Pre-PMF metrics: what to track before product-market fit
After launch, it is easy to celebrate downloads, signups, and early revenue. Those numbers show that people arrived, not that they received value. Before product-market fit, the more useful question is whether a specific group of users repeatedly does something that proves the product solves a real problem. This guide explains which early signals to watch, which ones to ignore, and how to avoid mistaking retention for fit.
What matters before product-market fit
Product-market fit means solving a specific problem for a specific audience with a solution that genuinely fits their needs. The goal is not to persuade people to keep using the app; it is to reach a point where they want to use it because it consistently delivers value.
Before that point, growth is not the main job. Some growth is necessary to generate enough data to learn from, but the core questions are different:
- Does the solution actually solve the problem?
- Who values it most?
- What makes those users return?
- What would they pay for?
A common mistake is to chase growth too early, often because of investor pressure or a few early wins. That can lead to paid acquisition, burned runway, churn, and a panic-driven push for even more users. Short-term spikes may follow, but sustainable growth usually does not. The better approach is to learn and look for strong signals that fit is emerging.
Metrics to ignore
Early-stage founders are often handed a long list of metrics and acronyms. Not all of them matter at this stage. The following are usually poor guides to whether users are getting value:
- Total downloads: a download does not prove that someone opened the app or benefited from it.
- Total signups: the same limitation applies.
- Social media followers: useful for awareness, but not proof of value.
- App store ranking: it may help growth slightly, but it says little about whether a real problem is being solved.
- Day-one download spikes: early users often behave differently from the eventual audience, so spikes should not be over-interpreted.
- Time in app or number of sessions without context: these can make free users look active without showing value.
The key question is whether a specific group of users gets repeated value. If a metric does not help answer that, it can be set aside.
Behavioral signals to track instead
The metrics that matter before product-market fit are behavioral. They show whether users did the things that indicate real value, not merely whether they showed up. Useful questions include:
- Are users coming back on their own?
- Are they using the core feature?
- Are they willing to pay?
- Are they telling others?
Activation is the central concept. Activated users are more likely to retain and pay. People will not pay for an app they have not truly experienced, and they have not reached the moment when value clicks. Look for patterns across at least two or three cohorts before drawing conclusions. With small numbers, consistency matters more than volume.
Time to first value and time to core value
Two activation metrics build on product-led growth thinking popularized by Wes Bush:
- Time to first value: did the user experience something valuable?
- Time to core value: did the user start building a habit?
In a meditation app, time to first value might be the time needed to complete a first meditation session. Time to core value might be reached when someone has meditated at least four times in a week and started a routine.
The goal is not to over-index on speed, because some actions naturally take time. In the early days, however, the process usually takes too long, so the team needs to help users reach the value moment faster. Relative timing matters more than absolute speed.
A food-scanning app offers a practical example. Users who scanned at least seven foods in a week were much more likely to stay than users who took two to three weeks to reach the same milestone. Scanning was the main way users could check whether a food was safe to eat. The team could therefore focus on helping more users complete that action within a defined period, such as through in-app nudges to scan or examples of what can be scanned.
Active users and word of mouth
Alongside activation, a measure of active users defined by meaningful behaviors is important. App opens alone are not enough. Tracking whether a higher percentage of users are activating, and doing so faster, gives a clear signal that the product is on the right track. That measure should be tied to the key behaviors identified as indicators of value.
Referrals are another strong signal. If 15% or more of new users come through referrals, that is a strong sign of product-market fit. Word of mouth takes time to build in a new app, but a growing share of users acquired through word of mouth is a clear positive signal.
What you cannot measure yet at launch
Some valuable metrics are hard to measure at launch:
- Retention curves need weeks.
- Realized lifetime value per customer can take months to pin down.
- Sustainable growth is similarly slow to assess.
These metrics should not be ignored entirely, but at launch they cannot yet be used to judge success. Focus instead on leading indicators, the behaviors that signal value early, rather than lagging indicators that reflect outcomes further down the line.
- Leading indicators: activation, early retention, qualitative feedback.
- Lagging indicators: lifetime value, revenue, long-term retention.
Leading indicators act as an early warning system. They show whether something may be off track before final results arrive, which lets the team be proactive rather than reactive.
