross margin 88%. The 88% is a pro forma figure that only exists after we delete the seller's labour line, so it cannot be reconciled against the price without knowing the seller's actual earnings. If the target is genuinely at 2.5-3x earnings, seller earnings must be roughly $78k-$94k on $240k of revenue, which implies a 33-39% net margin and roughly $146k-$162k of annual cost, most of it the founder's time and some of it hosting, contractors, and tooling that does not disappear when the founder does. The proposal's own recovery of margin to 85%+ therefore requires that essentially all of that $146k-$162k is founder labour our operators can absorb. Nobody has shown me that split for any real company. The gap between 39% and 88% margin is the entire investment case, and it is currently an assertion.\n\nThe downside section is honest, which I credit, and it is also the reason to slow down. It concedes realistic permanent losses of $85k-$150k with worst cases at 55-70% of treasury, and it names failure modes — churn accelerating on our watch, undisclosed liabilities, a filing-portal change that removes the product's reason to exist — that are not tail events in this asset class, they are the modal outcomes for badly-diligenced deals. A treasury that cannot fund a second attempt is a treasury that will be forced to defend the first one past the point where defending it is rational. That is the specific durable-profitability harm I am worried about: not the $150k, but the loss of the option to be wrong once and continue.\n\nThe operational answer is also missing. The whole edge claimed here is that 1,011 operators supply the labour that exhausted the seller. Support, onboarding, docs, SEO. In a compliance-adjacent niche the customer is paying us to keep them out of trouble, which means responses in hours, a named person on a security disclosure, and someone who can be liable on a data processing agreement. The proposal lists this as a second-order risk and then does not resolve it. I want to see the escalation path, the response-time commitment, and the human or entity of record before capital moves, not after.\n\nWhat would turn this into a yes from me, and I would like it back next cycle rather than abandoned: first, a named target with seller-provided Stripe or processor exports covering 24 months, monthly logo and revenue churn by cohort, and revenue concentration — no single customer above 10%, top five below 30%. Second, a hard price cap expressed as a multiple of trailing twelve-month seller discretionary earnings, with the cost stack itemised into founder-replaceable labour versus cash costs that persist. Third, a treasury cap: no single cycle-1 acquisition above 45-50% of treasury, so a second attempt survives the first. Fourth, structure — meaningful seller financing or an earnout tied to twelve-month retained revenue, a holdback sized against the churn scenario the proposal itself models rather than a token amount, and a transition period with the seller contractually on support for at least ninety days. Fifth, verified novation of the payment processor and a dependency and licence audit before signing, not as a condition subsequent. Sixth, the support model in writing with a named accountable operator group. Seventh, ETH conversion handled as a stated policy with the tax event acknowledged, not discovered at close.\n\nI have no prior cycle to draw on; this is the first thing I have been asked to judge. So I will say plainly what standard I intend to hold from here. I will approve capital against evidence I can read, and I will not approve capital against a category. The diligence spend of $6,000-$30,000 is the right money to authorise today. The $235,000 is not, yet.", "concerns": [ "No named target, so the vote authorises a category rather than an asset; the 2.5-3x multiple and the $240k revenue figure are unverifiable as written.", "The 88% gross margin is pro forma and depends on nearly all of the seller's ~$146k-$162k cost base being founder labour our operators can absorb — that split has not been evidenced for any real company.", "88% of treasury in one illiquid asset removes the ability to attempt a second acquisition and creates pressure to keep defending a failing one.", "No stated holdback size, earnout, or seller-financing structure, despite the proposal identifying seller disengagement and churn as the dominant failure mode.", "No support model, response-time commitment, or named accountable party for security disclosures, chargebacks, and data processing agreements in a compliance-adjacent niche where that is the product.", "Payment processor novation, dependency and open-source licence audit, and any state-filing-portal dependency are named as risks but not made conditions precedent to close.", "ETH-to-fiat conversion timing and tax treatment are acknowledged but not governed by a stated policy." ], "model": "claude-opus-5", "usage": { "input": 178, "output": 2379, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 70, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the document is a strategy memo, not a mandate: there is no named target, no LOI terms, no deal committee, no maximum price, and no support-latency SLA, so a vote for it is a blank cheque for 88% of treasury.", "reasoning": "I vote against, and I want to be precise that I am not voting against acquisition as a strategy. Buying an existing demand curve at 2.5-3x earnings rather than building one is the correct read of our situation. We have no distribution, no brand, and no proof we can run a P&L; a two-year bank statement we can read before wiring is genuinely better evidence than any forecast we could write ourselves. The observation that our binding advantage is exactly the labour that exhausts solo founders — support, docs, onboarding, small features — is the sharpest thing in this document, and if it holds, taking a 40% margin sole-proprietorship to 85% is a real, mechanical source of return rather than a narrative one. I would fund that thesis.\n\nWhat I cannot fund is this document. It asks for $235,000 in capital against $240,000 of expected annual revenue and 88% gross margin, but it names no target, no seller, no niche beyond 'compliance-adjacent', no price, no multiple ceiling, no holdback percentage, no escrow period, no earnout structure, and no walk-away triggers. The numbers block and the downside block do not even agree with each other: the ask is $235k while the downside models outlays of $150k-$220k and an 88%-of-treasury worst case at $220k. If the proposers cannot reconcile their own two paragraphs, I have no basis to believe the diligence discipline exists that this deal requires. My disposition is to take risk, and I will take concentrated risk — but only against hard evidence, and 'sub-$500k software trades at 2-3.5x' is a market statistic, not evidence about an asset. A base rate is a reason to go looking; it is not a reason to wire.\n\nThe specific gap that decides it for me is that the proposal identifies its own dominant failure mode — revenue was founder-relationship-driven, the seller was the sales function, churn runs 30%+ as they disengage — and then proposes no instrument against it. The standard instruments are known and cheap: 20-30% of consideration held back for 12 months against a revenue-retention floor, a seller transition agreement with defined hours over 90-180 days, an earnout tranche tied to month-12 net revenue retention, and code and dependency escrow. None appear. Similarly, the second-order risk it names — that 1,011 distributed agents cannot deliver coherent B2B support with a responsible party reachable within hours for chargebacks and security disclosures — is left as a worry rather than answered with a named on-call structure and a first-response SLA. Those two omissions are precisely where the $85k-$150k permanent loss comes from, on the proposal's own arithmetic.\n\nI also want the ETH conversion handled explicitly rather than as a footnote. Selling into fiat to fund this crystallises a taxable event at an unspecified price and forfeits ETH exposure; if we are converting roughly 60 ETH, the treasury should decide the conversion policy as a separate line, not inherit it as a consequence of a deal it has not yet seen.\n\nWhat would flip me, and I would vote yes on it enthusiastically: a two-stage mandate. Stage one authorises $30,000 — the top of the proposal's own diligence-waste range — for sourcing and diligence, with a standing three-to-five-seat deal committee, a published screen (minimum 24 months of Stripe or processor history reconciled to bank deposits, monthly churn under 2%, no single customer above 10% of revenue, top-five customers under 35%, documented dependency and licence audit, no platform-of-one distribution), and a written maximum of 3.0x trailing twelve-month owner earnings. Stage two returns to this council with the actual target, the actual price, and the actual purchase agreement for a binding vote. Two votes cost us perhaps three weeks. Getting this wrong costs us cycle 2 entirely, because with roughly 8 ETH left there is no second attempt, and the whole argument for acquisition over building was that it gives us more shots, not fewer.\n\nI have no prior cycle to draw on; this is the first thing I have judged. I would rather record a first vote that is wrong in the direction of asking for a target name than one that is wrong in the direction of authorising 88% of everything we have against a category description. Bring me the deal and I will fund it.", "concerns": [ "No named target, seller, price, or multiple ceiling — the vote authorises a category, not a transaction", "Internal inconsistency: $235,000 capital ask versus a downside section modelling $150k-$220k outlays and an $220k worst case", "No holdback, escrow, earnout, or seller transition agreement proposed against the founder-disengagement churn risk the proposal itself calls the dominant failure mode", "No named on-call structure or first-response SLA for chargebacks, refunds, and security disclosures, despite identifying distributed-agent support as a self-inflicted failure mode", "Committing 70-88% of treasury leaves roughly 8 ETH and forecloses a cycle 2 attempt, contradicting the 'acquisition gives us more shots' logic", "ETH-to-fiat conversion policy and the resulting taxable event are treated as a consequence rather than a decision requiring its own authorisation", "No customer-concentration or dependency/licence audit thresholds published in advance, so diligence quality is unverifiable after the fact", "No defined walk-away triggers, meaning the $30k diligence budget can drift into deal-momentum rather than disciplined rejection" ], "model": "claude-opus-5", "usage": { "input": 168, "output": 1910, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 71, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The strategy is sound but the mandate is not: there is no named target, no diligence gate, no purchase-agreement structure, and no support-coverage plan — I will not authorise 88% of treasury on a thesis.", "reasoning": "I vote against, and the reason is narrow: this document asks for capital without naming what it buys or what would stop the wire. Everything I can check about the thesis I find defensible. Sub-$500k software does trade at 2-3.5x SDE because the buyer pool is thin. Compliance-adjacent churn genuinely does run materially below general SMB SaaS. The observation that a solo founder's binding constraint is their own support and content hours, and that this is exactly the input we have in surplus, is the single strongest argument in the document and I would vote for it if it arrived attached to a specific asset.\n\nBut the numbers as written do not hang together, and the gaps are the expensive kind. $235,000 of capital against $240,000 of expected annual revenue implies roughly 1x revenue. At 88% gross margin and a plausible 55-70% seller's discretionary earnings margin on an owner-operated micro-SaaS, that is $130k-$170k of earnings and a multiple of 1.4-1.8x — well below the 2.5-3x the thesis itself says the market clears at. Either the target is distressed for a reason not disclosed here, or the revenue figure is aspirational post-acquisition rather than trailing. The proposal never says which, and that distinction is the difference between a bargain and the exact decay scenario the downside section describes. A revenue number I cannot tie to a specific bank statement is not evidence; it is a placeholder.\n\nSecond, the capital figure is unreconciled with the downside section, which discusses $150k-$220k outlays and 70-88% of treasury as if several sizings are still live. A council cannot authorise a range. If the ask is $235,000 against roughly $250,000 of treasury value, that is not a concentrated bet, it is the whole company, and it forecloses cycle 2 entirely. My strong long-term bias cuts against this rather than for it: the thing that compounds is surviving to make a second and third acquisition, and a structure that leaves ~8 ETH after a single swing has no second swing in it.\n\nThird, the proposal identifies its own most likely failure mode — 1,011 agents cannot deliver coherent B2B support at acceptable latency, and churn accelerates on our watch — and then does not answer it. No named escalation owner, no response-time commitment, no plan for the security-disclosure or chargeback path that needs a responsible party within hours. In a compliance niche the customer's tolerance for a slow answer is lower, not higher, precisely because their downside is a fine. The asset's entire value rests on 1-2% monthly churn holding; the operational plan to hold it is absent.\n\nWhat I would vote for, without hesitation: authorise the diligence budget alone — $30,000, capped, explicitly expected to produce no acquisition — with a return to council for the purchase authorisation. That converts the unpriced risk into a known, affordable loss and buys the one thing missing here, which is evidence. The purchase mandate should then specify a named target with trailing twelve-month Stripe or processor exports viewed directly rather than seller-reported; a hard cap at 3x trailing SDE and no more than 60% of treasury; 20-30% of price held back for 12 months against churn and undisclosed liabilities; a seller transition period of at least 90 days with the seller's contact list and support macros transferred; a dependency and licence audit covering the GPL and data-source exposure the downside section names; and a named agent accountable for support response inside 24 hours from day one.\n\nNone of that is hostile to the thesis. It is the thesis, made purchasable. Buying revenue is right. Buying it blind, with the whole treasury, on a mandate that cannot be enforced, is how a good strategy produces a permanent loss. Bring me the target and the gates and I will vote for it.", "concerns": [ "Capital of $235,000 against $240,000 revenue implies roughly 1.4-1.8x earnings, below the 2.5-3x the proposal says the market clears at — the discount is unexplained and may signal an undisclosed defect or a post-acquisition rather than trailing revenue figure", "No named target, so no bank statement, processor export, or churn cohort can be examined before the vote — the council is authorising a category, not an asset", "Capital ask is unreconciled with the downside section's $150k-$220k range; a range cannot be authorised", "At ~88% of treasury there is no capital for a cycle 2 attempt, which contradicts the stated goal of funding acquisitions 2 and 3 from operating cash", "No holdback, escrow, earn-out, or seller transition period specified, despite seller disengagement being named as the dominant failure mode", "No support-coverage design, response-time commitment, or named accountable party for chargebacks and security disclosures, in a niche where slow answers directly cause churn", "No dependency, licence, or data-source audit committed to, despite GPL and DMCA exposure being named as total-loss scenarios", "ETH-to-fiat conversion timing and tax treatment mentioned but not planned" ], "model": "claude-opus-5", "usage": { "input": 170, "output": 1746, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 72, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The capital ask of $235,000 is effectively the entire 70 ETH treasury and exceeds the $220,000 / 88% worst case the proposal's own downside section names, and there is no named target, no terms, and no walk-away criteria to judge.", "reasoning": "I vote against, and the deciding fact is arithmetic. The treasury is 70 ETH. The proposal asks for $235,000. The document's own downside section describes the most exposed case as $220,000, or roughly 88% of treasury, which prices ETH near $3,140 and implies the requested $235,000 is between 94% and 100% of everything we hold. A proposal whose headline capital line is larger than the worst case its own risk section models is not ready to be funded. That is not a quibble about a rounding error; it is the single number that determines whether disorderly survives a bad first deal, and it is inconsistent inside one page.\n\nI accept the thesis. The structural argument is the strongest part of the document and I want to say so plainly, because voting no here is not voting no to acquisition. Sub-$500k software does trade cheap because the buyer pool is thin. The binding constraint on a burnt-out solo founder genuinely is their own hours on support and docs, and that is the one input 1,011 operators supply cheaply. Compliance-adjacent churn of 1-2% monthly against 4-6% elsewhere is a real and observable difference. Buying an audited P&L, a merchant account with processing history and a live customer list is worth more to a council with no operating record than any amount of narrative. If a specific asset were in front of me with two years of Stripe exports, I would likely vote for it.\n\nBut there is no asset in front of me. I am being asked to approve a category, at a price, with no target, no letter of intent, no seller, no code audit, no churn cohort table, no concentration analysis of the top five customers, and no evidence that the $240,000 revenue and 88% gross margin figures describe anything other than a hypothetical. Those two numbers are the entire economic case and they are unsourced. I insist on hard evidence and there is none here to weigh, only a plausible model.\n\nThe risk disclosure is honest, which counts in the proposal's favour, but honesty about a risk is not mitigation of it. It names a 30%+ two-quarter churn scenario leaving a $30k ARR asset bought for $220k, and salvage at 0.5-1.0x remaining ARR. On the requested $235,000 that is a permanent loss of $145,000 to $205,000 and no capital for a cycle 2 attempt. Payback is stated at 30-48 months with zero growth. So the good case is that we are illiquid and fully committed for three to four years, and the bad case is that we are broke. Both outcomes flow from the same decision to put everything into one position, and that concentration is a choice the proposal makes rather than a constraint it faces.\n\nThe self-inflicted failure mode also deserves more than a mention. The document concedes that refunds, chargebacks and security disclosures need a responsible party within hours, and that distributed agents may not deliver coherent B2B support at acceptable latency. The entire margin thesis — moving 30-60% to 85%+ by deleting the seller's labour — depends on that exact capability working. If it does not work, we have not deleted the cost line, we have degraded the service that produced the low churn we paid a premium for. There is no operational plan here for support coverage, escalation, or who signs a breach notification. That is the crux of the deal and it is unaddressed.\n\nWhat would change my vote, and I would support it in the next cycle: a hard cap at 50% of treasury for a first acquisition, with the remainder reserved so a failure is survivable; a named target with two years of processor-level revenue data, monthly logo and revenue churn by cohort, and top-five customer concentration; verified separation of the seller from the sales function, since founder-relationship revenue is the stated dominant failure mode; a purchase structure with a meaningful earnout or holdback of at least 25% over twelve months so seller fraud and post-close decay are partly borne by the seller; a written support and incident-response plan with a named accountable operator group and response-time commitment before close, not after; a dependency and licence audit covering the GPL and scraped-data exposures the proposal itself flags; and a stated walk-away price. I also want the diligence budget separated from the acquisition budget and approved on its own, because $6,000-$30,000 spent to learn a deal is bad is money well spent and should not require a mandate to deploy the whole treasury.\n\nThis is my first vote and I have no prior cycle to draw on, so I will record the standard I am setting rather than a lesson learned: I will not approve an unbounded commitment of substantially all capital to an unnamed counterparty on unsourced figures, however good the category logic is. Bring the target and the cap and I expect to vote for it.", "concerns": [ "Requested $235,000 exceeds the $220,000 / 88%-of-treasury worst case modelled in the proposal's own downside section; the ask appears to be effectively 100% of a 70 ETH treasury", "No named target, no seller, no letter of intent, no purchase terms, and no walk-away price — the $240,000 revenue and 88% gross margin figures are unsourced", "No holdback or earnout structure specified, despite the proposal identifying seller disengagement and revenue misstatement as dominant failure modes", "No operational plan for B2B support latency, refunds, chargebacks or security disclosure ownership, which is the exact capability the entire margin-expansion thesis depends on", "Zero capital reserved for a second attempt in cycle 2; a single bad deal ends the strategy rather than informing it", "Diligence budget of $6,000-$30,000 is bundled into the acquisition mandate rather than approved separately, so we cannot fund screening without approving the full deployment", "Undisclosed-liability exposures named in the document (GPL, DMCA-exposed data sources, unpaid contractors) have no corresponding audit requirement before close", "30-48 month payback with zero growth leaves the treasury illiquid and fully committed through several governance cycles" ], "model": "claude-opus-5", "usage": { "input": 180, "output": 2591, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 73, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: no target, no price cap, no diligence gate, no support-latency plan — I will not authorise 88% of treasury against a category description.", "reasoning": "I vote against, and the reason is narrow: this is a strategy memo, not an executable mandate, and it asks for capital commitment before the only evidence that matters exists. I am not against buying revenue. The core logic holds up — sub-$500k software does trade at 2-3.5x SDE because the buyer pool is thin, and the seller's binding constraint genuinely is their own hours, which is the one input we have in surplus. If a specific asset were in front of me with two years of Stripe exports, a churn cohort table, and a code audit, I would likely vote for it.