How to measure product-market fit for an app
90% of mobile apps do not survive their first year, often after scaling without product-market fit. The signal that cannot be faked is a retention curve that stabilizes.
Article summary
- Median annual retention is 28%; the top of Productivity reaches 85.6%, Social and Lifestyle around 25%.
- The 40% test asks whether the user would be very disappointed to lose the app.
- 7 to 8 positive signals validate PMF; below 4, you need to iterate before scaling.
- For subscription apps, YC looks for monthly retention that stabilizes above 50%.
- Scaling assumes retention above your category median, an LTV/CAC ratio of at least 3, and a repeatable channel.
Summarize this article with
90% of mobile apps do not survive their first year. Not because the founders are bad, and not because the app is poorly built. Because they scaled before they had product-market fit, and they thought they had PMF because their downloads were going up.
firstapp sees it regularly: the apps that last are the ones whose founders learned to read the right signals before pressing the accelerator. Here is how to measure it concretely, with figures, not intuitions.
Product-market fit: a concrete definition for a subscription app
PMF is the moment when your app answers a real problem, for a defined population, in a way no other solution does as well.
Most articles on the subject stop there. In practice, for a subscription mobile app, PMF has a very operational translation: your subscribers renew without you having to convince them.
PMF is not a magic moment. It is a measurable configuration, specific to your model. A freemium app and a hard-paywall app do not have the same signals or the same benchmarks. The retention of a monthly subscriber and of an annual subscriber are not read the same way.
The most useful definition comes from YC: PMF is "the level of satisfaction that allows durable growth". Two conditions at the same time: a retention curve that stabilizes, and organic growth of new users month after month.
Newsletter
Once a week.
Last issue
We opened 10 apps to see how they sell a subscription.
+534 people read us each week
Retention: the only signal that does not lie
There are dozens of ways to measure PMF. NPS, "very disappointed" surveys, satisfaction rates. All of them can be skewed by a selection bias, a badly asked question, or an opportunistic moment.
The cohort retention is the only signal you cannot cheat.
The logic is simple: if your retention curve never touches zero and stabilizes, a fraction of your users keeps using your app without you pushing them. That is the operational definition of PMF.
The other way around, if the curve keeps falling toward zero, you do not have PMF, whatever your download growth. You will empty your user base faster than you fill it.
For subscription apps, two key retention metrics:
- Monthly retention : the % of subscribers still active after 1 month
- Annual retention : the % of subscribers still active after 12 months
💡 Key figure : Median annual retention is 28% according to RevenueCat SOSA 2026, down from 31% the previous year. That is your starting line, not your target.
Retention benchmarks by app category (2026 data)
A 25% retention rate can be excellent for a Social app and disastrous for a Productivity app. Always compare your retention to the median of your category, not to the global median.
| Category | Y1 annual retention | PMF signal threshold |
|---|---|---|
| Productivity top | 85.6% | > 60% |
| Travel | ~39% | > 30% |
| Health & Fitness | ~30% | > 25% |
| Social & Lifestyle | ~25% | > 20% |
| Global median | 28% | above the median |
Source: RevenueCat - State of Subscription Apps 2026 · 115,000+ apps · $16B of revenue analyzed
Productivity apps dominate because they create a daily habit with a high switching cost. When your app is part of the user's workflow, they stop leaving.
Retention by payment plan matters too. Monthly plans have a median retention of 18 to 24% at 12 months. Annual plans show ~5%, but that is structural: the churn of an annual subscriber is almost permanent once they leave. It is not an alarm signal. It is the nature of the plan.
To go further on the metrics to track after launch, read our guide on the first revenue of a mobile app.
PMF diagnostic: do you really have the fit?
Here is the reading grid firstapp uses in our audits. Answer each criterion honestly.
| Criterion | PMF signal | Alarm signal |
|---|---|---|
| Retention curve | Stabilizes above 0 | Falls toward 0 continuously |
| D7 retention | > 40% | < 20% |
| D30 retention | > 20% | < 10% |
| Annual retention | Above the category median | Below the median |
| Organic growth | Users from referral | 100% paid or direct |
| Reaction to bugs | Complaints, emails, 1 star | Total silence |
| The 40% test | > 40% "very disappointed" without the app | < 20% |
| Price resistance | A majority accepts an increase | Mass departure at the slightest test |
7–8 criteria: PMF validated - prepare to accelerate. Fewer than 4: no PMF - iterate before scaling. Source: firstapp · based on RevenueCat SOSA 2026 benchmarks and the Superhuman method (Rahul Vohra)
How to read the diagnostic:
- 7–8 PMF signals: you have the fit. Prepare to accelerate.
- 4–6 PMF signals: partial PMF. Identify the criteria in the red and iterate.
- Fewer than 4: you do not have PMF yet. Scale nothing. Iterate everything.
The knockout criterion is the retention curve. If it falls toward zero, the other signals do not count.
