Most failed startups take their evidence with them. A month of interviews, a pilot, and a clear answer about why something does not work disappears, and the next team pays for the same lesson. When a Circle closes here, it writes down what it believed, what it tried, what the evidence said, and what it would tell whoever comes next.
Closing is a normal outcome.
A month that produces a clear no is a month well spent. Nothing in the archive is framed as a failure or attached to a rating.
Publishing is a separate decision.
The debrief is private to the Circle by default. It becomes a public brief only when every active member approves it.
Evidence, not confession.
Briefs carry hypotheses, methods, and findings. Personal criticism, private conversations, customer data, and credentials stay out.
Credit survives the Circle.
Contributions stay attached to the people who made them, whether the Circle continued, reformed, or closed.
2 published briefs
Read the month before you repeat it.
Closed June 2026 · Trades & construction
Copperline
Quote software for residential copper rewire electricians.
Closed
What we believed
Residential electricians lose jobs to slow quoting and would adopt a mobile quote builder.
What we tried
Interviewed 24 electricians, prototyped a photo-based quote flow, and piloted it with six contractors across two cities.
What the evidence said
Five of six pilot users finished quotes faster, but none sent them to customers within the same week.
Where it broke
Quoting speed was not the bottleneck. Electricians only bid on jobs that arrive through referrals, and referral volume was unchanged by faster quotes.
Closed May 2026 · Food & hospitality
Third Place
Discovery tools for independent cafes hosting evening events.
Closed
What we believed
Independent cafes lose evening revenue because promoting events takes more time than they have.
What we tried
Built an events page generator, visited 19 cafes, and ran promoted events at five of them.
What the evidence said
Attendance rose at four of five promoted events, but all five owners said they would not pay monthly for the tool.
Where it broke
Owners already promote through regulars and Instagram stories for free. The tool saved minutes, not hours.
Building on a brief
A dead end for one team is a head start for another.
Every brief ends with the question the Circle did not get to answer. A forming Circle can take that thread as its starting opportunity instead of a blank page, and begin the month already knowing which twenty interviews have been done and which assumption has already been ruled out.
When a Circle builds on an archived brief, the original team is credited on the record. Preserved learning is the part of this network that compounds — not follower counts, not scores.