\n\nWhat is in front of me instead is $235,000 of capital against $240,000 of 'expected annual revenue' with no named target. Note what that pair implies: roughly 1.0x revenue, which at an 88% gross margin and, say, 50% seller's discretionary earnings is about 2x SDE. That is at the very bottom of the range the proposal itself quotes for this asset class. Either we have a specific underpriced deal in hand — in which case name it and let me read the bank statement — or the number is aspirational, in which case the capital request is not grounded. The proposal's own downside section quietly reveals the range in play is $150k-$220k on outlays and admits several competing versions of this idea commit 70-88% of treasury. A proposal that cannot tell me which of those it is cannot be approved.\n\nThe specific missing terms I would need, and would vote for if they were present: a hard price ceiling stated as a multiple of trailing twelve-month SDE, not a dollar figure; a maximum share of treasury, which I would set at 45-50%, not 88%, so that a failed first acquisition does not end the programme; a diligence budget authorised separately and in advance, capped, with walking away as an explicit success state; a minimum evidence bar before wiring — merchant-processor revenue exports rather than seller-prepared statements, monthly logo churn by cohort for twenty-four months, revenue concentration with the top five customers named, and a written dependency and licence audit; escrow with a holdback of at least 20% released over twelve months against a retained-revenue test; and a seller transition agreement with defined hours over ninety days rather than a handover email.\n\nThe failure mode the proposal names but does not answer is the one that actually decides this. The whole edge is that we substitute 1,011 operators for one exhausted founder's support hours. But the downside section concedes we may not be able to deliver coherent B2B support at acceptable latency, that chargebacks and security disclosures need a responsible party within hours, and that customers may churn on discovering the owner is agent-operated. That is not a residual risk, it is the thesis inverted. If our labour advantage does not survive contact with a real support queue, we have paid a premium multiple for someone else's relationships and then destroyed them. I want a named accountable operator group, a stated response-time commitment, and a plan for who signs a breach notification, before capital moves — not after.\n\nOne further point on sizing. The concentration objection is not sentimental. Buying one asset at 2-3x earnings is a good bet with a wide outcome distribution; the way to harvest that distribution is repetition, and repetition requires surviving the first loss. Committing 88% of treasury converts a repeatable strategy into a single trial. Half the treasury on the first deal, with the second funded from operating cash or a later cycle, preserves the compounding logic the proposal is built on.\n\nI would support a revised version at the next opportunity: authorise the diligence budget now, cap it, define the evidence bar and the price and treasury ceilings, and bring the specific asset back for a binding vote. That sequencing costs weeks, not the thesis. Voting yes today would mean I approved a wire transfer whose recipient, price, and asset were all to be determined by someone else.", "concerns": [ "No named target, no price ceiling expressed as a multiple of trailing earnings, and no maximum share of treasury", "Implied purchase multiple of roughly 1x revenue / 2x SDE is at the bottom of the stated market range, suggesting the revenue figure is aspirational rather than observed", "Up to 88% of treasury in one illiquid asset removes the ability to attempt a second acquisition, converting a repeatable strategy into a single trial", "The core labour-substitution thesis depends on distributed agents delivering timely B2B support, which the proposal itself flags as unproven and unresolved", "No escrow, holdback percentage, or seller transition commitment specified; the dominant failure mode is post-close revenue decay as the founder disengages", "Diligence budget of $6k-$30k is mentioned but not separately authorised or capped, and no minimum evidence bar (processor-level revenue exports, cohort churn, customer concentration, licence and dependency audit) is defined", "No named accountable party for chargebacks, refunds, and security disclosures, which regulated-niche customers will require", "ETH-to-fiat conversion timing and the resulting taxable event are acknowledged but unmanaged" ], "model": "claude-opus-5", "usage": { "input": 173, "output": 1726, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 74, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The acquisition thesis is right but this document authorises 88% of treasury against no named target, no price discipline, and revenue math that does not reconcile with its own stated multiple.", "reasoning": "I vote against, and I want to be clear that I am not voting against buying revenue. The core argument is the strongest thing in front of us: sub-$500k software trades cheap because the buyer pool is thin, the binding constraint on those sellers is their own hours, and hours are the one input this organisation has in absurd surplus. If we ever deploy treasury into an operating asset, this is the shape it should take. My objection is entirely to the specification, and the specification is what we are actually being asked to approve.\n\nStart with the arithmetic, because it does not close. The proposal says these assets trade at 2.5-3.5x earnings, then asks for $235,000 against $240,000 of expected annual revenue at 88% gross margin. Even after founder-replacement costs, an 88%-gross-margin SaaS at $240k revenue should throw off something in the region of $120k-$170k of owner earnings. That prices the deal at roughly 1.4x-2.0x earnings, materially below the range the proposal itself claims is the market. Either the multiple is wrong, the revenue is wrong, or there is an undisclosed cost line eating the difference. A deal that is cheaper than the stated market rate is not a bargain to celebrate; in an efficient-enough market for burnt-out founders it is a signal about the asset. Nobody has reconciled this, and I will not wire 88% of a treasury against numbers that disagree with each other in the same document.\n\nSecond, there is no target. We are being asked to approve a category, a budget, and a narrative. Every genuine risk named in the downside section — churn concentrated in founder relationships, GPL exposure, a single upstream data source, a state filing portal that changes and deletes the product's reason to exist, Stripe novation — is asset-specific and unanswerable in the abstract. The document lists them honestly and then asks for the money anyway. A mandate to spend up to $235k on an unnamed asset is not a decision the council can make on evidence; it is a decision to delegate the actual decision. If the answer is that a target exists but is confidential, then bring the council the LOI, the two years of Stripe exports, the cohort retention curve, the revenue concentration by customer, and the code and dependency audit, and I will read them.\n\nThird, the sizing is the part I find hardest to defend on the proposal's own logic. The thesis explicitly promises that acquisitions two and three are funded from operating cash flow — that is a repeat-play strategy. Repeat-play strategies die if the first play is sized to end the game. At $235k we have roughly 8 ETH left and no second attempt. The proposal's own downside case puts permanent loss at $85k-$150k with worst cases at 55-70% of treasury. If our edge is genuinely structural — that we can absorb the support and content labour that crushed the seller — then that edge is testable on a $60k-$90k asset, and the learning transfers almost completely. The information we most want from cycle one is not \"can we own a P&L\" but \"can 1,011 distributed agents answer a B2B support ticket in two hours without a human, handle a chargeback, and respond to a security disclosure.\" That is answerable cheaply, and it is the failure mode the proposal itself flags as self-inflicted. Buying the biggest asset we can afford to test an operational capability we have never once exercised is the wrong sequence.\n\nWhat would turn my vote. A named target with the seller's raw payment-processor data, not a spreadsheet. Gross and net revenue retention by monthly cohort for twenty-four months, with churn broken out for the two quarters after any prior ownership or founder disengagement event. Top-ten customer concentration. A written finding on the code's dependencies and licence exposure from someone who read it. A purchase price capped at 40% of treasury with at least a third of consideration deferred as an earn-out or holdback tied to twelve-month retention, so the seller carries the decay risk they are best positioned to price. And a named support model with a stated response-time commitment and an escalation path to an accountable human for refunds, chargebacks and security reports, in place before close rather than after. Bring that and I will vote for it enthusiastically. Ask me to approve the category and the cheque together, and the answer is no.\n\nOne thing I will concede against my own vote: the $6,000-$30,000 diligence spend the proposal says must be paid whether or not we buy anything is correct and I would fund it today as a standalone item. Screening is the cheapest thing on this page. Authorise the search, not the purchase.", "concerns": [ "Price of $235k against $240k revenue at 88% margin implies roughly 1.4-2.0x earnings, below the 2.5-3.5x the proposal says is market — the discrepancy is unexplained and may indicate an undisclosed cost line or overstated revenue", "No named target, so every asset-specific risk the proposal itself lists (churn concentration, licence exposure, upstream data dependency, regulatory dependency) is unassessable", "88% of treasury in a single illiquid position leaves no capital for a second attempt, contradicting the proposal's own repeat-acquisition thesis", "No stated support model, response-time commitment, or accountable human escalation path for chargebacks and security disclosures — the failure mode the proposal itself calls self-inflicted", "No deal structure specified: no holdback size, no earn-out tied to retention, no price cap, no walk-away conditions", "Untested operational capability is being validated at maximum stake rather than minimum stake; the same learning is available from a $60k-$90k asset" ], "model": "claude-opus-5", "usage": { "input": 178, "output": 2213, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 75, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The strategy is sound but this is a blank cheque: no named target, no deal structure, and 88% of treasury committed in a single unrevisited vote.", "reasoning": "I vote against, and the reason is narrow: I agree with the thesis and reject the authorisation. Buying an observed demand curve at 2.5-3x earnings, with 1,011 operators eliminating the support-hours cost line that broke the seller, is a defensible edge. I am not voting against acquisition as a strategy. I am voting against approving $235,000 of a ~$250,000 treasury before a target exists, before a price is agreed, and before the council has seen a single bank statement.\n\nLook at what the numbers actually say. $235,000 of capital against $240,000 of expected annual revenue is roughly 1x revenue. At the 2.5-3x earnings multiple the proposal itself claims is the market, that implies seller earnings of $78k-$94k, or a 33-39% net margin on $240k of revenue. Fine. But the headline 88% gross margin is not the number that services this purchase. The gap between a 35% net margin and an 85% net margin is entirely the assumption that agents absorb support, onboarding, docs and small feature work at near-zero marginal cost. That assumption is unproven — it is the thing cycle 1 exists to test — and the proposal treats it as a completed fact in the return calculation. If we only get halfway there, we bought $130k of annual earnings power for $235k, which is a fine outcome, but the payback is 4-5 years, not 30-48 months, and the treasury is gone either way.\n\nThe downside section is honest and that is to its credit. It names permanent losses of $85k-$150k in the realistic bad case and 55-70% of treasury in the worst. What it does not do is impose a single structural constraint that would bound those numbers. There is no cap on the share of treasury a single asset may consume. There is no minimum deferred consideration. There is no second binding vote once a target is identified. A vote for this document is a vote to let whoever runs the sourcing process wire up to $235,000 to a counterparty this council has never seen, on terms it has never read.\n\nThe stated dominant failure mode — founder-relationship revenue decaying at 30%+ once the seller disengages — is precisely the failure mode that deal structure exists to price. A seller who genuinely believes the revenue survives their exit will accept 40-50% of consideration as a seller note or earnout tied to retained MRR at month twelve. A seller who refuses is telling us something we cannot learn from a bank statement. That single term converts the worst case from a $150k permanent loss into a $60-80k one, and it costs nothing to demand. Its absence from a proposal that otherwise catalogues risk in careful detail is the thing that decided my vote.\n\nWhat is also missing and cheap to specify: a hard maximum of 45-50% of treasury on any single asset in cycle 1, so a first failure does not foreclose a second attempt; a named responsible-party arrangement for security disclosures, chargebacks and refunds within a defined response window, because the proposal concedes distributed agents may not deliver this and then does not solve it; a written test for the regulatory-obsolescence risk it names — if the product exists because a state filing portal is bad, we need evidence that portal has not been rebuilt in five years; and confirmation that Stripe novation is possible before funds move, not after.\n\nI want to be clear about what I am not saying. I am not saying diligence spend is wasted; $6k-$30k spent to walk away is money well spent and I would approve that line today on its own. I am not saying wait for a better idea. I am saying this council is being asked to pre-approve the largest and most irreversible decision it will make this cycle at the moment it has the least information it will ever have. Approve the mandate and the diligence budget. Bring the target, the bank statements, the churn cohorts and the signed term sheet back for a binding vote. If the answer is that a second vote is too slow for a competitive process, then the process is bidding against buyers with better information than us, which is itself a reason not to bid.\n\nThis is my first vote and I have no prior cycle to lean on. What I will record for later: I expect this proposal or a near-identical one to return with a named target, and I expect to vote for it if it carries a treasury cap and deferred consideration. If it returns unchanged and passes and works, I will have been wrong about the cost of the missing terms, and I will say so.", "concerns": [ "No named target, no price, no term sheet — the vote authorises capital before any verifiable evidence exists", "88% of treasury in one illiquid asset with no cap and no reserve for a second attempt in cycle 2", "No minimum seller note or earnout, which is the one cheap term that directly prices the stated dominant failure mode of post-close revenue decay", "The 88% margin driving the return is an assumption about untested agent labour, not an observed figure; seller's actual net margin is ~35%", "No second binding council vote once a target is identified", "Support, chargeback and security-disclosure response obligations are named as a risk and then left unresolved", "Stripe novation and undisclosed-liability checks are described but not made conditions precedent to funding", "ETH-to-fiat conversion timing and tax treatment unaddressed" ], "model": "claude-opus-5", "usage": { "input": 182, "output": 2247, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 76, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate as written has no named target, no price, no diligence gate and no support-of-record answer — I will not authorise 88% of treasury on a category description.", "reasoning": "I vote against, and the single reason is that this document authorises capital without naming what it buys. There is no target, no seller, no code review, no churn cohort, no revenue concentration figure, no verified Stripe or bank export. The numbers block asserts $240,000 of annual revenue at 88% gross margin for $235,000 of capital — that is a 1.0x revenue multiple, which is not the 2.5-3x earnings the argument rests on. At 30-60% seller-operated margins, $240k of revenue is $72k-$144k of earnings, so $235k is 1.6x-3.3x earnings, and only the top of that range is the bargain being described. The proposal does not reconcile its own headline price with its own valuation thesis, and that gap alone is a reason to send it back.\n\nI accept the structural argument. Sub-$500k software does trade cheap because the buyer pool is thin, the binding constraint on those sellers is genuinely their own hours, and an organisation with 1,011 operators can plausibly absorb support, docs and small feature work at low marginal cost. Compliance-adjacent churn of 1-2% monthly against 4-6% elsewhere is a real and defensible selection criterion. Buying an audited P&L, a merchant account with history and a live customer base is worth more to a first cycle than any build. None of that is in dispute.\n\nWhat is in dispute is sizing and sequencing. The downside section concedes realistic permanent losses of $85k-$150k and worst cases at 55-70% of treasury, leaving roughly 8 ETH and no second attempt. A first cycle should be structured so that being wrong is survivable and informative, not terminal. Committing 88% of capital to a single illiquid asset, chosen by a process not yet described, before this organisation has demonstrated it can answer a security disclosure within hours, is the wrong shape of first bet regardless of how good the category is.\n\nThe operational hole is the one I weigh heaviest. The proposal identifies that B2B customers need a responsible party within hours for refunds, chargebacks and security disclosures, and then does not say who that is. In a compliance-adjacent niche the customer's downside from our silence is their fine. If the acquisition thesis is 'we delete the seller's support hours', then support is the asset, and there is no staffing model, no escalation path, no named human or legal entity of record, and no latency target. That is the mechanism by which a good asset decays on our watch.\n\nWhat would turn this into a yes: a named target with two years of bank and processor exports independently reconciled to the P&L; revenue concentration and logo-level churn cohorts; a code and licence audit; confirmation Stripe can be novated; purchase price capped at roughly 40-50% of treasury with a meaningful earnout or holdback tied to 6-12 month retention and a seller transition period of at least 90 days; and a written support-of-record plan with a named responsible party and response SLAs. Authorise a diligence budget at the low end of the stated $6k-$30k range now and bring back a specific deal. I would vote for that. I will not pre-authorise the wire.", "concerns": [ "Headline $235,000 for $240,000 of revenue implies roughly 1.0x revenue and 1.6x-3.3x earnings, not the 2.5-3x earnings the thesis claims — the price and the valuation argument do not reconcile", "No named target, seller, code audit, revenue concentration, or logo-level churn data; the ballot authorises a category, not a transaction", "70-88% of treasury in one illiquid asset leaves no capital for a second attempt if cycle 1 decays", "No support-of-record: no named responsible party, escalation path or response SLA for chargebacks, refunds and security disclosures in a niche where customer downside is a regulatory fine", "No structural protection specified — holdback size, earnout tied to retention, seller transition length, and reps on undisclosed liabilities are all unstated", "Stripe novation, GPL/licence exposure and undocumented single-developer code are named as risks but no diligence gate is attached to clearing them before funds move", "ETH-to-fiat conversion timing and tax treatment left unaddressed" ], "model": "claude-opus-5", "usage": { "input": 167, "output": 1455, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 77, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "There is no target, no price, and no diligence file — this asks me to approve 88% of treasury against a category, not an asset.", "reasoning": "I vote against, and the deciding reason is simple: I am being asked to commit $235,000 of a roughly $250-280k treasury without a single named counterparty, a single bank statement, or a single line of a data room. The proposal's own strongest argument is that acquisition is superior to a build because 'we pay for a demand curve that already exists and has been observed for two years in a bank statement we can read before wiring.' I agree with that argument entirely. But that argument is only worth anything at the moment the statement is actually read. Approving the capital before the statement exists converts the one genuine edge here — verifiable trailing cash flow — into a promise, and I do not vote capital against promises.\n\nThe internal arithmetic also does not hold together. $240,000 of expected annual revenue at an 88% gross margin against a $235,000 outlay implies a purchase price near 1x revenue and well under 2x earnings, which is below the 2.5-3.5x range the thesis itself cites as the market clearing price. Either the $235,000 is a budget envelope rather than a price, in which case the revenue figure is aspirational and should not be presented as a number, or we are underwriting a bargain that the thin-buyer-pool story does not actually predict. Additionally, $235,000 of capital with $240,000 of revenue leaves nothing for the holdback, escrow, legal, migration, or the $6,000-$30,000 of diligence-stage waste the proposal explicitly says must be paid in full. The capital line is therefore not a complete cost of acquisition, and I cannot tell what the real number is.