Product-market fit is one of those terms that gets used to mean almost everything: traction, revenue, growth, morale. What it actually means is simpler and harder to fake. It is the point at which a defined group of customers wants your product badly enough that they pull it into their lives without much persuasion.
What product-market fit actually means
Marc Andreessen’s original framing is still the clearest: you can always feel when PMF is not there. Customers aren’t getting value, word of mouth isn’t spreading, usage isn’t growing. You can also feel when it is there. The company can’t hire fast enough, journalists write about you without being asked.
That description is useful as a sanity check but it’s not operational. “Feels right” doesn’t tell you when you have crossed the line, and it absolutely doesn’t tell you what to do when you haven’t.
The more useful definition: PMF is reached when a repeatable segment of customers derives enough value from your product that they retain, refer, and resist switching. All three verbs matter. Retention without referral might just be switching costs. Referral without retention is a leaky bucket. Resistance to switching separates a product people are glad to use from one they’re stuck with.
PMF is not a moment. It’s a set of behaviors that keep happening without you forcing them.
The signals that matter
The most widely cited PMF proxy is Sean Ellis’s survey question: how would you feel if you could no longer use this product? If 40% or more of your users say “very disappointed,” that is a meaningful signal. Below 30%, you almost certainly don’t have it yet.
That benchmark is worth using, but treat it as a directional indicator rather than a verdict. The shape of the 40% matters as much as the number itself. If your strongest respondents are a segment you didn’t plan for, that is useful signal too. You may have PMF with the wrong customer, and the question then becomes whether that customer is a viable market.
Retention curves are the harder, more honest signal. Plot your weekly or monthly active users by cohort. If the curve flattens above zero instead of declining to it, some customers are genuinely sticking. A flattened retention curve is evidence. A rising one is a green light. A curve that never flattens is a product still searching for its reason to exist.
Growth from word of mouth is the third signal. Not because virality is the goal, but because unprompted referral means the product is creating enough surplus value that customers feel compelled to tell someone. Paid acquisition can mask a PMF problem for a long time; referral can’t.
What customers say and how to read it
Quantitative signals tell you whether something is happening. Qualitative signals tell you why, and they’re harder to fake because you have to actually listen.
The language customers use to describe your product without prompting is one of the richest signals available. If they describe the problem it solves better than you do, you have found customers who feel the pain acutely. If they struggle to explain what it does, that is a positioning problem that may also be a product problem.
Listen for the moment of switch. Ask customers when they decided to look for a solution. The answers cluster around a triggering event: a scale threshold, a failed workaround, a broken status quo. If those trigger events are consistent across your best customers, you have a clear ICP. If they’re scattered, your best customers may not share a common problem.
Pay attention to who brings other people in. Referral paths inside a company, a community, or a professional network are the earliest signal of real pull. If your champion at a customer tends to recruit colleagues, or if buyers arrive already knowing your product from a peer, that behavior is measurable even at small scale.
When the signals are mixed
Mixed signals are the common case. You have strong retention in one segment and flat retention in another. The survey result is 38%, not 42%. Word of mouth is happening but slowly.
The mistake is to average everything out and conclude you’re close. Close is not a useful state. The more productive question is: which subset of your customers has the flattest retention curve, the strongest survey response, and the most referral activity? That subset is your real signal. If it is a coherent group with a describable profile, you may have PMF with a narrower market than you planned.
This is where market sizing earns its keep. If your true PMF segment is addressable enough to build a business on, the path is clear: go deep before going broad. If the segment is too thin, you have product and positioning work to do before the next push. Thinking through the go-to-market strategy for that narrower segment often clarifies what needs to change in the product itself.
Acting on what you find
Mixed or weak signals are not failure states. They are information. The question is what you do with them.
If retention is the problem, the issue usually lives in the product or in the expectations set by your acquisition message. Either the product isn’t delivering what it promised, or the promise attracted customers who were never going to stick. Brand health metrics can surface the latter: when customers who arrived with a specific expectation churn at higher rates, the problem is upstream of the product, not inside it.
If referral is weak but retention is strong, the product has value but the story is missing. Customers can’t explain it, so they don’t. That is positioning and messaging work, not product work.
If the survey score is low across the board, with neither retention nor referral compensating, you have a harder conversation to have about whether this product solves a real, felt need for anyone in particular.
How Strynal approaches product-market fit signals
Most of the PMF work we do isn’t called PMF work. It shows up as positioning audits before a launch, as retention analysis inside a brand launch plan, or as the question raised when a client’s growth is flat despite everything looking right on paper.
The signal-reading part is tractable. The harder part is having an honest conversation about what the signals require. A 30% survey score doesn’t mean the product is worthless. It means the current segment is wrong, or the messaging is wrong, or both.
We run this work through strategy and positioning, usually as a focused engagement before a campaign or relaunch. The output is a clear PMF hypothesis, a measurement plan, and the positioning adjustments needed to attract the customers most likely to stay. If you’re reading signals and not sure what they’re telling you, that is the conversation worth having.