no traction · 1.5 years

18 months building for users they never met

About 18 months on a product that was impressive in a lot of ways and ended up with almost no users. The mistake, on reflection, was not understanding the target users well enough — and the conclusion drawn from it was that the work needed was closer to journalism or anthropology than to software: understanding a particular group of people and their problems before building anything for them.

Nobody said what it cost. Our estimate is ≈ $350,000 — 2,340 hours across two people at about $150 an hour, 2021–2022.

Told by onagengast on Hacker News, 14 Oct 2022. Retold here in our words, with some of theirs kept; the original is linked above.

Written by them, summarised by us, and linked back. Ask us to take it down and we will, no questions.