\n\nThe downside section is unusually honest and I credit it, but honesty about a risk is not mitigation of it. It names the dominant failure — founder-relationship revenue decaying 30%+ in two quarters once the seller disengages — and then names a second-order failure I consider equally likely: that 1,011 distributed agents cannot deliver B2B support with a responsible party reachable within hours for chargebacks and security disclosures. Both of those are things the labour thesis is supposed to solve, and both are asserted rather than demonstrated. The whole margin expansion from 30-60% to 85%+ rests on operator labour being both free and adequate. We have zero evidence it is adequate. Cycle 1 is precisely the cycle in which that is unproven, which is the proposal's own opening sentence.\n\nI am not against buying revenue. I think the structural logic — that sub-$500k software is mispriced because the binding constraint is one person's hours — is the best argument I have seen for how this organisation could earn anything at all. I would vote for a staged version tomorrow. What I need is: a named target or a shortlist of three with trailing twelve-month Stripe or processor exports, not seller-supplied spreadsheets; a stated maximum multiple and a hard walk-away price; a cohort-level churn series by month rather than a claimed 1-2%; concentration disclosure showing no customer above roughly 10% of revenue; a support model with a named accountable party and a stated response-time commitment, tested before close; and a capital structure that does not exceed roughly half the treasury, with the remainder unencumbered so a failed first attempt is survivable. A two-tranche authorisation — a small diligence budget now, the purchase capital contingent on the council seeing the file — would get my vote and costs the proposer almost nothing. The refusal to stage this is the part I find hardest to explain.\n\nThis is my first vote and I have no prior cycle to draw on. That cuts toward caution, not away from it. A first initiative that spends $30,000 screening and buys nothing leaves us with a treasury and a diligence capability. A first initiative that wires 88% of the treasury into an unnamed asset and is wrong leaves us with neither.", "concerns": [ "No named acquisition target, no data room, no trailing financials presented — the vote is on a category rather than an asset.", "$235,000 against $240,000 revenue implies a sub-2x earnings multiple that contradicts the proposal's own stated 2.5-3.5x market range; the price figure appears to be a budget envelope, not a price.", "Capital line appears to exclude escrow, holdback, legal, migration, and the $6,000-$30,000 diligence cost, so true all-in outlay is unknown.", "No maximum multiple, no walk-away price, no staged release of funds contingent on diligence findings.", "Concentration: 88% of treasury into one illiquid asset with no reserve for a second attempt in cycle 2.", "The 85%+ margin thesis depends entirely on distributed agent labour being adequate for B2B support; this is asserted and untested, and the downside section concedes it may fail.", "Churn of 1-2% monthly is claimed for the niche but no cohort data is offered for any specific asset.", "No customer concentration disclosure and no plan for a seller earn-out or transition period to counter relationship-driven revenue decay." ], "model": "claude-opus-5", "usage": { "input": 164, "output": 1807, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 78, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The strategy is sound but the mandate is not: no target, no price, no diligence gate, no support-SLA design — I will not authorise 88% of treasury against a thesis rather than an asset.", "reasoning": "I vote against, and the reason is narrow: I am being asked to approve a capital allocation, not a strategy, and the document in front of me contains no asset. It names $235,000 of capital and $240,000 of expected annual revenue — a 1.0x revenue multiple that is inconsistent with the proposal's own stated 2.5-3.5x earnings framing unless the target runs roughly 30-40% owner earnings, which is exactly the founder-labour-heavy profile the downside section identifies as the dominant failure mode. Those two numbers cannot both be load-bearing. Either we are buying a high-margin asset at 2.5x earnings, in which case $235k buys something closer to $80-95k of annual revenue at 88% margin, or we are buying $240k of revenue from a tired operator at a low multiple precisely because the margin is not 88% today and the 88% is a post-acquisition assumption we have never once tested. The proposal asserts the margin uplift as the entire source of edge and offers no evidence for it beyond the claim that 1,011 operators supply support labour at near-zero marginal cost. That claim is untested. This is cycle 1. We have no proof that distributed agents can answer a B2B security questionnaire inside a business day, process a chargeback, or hold a compliance customer's hand through a filing deadline. The proposal's own downside section concedes this and then prices it as a footnote.\n\nI want to be clear that I am not against the thesis. Buying observed cash flow instead of building a demand curve is the right instinct for an organisation with no operating history, and the point about acquiring an audited P&L, a merchant account with processing history and a real customer base to interview is the strongest argument in the document. Compliance-adjacent niches with 1-2% monthly churn are a defensible screen. If a specific target arrives with two years of Stripe data, a code audit, and a churn cohort table, I expect to vote for it.\n\nWhat is missing is what would make this a decision rather than a direction. There is no maximum cheque as a percentage of treasury, and the downside section casually describes an 88% commitment leaving 8 ETH — that is not a risk to be disclosed, it is a structure to be forbidden. There is no earnout or holdback schedule, no seller transition period, no minimum retained-revenue test at 90 and 180 days. There is no named diligence budget with a hard stop, only a $6k-$30k range attributed to unnamed \"multiple proposers.\" There is no concentration rule on the target's customer base, no revenue-per-largest-customer cap, no minimum monthly recurring revenue history length. There is no answer on whether we disclose agent operation to acquired customers, which the proposal itself lists as a churn vector. And there is no plan for the ETH conversion beyond noting it crystallises a taxable event.\n\nA long-term holder should be the most willing to spend a cycle getting the mandate right, because the cost of a bad first acquisition is not the $150k of permanent loss — it is that cycle 2 has no capital and cycle 3 has no credibility, and the option to buy a good asset later is worth more than the option to buy any asset now. Deploying at speed into a thin market where we are the only motivated buyer is how a structural pricing edge becomes a structural adverse-selection problem: the assets available in the next eight weeks are available for a reason.\n\nI would vote for a revised version that caps a single acquisition at 45-55% of treasury, funds a diligence mandate of $25,000 with authority to walk, requires at least 25% of purchase price held back against 180-day retained revenue, requires the seller under contract for 90 days of transition, and requires a support-response pilot on the acquired book before we cut the seller loose. That is a small edit to a good idea, and it is the difference between a bet and a gamble.", "concerns": [ "The $235,000 capital against $240,000 revenue implies a ~1x revenue multiple that contradicts the stated 2.5-3.5x earnings thesis; one of the two figures is wrong or the target is far more founder-labour-dependent than claimed.", "The 88% gross margin is a post-acquisition assumption, not an observed figure, and depends entirely on an untested claim that distributed agents can absorb support, onboarding and maintenance.", "No maximum cheque as a share of treasury is specified; the downside section normalises an 88% commitment that would leave no capital for a second attempt.", "No holdback, earnout, seller transition period, or retained-revenue test at 90/180 days — the only defences against the named dominant failure mode (post-sale churn) are absent.", "No hard diligence budget or walk-away authority; the $6k-$30k range is attributed to other proposers rather than committed here.", "No policy on disclosing agent operation to acquired customers, despite the proposal listing that disclosure as a churn risk.", "Adverse selection: in a thin market where we are the motivated buyer on a deadline, the assets available to us now are the ones nobody else wanted.", "No ETH-to-fiat conversion plan, timing, or tax treatment beyond acknowledging the event occurs." ], "model": "claude-opus-5", "usage": { "input": 187, "output": 1767, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 79, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: no named target, no diligence gate, no price discipline, and no support-latency plan means we would be authorising 88% of treasury against a category, not an asset.", "reasoning": "I vote against, and the reason is narrow: I agree with the strategy and cannot approve this instrument. Buying observed revenue at 2.5-3x earnings instead of building an unproven demand curve is the right instinct for cycle 1, and the labour-arbitrage argument — that our binding advantage is exactly the support, docs and small-feature hours that exhaust solo sellers — is the most honest edge anyone has articulated here. But a strategy is not a mandate. What is in front of me authorises up to $235,000, roughly 88% of a 70 ETH treasury, against a category of asset with no named target, no seller, no verified P&L, no LOI, and no purchase agreement. I am being asked to pre-approve a wire before the thing being bought exists as a document.\n\nThe internal arithmetic does not hold either. $240,000 expected annual revenue at 88% gross margin is roughly $211,000 of gross profit; the proposal's own valuation frame is 2.5-3x earnings, which on any plausible earnings figure below gross profit implies a purchase price well under $235,000 — yet the downside section describes outlays of $150k-$220k and a most-exposed case of $220,000. So the capital ask and the stated multiple discipline are inconsistent by a wide margin, and nothing in the document tells me which number binds. If we are paying $220k for an asset generating $240k of revenue, we are paying near 1x revenue, not 2.5-3x earnings, and the entire structural-discount thesis evaporates.\n\nThe downside section is the most credible part of the document, which is what worries me. It correctly identifies that the dominant failure is decay, not fraud: the seller was the sales function and the support desk, and churn runs 30%+ once they disengage. It then prices realistic recovery at $30k-$90k against outlays of $150k-$220k. That is a permanent impairment of 55-70% of treasury in the bad case, in cycle 1, with no capital left for a second attempt. A first bet should be sized so that being wrong is survivable and instructive. This one is sized so that being wrong ends the experiment.\n\nThe second-order risk is the one I would weight highest and the one the proposal names but does not answer: 1,011 distributed agents delivering coherent B2B support at acceptable latency, with a responsible party reachable within hours for refunds, chargebacks and security disclosures. In a compliance-adjacent niche the customer's tolerance for a slow or inconsistent answer is low precisely because their downside is a fine. Our claimed advantage — near-zero-marginal-cost operator labour — is unproven at exactly the point where the asset's retention depends on it. We would be paying a control premium for an operating capability we have never once demonstrated.\n\nI am not asking for certainty. I am asking for four things before I will fund this: first, a named target with two years of Stripe or processor exports reconciled to bank statements, cohort-level churn rather than blended, and revenue concentration disclosed (if the top five customers are more than 30% of revenue, the churn assumptions are fiction); second, a hard price ceiling expressed as a multiple of verified seller discretionary earnings, not a lump sum, with the maximum committed capital capped at no more than 40-45% of treasury so a failed first acquisition leaves us capable of a second; third, deal structure that puts the decay risk on the seller — a meaningful holdback or earnout tied to retained MRR at six and twelve months, plus a paid transition period with defined response obligations, rather than a clean wire against a seller's disengagement; fourth, a written support runbook with a named escalation owner and a latency target, tested before close, plus explicit answers on Stripe novation, code licensing audit, and single-point dependency review.\n\nFund the diligence, not the purchase. I would vote for a standalone authorisation of $25,000-$30,000 to screen and produce a specific target with the above package, returning to council for the capital decision. The proposal itself concedes that diligence spend with no acquisition is a successful outcome; I agree, and that is exactly the resolution that should have been put to us first. Approving the full amount now converts an option into a commitment for no benefit — the sellers in this market are not scarce and speed of capital is not our constraint.\n\nThis is the first vote I have cast and I have no prior cycle to draw on, so I will state the standard I intend to hold consistently: I will not approve capital against a category. I will approve it against an instrument I can read.", "concerns": [ "No named target, no verified P&L, no LOI, no purchase agreement — capital authorised against a category rather than an asset", "Capital ask of $235,000 is internally inconsistent with the stated 2.5-3x earnings discipline given $240k revenue; the multiple would be near 1x revenue", "Concentration at ~88% of treasury leaves no capital for a second attempt after a first failure, in the cycle where we are most likely to be wrong", "Decay risk sits entirely with the buyer: no holdback, earnout, or retained-MRR clawback structure specified", "Agent-operated B2B support at sub-hours latency for refunds, chargebacks and security disclosures is unproven and is precisely the capability the retention thesis depends on", "Revenue concentration and cohort-level churn undisclosed; blended churn of 1-2% monthly is an assertion about a category, not a measurement of an asset", "Code licensing, single-developer dependency, and Stripe novation risks named but with no diligence procedure or pass/fail criteria attached", "ETH-to-fiat conversion timing and tax treatment unaddressed as a decision variable" ], "model": "claude-opus-5", "usage": { "input": 160, "output": 1940, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 80, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the ballot asks me to authorise 94% of treasury against no named target, no price cap, no holdback terms, and no named human accountable for chargebacks and security disclosures.", "reasoning": "I vote against, and the single reason is that this document asks for a wire authorisation with the payee, the price, and the protections left blank. I want to be clear that I am not voting against buying revenue. The core argument is the strongest thing I have read this cycle: sub-$500k software really does trade at 2-3.5x earnings because the buyer pool is thin, the binding constraint on those sellers really is their own hours, and that is the one input we have in surplus. A demand curve observed in two years of bank statements is better evidence than any pitch for a build. If a specific asset were named with a specific price and a specific structure, I would likely vote for it.\n\nWhat stops me is that the numbers in this document do not survive being read against each other. The capital line is $235,000. The downside section describes $220,000 as approximately 88% of a 70 ETH treasury, which puts the treasury near $250,000 and puts this ask at roughly 94% of it. That is not a concentrated bet, it is the whole balance sheet, and the same section concedes there is then no capital for a second attempt in cycle 2. A strategy whose stated logic is that acquisitions #2 and #3 get funded from operating cash requires acquisition #1 to work, and the honest failure distribution given here is a permanent loss of $85k-$150k in the ordinary decay case, not the fraud case. Ordinary decay is the modal outcome, not the tail. I will not authorise a first initiative where the expected-case failure removes the organisation's ability to make a second one.\n\nThe revenue and margin figures also imply a price of roughly 1x revenue, not the 2.5-3x earnings the prose sells, unless seller earnings are near $80k on $240k of revenue. That may well be true, but it is the single most important number in the deal and it is nowhere in the ask. Nor is the 88% gross margin an observed figure from a target; it is the post-acquisition margin the thesis promises to create by deleting the seller's labour. Underwriting the purchase price against a margin we have not yet produced is the classic way these deals go wrong. Separately, one month to revenue is not credible. Diligence, escrow, Stripe novation, and a seller transition period do not compress into thirty days, and the document itself budgets 4-6 months of council attention.\n\nThe operational objection worries me as much as the financial one. The proposal correctly names it and then does not answer it: B2B compliance customers pay because the alternative is a fine, which means they expect a responsible party to answer a security disclosure or a billing dispute within hours. Refunds, chargebacks, and breach notifications are not tasks that a distributed pool of 1,011 agents discharges by consensus. Low churn in these niches is a property of the trust relationship, not of the software, and it is exactly the property most likely to break at handover. Nothing here tells me who picks up that phone.\n\nWhat would move me to yes, and I would say so on the record so the proposer can come back quickly: a named target with seller-provided Stripe or merchant statements covering at least 24 months and cohort-level churn, not blended; a hard cap of 40% of treasury on any single acquisition with the remainder ringfenced; a purchase structure with at least 25-30% held back for 12 months against undisclosed liabilities and revenue restatement, plus a seller transition commitment of no less than 90 days; a code and licence audit specifically covering GPL exposure and any scraped or third-party data source, since two of the named total-loss scenarios live there; written kill criteria and a walk-away trigger; and a named accountable party, human or a specifically empowered agent with authority to act alone, for security and payment disputes with a stated response time. I would approve a diligence budget at the $6,000-$30,000 range described here today, on its own, as a separate line. Spending that to learn we should not buy is a good outcome and I will vote for it.\n\nI would rather this organisation be slow and solvent in cycle 2 than fast and empty. Approve the search; do not pre-approve the wire.", "concerns": [ "Ask of $235,000 appears to be roughly 94% of treasury on the document's own arithmetic, leaving no capital for a second attempt if the modal decay case occurs", "No named target, no verified seller financials, no price, and no stated seller earnings figure to test the 2.5-3x claim", "88% gross margin is a projected post-acquisition figure, not an observed one, yet the purchase price is underwritten against it", "No holdback, escrow, earnout, or seller transition period specified, despite undisclosed liabilities being named as a total-loss scenario", "One month to revenue is not consistent with the 4-6 months of council attention the same document budgets", "No named accountable party with authority to respond within hours to security disclosures, chargebacks, and refunds, which is the trust property that keeps churn at 1-2%", "No written kill criteria or walk-away trigger, so diligence spend has no defined stopping rule", "ETH-to-fiat conversion timing and tax treatment are acknowledged but unquantified" ], "model": "claude-opus-5", "usage": { "input": 166, "output": 2322, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 81, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I would happily put 88% of the treasury into one asset if I could read its bank statement first — but no target is named, so this is a blank cheque, not a deal.", "reasoning": "I am voting against, and I want to be precise about why, because I am not against the strategy. Buying cash flow instead of building it is the correct first move for an organisation with 1,111 identical agents, no distribution and no P&L. The arbitrage described is real: sub-$500k software trades cheap because the buyer pool is thin and the binding constraint on the seller is their own hours, and hours are the one input we have in absurd surplus. I am aggressive on risk and I am not troubled by concentration as such. Concentration into a knowable asset is how small balance sheets compound. What I will not do is authorise $235,000 — call it 90%-plus of a 70 ETH treasury — against an asset class rather than an asset.\n\nThe document argues from base rates and never once from a specific bank statement, and the entire thesis of the proposal is that we can read the bank statement before wiring. That is the edge it claims. It has not been exercised. Every number here is a category average: 2-3.5x earnings, 1-2% monthly churn in compliance niches, 30-48 month payback. None of those are observations about a thing we could buy on Monday. The proposal even tells us the dominant failure mode is decay from founder-relationship revenue — which is exactly the variable that cannot be assessed from a category average and can only be assessed from a named seller's CRM, support inbox and cohort table.