Building anVendor: lessons from our first year

A year ago we shipped the first version of anVendor. Here are the things we got wrong, the things we got right, and the things we'd tell ourselves on day one.

Lesson 1 — Pick a wedge sharper than you think you need

Our first positioning was "technology intelligence for go-to-market and research teams." Too broad. Customers couldn't repeat it back. Six months in we tightened to "find every company using [tool]" — one search box, one answer. Demos shortened. Trial-to-paid doubled.

Lesson 2 — Trust beats coverage

We obsessed over expanding the index. Customers cared more about whether the data was right. One wrong result lost more deals than ten missing tools. We rebuilt the confidence pipeline before we expanded coverage.

Lesson 3 — Pricing is product

The free tier started at fifty lookups. Conversion was flat. We dropped it to ten. Conversion doubled within a month. The number doesn't matter — what matters is whether a paying customer feels they're getting value past the wall.

Lesson 4 — Say no, then say no again

Every prospect asks for an integration. Salesforce, HubSpot, Outreach, Apollo, Clay. We built one (CSV export) and held the line. The team's energy stayed on the index. The integrations will come — after the wedge is undeniable.

Lesson 5 — Shipping matters more than strategy

The most consequential strategic decisions came from shipping fast and watching what stuck. The customer who unblocked a meaningful MRR account did it because we shipped CSV export on a Friday. Strategy is a hypothesis. Shipping is the test.