What went well

Four things worth repeating.

Deleting the onboarding checklist

We replaced fourteen steps with one import and cut go-live from nine days to four. The decision took an afternoon; we had been discussing it for months.

Median go-live: 9 d → 4 d

Owning the SMS bug publicly

We wrote to all 22 affected clinics before any of them wrote to us. Two of them referred us afterwards, which is not what we expected.

22 clinics contacted, 0 churned

The benchmark report

One piece of research produced fourteen trials in three days with no paid distribution. It also gave sales a reason to call people who had gone quiet.

14 trials, 4 multi-location

Referrals held up under growth

Half of new clinics still come from existing ones. We did nothing clever here — we just kept the product boring and reliable for the people already on it.

31 of 63 clinics by referral