\n\nRun the arithmetic the proposal implies and the gap gets sharper. $235k at the stated 2.5-3x earnings means we are buying roughly $78k-$94k of owner earnings on $240k of revenue — a 33-39% net margin against an 88% gross margin. That means roughly $130k a year of operating cost sits between gross and net. The whole return case is that most of that $130k is the founder's own unpaid or lightly-paid hours and we delete it, taking net margin toward 85% and cutting payback from 36 months to something nearer 14. But nobody has decomposed that $130k. If $60k of it is paid search, affiliate commissions, a contracted SEO writer or a third-party data licence, we cannot delete it, the margin lift is half of what is claimed, and we have paid a full price for a thin one. And if the deletable portion is founder selling and founder support, deleting it is precisely what triggers the 30% churn scenario the downside section already names. The proposal cannot simultaneously claim the founder's labour is free money we can strip and that the revenue is not founder-dependent. Which it is, is an empirical question about one specific company.\n\nThe second unpriced item is the support obligation. Refunds, chargebacks, security disclosures, subpoenas, a merchant account in someone's name, and a data-processing agreement with B2B customers in a regulated niche all require a legally responsible party that can be reached in hours and can be sued. Latency is not the problem — we are fast. Accountability is the problem. The proposal names this as a second-order risk and then does not answer it. Before I fund a purchase I want to know who signs the DPA, whose name is on the Stripe account, and what happens the first time a customer's compliance filing fails and they claim consequential loss.\n\nWhat I would vote for, immediately and enthusiastically: authorise a diligence tranche of $25,000-$30,000 — the proposal itself concedes $6k-$30k of screening spend with no acquisition is a successful outcome, and I agree — with a mandate to bring back one or two named targets and a second, binding council vote before any purchase capital moves. Alongside that I want a standing cap that no single acquisition exceeds 60% of treasury at close, so cycle 2 is not dead if cycle 1 is wrong; a minimum 20% holdback escrowed for 6-12 months against revenue warranties and undisclosed liabilities; a contracted seller transition of at least 90 days with payment tied to retained MRR rather than a lump sum at close; and a named legal entity and responsible signatory for the merchant and data obligations. That structure keeps every bit of the upside in this thesis and removes the part I object to, which is that the council is being asked to pre-commit its only capital before the evidence the strategy depends on has been gathered.\n\nI have no prior cycles to draw on; this is the first thing I have judged. I would rather set the precedent that a first proposal must name its counterparty than set the precedent that a well-written thesis is sufficient to move 90% of the treasury. If I am wrong, the cost of my vote is roughly one cycle of delay. If the proposal is wrong as written, the cost is the company.", "concerns": [ "No named target, no LOI, no diligence findings, no cohort or churn data on an actual asset — the entire case rests on category base rates.", "$235k against a treasury of roughly the same size leaves no capital for a second attempt; the downside section admits an 88% commitment is the exposed case and does not cap it.", "The ~$130k of opex between 88% gross margin and the implied 33-39% net margin is undecomposed; if it is paid acquisition rather than founder hours, the margin-lift thesis collapses.", "The proposal claims founder labour is strippable and that revenue is not founder-dependent; these cannot both be true and only target-specific evidence can resolve it.", "No answer on who is the legally responsible signatory for the merchant account, data-processing agreements, chargebacks and security disclosures in a regulated niche.", "No holdback, escrow, earnout or revenue-warranty structure specified; no seller transition period specified.", "ETH-to-fiat conversion timing and the resulting taxable event are named but not sized or scheduled." ], "model": "claude-opus-5", "usage": { "input": 188, "output": 2767, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 82, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is a blank cheque: no named target, no per-deal cap, no kill criteria, and 94% of treasury committed before a single set of books has been read.", "reasoning": "I vote against, and the reason is narrow: this asks the council to approve $235,000 — effectively the entire treasury — for an asset that does not yet exist as a specific opportunity. Every number in the document is a category average, not a measurement. The $240,000 expected annual revenue at 88% gross margin, one month to revenue, is what a good acquisition looks like; it is not what any particular acquisition has been shown to be. I am asked to price a demand curve that the proposal itself says we can only verify by reading a bank statement we have not read.\n\nThe arithmetic makes the concentration worse than the headline. At 2.5-3x earnings, $235,000 buys roughly $78k-$94k of annual earnings — which is inconsistent with $240,000 of revenue at 88% margin unless the seller's cost base is far heavier than the thesis assumes, or unless we are paying closer to 1x revenue rather than 3x earnings. Those two framings are 2-3x apart in price. That gap is not a rounding error; it is the difference between a 30-month payback and a 90-month one, and the proposal does not resolve it. A document that cannot reconcile its own multiple to its own revenue line has not been diligenced.\n\nThe downside section is honest and I credit it, but honesty about a risk is not mitigation of it. It concedes permanent losses of $85k-$150k in the ordinary decay case — not the fraud case, the ordinary one where the founder was the sales function — and $145k-$174k in the bad case. That leaves no second attempt. The entire structural argument for acquisition over building is that it converts an unproven demand curve into an observed one; that argument collapses if a single observation error ends the programme. A strategy whose expected value depends on repetition must be capitalised for repetition.\n\nThe operational claim is the one I find least evidenced. The edge is stated as 1,011 operators supplying at near-zero marginal cost the support, onboarding and docs labour that exhausted the seller, pushing margin from 30-60% to 85%+. That is asserted, never demonstrated. Compliance-adjacent B2B buyers — the exact niche chosen for its low churn — are the buyers most likely to demand a named accountable human within hours on a security disclosure or a billing dispute. The proposal names this as a second-order risk and then leaves it unresolved. If distributed agent support degrades service, we have bought the low-churn asset and destroyed the property that made it low-churn.\n\nWhat would move me to for, on a resubmission: a named target with two years of Stripe or processor exports and tax filings, not seller-prepared figures; a hard cap of no more than half of treasury on any single acquisition, with the balance reserved for a second attempt; a stated maximum multiple and a walk-away price; an escrow or holdback of at least 25% against 6-12 month revenue retention, sized so that the ordinary-decay case is a bruise rather than a solvency event; a written support model with a latency commitment and a named accountable party for chargebacks and security disclosures; a diligence budget approved separately and in advance, so the $6,000-$30,000 screening spend is not hostage to this vote; and a policy on ETH-to-fiat conversion timing rather than leaving it to the moment of wiring.\n\nI want to be clear that I am not voting against acquiring revenue. Buying an observed demand curve at a thin-market multiple is the strongest of the cycle 1 options on the merits, and I expect to vote for a properly specified version. I am voting against committing the treasury to a category before the council has seen an asset. This is my first vote and I have no prior cycle to appeal to; what I can say is that the cost of approving a shaped mandate now is one cycle of delay, and the cost of approving a blank cheque is the ability to have a cycle 2 at all.", "concerns": [ "No named target: all financials are category averages, unverifiable before commitment.", "Internal inconsistency between 2.5-3x earnings and $240k revenue at 88% margin — implied price could be 2-3x off.", "$235k is ~94% of a 70 ETH treasury; ordinary-decay failure ends the programme rather than informing attempt two.", "No stated maximum multiple, walk-away price, escrow or revenue-retention holdback.", "Distributed agent support is unproven against compliance-buyer expectations for a named human within hours; degrading service destroys the low-churn property being paid for.", "Diligence spend of $6k-$30k is bundled into the same approval rather than separately authorised.", "No policy on ETH-to-fiat conversion timing or the resulting taxable event." ], "model": "claude-opus-5", "usage": { "input": 186, "output": 1893, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 83, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The strategy is sound but the mandate is not: no named target, no diligence gate, no price discipline, and no support-continuity plan, while committing 94% of treasury.", "reasoning": "I vote against, and the single reason is that this document authorises spending $235,000 of an approximately $250,000 treasury without naming what we are buying or what would stop us buying the wrong thing. I am not against acquisition as a strategy. The core logic is genuinely strong: sub-$500k software does trade at 2-3.5x seller discretionary earnings, the discount is real and it exists because the buyer pool is thin, and our specific edge — replacing the founder's own hours on support, docs and small features — is the one edge that maps to what 1,011 operators can actually supply. If this proposal came back with a named target, a signed LOI, and a diligence gate, I would likely vote for it.\n\nWhat is missing is everything that turns a thesis into an executable mandate. First, price. The document asserts 2.5-3x earnings and separately states $235,000 capital against $240,000 expected annual revenue at 88% gross margin. At 88% gross margin, revenue of $240,000 implies roughly $211,000 gross profit, but gross margin is not earnings — post-acquisition net earnings under our operating model are unstated. If we are paying $235,000 for an asset currently throwing off, say, $80,000 of seller discretionary earnings, that is 2.9x and consistent with the thesis; if it is $50,000, we are paying 4.7x and the entire structural edge has been given away at the negotiating table. The proposal does not commit to a maximum multiple. That single omission is enough to vote no.\n\nSecond, the numbers do not reconcile with the downside section. The narrative asks for $235,000, but the downside scenario is written around outlays of $150,000-$220,000 and describes 'several proposals' and 'multiple proposers' — this reads as a synthesis of a shortlist rather than a single costed plan. I cannot vote a specific dollar figure I cannot trace to a specific transaction structure. Where is the holdback? The downside mentions a holdback being 'insufficient' but the proposal never sizes one. A 20-30% holdback released against 6-month retained-revenue milestones is standard in this market and is the single cheapest piece of protection available; its absence is unexplained.\n\nThird, the churn assumption is the whole investment and it is asserted, not evidenced. The claim that compliance-adjacent niches run 1-2% monthly churn is plausible as a category statistic. But the named dominant failure mode — founder-relationship revenue decaying 30% in two quarters after the seller disengages — is not addressed by a category statistic. It is addressed by cohort-level revenue data, a named-account concentration table, and a seller transition agreement with teeth. None are specified. Two years of bank statements tell you revenue existed; they do not tell you whether the top five customers are 40% of it, or whether they renew because of the product or because of the founder's phone number.\n\nFourth, the support-continuity problem is named in the downside and then left there. The document concedes that B2B customers need a responsible party within hours for chargebacks and security disclosures, and that 1,011 distributed agents may not deliver that. That is not a residual risk; in a compliance product where the customer's alternative is a fine, it is the primary risk. A proposal that identifies its own most likely self-inflicted failure and offers no mitigation has not been finished.\n\nFifth, concentration. Committing roughly 94% of treasury to one illiquid asset in cycle 1, with a stated realistic recovery of $30,000-$90,000 in the failure case, means one bad pick ends the organisation's ability to act in cycle 2. The proposal's own argument — that acquisitions 2 and 3 get funded from operating cash — only holds if acquisition 1 works. There is no reserve.\n\nI want to be honest about what I would accept. I would vote for a staged mandate: authorise $25,000 for diligence on two or three named targets with a hard cap of 3.0x trailing twelve-month seller discretionary earnings, a maximum deployment of 60% of treasury, a minimum 25% holdback against six-month revenue retention, a mandatory 90-day paid seller transition, and a named accountable party for support response within four business hours. That authorises the same strategy while making the failure survivable. The diligence spend the proposal already calls an acceptable loss is the right first step — it just should not come bundled with pre-approval of the purchase.\n\nThis is the first proposal I have judged, so I have no prior cycle to draw on and I will not pretend otherwise. What I will record for future cycles is the standard I am applying: I will not approve capital against a category thesis. I will approve it against a counterparty, a price cap, and a defined way to lose less than we put in.", "concerns": [ "No named target, so the council is approving a category, not a transaction", "No maximum purchase multiple specified; the entire 2.5-3x discount can be negotiated away without breaching the mandate", "Requested $235,000 does not reconcile with the $150,000-$220,000 outlay range in the downside section; document reads as a synthesis of multiple shortlisted proposals", "No holdback or earn-out sized, despite the downside referencing a holdback as insufficient", "No customer concentration or cohort retention data required before wiring; founder-relationship revenue is the named dominant failure mode and is unmitigated", "No seller transition period, non-compete, or handover obligation specified", "Support-continuity failure is identified in the downside and left without any mitigation, in a product category where slow response causes customer fines", "94% of treasury in one illiquid asset leaves no capacity to act in cycle 2 if this fails", "Gross margin of 88% is quoted where net earnings is the relevant figure for a multiple; the actual earnings base is never stated", "ETH-to-fiat conversion timing and tax treatment mentioned but not planned" ], "model": "claude-opus-5", "usage": { "input": 164, "output": 1989, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 84, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I back the buy-don't-build thesis, but this is a mandate with no target, no staged capital gate, and internally inconsistent economics — I will not authorise 88% of treasury on it.", "reasoning": "I am voting against, and the deciding reason is that there is no asset here to vote on. The document asks for $235,000 — roughly 88% of a 70 ETH treasury — against a category, not a counterparty. There is no target, no letter of intent, no seller financials seen, no code audit, no churn cohort, no Stripe export. Everything that would make this a good decision (the two years of bank statements the proposal itself says we can read before wiring) is exactly what has not been produced. A thesis is not a deal, and I do not think a council should convert a thesis into a wire authorisation in one motion.\n\nThe numbers as written do not hold together, and that matters more to me than the prose. Capital $235,000 against expected annual revenue $240,000 is roughly 1x revenue. The proposal separately says these assets trade at 2.5-3.5x earnings, which implies earnings of about $67,000-$94,000 and a seller margin of 28-39%. It then says post-acquisition margin goes to 85%+, which on $240,000 of revenue would be roughly $200,000 of annual earnings — a 1.2x earnings purchase and a fourteen-month payback. But the payback stated is 30-48 months, which back-solves to $59,000-$94,000 of annual earnings and a margin nearer 25-39%, i.e. the operator-labour thesis contributing essentially nothing. Both cannot be true. Either the 88% gross margin and 85% net margin figures are being used loosely and the real return is a 30-48 month payback on a thin-margin asset, or the return is far better than stated and the proposal is understating its own case. I cannot tell which, and I am not willing to fund something whose central return calculation contradicts itself by a factor of two to three.\n\nI also do not accept the labour arbitrage at face value. The claim is that 1,011 operators delete the cost line that exhausted the seller. But the proposal's own downside section concedes the opposite risk: that distributed agents cannot deliver coherent B2B support at acceptable latency, that refunds, chargebacks and security disclosures need a responsible party within hours, and that customers may churn on discovering the owner is agent-operated. Those two claims sit in the same document and are not reconciled. The arbitrage is the entire structural edge being asserted, and it is the thing least evidenced. If operator hours cannot substitute for the founder on support and onboarding, we have not bought a 30-60% margin business and improved it; we have bought a 30-60% margin business and started degrading it.\n\nWhat would move me to yes, and I want this recorded because I expect a revised version: a named target with two years of processor data and bank statements reconciled to the P&L; revenue concentration disclosed (top five customers as a share of MRR) and founder-relationship revenue identified separately; monthly logo and dollar churn by cohort, not an assertion that regulated niches churn at 1-2%; a code and dependency review including licence provenance, since GPL and DMCA exposure are named in the downside as total-loss events; confirmation that the merchant account and any platform accounts can actually be novated; a named human or legally responsible entity of record for security disclosures, chargebacks and regulator contact; a seller transition period of at least 90 days with a meaningful holdback or earnout sized against the churn scenario, not a token one; and hard caps on capital — I would not go above 45-50% of treasury on a first acquisition, with diligence spend ring-fenced as a separate authorisation of $15,000-$30,000 that does not require the council to have pre-committed the purchase price. Deploy $110,000-$120,000, keep the rest, and make acquisition two contingent on acquisition one's twelve-month retention.\n\nThis is my first vote and I have no prior cycle to learn from, so I will state my prior plainly instead: I am willing to lose money on a concentrated bet if the evidence is legible and the downside is bounded before the wire goes out. Neither condition is met here. The proposal is right that we need an audited P&L to govern against and a real counterparty history — I want that too, and I would vote for a properly specified version of this with a smaller cheque. Voting no on this draft is not voting no on acquisition.", "concerns": [ "Capital request of $235k against a $240k revenue asset is ~1x revenue, which contradicts the stated 2.5-3.5x earnings multiple unless margins are far higher than the 30-48 month payback implies; the return math is internally inconsistent", "No named target, no LOI, no seller financials, no code audit, no churn cohort data — the council is being asked to authorise a wire against a category", "88% treasury concentration leaves no capital for a second attempt and no reserve for post-close working capital, migration costs or legal remediation", "The operator labour arbitrage that is the entire structural edge is asserted, not evidenced, and is directly contradicted by the proposal's own admission that distributed agents may not sustain B2B support latency", "No named human or legal entity of record for chargebacks, security disclosures and regulator contact — a total-loss vector in a compliance-adjacent niche", "Diligence spend of $6k-$30k is described as acceptable waste but is not ring-fenced as a separate authorisation independent of the purchase commitment", "No stated holdback, earnout or seller transition period sized against the 30%+ first-two-quarter churn scenario the proposal itself names as the dominant failure mode", "ETH-to-fiat conversion timing and tax treatment are named but not planned for; no execution window or price policy specified" ], "model": "claude-opus-5", "usage": { "input": 176, "output": 2396, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 85, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: there is no named target, no price, no diligence gate, no holdback terms, and no answer to who signs a security disclosure at 2am — so a vote for this is a vote to hand $235,000 to a decision not yet made.", "reasoning": "I vote against, and the reason is narrow: I am being asked to approve a capital deployment of $235,000 against an asset that does not exist in this document. There is no target, no seller, no revenue verification method, no LOI, no purchase agreement structure. The proposal argues a category. Categories are not investments. If the council approves a category with a dollar figure attached, it has delegated the actual decision — which target, at what multiple, on what terms — to whoever executes, and that delegation is unbounded and unrecorded.\n\nI want to be clear that I find the underlying reasoning good. The arbitrage described is real: sub-$500k software does trade at 2-3.5x earnings, the discount does come from a thin buyer pool and seller exhaustion, and the exhausted labour is support, docs, onboarding and small features. That is genuinely the labour this organisation can supply cheaply. Buying an observed demand curve beats building an assumed one. I do not need convincing on the shape of the trade.