Product-market fit signals: what to measure and trust

What product-market fit actually means

Product-market fit is the point at which a specific group of customers wants your product badly enough that demand pulls you forward. Before it, every customer feels like a push: long conversations, heavy discounts, constant reminders to log in. After it, the work changes from convincing people to keeping up with them.

Two words in that definition matter. 'Specific' because fit is almost always found with a narrow segment first, not with everyone. 'Group' because a single enthusiastic customer is not fit; you are looking for a pattern that repeats among people who share a problem.

Fit is also not permanent or binary. You can have strong fit with one segment and none with another, and markets move. Treat it as something you measure continually, not a milestone you pass once.

Leading and lagging signals

Lagging signals, such as revenue growth and long-term retention, are the most trustworthy but take months to show. Leading signals appear sooner and are noisier. You need both: leading signals to steer week to week, lagging ones to confirm you were steering in the right direction.

The retention curve that flattens

Group users by the week or month they started, then plot what share of each group is still active over time. Every product loses people early. The question is what happens next. If the curve keeps sliding towards zero, users are trying the product and leaving. If it drops and then flattens into a stable line, a core group has found lasting value.

An illustrative example: of 100 users who sign up in March, 60 are active after the first week, 40 after a month, 35 after two months and 34 after three. The early drop is large, but the flattening between month two and three suggests a stable core of around a third. If April and May cohorts flatten at a similar or higher level, that is a meaningful signal. If each new cohort flattens lower, something is getting worse.

Where the curve flattens matters less than the fact that it does, and whether it improves over successive cohorts. What counts as good varies too much by product type and price point for a universal benchmark to be useful.

Organic pull

Pull is when customers arrive without you pushing them there: through word of mouth, by searching for you by name, or because an existing customer brought a colleague. Track how each new customer found you, and ask them directly during onboarding. A rising share of customers who came from referrals or direct searches is one of the clearest signs that something is working.

Other forms of pull are qualitative. Customers complain loudly when something breaks, because they depend on it. They ask for features that extend the product rather than fix it. They tell you about competitors trying to win them over. They pay invoices without chasing.

The very disappointed survey

A widely used method, associated with the growth marketer Sean Ellis and often called the Sean Ellis test, asks active users one question: 'How would you feel if you could no longer use this product?' with the options very disappointed, somewhat disappointed, and not disappointed. The share answering 'very disappointed' is the measure. A figure of 40 percent is commonly used as a benchmark for fit, though it is a rule of thumb rather than a law.

Run it on people who have actually used the product recently and more than once, otherwise you are measuring curiosity. Look at the answers by segment: the overall figure may be middling while one group is far above it, and that group tells you where your fit is. Follow-up questions such as 'What is the main benefit you get?' and 'What would you use instead?' often matter more than the headline number.

What to do while you are still searching

Most early products do not have fit yet, and pretending otherwise is expensive. Until the signals line up, the job is narrowing, not scaling. Find the users who would be most disappointed to lose the product, study what they have in common, and rebuild your positioning and roadmap around them. Deliberately deprioritise requests from users who are lukewarm, even if they are numerous.

Hold off on the things that amplify whatever you already have: paid acquisition, hiring a sales team, expanding into new segments. Amplifying a product without fit mostly accelerates churn and burns runway. When retention flattens, referrals start to appear and the disappointment survey climbs within a segment, that is the point at which spending to grow starts to make sense.

False signals

• Press coverage and launch-day traffic — Attention is not demand. Spikes fade.
• Sign-ups without activation — A long waiting list can coexist with no fit at all.
• Revenue from a few large favours — One customer paying a lot because they know you is a relationship, not a market.
• Growth bought with spend — Paid acquisition can grow numbers while retention quietly collapses.
• Compliments and feature requests — People are kind, and requests often describe a different product.
• Strong averages hiding weak segments — Mixed results across segments may mean strong fit in one and none elsewhere.

Questions