How to measure PMF before you have solid data
In the early stage, you do not yet have cohorts over 12 months. Three proxies you can use from the first weeks.
The "40%" test
Ask your active users a single question: "How would you feel if you could no longer use our app?" If more than 40% answer "very disappointed", you are close to PMF. This test, popularized by Rahul Vohra (Superhuman), has its limits, but it gives a fast direction without waiting 12 months of data.
Retention at D7 and D30
Even without long-term data, D7 and D30 are reliable early indicators. Above 40% at D7 and 20% at D30: a positive signal. Below: an onboarding or perceived-value problem to solve before any acquisition.
Interviews about not uninstalling
Ask your active users why they have not uninstalled your app. The answers reveal whether you solved a real problem or simply created a temporary distraction. To validate your app idea before launch, these interviews are the most effective shortcut.
PMF or the illusion of PMF: the 4 classic traps
The Product Hunt launch trap
A download spike at launch creates the illusion of strong demand. Look at retention at D14 and D30. If the curve collapses, you have buzz, not mobile app product-market fit. Those are two very different things.
The paid growth trap
Meta or Apple Search Ads campaigns can keep your base up in volume even without organic retention. Always isolate your organic cohort to measure real PMF. Paid growth masks the absence of PMF. It does not create it.
⚠️ Warning : The trap of a good review in a bad market
Your first 20 users love the app. But they are your friends, tech early adopters, or users recruited with special attention. PMF is measured on a population representative of your target market, not on a biased sample.
The secondary-metric trap
A 70% activation rate is not PMF. Long sessions are not PMF. 5-star reviews are not PMF. Always come back to the cohort retention curve as the primary signal. Everything else is noise.
PMF found: when should you accelerate growth?
YC says it clearly: "Growth is a natural result of doing the right things. It is not an end in itself."
Three conditions to meet at the same time before accelerating:
1. Retention above the median of your category
You do not need to be in the top 25%. But being below the median means you will be filling a leaky bucket. Every euro spent on acquisition will be partly wasted on churn you do not control.
2. Healthy unit economics
Your LTV covers your CAC with a ratio ≥ 3 before you inject paid budget. Scaling with a ratio below 1 only accelerates your destruction of value. To calculate these metrics precisely, our guide LTV and CAC of a mobile app details the method.
3. A repeatable acquisition channel
A one-off organic growth spike is not a channel. You need a repeatable mechanism, ASO, paid, referral, SEO, with a predictable cost and a measurable output. Without that, acceleration stays random.
If all three conditions are met, accelerate. Otherwise, keep iterating the product.
What 20 years of YC startups teach about mobile PMF
YC analyzed more than 3,000 portfolio companies over 20 years to define quantifiable PMF metrics by product type. For subscription apps, the validated PMF signal is a monthly retention curve that stabilizes above 50%, meaning more than half of subscribers are still active after several months.
That figure can look high next to the 28% annual median. The nuance: the annual median includes every app, including those without PMF. The YC benchmark describes the apps that found their market.
The main lesson does not change: grow after PMF, never before. Calyra (ex-Nutria), which we support at firstapp, reached €30,000 in revenue in the first month with 2% monthly churn, because the onboarding and the product had been iterated until retention was solid before any acquisition. Not the other way around.
FAQ
What retention rate shows that a mobile app has PMF?
There is no universal figure. It all depends on your category. Median annual retention is 28% according to RevenueCat SOSA 2026, but Productivity apps reach 85.6% and Social & Lifestyle apps sit around 25%. Always compare your retention to the median of your category. The decisive signal is not the absolute figure: it is that the curve stabilizes rather than falling toward zero.
Can you have PMF without being profitable?
Yes. Product-market fit is about product-market fit, not profitability. An app can have solid retention and organic growth while still having a business model to optimize. Profitability and PMF are two distinct steps. Aim for retention first, then unit economics, in that order.
How long does it take to find PMF for a mobile app?
The apps that find PMF quickly (3 to 6 months) are the ones whose founders ran many user interviews before building. The ones that take 12 to 24 months often built without listening, then pivoted. Early validation of your app idea is systematically the most effective shortcut.
How do you measure PMF for a freemium app?
For a freemium app, the primary signal is the retention of free active users, not the conversion rate to paid. If your free users come back regularly, you have the base of PMF. Conversion measures your monetization, not your product-market fit. The two are worked in parallel, but they are independent of each other.
Your retention is flat and you do not know if you have PMF? Our audits identify the precise levers to improve it and decide calmly when to accelerate.Talk about your project ↗
How do I know if I have product-market fit for my mobile app?
The most reliable signal is a cohort retention curve that stabilizes above zero. Complete it with the 40% test ("very disappointed if you could no longer use it") and three qualitative signals: complaints when it breaks, organic growth by referral, resistance to a price increase. If all three are present with a stable retention curve, you have PMF.