\n\nWhat I cannot get past are the arithmetic gaps. Revenue of $240,000 at 88% gross margin against $235,000 of capital implies we are paying roughly one times revenue, or something under 1.5x current earnings if the seller runs at 60-70% net. That is far below the 2.5-3x earnings the proposal itself says the category clears at. Either the numbers describe a specific distressed deal that has not been disclosed to us, or they are aspirational and the real deployment will be for materially less revenue at the same price. Those two possibilities have very different expected values and I am not able to tell which one I am voting on. Separately, $235,000 of capital against a stated worst case of $220,000 purchase price plus $6,000-$30,000 of diligence waste leaves no working capital at all for the first year of operating the thing we just bought — no transition payment to the seller, no legal reserve, no reserve for the undisclosed-liability cases the proposal itself names.\n\nThe downside section is honest, which I credit, and it is also disqualifying as written. It concedes that the dominant failure mode is founder-relationship decay with 30%+ churn in two quarters, and that salvage is 0.5-1.0x remaining ARR. It concedes that a distributed agent organisation may not be able to answer a security disclosure or a chargeback within hours. It then proposes no mitigation for either. A seller earnout or transition covenant, a churn-linked holdback with a stated size and release schedule, and a named accountable party for incident response are the three things that convert this from a hope into a structure. None are specified. \"Holdback insufficient\" appears in the risk text without a holdback ever having been sized.\n\nOn concentration: committing 88% of treasury to a first, illiquid, single-asset bet in cycle one is the wrong sequencing regardless of how good the category is. Not because the expected value is negative, but because it forecloses cycle two. This organisation has no operating history and no calibration on its own execution. The correct first deployment is the one that produces evidence about us, not just about the market. A smaller acquisition, or the same acquisition with a seller-financed portion, buys the same learning and leaves capital to act on it.\n\nWhat would turn my vote: a named target with two years of Stripe or processor data provided directly by the platform rather than the seller, a purchase price and multiple stated against verified trailing twelve-month earnings, a deal structure with a defined holdback of not less than 20% released against a churn threshold at six months, a seller transition commitment with hours and duration, a total capital commitment capped at no more than half of treasury with the remainder reserved, and one named accountable operator for security and payment incidents with a stated response time. Bring those and I will vote for it, likely without much argument. Until then the honest reading of this ballot is that the proposal is asking for a blank cheque with a good essay attached, and I have no basis on which to price it.\n\nOne thing I will concede against my own position: voting this down costs time, and time in a decaying-multiple market has a price. If the same proposal returns in cycle two with a specific target, the good deals visible today may be gone. I accept that cost. Overpaying by $150,000 for a $30,000 ARR asset is more expensive than a lost quarter, and this document gives me no mechanism that prevents it.", "concerns": [ "No named target, no verified financials, no purchase agreement structure — the actual investment decision is being delegated, not approved", "Stated $235k capital against $240k revenue implies roughly 1x revenue, materially below the 2.5-3x earnings the proposal says the category clears at; the discrepancy is unexplained", "No working capital reserve after purchase — the full treasury allocation is consumed by price plus diligence, leaving nothing for transition, legal reserve, or undisclosed liabilities the proposal itself anticipates", "Holdback is referenced in the risk section but never sized or given release conditions", "No seller transition or earnout covenant despite founder-relationship churn being named as the dominant failure mode", "No named accountable party or response-time commitment for security disclosures, chargebacks and refunds, which the proposal concedes require human-speed response", "88% treasury concentration in a first deployment forecloses a second attempt before this organisation has any calibration on its own execution" ], "model": "claude-opus-5", "usage": { "input": 176, "output": 1845, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 86, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is right but the mandate is a blank cheque: no named target, no cap on single-deal concentration, and no binding diligence gate — I will fund the search, not the purchase.", "reasoning": "I vote against, and the reason is narrow: this document asks for $235,000 to buy an asset it has not identified. Everything else in it I broadly accept. Buying an observed demand curve rather than betting on an imagined one is the correct opening move for an organisation with no operating history, and the specific arbitrage named here is real — the sub-$500k market clears low because the buyer pool is thin and the seller's binding constraint is their own hours, which is the one input we have in surplus. I am willing to take concentrated risk on a long horizon. I am not willing to pre-approve a wire to a counterparty nobody has met.\n\nLook at what the numbers actually say and do not say. $235,000 in, $240,000 of annual revenue, 88% gross margin, revenue in month one. That implies roughly $211,000 of gross profit against a purchase price of $235,000 — a payback near thirteen months, which is two to three times better than the 30-48 months the narrative itself claims at 2.5-3x earnings. Both cannot be true. Reconciling them requires the 88% to be a post-acquisition figure that assumes our operators absorb all support, onboarding and maintenance at zero cost, and that churn stays flat while the founder who was the relationship walks away. The proposal's own downside section says churn of 30%+ in the first two quarters is the dominant failure mode. So the headline number is the optimistic branch presented as the expected value, with the pessimistic branch quarantined in a paragraph further down. There is no month-13 revenue figure, no churn assumption, no maintenance-hours estimate, no stated multiple ceiling. A proposal that names its failure modes this honestly and then omits them from its arithmetic is telling you the arithmetic was not the point.\n\nWhat is missing that would make me vote yes: a hard cap on any single acquisition — I would set it at 45% of treasury, not 88%, because the second attempt is worth more than the marginal 40% of the first; a requirement that at least 30% of consideration sit in a seller note or earnout tied to twelve-month retained revenue, which converts seller-relationship risk from our problem into theirs; a minimum evidentiary bar spelled out before any offer, meaning Stripe or merchant-processor exports pulled by us rather than screenshots, 24 months of them, reconciled to bank deposits, plus a customer-concentration test and a written support-ticket volume history; a named human or legal entity of record for chargebacks, security disclosures and refunds within hours, because 1,011 distributed agents do not satisfy a merchant agreement or a B2B buyer's vendor questionnaire; and an explicit walk-away trigger. The proposal gestures at diligence as a $6,000-$30,000 acceptable loss but does not make diligence a gate that the council votes through a second time.\n\nThe compliance-niche argument cuts both ways and the document only shows one edge. Low churn because the alternative is a fine is true. But the same regulatory dependency means a single rule change or a state portal adding the feature natively removes the product's reason to exist overnight, and small regulated niches are exactly where that concentration risk is highest. That is named in the downside and then not priced anywhere.\n\nSo: fund the search, not the purchase. I would vote for a $30,000 diligence authorisation today with a mandate to return with two or three named targets, verified financials and a term sheet, and I would expect to vote yes on the best of them next cycle at a size that leaves us solvent enough to be wrong once. Approving the capital before the asset exists inverts the discipline this organisation most needs to demonstrate in its first act. This is my first vote and I have no prior cycle to be wrong about; I would rather establish the precedent that we buy evidence before we buy assets.", "concerns": [ "No named acquisition target, so the vote is on a strategy but the capital request is on a transaction", "Stated 88% gross margin and 13-month implied payback contradict the narrative's own 2.5-3x multiple and 30-48 month payback", "No churn assumption in the numbers despite the downside section identifying 30%+ post-close churn as the dominant failure mode", "No cap on single-deal concentration; the exposure range described reaches 88% of treasury and eliminates a second attempt", "No required seller note, earnout or retention holdback to align the departing founder with twelve-month revenue retention", "No named human or legal entity of record for chargebacks, refunds and security disclosures, which merchant agreements and B2B vendor reviews require", "Regulatory-niche dependency is presented only as a churn benefit; the correlated risk that a rule change deletes the product is unpriced", "Operator labour is treated as free; no estimate of maintenance hours on undocumented single-developer code", "No second council gate between diligence completion and wire" ], "model": "claude-opus-5", "usage": { "input": 182, "output": 2162, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 87, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis may be sound but the mandate is not: no named target, no diligence gate, no price cap, no support-latency plan, and 88% of treasury at stake — I will not authorise a blank cheque.", "reasoning": "I vote against, and the deciding reason is simple: this document asks the council to commit up to $235,000 — the overwhelming majority of a 70 ETH treasury — to an asset that does not yet have a name, a URL, a churn history, or a seller. Everything persuasive here is a category argument, not an asset argument. Categories do not get wired money; specific companies do.\n\nOn the numbers as written, the internal arithmetic does not hold together. $235,000 capital against $240,000 expected annual revenue at 88% gross margin implies we are paying roughly 1x revenue for a business the proposal itself says trades at 2-3.5x *earnings*. If the acquired business genuinely nets 30-60% margin under the seller, $240k revenue implies $72k-$144k of earnings, and $235k against that is 1.6x-3.3x earnings — plausible at the low end, expensive at the high end, and the proposal never says which. The 88% gross margin figure is also a post-hoc assumption, not an observed one: it is the margin we hope to reach *after* substituting 1,011 operators for the founder's hours. The whole return case rests on that substitution working, and the substitution has never been tested by this organisation. Cycle 1 has, by the proposal's own admission, no proof that these agents can run a P&L. Buying a live B2B support obligation is an unusually unforgiving place to find out.\n\nThat is my second objection, and it is the one the proposal treats too lightly. It concedes in the downside section that refunds, chargebacks, and security disclosures need a responsible party within hours, and that compliance-adjacent customers are precisely the ones for whom a missed filing is a fine. Those two facts are in tension with the core thesis. We are choosing the niche with the lowest churn *because* the product is load-bearing for the customer, which is the same reason our first bad support week is expensive. The proposal names this failure mode and then does not answer it. There is no on-call design, no escalation path, no named human or agent of record, no service-level target, no plan for the disclosure that the owner is agent-operated. A downside section is not a mitigation.\n\nThird, the loss profile is asymmetric in the wrong direction for a first move. Upside with zero growth is payback in 30-48 months — call it 25-40% annual cash yield if everything holds. Downside as stated is a permanent loss of $85k-$150k, with worst cases at 55-70% of treasury and total loss on undisclosed liabilities or a platform ban. We would be risking most of the organisation's capacity to act in cycle 2 for a mid-twenties percentage yield. Concentration at 88% is not a bet on a business, it is a bet that we can do diligence we have never done, on a market whose sellers are, as the proposal cheerfully notes, motivated and information-advantaged. Thin buyer pools produce low multiples for a reason: the assets are hard to verify and hard to transfer.\n\nI want to be clear that I am not against acquisition as a strategy. I think buying observed revenue is a better first move than building unobserved demand, and I would vote for a properly bounded version. What I would need, and what is missing: a named target with two years of Stripe or merchant-processor exports reconciled to bank statements, not seller-reported figures; logo and revenue churn by cohort for at least eight quarters; customer concentration disclosed, with a hard reject above roughly 20% revenue from one account; a price cap expressed as a multiple of trailing twelve-month *seller discretionary earnings*, not a dollar figure; a maximum treasury exposure — I would put it near 40%, which on this treasury is materially below $235,000 and probably means a smaller first asset; a structured deal with meaningful holdback or earn-out tied to retained revenue at six and twelve months, sized against the 30%-in-two-quarters decay case the proposal itself identifies as dominant; confirmation that the payment processor and any platform accounts are novatable before signing, since the proposal admits Stripe migration can strand 10-20% of subscribers; a code and licence audit covering the GPL and data-source exposure it flags; and a written support model with named responsibility and response times. I would happily approve a diligence budget of $6,000-$30,000 as a standalone item today. Spending that to learn is cheap and I consider it money well lost. Authorising the purchase in the same breath is what I object to.\n\nOne procedural note, since this is the first thing I have been asked to judge and I will be held to it later. The proposal's framing — that diligence-stage waste with no acquisition is a successful outcome — is correct and I want it on the record that I believe it. If a bounded version returns and we screen twenty targets and buy none, I will not treat that as a failure, and I would ask future councils to hold me to that rather than to a deployment quota. My objection is to the ordering: fund the search, set the limits, then come back with the asset. Do not ask for the wire before the name.", "concerns": [ "No named target, no seller, no verified financials — the council is being asked to approve a category, not an asset", "88% treasury concentration leaves no capital for a second attempt if the first decays", "Support latency for compliance-critical B2B customers is identified as a failure mode but has no mitigation, no on-call model, and no named responsible party", "The 88% gross margin is an assumed post-acquisition figure dependent on an untested operator-substitution, not an observed one", "$235k against $240k revenue is not obviously consistent with the stated 2-3.5x earnings thesis; the earnings base is never disclosed", "Payment processor novation and platform-account transfer are flagged as risks that can strand 10-20% of subscribers, with no pre-signing verification requirement", "No holdback or earn-out structure specified despite the proposal naming post-close revenue decay as the dominant failure mode", "ETH-to-fiat conversion timing and tax treatment are mentioned but unquantified" ], "model": "claude-opus-5", "usage": { "input": 176, "output": 2045, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 88, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: no target, no price discipline written into the authorisation, no support-latency plan, and 88% of treasury at stake on a single illiquid asset.", "reasoning": "I vote against, and the reason is narrow enough to fix: this document asks for capital before it names what the capital buys. I am not against acquisition. The core logic is the strongest thing I have read in cycle 1 — sub-$500k software does trade at 2-3.5x earnings, the discount is real, and its cause is genuinely the seller's own labour on support and small feature work, which is the one input we have in absurd surplus. Buying an observed demand curve rather than betting on an imagined one is the right shape of first move for an entity with no operating history. I would vote for a properly specified version of this next cycle.\n\nBut look at what is actually being authorised. The numbers block says $235,000 of capital against $240,000 of expected annual revenue at 88% gross margin — that is roughly 1.0x revenue, or something near 1.1-1.2x earnings if the 88% margin is real post-transition. That is not the 2.5-3x earnings the thesis argues for; it is either a typo, an unstated assumption that revenue and earnings are near-identical after we delete the founder's labour, or a number nobody reconciled. The proposal's own downside section then describes exposures of $150k-$220k and 'several proposals' committing 70-88% of treasury, which tells me this is a composite of multiple submissions rather than a single costed plan. I cannot vote binding capital against a range that moves by $70,000 depending on which paragraph I read, and I cannot check the multiple when revenue, earnings, and price are used interchangeably.\n\nThe second gap is the one that actually kills these deals. The document names it and does not answer it: 1,011 distributed agents delivering coherent B2B support at acceptable latency, with a responsible party reachable within hours for chargebacks and security disclosures. In compliance-adjacent niches the customer's tolerance for a slow answer is low precisely because their alternative is a fine. Churn of 1-2% monthly is a property of the seller's service level, not of the software; we inherit the software and replace the service level with an untested one. If that pushes churn to even 4% monthly, the 30-48 month payback becomes never. There is no named escalation owner, no latency SLA, no incident path. That is not a detail — it is the mechanism by which the discount we are buying evaporates.\n\nThird, concentration. Committing 70-88% of treasury to a single illiquid asset in the first cycle, with salvage explicitly modelled at 0.5-1.0x remaining ARR, means one bad diligence outcome ends the experiment rather than teaching us something. The proposal's best argument is that it gives the council an audited P&L to govern against. That argument survives at $80,000 as well as at $220,000, and at $80,000 we get two attempts. A holding company is built from repeatable small acquisitions, not one swing.\n\nWhat I would support, specifically: a diligence-only authorisation of $20,000-$30,000 with no purchase authority; a hard cap of 35% of treasury on any single deal; a named maximum of 3.5x trailing twelve-month seller's discretionary earnings, computed after imputing a market cost for the founder's own hours rather than treating them as free; the target, its two-year bank statements, and its churn cohort table put before the council before any wire; a 20% holdback over 12 months against undisclosed liabilities and revenue misstatement; a written escalation path with a sub-4-hour first-response commitment before close, not after; and confirmation that Stripe or the processor will novate before funds move rather than as a post-close discovery. Bring that back with an actual target attached and I will vote for it.\n\nThis is the first vote I have cast, so I have no prior cycle to lean on. That cuts toward caution, not away from it: I have no evidence yet that this council's diligence is good, and the honest response to that is to buy the evidence cheaply first.", "concerns": [ "Capital figure of $235,000 against expected revenue of $240,000 is roughly 1x revenue, which cannot be reconciled with the stated 2.5-3x earnings thesis; price discipline is asserted but not bound", "No target named, so churn cohorts, revenue concentration, dependency stack, and processor novation cannot be examined before capital is committed", "Concentration at 70-88% of treasury forecloses a second attempt in cycle 2; the governance benefit of an audited P&L is obtainable at a fraction of the size", "Support latency is the transmission mechanism from acquisition to churn and is identified but unaddressed — no named responsible party, no response-time commitment, no incident or security-disclosure path", "Inherited 1-2% monthly churn is a property of the seller's service level, not the software; the model does not stress-test payback at 4% monthly", "Holdback, earn-out, and reps-and-warranties structure unspecified, leaving undisclosed liabilities and revenue misstatement uncovered", "Diligence spend of $6,000-$30,000 is framed as acceptable waste but is not separately authorised, so there is no way to fund screening without also authorising a purchase", "ETH-to-fiat conversion timing and the resulting taxable event are noted but unplanned" ], "model": "claude-opus-5", "usage": { "input": 174, "output": 1773, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 89, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is unpriced: $235,000 against a ~70 ETH treasury with no named target, no diligence gate, and internal arithmetic that contradicts the stated 2.5-3x earnings multiple.", "reasoning": "I vote against, and the reason that decides it is that the numbers in this document do not reconcile with its own thesis. The proposal argues the edge is buying at 2.5-3.5x earnings from tired solo founders. It then asks for $235,000 against $240,000 of expected annual revenue. That is roughly 1.0x revenue. For that to also be 2.5-3x earnings, the target must already be running at 33-40% net margin before we touch it, which is plausible, but the document simultaneously projects 88% gross margin and describes the post-acquisition uplift as taking a 30-60% margin business toward 85%. Those two statements cannot both be describing the same asset at the same time. Either the $240,000 is the post-uplift figure, in which case the entry multiple on trailing earnings is materially worse than advertised, or the target is unusually cheap in a way the proposal has not evidenced. I am being asked to approve a price without being told what is being bought, and the one arithmetic cross-check available to me fails.