How to find product-market fit faster

Product-market fit is described as a feeling.

"You'll know it when you have it." "It feels like a pull." "The product starts growing on its own."

These descriptions are accurate for a certain stage of fit — the stage where the signal is so strong it's unmissable.

But they're not useful for finding fit faster. You need to know when you're moving toward it, not just when you've arrived.

What product-market fit actually measures

PMF is the degree to which a product satisfies a strong market demand.

More precisely: a sufficient number of users rely on your product enough that losing it would be a real problem for them.

The Sean Ellis test operationalizes this: ask your users "how would you feel if you could no longer use this product?"

If 40%+ say "very disappointed" — you have PMF. If less than 40% — you don't, and the distribution of answers tells you who does.

This isn't a binary state. You can have PMF for a segment of your users before you have it for the whole market. Finding that segment is often the work.

The metrics that tell you if you're moving toward it

Metric 1: Week 2 retention

After a user's first week, what percentage come back for a second week?

Below 20%: the initial value proposition isn't converting to habit. 20-40%: you have something real for a subset of users. Above 40%: strong signal, focus on expanding who gets here.

Week 2 retention is the earliest meaningful PMF signal. If it's low, every other metric is noise. For the data infrastructure that makes tracking this straightforward, how to use data to make faster product decisions covers the minimal setup you need.

Metric 2: The "very disappointed" score

Once a month, send the Ellis survey to users who have been active for at least 30 days. Track the percentage who would be "very disappointed."

Watching this number move over time tells you whether your iterations are moving toward or away from PMF.

Metric 3: Organic word-of-mouth

Are new users arriving because existing users told them about the product? Not because of paid ads or press — because a user recommended it.

Organic referral is the strongest PMF signal because it requires existing users to stake their credibility on your product. People don't refer products they're lukewarm about.

Metric 4: Qualitative urgency

Are users who you try to churn (expired trials, cancelled subscriptions) fighting to stay? Do they reach out asking for discounts, extended trials, payment plans?

Customers who fight to keep a product have PMF with that product. Customers who quietly churn don't.

The fastest path to PMF: narrow before you expand

Most teams try to find PMF by building more features. More features for more users with more use cases.

This is backward.

PMF is found by going deeper for fewer users, not shallower for more. The 40% threshold is not achieved by making 40% of all users happy. It's achieved by finding the specific segment where 80%+ are "very disappointed" — and then making that segment larger.

The narrowing process:

Take your current users. Segment them by role, industry, company size, use case. Within those segments, identify who has the highest retention and the highest "very disappointed" scores.

That's your PMF segment. It might be small (50 users). That's fine.

Now: understand what they have in common. What problem are they solving? How do they use the product? What feature do they rely on that others don't?

Build more of what that segment needs. Reach more people like them. Ignore, for now, everything that isn't them.

What slows the path to PMF

Building features instead of talking to users. Features built without user input often miss the actual blocker to PMF. Before building the next feature: talk to 5 users who are retained and 5 who aren't. The contrast tells you what the retained users have that others don't. The habit of getting out of your head and into the market is the weekly practice that prevents this mistake from recurring.

Optimizing acquisition when retention is broken. If your week-2 retention is 15%, growing faster just fills a leaky bucket. More users → more churn → same underlying problem, larger scale. Fix retention first. Acquisition follows.

Measuring the wrong users. Your 100 most active users may not represent your PMF segment. They may be power users who found creative uses for the product that don't generalize to new users. Measure users who found the product through normal channels and used it normally.

Moving on from a feature before users deeply adopt it. Shipping a feature and measuring adoption at 30 days isn't enough. Deep adoption often takes 60-90 days — users need to build it into their workflow. Give new features time before concluding they're not working.

PMF is a spectrum

You don't go from no PMF to full PMF in one moment.

You move along a spectrum:

• No fit: most users don't come back after day 7
• Weak fit: some users stay, but wouldn't be "very disappointed" to leave
• Partial fit: strong PMF with a specific segment, unclear for others
• Strong fit: broad user base, high "very disappointed" scores, organic growth

Most teams give up somewhere between weak and partial fit — when the signal is there but not yet strong enough to feel like success.

The move from partial to strong fit is usually about reach, not product. You have the right product for someone. You need to reach more of them.

That's a distribution problem, not a product problem. Recognizing the difference is how you find PMF faster. When the data consistently points to a mismatch rather than a distribution problem, how to decide between pivoting and persisting gives you the framework to make that call clearly.

Find out if your idea has the right fit →