The retention trap
Retention is one of the most commonly recommended early metrics. While retention curves take time to understand, 7-day and 30-day retention are reasonable starting points. The risk is over-indexing on retention too early.
Product-market fit is not the same as keeping customers. It is possible to keep people around without solving their problem. Good retention can make an app feel as if it has product-market fit before it really does. Four patterns can create this illusion:
- Gamification over value. Streaks, reminders, and badges can drive retention without delivering real value. A user may return because of the mechanics rather than because the product is genuinely useful.
- Retention driven by a small group of power users. This is not inherently bad, but it needs watching. If growth only happens within a small group, the product-market fit may not be ready to scale. The app needs to be specific enough to stand out without becoming so niche that it serves only a handful of users who are not representative of a larger market.
- Pricing that masks weak product-market fit. Heavy discounts or extended trials can attract bargain hunters. Strong retention numbers may come from users who signed up for an annual subscription at a very low price but are not actively using the app.
- Annual subscriptions that delay churn rather than prevent it. Annual plans are good for cash flow and have many benefits, but locking users in does not mean they value the product. A long commitment is not proof of genuine attachment: someone who stays because an annual plan is already paid for is not the same as someone who would be genuinely upset to lose the app.
To avoid the trap, pair retention data with engagement data. If users renew but do not use the app, dig deeper before celebrating. Also rely on qualitative signals to understand users' experience and motivations.
Qualitative signals to focus on
In the early days, sample sizes are small, and it is hard to separate noise from signal. That is why qualitative signals are extremely valuable. Product-market fit is qualitative first and quantitative second. Founders often feel it before they can measure it: users reach out unprompted with feedback, say they love the app, ask when features are coming, or refer friends without being asked. Growth also starts to feel easier. These moments can show that the team is on the right track long before statistical significance is reached.
The Sean Ellis test
Sean Ellis studied hundreds of startups to find what separated those that succeeded. He found that successful startups typically had at least 40% of users who would be very disappointed if the product no longer existed. He created a simple PMF test to measure this.
Ask users: "How would you feel if you could no longer use [app name]?"
- Very disappointed
- Somewhat disappointed
- Not disappointed
- I no longer use this
Follow up with: "Could you explain your answer?" That second question is critical. It helps the team understand what actually drives product-market fit, even when the sample is too small for statistical significance.
Practical tips:
- Gather enough responses to get meaningful insight: at least 100 for a general sense, and 500–1,000 if segmenting by signup reason, main feature used, or other factors.
- Survey the right users at the right time: ask those who should have reached their aha moment.
- Do not survey day-one users if it normally takes seven days to get value.
- Do not limit the survey to users who have been around for months, or the results may be skewed toward optimism.
The test gives a qualitative measure of product-market fit and helps identify both the level of engagement and the reasons behind it.
NPS and user interviews
Net Promoter Score correlates strongly with the Sean Ellis test. For example, Ladder found that users who said they would be somewhat disappointed had much lower NPS scores than those who said they would be very disappointed. It can be valuable to ask, alongside the PMF question, why users would or would not promote the app. This shows not only how much people value the product but also what drives advocacy and where improvements could turn more users into promoters.
If there are not enough users for the Sean Ellis test or NPS, the best recommendation is to conduct user interviews and hear directly from customers about why they love or do not love the product. Starting with even five to ten Jobs-to-be-Done interviews can teach a great deal about what drives users to stay and what matters most to them. Prioritize users with whom product-market fit is suspected, as well as recently churned users with a similar need. That comparison shows what the difference is and what to prioritize. It also helps avoid distraction from noisy users and keeps focus on the behaviors and needs that indicate product-market fit.
Finding your North Star Metric
As the team narrows down which metrics predict value and gains insight from qualitative research, an overarching metric is helpful. This is the North Star Metric: the single best metric indicating whether users are getting value, whether the business is getting value, and whether a sustainable business is being built.
After product-market fit, North Star Metrics often look like active subscribers or monthly recurring revenue. Before product-market fit, the metric should be behavioral and focused on actions that show users are truly experiencing value. Examples from established companies include Spotify focusing on time spent listening, Slack focusing on messages sent, and Dropbox focusing on files uploaded. These companies have already found product-market fit, so the goal is not to copy their metric but to find the equivalent for the app at its current stage.
Start with a core hypothesis:
- What is the core problem being solved?
- What behaviors would indicate that it is being solved?
