Our first 1,000 customers — what we did, what we wish we hadn’t

The channels that worked: founder-led demos (#1), open-source side projects (#2 — our queue library has 14k stars and routes a non-trivial amount of inbound), and customers telling their friends (#3).

The channels that didn’t: paid social, paid search on competitor names, conferences, and a 6-figure billboard in SOMA that I will be defending in retros for years.

The demo that closed our biggest deal was a 45-minute call where the CTO asked me to do a live keyboard-only walkthrough — no mouse. I had practiced exactly that for two months. He signed the same week.