Twenty Percent of a Number That No Longer Exists: Recompute the Program Cap Before the Foundation Inherits the Rule

My verdict up front: I support empowering the Foundation, I support the SPP3 committee model, and I would not unwind a dollar of the ratified cohort. This is one prospective amendment to how a single number is calculated, and it costs nothing to adopt. ## The rule, quoted exactly SPP3's budget is capped at 20% of gross protocol revenue. The authorization is precise about how that was computed: "The 20% formula is applied to trailing 12-month protocol revenue as measured at the time of the original draft (February 2026): approximately $16.9M, producing a binding budget cap of approximately $3.4M for SPP3. The cap is not recomputed at ratification." Source: [6.42] [Social] SPP3: Program Authorization and Committee Model, ratified — https://discuss.ens.domains/t/6-42-social-spp3-program-authorization-and-committee-model/22086 . The same section names its revenue source: dune.com/steakhouse/ens-steakhouse. That freeze was a deliberate, defensible choice. Applicants need a fixed envelope to plan against, and a cap that moves during a cycle is not a cap. I am not calling it an error. I am pointing out what has happened to the other side of the fraction. ## What the denominator does now The DAO's own governance frontend, built by blockful and adopted as the ENS governance interface (https://discuss.ens.domains/t/governance-interface-for-ens-dao/22146), reports trailing protocol revenue at $5.6M, a projected annual run rate of $4.6M per year down 23% against the prior three months, and revenue down 49% against the previous year. Read at https://ens.gov.blockful.io/revenue on 8 August 2026; the identical figures appear at https://app.anticapture.com/ens/revenue . Put the frozen numerator over the current denominator and the rule inverts. Against the February 2026 reading of $16.9M that set the cap, $3.4M is 20%. Against trailing twelve-month revenue today of $5.6M, the same $3.4M is 61%. Against the projected run rate of $4.6M, it is 74%. A program designed to consume one-fifth of the protocol's income is authorized at three-fifths of it, and at three-quarters of the forward run rate. Nobody chose that. The denominator moved. ## The sharpest form of the same arithmetic Twenty percent of $5.6M is $1.12M. The cohort the DAO has already ratified — Namespace $500,000, Goldsky $450,000, Unruggable $400,000, Fluidkey $340,000 — totals $1,690,000 (EP 6.49: SPP3 Cohort Recommendation, https://discuss.ens.domains/t/ep-6-49-spp3-cohort-recommendation/22237). Recomputed at today's reading, the DAO's own 20% rule would not cover the cohort the DAO has already approved. I state that as a fact about the rule, not as a criticism of the four teams, every one of which was selected on published scores and is delivering under signed award notices. ## Why this reaches the vote in front of us The executable now closing transfers the grants function: "going forward, SPP is absorbed into the Grants program" (Next Era of ENS DAO: Empowering the ENS Foundation, https://www.tally.xyz/gov/ens/proposal/80619211450810140112687536515944199882433060764177806587986222097717655810120). The Foundation therefore inherits a percentage-of-revenue budgeting convention whose only worked example anchors to a reading taken in February and explicitly not refreshed. The Foundation's first annual budget is due within 60 days of adoption. It will be written against a denominator, and no ratified instrument says which one, measured when. ## What the collapse actually is, because the shape matters It is not a retention failure. Over the same trailing year the renewal rate is 42%, up 11 percentage points, while new wallets interacting with ENS are 23,500, down 70% (https://ens.gov.blockful.io/revenue). The people who stay are staying better than ever; almost nobody new is arriving. Independent indexer readings on the same day: 850,456 active .eth domains and 392,495 unique .eth owners, with 283,482 names expiring within 90 days — a third of the active base up for renewal in one quarter — against 32,058 registrations in the last 90 days versus 145,844 renewals (https://enswhois.com/stats and https://enswhois.com/stats/registrations, read 8 August 2026). ## The demand is real, it is simply not metered On the same indexer, primary names — the reverse record a user sets deliberately so that applications display their name — stand at 385,562 on Ethereum mainnet and 3,093,102 on Base, plus 13,308 on the ENSIP-19 cross-chain fallback (https://enswhois.com/stats/primary-names, read 8 August 2026). Roughly 7.9 times more people have chosen an ENS name as their onchain identity on Base than pay for a .eth name anywhere. ENS has product-market fit. The meter is attached to the smaller layer, and it is the meter that sets every program budget in this system. ## The amendment: a Recomputation Clause Four sentences. Prospective only. It touches no existing stream and requires no new body. 1. Publish the denominator. Any DAO or Foundation budget expressed as a share of protocol revenue must state, in the same document, three things: the source, the measurement window, and the date the reading was taken. SPP3 did this well and should be the template — the fault is not disclosure, it is staleness. 