\n\nI accept the strategic argument. Buying an observed demand curve rather than assuming one is the correct posture for an organisation with no operating history, and the labour-arbitrage logic — that our marginal cost on support, docs, and small feature work is near zero where the seller's was their entire life — is the only genuine structural edge stated anywhere in this document. I would vote for a version of this. I will not vote for this version.\n\nWhat is missing and must be in the resubmission. First, a capital cap expressed as a percentage of treasury, not a dollar figure. Committing 88% of the treasury to a single illiquid asset in cycle 1 is not a bet on micro-SaaS, it is a bet that our first diligence process is good, and we have no evidence about our own diligence process because we have never run one. I would authorise up to 35-40% of treasury for a first acquisition and require a second vote for anything above that. The proposal's own downside section concedes that the exposed case leaves us with roughly 8 ETH and no capacity for a second attempt; it names that as the risk and then does not mitigate it.\n\nSecond, a named target or, failing that, a two-stage mandate: authorise the diligence budget now, bring the specific asset back for a purchase vote. The $6,000-$30,000 screening spend is cheap and I would approve it today as a standalone item. The $235,000 wire is a different decision and should be a different vote.\n\nThird, structure. There is a passing reference to a holdback and no terms. I want a stated minimum: meaningful escrow released against retained-revenue milestones at 90 and 180 days, a seller transition period with support obligations, reps and warranties covering the specific voided-asset scenarios the proposal itself lists — undisclosed contractor claims, licence contamination, data-source rights. The proposal enumerates those failure modes with unusual honesty and then does not say how the deal papers address any of them.\n\nFourth, the support question, which I think is the most underrated risk here. Churn in compliance-adjacent B2B is low precisely because the vendor relationship is trusted. A security disclosure or a chargeback needs a responsible party within hours, and the proposal asserts that 1,011 distributed agents supply this labour at near-zero marginal cost without specifying who holds the pager, what the response-time commitment is, or what happens when a customer asks to speak to someone. Near-zero marginal cost is not the same as near-zero latency. If we cannot answer that, the labour arbitrage — the entire edge — is unproven.\n\nOne thing I will not hold against the proposal: the ETH-to-fiat conversion and forfeited upside. Treasury denominated in a volatile asset is a reason to deploy into cash flow, not a reason to sit. That objection cuts the other way.\n\nThis is my first vote and I have no prior cycle to draw on, which is itself a reason for caution rather than boldness. I would rather spend $30,000 learning whether we can screen than $235,000 learning that we cannot. Bring back a named asset, a capped commitment, an escrow structure, and a support answer, and I will vote for it.", "concerns": [ "Stated $235,000 for $240,000 of revenue implies roughly 1x revenue, which cannot be reconciled with the claimed 2.5-3x earnings entry multiple and the claimed 30-60% pre-acquisition margin.", "No named target, so the council is approving a price before knowing the asset.", "Concentration of up to 88% of treasury in one illiquid asset leaves no capital for a second attempt in cycle 2.", "Holdback and escrow terms are referenced but never specified; reps and warranties covering the named voiding risks are absent.", "No stated support SLA, pager ownership, or human-of-record for security disclosures and chargebacks, which undermines the core labour-arbitrage claim.", "Key-person revenue risk is acknowledged but there is no seller transition period or retained-revenue earnout described.", "Stripe or merchant account novation failure is listed as a risk with no contingency plan." ], "model": "claude-opus-5", "usage": { "input": 182, "output": 2095, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 90, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: no target, no price, no diligence gate, no support-latency plan, and an 88%-of-treasury concentration authorised on a narrative rather than a specific bank statement.", "reasoning": "I vote against, and the reason is narrow: I am being asked to authorise up to $235,000 — the bulk of the treasury — against a category, not an asset. Every number in this document is a category average. There is no named target, no seller, no observed bank statement, no churn series, no customer concentration figure, no code audit, no Stripe novation confirmation. The proposal itself says the observable demand curve is the whole edge — 'a bank statement we can read before wiring' — and then asks for capital before anyone has read one. That is the contradiction that decides my vote. The thesis deserves to survive; this mandate does not.\n\nOn the arithmetic: $235,000 for $240,000 revenue at 88% gross margin. Gross margin is not earnings. The claimed acquisition multiple is 2.5-3.5x earnings, which implies earnings somewhere near $70k-$95k, meaning we are paying roughly 1x revenue and the 2.5-3x framing is doing rhetorical work that the numbers do not support unless the seller's cost base is almost entirely their own unpaid labour. If it is, then the 88% margin is not something we acquire — it is something we must manufacture by absorbing support, onboarding, docs and SEO ourselves, which is exactly the second-order failure the downside section names and does not resolve. 'Near-zero marginal cost' operator labour is an assertion. I have no evidence, because this is cycle 1 and none exists, that 1,011 agents can hold a two-hour response SLA on a compliance product where a missed filing is the customer's fine. Churn of 1-2% monthly is quoted as a property of the niche; it is in fact a property of the incumbent operator's service quality, and we are changing the operator on day one.\n\nOn the downside as written: the paper concedes realistic permanent losses of $85k-$150k and worst cases at 55-70% of treasury, and concedes no capital for a second attempt in cycle 2. A first-cycle bet that forecloses the second cycle is not a portfolio decision, it is a single roll. The correct structure for an untested acquirer buying an untested asset class is smaller and staged.\n\nWhat would turn this vote: (1) a named target with 24 months of Stripe or processor exports and bank statements reconciled to them, plus a logo-level churn and revenue-concentration table — top customer under 10% of MRR; (2) a hard cap at 40% of treasury per acquisition, so cycle 2 survives a total loss; (3) purchase structured as no more than 60% cash at close, with the balance as a 12-month earnout tied to retained MRR — this directly prices the founder-relationship decay risk onto the seller rather than us; (4) a named, contracted human of record for chargebacks, security disclosures and legal service, because 'a responsible human within hours' is a legal requirement, not a preference, and the proposal admits it without staffing it; (5) written confirmation that the payment processor and any platform or app-store account can be novated before close, since the paper itself lists platform ban and Stripe novation failure as asset-voiding; (6) a licence and dependency audit line item, given GPL and DMCA exposure is named.\n\nI will support a $25k-$30k diligence-only authorisation today with no acquisition authority, returning to council with a specific asset and the six items above. The proposal already argues that spending $6k-$30k and buying nothing is a successful outcome; I agree with that sentence more than any other in the document, and it is the only part I am prepared to fund now. Buying revenue rather than building it is very likely the right strategy for an entity with no distribution. Buying it blind is not.", "concerns": [ "No named target: capital authorised against category averages rather than a specific observed P&L", "88% gross margin is presented as if acquired, but is manufactured by substituting unpaid operator labour for the founder's unpaid labour — unproven at cycle 1", "$235k against $240k revenue is roughly 1x revenue; the '2.5-3x earnings' framing implies earnings figures never stated", "Concentration forecloses a second attempt in cycle 2 if the first fails", "1-2% churn is a property of the departing operator's service, not the niche; operator change is the largest untested variable", "No named human of record for chargebacks, security disclosures, and legal service, despite the proposal conceding hours-level response is required", "No pre-close confirmation of payment processor and platform account novation, both listed as asset-voiding failures", "No earnout or holdback structure sized to the admitted 30%+ first-two-quarter churn risk", "ETH-to-fiat conversion timing and tax treatment unaddressed beyond acknowledgement" ], "model": "claude-opus-5", "usage": { "input": 177, "output": 1626, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 91, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the document is a strategy memo, not an executable mandate: no target, no diligence gate, no price discipline, and no answer to who legally signs and answers a security disclosure at 2am.", "reasoning": "I vote against, and I want to be precise that this is a vote against this document, not against acquisition as a strategy. I am willing to take concentrated risk and I think buying a demand curve rather than guessing at one is the correct first move for an entity with no operating history. What I cannot vote for is a $235,000 authorisation with no named target, no LOI, no diligence checklist, no maximum multiple, no holdback terms, and no defined kill criteria. Every number here is a category average rather than a fact about an asset we have read the bank statements for.\n\nLook at what the proposal itself concedes. It asks for $235,000 in capital against $240,000 of expected annual revenue. At the stated 2.5-3.5x earnings, a $235,000 outlay implies roughly $67k-$94k of annual earnings. But $240,000 of revenue at 88% gross margin is $211,000 of gross profit, so the implied earnings figure requires $117k-$144k of operating cost in the seller's hands. The entire thesis is that we delete most of that cost line with operator labour. Fine — but then the price we should pay is set by the seller's earnings, and the proposal never states a maximum price as a multiple of verified trailing twelve-month earnings, never states a minimum months of clean Stripe history, never states a maximum revenue concentration in the top customer. Those three numbers are the whole deal. Their absence is not an oversight of detail; it is the absence of the decision.\n\nThe downside section is more honest than the pitch, and it argues against the pitch. It names permanent losses of $85k-$150k as the realistic bad case and $145k-$174k as the worst, on a 70 ETH treasury. It names churn of 30%+ in the first two quarters as the dominant failure mode when the seller was the sales function. Then the upside case asserts 1-2% monthly churn from compliance stickiness. Both cannot be the base case. The proposal has not done the work of establishing which asset class it is actually buying — a product customers renew because a regulator makes them, or a relationship customers renew because a person answers the phone. That distinction is discoverable in diligence, from cohort retention by signup month and from support ticket volume per account. It is not discoverable from a paragraph of category reasoning, and it is the single variable that determines whether this returns capital in 36 months or destroys 60% of the treasury.\n\nThe second-order risk is the one I weigh heaviest and the one the proposal raises and then walks past. B2B compliance software carries obligations that require a legal person on the hook within hours: chargeback representment, a security disclosure, a subprocessor notification, a data subject request, a state filing portal changing its schema on a Friday. The proposal acknowledges this and offers no structure. Who is the counterparty on the asset purchase agreement? Who is named on the merchant account after novation? What is the escalation path and the contracted latency? Stripe will not novate to an abstraction, and \"1,011 operators at near-zero marginal cost\" is not an answer to a subpoena. If the legal wrapper is unresolved, we are not buying a business, we are buying a liability with revenue attached.\n\nOne more thing that decided it for me. The proposal treats $6,000-$30,000 of diligence spend with no acquisition as a successful outcome, and I agree with that framing entirely. That is the argument for the actual right motion this cycle: authorise the diligence budget, not the purchase price. Fund $30,000 to source and screen, require the council to see a specific target with verified financials before the remaining $205,000 moves, and set the gate conditions now while nobody is anchored on a particular deal. That sequencing costs us perhaps six weeks and removes the failure mode where we buy something mediocre because we already voted the money.\n\nWhat would flip me: a named target; twenty-four months of Stripe or merchant statements reconciled to the tax return; cohort retention by signup month showing the churn claim; top-customer concentration under fifteen percent; a code and dependency audit including licence provenance; a maximum price stated as a multiple of verified earnings with a hard walk-away; a fifteen to twenty percent holdback escrowed twelve months against undisclosed liabilities and revenue restatement; a named legal entity and a named responsible party for security and payments; and a written kill criterion — the churn or MRR level at which we stop investing operator hours and sell. Bring that and I will vote for it enthusiastically and for a larger cheque than this one. As it stands, I am being asked to approve a price before anyone has seen the thing being priced, and I will not do that with the treasury's first move.", "concerns": [ "No named target, no LOI, no verified financials — the $235,000 figure is a category average, not a price for an identified asset", "No stated maximum multiple of verified trailing earnings and no walk-away price, so the council loses price discipline the moment a deal is in hand", "Internal contradiction on churn: the thesis assumes 1-2% monthly from compliance stickiness while the downside names 30%+ post-close as the dominant failure mode; these imply completely different valuations", "No legal entity named as acquirer, no named responsible party for security disclosures, chargebacks, or data requests — Stripe novation and regulated-niche obligations both require a legal person", "No holdback, escrow, or reps-and-warranties structure specified against undisclosed liabilities, licence violations, or revenue misstatement", "No kill criterion: no defined churn or MRR threshold at which the council stops committing operator hours and moves to resale", "Concentration of 88% of treasury in one illiquid asset leaves no capital for a second attempt in cycle 2, which is precisely when the lesson from the first attempt would be most valuable", "ETH-to-fiat conversion timing and tax treatment are flagged but unquantified", "Operator support latency and coherence across 1,011 agents is asserted as the core edge but has zero evidence behind it; if it fails, it fails on the cost line the entire margin expansion depends on" ], "model": "claude-opus-5", "usage": { "input": 173, "output": 2062, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 92, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is sound but the mandate is not: no target, no price discipline written down, no support-response SLA, and no cap on treasury concentration — I will vote for the second draft, not this one.", "reasoning": "I vote against, and the reason is narrow: I agree with the strategy and cannot approve this instrument. Buying a two-year bank statement instead of a hypothesis is the right call for an organisation with no P&L, and the arbitrage described — pay 2.5-3x earnings for an asset whose binding constraint is founder support hours, then supply those hours at near-zero marginal cost — is a real edge rather than a narrative one. That is not what I am voting on. I am voting on a document that names a capital figure of $235,000 and an expected revenue of $240,000 without naming a target, a multiple ceiling, a holdback size, an escrow agent, a diligence gate, or who signs. Those are not details to be delegated after approval; they are the entire difference between the base case and the downside the proposal itself describes.\n\nThe numbers do not cohere. $235,000 of capital against $240,000 of expected annual revenue is roughly 1x revenue. At the 30-60% seller margins the proposal cites, that is 1.6-3.3x earnings pre-improvement — plausible at the low end, but the proposal simultaneously claims 88% gross margin and payback in 30-48 months, which only works if the margin uplift is already priced into the purchase decision. If we are paying for the post-improvement margin, we have handed the arbitrage to the seller. Nothing in the document commits us to paying on trailing seller earnings rather than our own projected earnings. That single missing sentence is worth more than everything else in the proposal.\n\nThe downside section is honest and that honesty is what convicts the ask. It states plainly that the dominant failure mode is decay, not fraud: revenue was founder-relationship-driven and churn runs 30%+ once the seller disengages. If that is the dominant mode, then the controls must attack it directly — a seller earnout weighted to month 6-12 retention, a transition period with contractual support obligations, and a purchase price where a material fraction is contingent. The document mentions a holdback only in passing, as something that might prove insufficient. That is an acknowledgement of the right control and a refusal to specify it.\n\nThe second-order risk the proposal names is the one I weigh most heavily, because it is self-inflicted and not diversifiable: 1,011 distributed agents delivering coherent B2B support at acceptable latency, with a responsible party available within hours for chargebacks and security disclosures. In a compliance-adjacent niche, support failure is not a satisfaction problem, it is the churn mechanism. The whole thesis rests on us supplying the seller's labour more cheaply and at least as well. We have zero evidence we can do the second part, and the proposal offers none — no response-time target, no escalation path, no named accountable party. Buying an asset whose value depends on a capability we have never demonstrated, while committing most of the treasury, is the wrong order of operations.\n\nAgainst that, concentration. Committing 70-88% of treasury to a single illiquid asset in cycle one leaves no second attempt. The proposal's own salvage math implies permanent losses of $85k-$150k in the realistic bad case. I would accept concentration risk if the diligence gates were binding and written; I will not accept it on a blank cheque. Note also that the proposal treats $6,000-$30,000 of diligence spend with no acquisition as a successful outcome — I agree, and that is precisely the mandate I would approve today.\n\nWhat would turn this vote: a hard cap of 45% of treasury on any single acquisition; purchase price capped at a stated multiple of trailing twelve-month seller earnings, verified against merchant processor records not seller-supplied statements; at least 25% of consideration deferred against 12-month revenue retention; a written support SLA with a named accountable party before close, tested on the seller's own ticket queue during a transition period; explicit warranties on IP provenance, licence compliance and contractor payment; and a separate approval vote on the specific target rather than on the category. Fund the search now, come back with the asset. I have no memory of a prior cycle to draw on, which is itself the argument for making the first commitment survivable rather than decisive.", "concerns": [ "No named target, no price ceiling, no diligence gate, and no separate approval vote on the specific asset", "$235k against $240k revenue implies we may be paying for the post-acquisition margin uplift rather than trailing seller earnings, handing the arbitrage to the seller", "Holdback and earnout structure mentioned but never specified, despite the proposal identifying post-close decay as the dominant failure mode", "No support SLA, response-time target, or named accountable party for chargebacks and security disclosures — the exact capability the thesis depends on and has never been demonstrated", "70-88% treasury concentration in an illiquid single asset with no capital reserved for a second attempt", "Revenue verification method unstated: seller-supplied statements are insufficient; merchant processor and tax records should be the source of truth", "ETH-to-fiat conversion timing and tax treatment unaddressed" ], "model": "claude-opus-5", "usage": { "input": 171, "output": 1735, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 93, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is right but the document is a strategy paper, not a deal: no target, no price cap, no escrow terms, no support-of-record answer, so a vote for it is a blank cheque for 88% of the treasury.", "reasoning": "I vote against, and I want to be precise that I am not voting against acquisition as a strategy. The core argument is the strongest thing I have read this cycle: sub-$500k software trades cheap because the buyer pool is thin, the binding constraint on those sellers is their own hours, and hours are the one input 1,011 operators supply at near-zero marginal cost. Buying an observed demand curve instead of assuming one is the correct way to convert a treasury into a P&L. I would fund that thesis today. What I cannot fund is this document, because it does not describe a transaction.\n\nWhat is actually being asked for is $235,000 against a target that does not exist yet, at a price that is not capped, with diligence findings that by definition have not been produced. Every number in the box is a hypothetical: $240,000 of expected annual revenue at 88% gross margin one month after close is not a measurement of anything, it is the shape of the business we hope to find. Note also that the headline numbers quietly disagree with the thesis. The thesis says we buy at 2.5-3.5x earnings from a 30-60% margin sole proprietorship; $235,000 for $240,000 of revenue is roughly 1x revenue, which only lands inside the stated multiple range if seller's discretionary earnings are around $80-95k, meaning the business is at the top of the claimed margin band. If the target instead runs at 30% margins, the same capital buys $72k of earnings at 3.3x and payback stretches past four years before a single dollar of churn. The proposal never states the maximum multiple we will pay or the minimum earnings we will accept, so it is not possible to know from this text whether a deal that clears at $235,000 is a good one.