How to know if you actually have product-market fit (5 real signals and 3 that are lying to you)

If you're not sure whether you have product-market fit, you don't.

Sit with that one. It's the cheapest thing in this post.

Fit has a feel to it. The product gets pulled instead of pushed, the support queue fills with people who want more of it rather than people who can't work out what it's for, and growth turns up in places nobody planned.

The first fifty customers are where we get this wrong. They bought a vision, forgave rough edges, and gave the team the benefit of every doubt.

That enthusiasm is a loan. It's conviction early customers lend us before the product has earned any of it, and from the inside it looks identical to the real thing.

We read it as a weak version of fit. It isn't a weak version of anything.

Fit is what's left standing when the loan gets called. A stranger with no warm introduction and no early-adopter energy finds the product, uses it, gets something out of it, comes back, and tells somebody else.

That's harder to see than it sounds, because we're standing inside the thing we're trying to measure.

Curiosity in a dashboard

The loan is easiest to miss when it shows up as a number.

Twenty people join a waitlist and we call it fit. Beta testers say generous things and we decide the product is ready to scale. A launch spikes trial sign-ups and we read it as pull.

Every one of those measures curiosity. Curiosity doesn't pay and it doesn't stay.

The opposite error costs just as much. A company can have real fit inside a narrow segment and read its own methodical growth as absent fit.

The pain is acute, retention holds, and the market is simply smaller than the plan assumed. That's a distribution problem wearing a fit costume.

One mistake scales too early and burns runway on people who leave. The other starves an engine that had already caught.

The empty-room test

Every signal worth trusting is one we can only read when we're not in the room.

That's the whole diagnostic. Call it the empty-room test. Our attention distorts everything it touches, so any measurement taken while we're standing there is mostly a measurement of us.

Five signals survive our absence.

Retention that holds without heroics

The most reliable indicator is what happens after acquisition. Do users stay when nobody intervenes to keep them there?

Early on we compensate for weak retention with personal attention. We're in their Slack. We're running check-in calls.

We're building a custom feature for one account so it doesn't churn. That's founder-led life support, and it works right up until we stop.

Real fit shows up in the shape of the retention curve. After the first thirty days or so, the curve flattens. The people who made it through onboarding stay, at a stable rate, with nothing from us.

What counts as high depends on the category and the billing model, and the shape matters more than any absolute number. A curve that keeps sliding at any stage means the search is still on.

Referrals nobody asked for

When someone recommends the product to a peer unprompted and uncompensated, that's fit.

Referral programs don't count. Neither does the share prompt in onboarding. This is a person putting their own reputation behind us because their experience was good enough to risk it.

Ask the last ten customers how they found you. If any of them name a person they trust, and no referral program was running, something real happened.

Work rebuilt around the product

The strongest behavioral signal is when people reorganize their week around the thing.

They build their own templates. They write internal documentation for teammates. They get angry when the product goes down, because their actual work stopped with it.

Ask the best customers what they'd do if they had to stop tomorrow. "Find another tool like yours" is a compliment. "We'd have to rebuild a whole process" is fit.

Demand that arrives on its own

Before fit, we push. Outreach, events, the network, every channel we can pry open, because there's no pull to ride.

After fit, the direction changes. Someone searches for their specific problem and finds us. Someone asks in a community for a tool and two people say our name.

Watch direction before volume. Modest inbound we didn't manufacture through paid acquisition or a PR push is worth more than a large number we bought. If the first fifty customers came from hustle and the next twenty arrived on their own, that shift is the signal.

The forty percent answer

Sean Ellis published this test in 2009, after using it while growing Dropbox, LogMeIn, and Eventbrite. Ask active users how they'd feel if they could no longer use the product. They answer very disappointed, somewhat disappointed, not disappointed, or not applicable.

When 40% or more say very disappointed, the product has fit. Ellis found that correlation held across hundreds of startups.

The number is old and it still holds, because it measures how a person feels about losing something rather than a market condition that moves. Buying behavior changes every year. Dependence doesn't.

Survey at least thirty active users, and define active as having used the product in the last two weeks. Under 40%, ask the very disappointed group what they value most. Their sentences are your positioning, already written.

Signals that flatter us

These three feel like fit. All three get measured with us in the room.

Sign-up volume

Sign-ups measure how good the homepage copy is and how well the channel worked. Whether the product solves a real problem for the right people is a separate question.

Three thousand sign-ups in a week from a launch is a successful launch. What matters is what those accounts do next.

As explored here, free-to-paid conversion and day-30 retention carry the information sign-up volume doesn't. Most free accounts never become paying ones, and when the positioning is loose and the ICP is broad, almost none of them do.

The curious and the committed look identical in the sign-up dashboard. They look nothing alike thirty days later.

A strong NPS

NPS is a satisfaction metric. It tells us how happy people are, and happiness isn't dependence.

Happy customers leave when a budget gets cut or priorities shift. Indispensable products survive both, because there's no obvious replacement and leaving costs more than staying.

Praise from beta users is softer still. Early adopters are forgiving by temperament. They know the software is unfinished and they're rooting for us, which makes their enthusiasm genuine and unrepresentative at the same time.

Use NPS to find promoters and detractors. Don't use it to conclude anything about fit.

Deals we personally closed

If the first twenty customers all came through a call we made or a conference where we pitched, we've learned how well we sell.

Can the product sell when we're not in the room? Can a stranger find it, see themselves in it, get value quickly, and stay without anyone shepherding them through?

Founder-led sales is the right move early. It also hides the absence of fit behind personal conviction, and the hiding stays free until the day we hand the motion to a rep.

When the answer is no

Start by defining the ideal customer profile tighter than feels comfortable. Most companies without fit are building for an audience too broad to share a single pain.

If retention is weak, resist the feature backlog. Ask first whether the right people are activating in the right way, because most early retention problems turn out to be onboarding problems.

If the right people are activating and still leaving, the product is the problem. Go back to conversations. Ask what it doesn't do, what's frustrating about the parts it does, and what they reach for when it falls short.

And when there's genuine fit inside a narrow segment, the work changes shape entirely. That's a GTM motion problem, and scaling distribution before confirming repeatability is where it usually goes wrong.

Fit is what keeps happening after we stop helping.

Leave the room. Then look.

What to do next

If the empty-room test came back negative and the product is genuinely good, the gap is usually who it's for and what we say about it.

That's a positioning audit. We look at the segment, the words on the site, and the moment a buyer decides, then tell you which of the three is costing you the retention you thought you had.

If that's where you are, start here. The first conversation is free.