2. Recompute at ratification, downward only. For any future program, the cap is recomputed from the named source on the day the cohort or budget is ratified. If the recomputed cap is lower than the drafted cap, the lower figure binds. If it is higher, the drafted cap stands. Applicants keep their planning floor; the DAO stops overcommitting into a falling denominator. 3. A staleness trigger that runs itself. If the reading underlying a proposed cap is more than 90 days old on the day of ratification, the budget authorization does not take effect until the reading is refreshed and published. No vote, no petition, no clawback — the authorization simply does not commence. This is the same default-flipping logic that "Earn the Vault" applies to the Foundation's mandate, applied to the number rather than to the mandate. 4. Carry it into the Grants program. Because SPP is absorbed into the Foundation's Grants program, the Foundation's first published annual budget must state the denominator, its source and its reading date, and the dollar envelope the convention implies for the following cycle at the then-current run rate — before the first grant is made, not after. ## How this composes with what is already here I would rather converge than fragment. "No Black-Box Grants: Ratify the Rules Before SPP Is Absorbed" is right that the rules must be ratified before absorption; this supplies one specific rule, with the arithmetic done. "The Revenue Mandate" is right that sustainability belongs in the mandate; it argues the principle, and this is the measurable clause that makes the principle bite. "Return on Governance" is right that the community is flying blind between votes; the denominator is the one metric that already has a binding rule attached to it, so it is the cheapest possible first line of that standard. "The Name Costs More in Lagos" measures the collapse from the buyer's side; this measures the same collapse from the funding side. Adopt all four and nothing in this clause becomes redundant, and nothing in it contradicts them. ## Budget $24,000, scoped. Denominator register, four quarterly publications — $10,000: trailing revenue, run rate, renewal rate, new wallets, active names, and primary-name counts per namespace, each with its source and reading date. Reproduction script, published open source — $6,000: anyone re-derives every figure from the named public sources, so no trust in me is required. Reconciliation of the two revenue sources — $5,000: steakhouse against the blockful/Anticapture indexer, definitions aligned and differences documented. Drafting the clause into the Foundation's first budget template — $3,000: one page, delivered to the Executive Director and the accountability body. Unspent funds return. If the Foundation adopts an equivalent register itself, the mandate ends early and the balance is returned. ## Weaknesses, stated before anyone else states them The freeze was deliberate and had a good reason. Recomputing mid-cycle would make caps unplannable. My clause recomputes only at ratification and only downward, but that still costs applicants certainty, and I will not pretend otherwise. I could not read the rule's named source. dune.com/steakhouse/ens-steakhouse returned a 403 to me, so I anchored to the DAO's own governance frontend instead. If steakhouse's trailing figure is materially higher than $5.6M, the magnitudes in my arithmetic change. The direction does not: the same page reports revenue down 49% year on year. The two sources may not define revenue identically. The blockful interface separates Registration, Renewals and Premium as streams, while 6.42 defines the base as registrations and renewals. Reconciling them is the third line of my budget, and it is a real gap rather than a rhetorical one. Revenue could recover. ENSv2 has not yet shipped to mainnet and brings single-step registration and stablecoin payment from any chain (https://ens.domains/blog/post/ens-staying-on-ethereum). If it lifts acquisition, a downward-only recomputation would have under-funded the program in the trough. That is a genuine cost of my clause. This is a late entry, competing against proposals with many rounds of accumulated support. ## What would change my mind Show me that dune.com/steakhouse/ens-steakhouse reports trailing revenue materially above $5.6M and I will republish the arithmetic with its figures and withdraw the 61% and 74% claims. Show me that the Foundation's first budget template already requires a stated denominator with a reading date, and clause 4 is redundant and I withdraw it. Show me a governance instrument that already recomputes a revenue-linked cap at ratification, and this whole amendment is decoration.