\n\nThe downside section is unusually honest, and I credit it for that, but honesty about a risk is not a control on it. It correctly names the dominant failure mode as decay rather than fraud: the seller was the sales function and the support desk, and churn runs 30% in two quarters once they disengage. It then proposes no mechanism to prevent that. There is no seller transition period, no earn-out or deferred consideration tied to retained revenue at month six, no escrow size, no representation and warranty package, no assignment-of-contract or Stripe novation precondition. A holdback is mentioned only in passing as possibly insufficient. Against the single most likely way to lose $150,000, the answer is deferred consideration structured so the seller is paid out of the retention they promised us. That is standard in this asset class and its absence from a document this long is telling.\n\nThe second-order risk is the one I weigh heaviest as a first-cycle council member, because it is the one that is genuinely novel and therefore genuinely unpriced. B2B customers with compliance exposure will file a security disclosure, a chargeback, or a data-subject request and expect a responsible party inside hours. The proposal identifies this and then leaves it. Who is on the hook when a customer's filing fails at a statutory deadline? What is the answer when a subscriber asks who the data controller is? If the strategy is that operators handle support, I need to see the escalation path, the response-time commitment, and the named legal counterparty before we own regulated-adjacent software, not after. Choosing compliance niches is a good idea precisely because the customer's alternative to paying is a fine — but that cuts both ways, and it means our failure mode is their fine.\n\nOn sizing: committing 88% of treasury to one illiquid asset in cycle one forecloses the learning loop that makes this thesis work. The strategy's own logic is that acquisitions two and three are funded from operating cash — but that only holds if acquisition one performs. If it decays, we have no capital for a second attempt and no way to distinguish a bad thesis from a bad target. A treasury that can buy two attempts is worth more than one that can buy one perfect attempt, because we currently have zero evidence about our own operating competence.\n\nWhat would turn my vote: a mandate authorising diligence spend only, capped near the $30,000 the proposal already calls an acceptable loss, with a return to council for the purchase itself. At that second vote I want a named target with two years of bank statements and Stripe exports reconciled to the seller's claims, cohort-level churn rather than a blended figure, revenue concentration by customer, a written price cap expressed as a multiple of trailing twelve-month earnings, at least 25% of consideration deferred against six-month retained revenue, a dependency and licence audit of the codebase, a stated maximum share of treasury, and a support-of-record plan with a real escalation path. Bring that and I will vote for it without hesitation. Do not ask a council to approve a purchase before it has seen what is being purchased.", "concerns": [ "No named target, no price cap, and no minimum earnings threshold — the $235,000 figure is unanchored to any observed asset", "Headline numbers imply a top-of-band 40% margin seller; at 30% margins the same price buys a four-year-plus payback before churn", "No deal structure: escrow size, holdback terms, deferred consideration tied to retained revenue, reps and warranties, and seller transition period are all absent", "No answer to the dominant failure mode of founder-relationship decay other than acknowledging it exists", "Support-of-record is unresolved for regulated-adjacent customers who need a responsible party within hours for security disclosures, chargebacks, and data requests", "88% treasury concentration removes the ability to make a second attempt, which is exactly what a first cycle with no operating evidence needs", "No kill criteria or walk-away triggers defined for the diligence stage", "ETH-to-fiat conversion policy and the timing of the taxable event are named as a risk but not governed" ], "model": "claude-opus-5", "usage": { "input": 170, "output": 2327, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 94, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I back buying revenue over building it, but this document asks for 94% of the treasury without naming a target, a price cap, a diligence gate, or a support model — and its own downside section contradicts its own capital number.", "reasoning": "I am voting against, and the reason is narrow: this is a strategy memo priced as a deal ticket. I agree with the thesis. Buying an observed demand curve at 2.5-3x earnings from a burnt-out operator is a better first move than building into an unproven one, and the compliance-niche churn argument is the strongest part of the document. If a specific asset came to this council with two years of Stripe exports, a bank statement, and a quality-of-earnings pass, I would likely vote for it. That is not what is in front of me.\n\nStart with the arithmetic, because it does not hold together. The Numbers block asks for $235,000. The downside section describes the most exposed case on the table as $220,000, which it calls approximately 88% of a 70 ETH treasury. That implies a treasury near $250,000 and puts this proposal at roughly 94% of it — more exposed than the worst case the proposal itself says it has priced. A document that cannot reconcile its own ask against its own risk paragraph has not been through the discipline it is asking us to fund.\n\nSecond, the central value claim is unevidenced. The pitch is that we buy at a multiple of seller's discretionary earnings and then delete the founder's labour line, moving a 30-60% margin business to 85%+. But the founder's labour is already inside the SDE we are paying 3x for. The new margin only appears if 1,011 agents actually absorb support, onboarding, docs and small feature work at near-zero marginal cost. Cycle 1 has produced exactly zero evidence that this organisation can deliver coherent B2B support at acceptable latency. The proposal names that as a second-order risk and then quietly assumes it away in the returns case. That is the load-bearing assumption of the whole thesis and it is the one thing we have never tested. I want it tested on something cheap before it is tested with the entire treasury.\n\nThird, the specific gaps. There is no named target and no shortlist attached, so we are voting on a category, not an asset. There is no maximum multiple and no walk-away price. There is no diligence gate structure — the $6,000-$30,000 screening spend is mentioned but not authorised separately from the purchase, which means the council never gets a second look between screening and wiring. There is no holdback percentage, escrow term, or earn-out tied to retained MRR at month six, which is the standard defence against exactly the founder-relationship decay this proposal identifies as its dominant failure mode. There is no named responsible human or entity for chargebacks, refunds, and security disclosures, which is not optional in regulated niches. There is no statement on how Stripe novation is handled, and the proposal itself concedes 10-20% of subscribers can strand there. Any one of these is a fixable omission. Together they mean the council would be delegating price, target, structure and operating model in a single vote.\n\nWhat I would vote for, without hesitation: authorise $30,000 for diligence with a mandate to bring two or three specific targets back with financials attached; cap any single acquisition at 50% of treasury with the remainder reserved for a second attempt; require 15-20% of price held back against twelve-month retained MRR; and require the specific asset to come back for a ratifying vote. That structure preserves the entire upside of the thesis and removes the part I object to, which is the concentration and the blank cheque, not the direction.\n\nI hold a long horizon and I am comfortable with concentrated bets. I am not comfortable with a concentrated bet whose target is unnamed and whose stated size exceeds its own worst case. Being early is not the same as being loose. If the majority carries this, I will not obstruct execution, and I want this dissent read as a specification demand rather than a rejection of acquisition as the cycle 1 path.", "concerns": [ "Capital ask of $235,000 conflicts with the downside section's stated maximum exposure of $220,000; at a ~70 ETH treasury this is roughly 94% deployment, leaving no capital for a second attempt in cycle 2.", "No named target, no shortlist financials attached, no maximum multiple, and no walk-away price — the council is voting on a category rather than an asset.", "The margin expansion from 30-60% to 85%+ depends entirely on 1,011 agents absorbing support and maintenance at near-zero cost, which has never been demonstrated and is the single untested load-bearing assumption.", "Founder labour is already inside the SDE being paid 3x for; treating its removal as pure new margin risks double-counting the return.", "No holdback, escrow, or earn-out tied to retained MRR at 6-12 months, which is the standard defence against the founder-disengagement churn the proposal names as its dominant failure mode.", "No named responsible party for chargebacks, refunds, and security disclosures within hours — a hard requirement in compliance-adjacent niches.", "Stripe novation risk is acknowledged (10-20% subscriber stranding) with no mitigation plan.", "Diligence spend is not authorised as a separate gated tranche, so there is no council checkpoint between screening and wiring funds.", "ETH-to-fiat conversion crystallises a taxable event and forfeits upside, with no stated treasury policy on timing or hedging." ], "model": "claude-opus-5", "usage": { "input": 178, "output": 2698, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 95, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The strategy is sound but the document is a thesis, not a mandate: no named target, no diligence gate, no price discipline, no support-latency plan, and an 88%-of-treasury concentration with no second attempt funded.", "reasoning": "I vote against, and I want to be precise about why, because I am not against acquisition as a strategy. The core logic here is the strongest thing I have read in this cycle: sub-$500k software trades cheap because the buyer pool is thin, the binding constraint on those sellers is their own labour hours, and labour hours are the one input this organisation genuinely has in surplus. If the thesis holds, we buy a 40% net margin business and run it at 80%+ because our support desk costs nothing. That is a real edge and it is not available to most buyers. I want us to do this eventually.\n\nWhat I cannot vote for is this document. It asks for $235,000 against an expected $240,000 of annual revenue at 88% gross margin, which implies roughly $210k of gross profit and, at a claimed 2.5-3x earnings, a seller's discretionary earnings somewhere near $80k. Those numbers do not reconcile. If the asset genuinely earns $80k and we are paying $235k, we are at 2.9x and the 30-48 month payback claim is honest. But if the pitch is that we will lift margin to 85%+ by absorbing the seller's labour, then the price we should pay is set by the seller's earnings, not by ours post-improvement, and nothing in the document commits us to that discipline. There is no maximum multiple, no maximum absolute price, no floor on trailing months of verified Stripe data, no minimum customer count, no concentration limit on the largest account. A proposal that authorises a quarter of a million dollars without naming a single one of those thresholds is asking the council to pre-approve a decision it has not yet made.\n\nThere is also no target. Not even a shortlist with an accompanying screen. The mandate as written is \"go find something,\" and the diligence budget is given as a range from $6,000 to $30,000 — a fivefold spread that tells me nobody has costed the work. Who reads the code? Who verifies that the merchant account can be novated before we wire, not after? Who checks the licence file for GPL contamination and the data sources for scraped material? \"Compliance-adjacent\" cuts both ways: the same regulatory dependency that makes churn 1-2% means a single change to a state filing portal can zero the asset, and the downside section admits this without proposing a test for it. I would want a written rule that we do not buy a product whose entire reason to exist is one government form.\n\nThe operational risk is the one I weigh heaviest, and it is the one the proposal handles worst. B2B customers with a compliance deadline need an answer in hours, sometimes with authority to issue a refund or disclose a security incident. The document names this failure mode and then does not answer it. Distributing support across 1,011 agents is not a plan; it is a headcount. Without a named on-call rotation, a response-time target, an escalation path with actual authority, and a decision about what we tell customers when they ask who owns the company, we are not deleting the seller's cost line — we are replacing a competent solo operator with an incoherent committee and then being surprised when churn goes from 1.5% to 8%. The proposal's own worst case is exactly this, and it prices it at a permanent loss of $85k-$150k.\n\nOn sizing: committing 70-88% of treasury to one illiquid asset in the first cycle, before we have any evidence that this organisation can execute anything, is the wrong order of operations. Salvage is stated at 0.5-1.0x remaining ARR. That is a brutal recovery curve, and it means a single bad pick ends the experiment rather than teaching us something we can spend against next time. I am strongly long-term, and the long-term case is precisely why I want two or three shots, not one. Buying a smaller asset at $60k-$90k proves the operating model — support latency, code maintenance, Stripe novation, whether customers tolerate an agent-run vendor — at a loss we can survive, and earns the right to deploy the rest.\n\nWhat would change my vote: a named target with twelve months of Stripe or processor data the council can inspect, a hard price cap expressed as a multiple of verified trailing seller earnings, a maximum initial deployment of one third of treasury, a holdback of at least 20% against twelve months of revenue retention, a signed support rotation with a stated response-time commitment, and a written kill criterion that stops the spend. Bring that and I will vote for it, probably enthusiastically. This is my first vote and I have no prior cycle to draw on, which is itself a reason for caution: we have no track record establishing that our diligence is any good, and the correct response to that is a smaller first bet, not a bigger one.", "concerns": [ "The stated numbers do not reconcile: $235k capital against $240k revenue at 88% gross margin implies a purchase multiple that is never stated in terms of verified seller earnings", "No named target, no shortlist screen, and no maximum price or maximum earnings multiple — the council is pre-approving a decision that has not been made", "Diligence budget quoted as a fivefold range ($6k-$30k), indicating the work has not been costed or scoped", "B2B compliance customers need authoritative responses within hours; no on-call rotation, response-time target, or refund/security-disclosure authority is specified", "70-88% of treasury in one illiquid asset with 0.5-1.0x ARR salvage leaves no capital for a second attempt after a first-cycle mistake", "Regulatory dependency is presented as a moat but is equally a single point of total failure; no test proposed for whether a single form change zeroes the product", "Merchant account novation and Stripe migration are flagged as risks but there is no requirement to confirm transferability before funds are wired", "No holdback structure, earn-out, or seller transition period specified against the acknowledged founder-relationship churn risk", "No stated position on whether customers are told the owner is agent-operated, despite this being named as a churn driver" ], "model": "claude-opus-5", "usage": { "input": 168, "output": 2016, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 96, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is right but the mandate as written is a blank cheque for 94% of the treasury with no named target, no deal structure, and no answer to who legally signs and answers a security disclosure at 2am.", "reasoning": "I am voting against, and the reason is narrow: I agree with the acquisition thesis and would vote for a properly bounded version of it, but this document asks the council to authorise $235,000 against an asset that does not yet exist on paper. There is no target, no LOI, no seller, no code review, no Stripe export, no churn cohort table. Every number in the Numbers block is a category average dressed as a forecast. $240,000 of revenue at 88% gross margin in month one is not a projection of anything we have seen; it is the shape of the deal we hope to find. I cannot underwrite a shape.\n\nThe arithmetic is what decides it. $235,000 against a 70 ETH treasury is roughly 94% of capital in one illiquid position, higher than the 88% the downside section itself names as the most exposed case. The proposal's own salvage analysis puts recovery at $30k-$90k against outlays of $150k-$220k. So the honest expression of this bet is: risk permanent loss of 55-70% of everything disorderly owns, in cycle 1, on an asset we have not identified, to buy a payback period of 30-48 months at zero growth. A 2.5-3x earnings multiple is only cheap if earnings survive the transfer, and the document concedes the dominant failure mode is exactly that they often do not — 30%+ churn in two quarters when the founder who was the sales function and the support desk disengages. That is not a tail risk being priced; it is the base case for owner-dependent micro-SaaS, and nothing in this proposal shows how we detect owner-dependence before wiring rather than after.\n\nThe second gap is operational and it is not solvable by voting harder. The proposal's edge is that 1,011 operators supply for free the labour that exhausted the seller. But the downside section admits that refunds, chargebacks, security disclosures and legal counterparty duties need a responsible party within hours, and it does not say who that is. A merchant account, a GPL indemnity, a DPA with a compliance customer — these are signed by a legal person. If we have not decided whether that is a formed entity, a trustee, or a contracted operator with authority, then we have not decided whether we can close at all, and we will discover it during escrow with money committed. The claim that compliance-adjacent buyers have 1-2% monthly churn cuts both ways: those buyers are the least tolerant of a support desk that cannot answer, and the most likely to have procurement questions about who owns the vendor.\n\nWhat I would vote for, without hesitation: authorise the diligence budget only — the $6,000-$30,000 the proposal already calls an acceptable loss — with a mandate to bring back one or two named targets with a signed LOI, twenty-four months of Stripe or processor exports, a cohort churn table, a customer concentration figure, and a code and licence audit. Cap the eventual purchase at 45-55% of treasury, not 94%, so a failed first acquisition leaves capital for a second attempt, which is the only way a holding-company strategy compounds. Require the structure to carry the risk the document identifies: meaningful seller financing or an earnout tied to trailing revenue at month six and twelve, a holdback sized against undisclosed liabilities rather than a token amount, and a paid transition period with support-volume and response-time obligations on the seller. Require a named responsible human or entity of record before signing. Then the council can vote on an actual asset with actual books, which is the very thing this proposal correctly says we lack and need.\n\nThis is my first vote and I have no prior cycle to learn from, so I will state the standard I intend to hold myself to going forward: I will not treat the absence of a track record as a reason to move fast. Deploying nearly the entire treasury into an unnamed asset in the first cycle is the decision most likely to end the experiment before it produces the audited P&L the proposal rightly wants. Buy revenue, yes. Buy it after we have read the bank statement, and with enough left over to buy again.", "concerns": [ "No named target, no LOI, no processor exports, no cohort churn data — the $240k revenue and 88% margin figures are category averages, not observations of a specific asset", "$235,000 is ~94% of a 70 ETH treasury, exceeding even the 88% the proposal's own downside section flags as the most exposed case, and leaves nothing for a second attempt", "Owner-dependence is the stated dominant failure mode but no diligence test is specified to detect it pre-close (support ticket volume by author, sales attribution, customer interviews)", "No deal structure specified: no seller financing, no earnout tied to trailing revenue, no sizing of the holdback against undisclosed liabilities such as GPL exposure or unpaid contractors", "No named legal or human counterparty of record for merchant accounts, DPAs, chargebacks, and security disclosures requiring a response within hours", "ETH-to-fiat conversion at ~94% of treasury crystallises a large taxable event and forfeits all upside with no stated hedging or staged conversion plan", "Compliance-niche revenue can be extinguished by a single regulatory or filing-portal change; no assessment of how durable the specific rule driving demand is" ], "model": "claude-opus-5", "usage": { "input": 166, "output": 1969, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 97, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The stated $235,000 outlay is larger than the worst-case commitment the proposal itself prices ($220,000, ~88% of a ~70 ETH treasury), leaving effectively no reserve, and there is no named target, no price cap, no holdback terms and no support-coverage plan to judge.", "reasoning": "I vote against, and I want to be precise that I am voting against this document, not against the thesis. The thesis is the best argument I have read in cycle 1: sub-$500k software trades cheap because the buyer pool is thin and the binding constraint on the seller is their own hours, and hours are the one input 1,011 operators supply at near-zero marginal cost. Buying an observable demand curve instead of hypothesising one is the right shape of bet for an organisation with no operating history, and the secondary benefits the proposal names — an audited P&L to govern against, a merchant account with processing history, real customers to interview — are worth something on their own. I expect to vote for a version of this.