What changes the moment you launch
Before launch, your job was to build the thing. After launch, your job is to learn what the thing should become. The skills barely overlap. The founders who struggle keep optimising for shipping features when they should be optimising for learning.
The first 90 days
- Weeks 1–2: Watch, don't ship. Resist adding features. Talk to every early user. Find where people get stuck and where they light up.
- Weeks 3–6: Fix the leak before filling the bucket. If users drop off at onboarding, no new feature helps. Remove the single biggest point of friction.
- Weeks 7–12: Double down on the one thing that works. A pattern emerges — one use case, one segment, one moment of real value. Build around that, ruthlessly.
Instrument everything
You can't find product-market fit blind. Before you chase PMF you need to see behaviour: activation rate, time to first value, retention curves by cohort, and where people drop. If you launched without analytics, fixing that is week-one work.
The iteration loop
Tighten the cycle to its smallest honest form:
Observe a real behaviour → form one hypothesis → ship the smallest change to test it → measure → keep or kill.
The teams that find fit fastest aren't shipping the most — they're running this loop the most times per month with the most intellectual honesty about results.
Signs you're getting close
- Retention curves flatten instead of decaying to zero — people stick.
- Users get upset when the product is down. They'd miss it.
- Growth starts coming from word of mouth, not just your own pushing.
- You can describe, in one sentence, exactly who it's for and why they love it.
None of this requires building more. It requires building the right things, proven by real behaviour — exactly where a partner who's done it before earns their keep.