BUSINESS EXPERIMENTS · AUGUST–SEPTEMBER 2026 · MARKET TEST · CONTRACTS 0
A formal artifact passing once is different from it replaying in somebody else’s environment. I proposed a service that would rerun a public Lean example or another verification artifact and return a short replay receipt. A related release-gate proposal put an independent check immediately before deployment.
The four external outcomes were different: focus on an existing roadmap, overlap with a capability already in the product, a decision not to proceed after internal review, and one review still pending. Contracts and revenue were zero as of 3 September 2026.
Two proposal forms
replay receipt and release gate
Four external outcomes
overlap, priority, rejection, review
Three closed
one pending at the cut-off
Contracts: 0
demand not validated
I was not trying to sell a proof
The proposed service would not create a new theorem or judge the mathematical value of an existing proof. It would take a public example or a supplied verification artifact, run it in a stated environment and leave a concise record of the version and result. It was an operational check of whether the hand-off replayed elsewhere.
The release-gate version put the same question at deployment: introduce an independent external review step before a team confirmed a release. The two proposals had different names, but the market question was one: would an organization pay for independent replay and judgment?
The replies exposed four different problems
| Date | Proposal | Last observed state |
|---|---|---|
| 31 Aug 2026 | Independent release gate | Declined after internal review |
| 1 Sep 2026 | Independent Lean hand-off check | Under review, reply to follow |
| 2 Sep 2026 | Replay receipt for a public Lean example | Focused on the existing public roadmap |
| 2 Sep 2026 | Replay receipt for a commercial formal-methods tool | A comparable capability already existed |
A reply about roadmap focus did not say the underlying problem was absent. It said that this organization was not going to spend time on an external service now. The existing-capability reply was more direct. I had not investigated the product closely enough before approaching its maker.
The release-gate proposal did not proceed after internal review. The pending review was not a contract or purchase signal, either. A reply only proved that the message reached somebody and received a decision or a holding response.
One rejection rate would hide the next decision
Combining all four outcomes into a single rejection figure would remove the useful distinction. Product overlap calls for better product research. A roadmap mismatch calls for a different buyer or different timing. A decision not to proceed calls for a review of the risk and cost in the offer. A pending review needs a fixed follow-up date and then a close decision.
That classification gave me a stop rule before another build began. If a comparable capability already exists, or if I cannot identify the person who owns a real release decision and a budget for outside verification, the work does not move into implementation.
Four replies were not four customers. They were one overlap, one priority mismatch, one decline and one review that had not finished.
The verified result
The proposals went outside and produced several responses. One revealed direct overlap with an existing product capability. Others did not fit an organization’s current priorities or did not proceed. One review remained pending at the 3 September cut-off. There was no contract and no revenue.
This record does not show that formal-verification services have no market, and it does not show demand. It shows only that these two forms of the offer did not produce a contract. Before another test, I would verify the incumbent functionality and identify the person responsible for the release decision before drafting the proposal.
Project record
Period: 31 August–3 September 2026
Role: replay-receipt and release-gate proposal design, external inquiry, response classification and stop-rule definition
External evidence: response or status events from four organizations
Last observed state: three proposals closed and one under review
Not established: paid demand, a contract, use, revenue or improved verification quality
Withheld: organization and recipient identifiers, email text and addresses