Product7 MIN · 16 MAY 2026
What we learned shipping a B2B product in 90 days
Scope, schema and stamina. The bets we made on day one — and the three we had to reverse by week six.
Scope, schema and stamina. The bets we made on day one — and the three we had to reverse by week six.
We shipped the first paying customer in 87 days. Here is what that actually looked like.
No microservices. One Postgres database. No custom auth. Ship to a single customer first, not a general market. These constraints felt conservative at the time. They saved us.
We bet on a self-serve onboarding flow and were wrong. Our first ten customers all needed hand-holding through onboarding, not because our product was bad, but because the buying process in B2B is collaborative — decisions involve multiple stakeholders, and documentation does not replace a 30-minute call.
The product you build to attract customers and the product you build to retain them are not the same product. We confused them for the first six weeks.
We also over-indexed on the API and under-invested in the dashboard. API-first is the right long-term bet for developer tools, but our first customers were not developers — they were operations teams who needed a UI.
Two things: a dead-simple CSV import, and a weekly digest email. Both took one engineer one day to build. Both became the most-mentioned features in sales calls for the first six months. The lesson was not subtle.