- How often should that behavior occur?
For instance, a budget-tracking app might hypothesize that users who categorize at least five transactions per week are deriving value. Ideally, the metric includes both the specific action and the time frame.
After product-market fit, the North Star Metric should remain stable unless there is a major pivot. Before product-market fit, it is acceptable to refine the metric as more is learned. The team probably will not know with confidence exactly what the action or metric should be, and that is fine.
Defining PMF for your app
Write down what the PMF definition looks like in behavioral terms. A useful template is:
PMF is approaching when [specific user type] repeatedly [specific behavior] because the app helps them [specific outcome].
For example:
PMF is approaching when freelance designers repeatedly open the app on Monday morning to queue the week's client invoices because it helps them stop losing track of billable hours.
Or:
PMF is approaching when new runners repeatedly log a run at least three times a week because the app turns their training history into a clear picture of progress.
PMF signals are directional. At this stage, they might appear in just one segment, one use case, or one geography, and that is fine. It is normal to start niche and expand. If PMF appears in an unexpected segment, that is valuable information. It may mean doubling down on that audience rather than trying to force-fit the original target.
Stress testing your PMF definition
To test whether the definition is strong enough, ask:
- Do activated users who engage in that behavior retain significantly better?
- Does the pattern hold across cohorts?
- Does improving the metric improve downstream outcomes?
It is easy to rush as a startup, so forcing reflection on these decisions is important.
When to move on from this phase
There is no perfect moment, but several signals suggest readiness to shift focus:
- 40% or more of users say they would be very disappointed without the app.
- The retention curve starts to flatten after a few weeks rather than dropping off completely.
- Users come back on their own without push notifications or reminders.
- Organic referrals appear, with people telling others without being asked.
- Users are willing to pay without heavy discounts or extended trials.
Not all of these are required. If a few appear consistently, especially within a specific segment, that is a good sign that something worth building on has been found.
From there, the focus moves to product-model fit: making sure the monetization model matches how users experience value. That is where pricing, packaging, trials versus paywalls, and funnel decisions come in. Channel metrics also become more important, as the team works out which channels will drive scale. Those decisions are much easier once it has been validated that the product solves a real problem for a specific group of people.
A pre-PMF metrics checklist
Use this checklist to keep the pre-PMF phase on track:
- Shift the mindset from growth to learning.
- Identify the vanity metrics to ignore.
- Define what time to first value and time to core value look like for the app.
- Set a behavioral North Star hypothesis.
- Write the PMF definition in behavioral terms.
- Talk to users through PMF surveys, NPS, and user interviews.
- Focus on leading indicators, not lagging ones.
It is hard to resist the pull of growth, especially right after launch when the numbers are moving. But if product-market fit is nailed first, everything that comes after, including the funnel, pricing, and ads, will work much better.
| Ignore | Focus on |
|---|---|
| Total downloads | Users returning on their own |
| Total signups | Core feature usage |
| Social media followers | Willingness to pay |
| App store ranking | Users telling others |
| Day 1 spikes | Repeat usage in the first week |
| Time in app (without context) | Active users (defined by behavior) |
This table is a directional comparison from the original article, not a measurement of any specific app's performance. Use it to choose which signals to prioritize, then validate with your own cohort data.
Frequently asked questions
What is the difference between time to first value and time to core value?
Time to first value asks whether the user experienced something valuable. Time to core value asks whether the user started building a habit. In a meditation app, the first might be completing a first meditation session, while the second might be meditating at least four times in a week and starting a routine.
Why can high retention be misleading before product-market fit?
Retention can be driven by gamification, a small group of power users, heavy discounts, or annual subscriptions that delay churn rather than prevent it. In those cases, users may stay without receiving real value. Pair retention data with engagement data and qualitative feedback before treating it as proof of fit.
How many responses do I need for the Sean Ellis test?
At least 100 responses provide a general sense. If segmenting by signup reason, main feature used, or other factors, 500–1,000 responses are more useful. Survey users who should have reached their aha moment, not day-one users if value normally takes seven days.
What is a good North Star Metric before product-market fit?
It should be behavioral and tied to the core problem being solved. Include both the specific action and the time frame. For example, a budget-tracking app might hypothesize that users who categorize at least five transactions per week are deriving value. Refine the metric as more is learned.