What to Do After Your MVP Launches: The Next 90 Days
Launch is the start, not the finish. The metrics to watch, how to read the signal, and how to decide whether to double down, iterate, or pivot.
TL;DR
- Launch is the start of the learning, not the end of the work. The MVP's whole purpose is the data that arrives after it goes live.
- Watch three things first: activation (do users complete the core workflow?), retention (do they come back?), and qualitative signal (what do they say and where do they get stuck?).
- Use Sean Ellis' test to read product-market fit: if 40%+ of users would be "very disappointed" without your product, you likely have it.
- The decision after 90 days is one of three: double down (you have signal, scale it), iterate (partial signal, fix the gaps), or pivot (no signal, change the bet).
- Scope version two from evidence, not your original roadmap. The best post-launch builds fix what the data exposed, not what you assumed at the start.
What should you do after an MVP launches?
After an MVP launches, the job is to measure real user behaviour (activation, retention, and qualitative feedback), read whether the product is finding demand, and decide whether to double down, iterate, or pivot. The next build phase is scoped from that evidence, not from the pre-launch roadmap.
The short answer
Stop building for a moment and start watching. Measure whether users complete your core workflow, whether they come back, and what they tell you, then decide, from that evidence, whether to scale the product, fix it, or change the bet. The MVP existed to generate exactly this data; the mistake is to keep building on the original plan as if launch taught you nothing.
If you are still on the clock before launch, the 21-day MVP development process shows how Day 21 hands you this learning loop. If you need the definition of what you just shipped, see what a minimum viable product is.
The three signals to watch first
Do not drown in a dashboard. Three signals tell you almost everything in the first 90 days.
1. Activation: do users complete the core workflow?
The single most important early metric. Of the people who sign up, how many actually complete the one core workflow your MVP was built around? A low activation rate means users cannot or will not get to the value, usually a friction or onboarding problem, not a demand problem. Fix activation before you judge anything else, because every downstream metric is polluted by users who never reached the product's point.
2. Retention: do they come back?
Usage without return is a leaky bucket. Plot how many users are still active 1, 7, and 30 days after signup. A retention curve that flattens, where a stable share keeps coming back, is the clearest sign of real demand. A curve that decays to zero means people tried it and left, which is a signal no amount of new features will fix. Y Combinator and Lenny Rachitsky both treat the retention curve as the truest early read on product-market fit.
3. Qualitative signal: what do they say, and where do they stick?
Numbers tell you what; conversations tell you why. Talk to your first users, the ones who stayed and the ones who left. Watch session recordings for where they hesitate. The richest version-two ideas come from watching a real user get stuck on something you thought was obvious. Steve Blank's rule applies: the answers are outside the building, with users, not inside it with your assumptions.
How to read product-market fit
The cleanest early test comes from Sean Ellis, who ran growth at Dropbox and LogMeIn: ask your active users how they would feel if they could no longer use the product. If more than 40% say "very disappointed," you likely have product-market fit and should pour fuel on it. Below 40%, you do not yet, and the job is to find out why. Superhuman built its entire early roadmap around this survey, segmenting the "very disappointed" users and building for them specifically.
Pair the survey with the retention curve. A flattening retention curve plus a 40%+ "very disappointed" score is as strong an early signal as you will get. Weak on both means the bet needs to change, not the feature list.
The 90-day decision: double down, iterate, or pivot
After roughly 90 days of real usage, the data points to one of three moves.
| The signal you see | The move | What it means for version two |
| Strong activation + retention flattening + 40%+ PMF | Double down | Scale what works: growth, polish, and the next-most-wanted features |
| Partial signal, some retention, clear friction points | Iterate | Fix the specific gaps the data exposed; re-measure before expanding |
| Weak activation + decaying retention + low PMF | Pivot | Change the core bet: new workflow, new audience, or new problem |
Most first MVPs land in the middle: partial signal. That is not failure; it is the normal result of a real experiment, and it is precisely why you built an MVP instead of the whole product. You now know which parts to keep and which to change, which is knowledge the full-build founder pays for with a year and a budget.
Scope version two from evidence, not the roadmap
Here is the discipline that separates founders who compound from founders who thrash. Version two should fix what the data exposed, not resurrect the feature list you cut before launch. Those cut features were assumptions; the launch replaced assumptions with evidence. Re-scope from the evidence.
Practically: take the same core-workflow discipline you used for the MVP and apply it to the next build. What is the one thing the data says will most improve activation or retention? Build that, measure again, repeat. Our guide on how to scope an MVP works just as well for scoping version two: the method does not change, only the inputs do. When the next phase is real product engineering rather than a quick fix, that is where custom software development picks up, or AI and machine learning if the signal points to intelligent features.
Many founders come back to 21-Day MVP Development for a tight iterate cycle: one workflow fix, locked scope, measured again in weeks rather than quarters.
MVP live and not sure what the data is telling you?
Book a free scoping call. We will help you read your activation and retention, decide whether it is a double-down, an iterate, or a pivot, and scope version two from the evidence, not the old roadmap.
Frequently Asked Questions
What should you do after launching an MVP?
Measure real user behaviour, activation, retention, and qualitative feedback, read whether the product is finding demand, and decide whether to double down, iterate, or pivot. Scope the next build from that evidence rather than from your original pre-launch roadmap.
What metrics matter most after an MVP launch?
Activation (do users complete the core workflow?), retention (do they come back at 1, 7, and 30 days?), and qualitative signal (what users say and where they get stuck). A flattening retention curve is the clearest early sign of real demand.
How do you know if your MVP has product-market fit?
Use Sean Ellis' test: ask active users how they'd feel without the product. If more than 40% say 'very disappointed,' you likely have product-market fit. Pair that with a retention curve that flattens rather than decaying to zero.
When should you pivot after an MVP?
When the data shows weak activation, decaying retention, and a low 'very disappointed' score after real usage. That combination means the core bet isn't landing, changing the workflow, audience, or problem is more productive than adding features.
How do you decide what to build in version two?
Build what the data says will most improve activation or retention, not the features you cut before launch. Those were assumptions; the launch replaced them with evidence. Apply the same one-core-thing scoping discipline you used for the MVP.
How long should you wait before making post-launch decisions?
Give it around 90 days of real usage, enough to see a retention curve form and gather qualitative feedback, but not so long that you're building on the original plan instead of the evidence. Fix activation first, since it pollutes every downstream metric.