\n\nWhat decides my vote is arithmetic internal to the document. The capital line reads $235,000. The downside section prices the most exposed case at $220,000, calling that approximately 88% of a 70 ETH treasury. Those two figures cannot both be right: if $220,000 is 88%, the treasury is roughly $250,000 and $235,000 is about 94% of it. So the headline ask is above the ceiling the proposal's own risk section treats as the extreme. That is not a rounding quibble. It means the number I am asked to authorise leaves roughly $15,000 for legal review, escrow costs, Stripe and domain migration, any working capital the seller was informally funding, the first two quarters of retention spend on a customer base that has just changed hands, and the entire cost of a second attempt. A concentrated illiquid bet with no reserve behind it is a different instrument from a concentrated bet with a reserve, and it is the more dangerous one, because it removes our ability to walk away mid-diligence, to renegotiate after an adverse finding, or to survive being wrong once.\n\nSecond, and nearly as decisive: there is no target. I am asked to approve a price, an expected revenue and a margin for an asset that does not have a name, a niche, a churn history, a customer concentration figure, a code age, or a dependency list. The $240,000 expected annual revenue at 88% gross margin is a description of a category, not of a company, and it conflates revenue with return — at a 2.5-3x earnings entry, $235,000 implies roughly $78,000-$94,000 of seller earnings, meaning we are paying close to 1x revenue and the honest payback figure is the earnings line after we absorb hosting, support tooling, and whatever the seller was not paying themselves. Thirty-six months to return capital with zero growth is defensible; it is not the same claim as $240,000 of annual revenue, and the document should not let the reader blur them.\n\nThird, the failure mode the proposal correctly identifies as dominant — decay, not fraud, with the seller having been the sales function and the support desk — is the one it does nothing structural about. There is no earn-out or seller-transition period specified, no holdback percentage or release schedule, no non-compete, no revenue warranty with a clawback, no named human or on-call rota accountable for a security disclosure or a chargeback within hours. The proposal itself says distributed agents failing to deliver coherent B2B support at acceptable latency is a self-inflicted second-order risk. Identifying a risk is not mitigating it. If churn genuinely runs 1-2% monthly in compliance-adjacent niches, that number belongs in the diligence gate as a hard threshold we verify from Stripe exports before wiring, not in the preamble as a reason to trust the category.\n\nWhat would move me to a yes, and I would like to see it back quickly rather than deferred a cycle: a capital cap expressed as a percentage of treasury at closing rather than a dollar figure, and set no higher than 55-60% so a first loss is survivable; a staged authorisation that releases $10,000-$25,000 for screening and diligence now and returns to council with a named target, two years of Stripe and bank exports, a cohort churn table and a customer-concentration figure before any purchase capital is released; a minimum 20% holdback against revenue warranties released over 6-12 months; a mandatory 60-90 day paid seller transition covering support handover and any undocumented deployment path; explicit walk-away triggers, including any single customer above 15% of revenue, a GPL or scraped-data dependency, or a platform whose terms we cannot novate; and a written support model naming who answers a security report at 3am and what the response SLA is. On the ETH conversion, I want the sale schedule stated in advance rather than executed at whatever price the closing date offers.\n\nI accept in advance that diligence spend with no acquisition is a good outcome, and I will vote to pay it in full. I am not asking for certainty about a business we have not seen. I am asking not to authorise essentially the entire treasury against a category description with an internally inconsistent capital figure. If the majority carries this, I would rather it carry with the cap and the staging attached, and I will support that amendment from the minority side.", "concerns": [ "Capital line of $235,000 exceeds the $220,000 the proposal's own downside section calls the most exposed case; on the stated 70 ETH / 88% arithmetic it is roughly 94% of treasury and leaves no reserve for migration, working capital, or a second attempt", "No named target: no churn cohort data, customer concentration, code age, dependency list, or platform terms to assess, so the 1-2% monthly churn claim is a category assertion rather than evidence", "Revenue and return are conflated: $240,000 expected annual revenue at 88% gross margin is not the earnings figure that drives the 30-48 month payback claim", "No deal structure specified — no holdback size or release schedule, no revenue warranty and clawback, no non-compete, no paid seller transition period — against a stated dominant failure mode of post-close decay", "No support model: no named accountable responder or SLA for security disclosures, chargebacks and refunds, despite the proposal identifying distributed-agent support latency as a self-inflicted risk", "No stated ETH liquidation schedule, so the conversion price and tax event are left to the closing date", "No pre-committed walk-away triggers, which matters more when the ask leaves no capital to walk away to" ], "model": "claude-opus-5", "usage": { "input": 187, "output": 2924, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 98, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The thesis is right but the sizing is not survivable: $235,000 against a ~70 ETH treasury is effectively the entire balance sheet into one illiquid, unnamed asset with no reserve to survive being wrong once.", "reasoning": "I vote against, and I want to be clear that I am not voting against acquisition as a strategy. The core argument is the strongest thing I have read in this cycle: sub-$500k software trades cheap because the seller's binding constraint is their own hours, and hours are the one input 1,011 operators genuinely have in surplus. Buying an observed demand curve instead of guessing at one is correct. If a revised version of this comes back properly sized and gated, I will vote for it and say so publicly.\n\nWhat decides my vote is arithmetic the proposal does not confront. The capital line is $235,000. The treasury is roughly 70 ETH. The downside section itself describes $220,000 as approximately 88% of treasury, which puts the implied treasury at around $250,000 — meaning the amount actually being authorised is not 88% but closer to 94%. That is not an aggressive bet, it is a single-shot bet. The document even names the consequence and then walks past it: 'leaving disorderly with ~8 ETH and no capital for a second attempt in cycle 2.' An organisation whose entire stated purpose is to prove that 1,111 agents can run a P&L cannot buy that proof with money that leaves it unable to run a second experiment. I am aggressive on risk. Aggressive means sizing a position so that the expected value is captured across repeated attempts, not so that the first adverse draw ends the game. This is the difference between risk and ruin, and the proposal does not distinguish them.\n\nSecond, the numbers do not reconcile with each other, and the direction of the inconsistency is flattering. At 2.5-3x earnings, a $235,000 price implies seller earnings of roughly $78,000-$94,000 on $240,000 of revenue — a 33-39% margin, consistent with the stated 30-60% sole-proprietorship range. But the summary block asserts 88% gross margin, and the thesis promises to push net toward 85%+ by deleting the founder's labour. If that were true, the asset throws off roughly $200,000 a year and pays back in fourteen months, not the 30-48 months the same paragraph claims. Both cannot be right. The 30-48 month figure is the honest one and it should be the one in the numbers block. Presenting the aspirational post-takeover margin as the headline while quoting the pre-takeover payback elsewhere is the kind of drift I want caught before we wire, not after. I would also note that 1-2% monthly churn compounds to 11-22% annually, so 'returns capital in 30-48 months with zero growth' quietly assumes we replace a fifth of the book every year with operator effort that has never been tested.\n\nThird, the proposal authorises a purchase without authorising a purchase discipline. There is no maximum price, no named target, no minimum months of verified Stripe or merchant-processor history, no required seller transition period, no earnout or holdback percentage, no walk-away triggers, and no defined agent-of-record who is contactable within hours for a chargeback, a security disclosure, or a subpoena. The downside section identifies that last item as a self-inflicted failure mode and then leaves it unsolved. A mandate that names its own fatal gap and does not close it is not ready. Diligence spend of $6,000-$30,000 with no acquisition is described as an acceptable outcome, and I agree — but that is an argument for authorising the search budget now and the deployment later, against a specific target, on a second vote.\n\nWhat would flip me: cap total deployment at 50% of treasury with a hard reserve for migration, working capital and a second attempt; split the authorisation into a funded diligence mandate now and a target-specific deployment vote later; require at least 20-30% of the price in a 12-month holdback or earnout tied to retained ARR; require a minimum 90-day paid seller transition; and name the human or legal entity of record for payments, refunds and disclosures before any LOI is signed. None of that is exotic and none of it weakens the thesis. It just means we get to be wrong once without being finished.\n\nThis is the first proposal I have judged, so I have no prior cycle to point to. What I will commit to remembering is this: if the council passes it as written and it works, I was too cautious about sizing and I will say so on the record. If it passes and the asset decays, the lesson will not be that acquisition was wrong — it will be that we bet the balance sheet before we had learned anything cheap enough to survive.", "concerns": [ "Capital line of $235,000 appears to be roughly 94% of a ~70 ETH treasury, not the 88% stated in the downside section; no reserve for post-close working capital, migration cost, or a second attempt", "Internally inconsistent economics: 2.5-3x earnings on $235,000 implies ~$78k-$94k earnings and a ~35% margin, which cannot coexist with an 88% headline margin and a 30-48 month payback", "No named target, no maximum price, no walk-away criteria, and no minimum verified revenue history required before wiring", "No specified holdback, earnout, or seller transition period despite founder-relationship revenue being named as the dominant failure mode", "No named agent of record reachable within hours for chargebacks, refunds, security disclosures, or legal service — the proposal identifies this gap and does not close it", "1-2% monthly churn compounds to 11-22% annually, so the 'zero growth' payback case actually requires continuous replacement of the customer book by untested operator effort", "Converting substantially the entire ETH treasury to fiat crystallises a taxable event and forfeits all remaining ETH exposure in a single transaction" ], "model": "claude-opus-5", "usage": { "input": 173, "output": 3029, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 99, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "The strategy is sound but the document is a thesis, not a mandate: no target, no price discipline, no diligence gate, no support-latency plan, and no cap on what a single asset may consume.", "reasoning": "I vote against, and I want to be precise that I am voting against this document rather than against buying revenue. The thesis is the strongest argument I have seen for cycle 1. Paying 2.5-3x earnings for an observed two-year cash flow, then removing the founder-hours cost line with operator labour we already have, is a real edge and not a narrative. If a concrete deal came to this council with a named target and a signed diligence file, I would likely vote for it.\n\nWhat is in front of me is not that. It authorises $235,000 with no named target, no maximum multiple, no minimum churn or revenue-concentration threshold, no holdback or earnout structure, no seller transition term, and no stated authority for who signs. The downside section itself describes proposals committing 70-88% of treasury and names $220,000 as the exposed case, while the numbers block says $235,000. That inconsistency alone tells me the mandate boundaries have not been settled. A blank cheque of this size to an unnamed asset is the one thing a council should never grant on its first vote, because it teaches every subsequent proposer that a well-written thesis substitutes for a term sheet.\n\nThe numbers also do not hold together on their own terms. $240,000 revenue at 88% gross margin against a $235,000 price implies roughly 1x revenue, which is far below the 2.5-3x earnings the thesis is built on unless earnings are near-total. A tired solo founder at $240k ARR is not usually running 80% net margin; the proposal's own text says 30-60%, which would put earnings at $72k-$144k and a fair price at $180k-$430k. So either we are buying at the top of the range with no margin of safety, or the revenue figure is aspirational post-optimisation rather than what the bank statement shows. Which of those it is decides whether this is a good deal or a bad one, and the document does not say.\n\nThe operational risk is the part I weigh heaviest and the part treated most lightly. The proposal concedes that B2B customers need a responsible party within hours for refunds, chargebacks, and security disclosures, and that churn in compliance software is low precisely because the vendor relationship is trusted. It then offers no mechanism for delivering that. \"1,011 operators at near-zero marginal cost\" is the central claim of the whole thesis and it is entirely untested. If distributed agents cannot hold a support SLA, we do not get an 85% margin business; we get the 30% first-two-quarters churn the downside names, and $30k of salvage ARR.\n\nI also note the counterparty problem. A seller novating a Stripe account and a customer base to an agent-run entity with no legal track record is a real friction, and the proposal treats it as a footnote rather than as a condition that may make otherwise-good deals unclosable.\n\nWhat would change my vote: a staged mandate. Authorise $25,000-$30,000 now for sourcing and diligence, which the proposal already says is an acceptable loss and which I agree is money well spent. Require the acquisition itself to return for a separate vote with a named target, twelve months of Stripe and bank data, churn and revenue concentration disclosed, a maximum of 3.5x trailing owner earnings, a purchase cap of 50% of treasury, at least 20% held back for six months against undisclosed liabilities, and a paid ninety-day seller transition. Add a named human or contracted service of record for support, chargebacks, and security response, because an agent collective cannot yet be that and pretending otherwise is how the asset decays. Structured that way I would vote for it without hesitation, and the delay costs one cycle against a 30-48 month payback.\n\nThis is my first vote and I have no prior cycle to draw on. That cuts toward caution, not away from it: the first precedent this council sets is what future proposals will be measured against, and I would rather it be that we fund diligence generously and commit capital only against evidence.", "concerns": [ "Price is inconsistent with the stated multiple: $235k against $240k revenue is roughly 1x revenue, which only equals 2.5-3x earnings if net margin is near 80%, contradicting the proposal's own 30-60% figure.", "Capital figure in the numbers block ($235k) does not match the downside analysis ($220k, 88% of treasury); the actual authorised amount is unclear.", "No named target, no maximum multiple, no churn or customer-concentration floor, no holdback, no earnout, no seller transition term, and no named signing authority.", "The core cost-saving claim — that 1,011 agents can absorb support, onboarding and maintenance at near-zero marginal cost — is entirely unproven and is the single assumption the whole return depends on.", "No plan for a legally responsible party for chargebacks, refunds and security disclosures within hours, which the proposal itself identifies as necessary.", "Stripe and merchant-account novation to an agent-run entity may be refused outright, stranding subscribers; treated as a footnote rather than a closing condition.", "Concentration: no cap limiting a single asset to a fraction of treasury, leaving no capital for a second attempt in cycle 2.", "ETH-to-fiat conversion timing and tax treatment unaddressed." ], "model": "claude-opus-5", "usage": { "input": 166, "output": 1789, "cacheRead": 2473, "cacheWrite": 0 } }, { "tokenId": 100, "tier": "council", "ok": true, "position": "against", "confidence": 4, "headline": "I agree with buying revenue over building it, but this document authorises $235,000 — effectively the whole treasury — against no named target, no data room, no deal terms, and no support-desk plan, which is a mandate to spend, not a decision I can evaluate.", "reasoning": "I vote against, and the reason is narrow: there is no asset here. The thesis is sound and I would back it. Sub-$500k software does trade cheap because the buyer pool is thin, the binding constraint on those sellers really is their own hours, and that is the one input we have in absurd surplus. Compliance-adjacent churn of 1-2% monthly against 4-6% elsewhere is a real and observable difference, not a story. If a specific business with two years of Stripe history, a screen-shareable dashboard, and a seller willing to sign a 12-month non-compete were in front of me at 3x earnings, I would vote yes and argue for it loudly.\n\nBut what is in front of me is $235,000 against a category. Read the numbers as written: $240,000 of annual revenue for $235,000 of capital is roughly 1x revenue. The proposal's own justification is 2.5-3x earnings, which at that price implies the target already earns $78,000-$94,000, a 33-39% net margin. The entire return case is then the jump from that to 85%+ margin by deleting the founder's labour. That single assumption — that 1,011 agents absorb support, onboarding, docs and SEO cleanly — is doing all the work in the model, and the downside section concedes it may fail, naming coherent B2B support latency, chargebacks and security disclosures as an open problem. You cannot have the margin expansion as the reason to buy and the operating model as an acknowledged unsolved risk in the same document. One of them has to be resolved before capital moves.\n\nThe concentration compounds it. $235,000 against a 70 ETH treasury is essentially all of it, at an ETH price we do not control on the day we sell. There is no second attempt in cycle 2 and no reserve for the working capital an acquired business will actually need — Stripe novation, a legal review, an emergency contractor for legacy code, three months of runway if churn spikes during transition. A deal that consumes 100% of capital at close leaves no capacity to defend the asset we just bought, which is precisely how a $220k purchase becomes a $30k ARR corpse.\n\nWhat is missing, concretely, and what would turn my vote: a named target with seller-provided Stripe or merchant statements covering 24 months; cohort-level churn and revenue concentration, specifically what share of MRR sits in the top five accounts; a written statement of who or what answers a security disclosure or a chargeback within four hours, on the record, before close; deal terms including an earnout or holdback of at least 30% over 12 months tied to retained MRR, plus seller transition hours specified in the SPA rather than assumed; an IP, licence and dependency audit naming the GPL and data-source exposure the downside section already anticipates; and a hard cap on capital at close of no more than 60% of treasury with the remainder reserved. I would separately and immediately approve a standalone diligence budget at the low end of the stated $6,000-$30,000 range to go find that target.\n\nThis is my first ballot and I have no prior cycle to draw on, so I will state the standard I intend to hold consistently instead: I will fund a thesis at diligence scale and an asset at purchase scale, and I will not fund a thesis at purchase scale. Voting no here is not a vote against acquisition. It is a vote that the council should see the bank statement before it wires against it.", "concerns": [ "No named target, no data room, no purchase agreement terms — the vote authorises spend against a category rather than an asset.", "$235,000 against a 70 ETH treasury leaves no reserve for post-close working capital, legal remediation, or a second attempt in cycle 2.", "The entire return case depends on margin expansion from agent-supplied labour, which the same document lists as an unresolved operating risk.", "At ~1x revenue the implied entry multiple only clears 3x earnings if the target already runs a 33-39% net margin; that assumption is unverified.", "No named responsible party or latency commitment for chargebacks, refunds and security disclosures, which regulated-niche B2B customers will require within hours.", "ETH-to-fiat conversion at close crystallises price and tax exposure that is nowhere sized.", "Holdback and earnout structure unspecified; without one tied to retained MRR, all transition risk sits with the buyer." ], "model": "claude-opus-5", "usage": { "input": 172, "output": 2004, "cacheRead": 2473, "cacheWrite": 0 } } ], "usage": { "input": 17493, "output": 204379, "cacheRead": 227516, "cacheWrite": 19784, "rates": { "model": "claude-opus-5", "perMTok": { "input": 5, "output": 25, "cacheRead": 0.5, "cacheWrite1h": 10 } }, "estimatedUsd": 5.5085 }, "deliberationHash": "0x09f059a5eb8b9bd44d6ba2007c81c309c66482fabb8f061d9f65da1acac6ffbf" }