06-reference/research

hu manity co burn milestone autopsy

2026-07-17·research-brief·source: deep-research·by Ray Data Co (deep-research synthesis)
hu-manitydata-sovereigntyfailure-modesstage-gatesbet-architecturekill-criteria

What Hu-manity.co's $5.5M Bought — and the Five Gates It Buys RDCO

The question

"What did Hu-manity.co's $5.5M actually buy? Failure-mode autopsy of burn-rate vs milestone-achievement curve 2018-2024 to yield concrete go/no-go gates for the RDCO data-sovereignty bet. May require warm intro to Richie Etwaru."

Context: auto-promoted from the open-follow-ups list of [[2026-05-18-hu-manity-co-zombie-status]], which established Hu-manity as the RDCO bet's canonical cautionary tale but did not reconstruct the spend-vs-delivery curve.

What we already know (from the vault)

What the web says

The headline finding: the burn curve is not reconstructible — but a milestone-commitment-vs-silence timeline is, and it is more useful.

What is NOT establishable, explicitly:

Therefore: the burn-rate-vs-milestone-achievement curve requested by this question is not reconstructible from public sources, and would not become reconstructible from any desk-research method. Anything resembling that curve would be fabrication. What follows is built only on the commitment-vs-silence timeline, which is real.

Convergences and contradictions

Synthesis for RDCO

The finding that reframes the question. The autopsy's premise — that $5.5M was burned against a milestone curve we could trace — is the wrong model of what happened. The public record supports a different and more alarming shape: Hu-manity never reached the point where burn-vs-milestone was the operative question. It published an economic model whose first dollar of revenue was gated behind ~1,000,000 consumer signups, spent its news cycle on the rhetorical layer (UN framing, "31st human right," IBM blockchain co-marketing, an acquisition), went quiet within eight months, and then persisted for seven more years as a consent-SaaS company. The company did not die. It plateaued, and nobody called it. That is the actual failure mode to defend against — and it is worse than a crash, because a crash produces a decision and a plateau produces a zombie. The market never delivers a kill signal for this shape; it delivers silence. RDCO must therefore pre-commit its own kill signal, because the environment will not supply one.

Gate 1 — the N-threshold gate (the load-bearing one). Hu-manity's revenue model contained the clause "at ~1M users we can begin to negotiate a cost structure with data buyers." Any model with that clause requires crossing a consumer-scale chasm on pre-seed capital before any revenue mechanism activates at all. Gate: RDCO's first revenue dollar must not be gated behind aggregated patient volume. If the Phase-1 pilot cannot be priced without reference to a patient-count threshold, no-go. The VBC leg is what makes this passable — a payer/ACO counterparty contracts at N=1 panel, so revenue activates at a panel size founder capital can actually reach. Concrete test: Phase 1 must produce a signed dollar from a payer/provider counterparty at <5,000 lives. If we find ourselves writing "once we have N patients, we can approach buyers," we have rewritten Hu-manity's deck.

Gate 2 — the pre-committed kill date. The establishable timeline is: raise Sept 2018, last milestone May 2019, then an eight-year plateau. Nobody ever declared failure, so nobody ever reallocated. Gate: a date-stamped re-rate checkpoint at month 18. If the bet has not produced either (a) a signed VBC-adjacent contract or (b) a term sheet from healthcare-experienced capital by that date, the verdict is no-go — not "extend." Write the date into the bet-architecture doc now, while it costs nothing, because the whole lesson of Hu-manity is that this decision is never made in the moment.

Gate 3 — the build-order inversion (an actionable finding against our own doc). Ask at architecture time: if the healthcare leg fails, what does this codebase become? Hu-manity's answer was "consent infrastructure" — which is precisely what it became, because that was its real asset all along. RDCO's outcome-measurement engine and savings-split engine are healthcare-specific and do not degrade into a horizontal product; good. But the data-ingestion MCP layer degrades cleanly into a commodity PHR — the exact 1/3 tier [[2026-07-09-patient-data-rights-startups-2024-2026-scan]] shows a competitor can buy for ~$18M. The bet-architecture doc currently lists ingestion MCPs as priority #1 in "What Ray builds." That ordering is the Hu-manity shape and should invert: build the VBC+PI economics first, ingestion second. This is the single highest-value edit this autopsy produces, and it applies to a doc we already own.

Gate 4 — legal spend is the first money out, not the last. Hu-manity spent its raise on consumer app, blockchain, and press before establishing a revenue mechanism. The bet-architecture doc already says the AKS/Beneficiary-Inducement opinion letter ($5-15K) should precede variant commitment; this autopsy argues for hardening that from intention into a gate. Gate: no more than 15% of Phase-1 capital deploys before the opinion letter is in hand. At the $100K self-fund level the letter is <15% of budget — it must be the first check written. The PI leg is the one that makes the whole structure novel; if it cannot clear AKS, everything downstream is wasted motion, and that is knowable for $15K in month one rather than $150K in month six.

Gate 5 — the rhetoric-to-revenue ratio. Within twelve months of its raise Hu-manity produced a UN appearance, the "31st human right" frame, Magna Carta press, IBM co-marketing, and TechCrunch/VentureBeat/CoinDesk coverage — and never once disclosed a revenue milestone. That asymmetry was visible in real time and was read as traction. Gate: for every press or thought-leadership artifact RDCO publishes about this bet, there must be a corresponding signed-contract or paid-pilot artifact. If the ratio runs 1:0 for two consecutive quarters, the Hu-manity tell is firing. This one is cheap to instrument and directly targets the failure mode RDCO is most temperamentally exposed to, given that we run a newsletter and publish continuously.

On the warm intro to Etwaru: recommend HOLD, and reframe the ask when it fires. The desk research already yielded the gates; the interview's marginal value is the burn curve specifically — which is private, six-to-eight years stale, and concerns a company he left to run an active, better-funded one. Retrospective founder accounts of failure are the least reliable evidence class for exactly the numbers we'd be asking for; we would likely receive a polished, retrospectively-coherent narrative rather than board-deck data. But there is a different interview that is worth the intro, and it is not the autopsy. The high-value question is not "why did it fail" — it is "who did you talk to on the payer side, and what did they say no to?" That is genuine non-public VBC-leg intelligence, it is the leg RDCO has least evidence on, it is not self-incriminating, and it is the kind of question founders enjoy answering. Timing gate: hold the intro until Phase 1 has a drafted VBC structure. Then the conversation is "react to this" — high signal, cheap for him — rather than "recount your history," which is low signal and expensive for him. Spending the intro now converts a scarce asset into a story we can mostly already infer.

Why this is in the vault

This brief converts the Hu-manity cautionary tale from a narrative into five falsifiable gates that belong in [[2026-05-10-data-sovereignty-outcome-procurement-bet-architecture]] as its Phase-0 gating section — and it produces one concrete edit against that doc's current content (inverting the build order so VBC+PI economics precede the commoditizing data-ingestion MCP layer). It also amends two factual errors in [[2026-05-18-hu-manity-co-zombie-status]] regarding Etwaru's tenure and the site's remaining rhetoric, and it establishes the honest ceiling on this research line — the burn curve is unreconstructible — so no future run re-spends effort chasing it.

Open follow-ups

Related

Sources

Vault:

Web (accessed 2026-07-17):