How to know if you have product-market fit: 6 signals (and the one you can actually measure)

Marc Andreessen, who coined the phrase, described product-market fit by how it feels: customers buying as fast as you can make it, servers melting, money piling up, you hiring sales and support as fast as you can. "You can always feel product-market fit when it's happening," he wrote.

Which is a wonderful description and a terrible diagnostic. Because the founders who most need to know whether they have it are exactly the ones for whom it isn't obvious — the ones with some traction, some churn, a few loud fans, and a nagging question about whether any of it is real.

So let's get specific. Here are the six signals that actually tell you, roughly in order from "slowest to confirm" to "fastest to measure."

1. Your retention curve flattens

This is the most honest signal there is, because users vote with their behavior, not their words. Plot the percentage of a signup cohort still active over time. One of two things happens.

Without fit, the curve decays toward zero — everyone eventually leaves. With fit, it flattens: it drops at first, then levels off at a stable plateau of people who stick around indefinitely. Brian Balfour calls a flat or "smiling" retention curve one leg of the product-market fit trifecta, alongside meaningful usage and real top-line growth. A curve that never flattens is the clearest proof you don't have it yet, no matter what your signup numbers say.

The catch: retention curves need months of data to read. You can't run this on a product that launched six weeks ago.

2. At least 40% of users would be "very disappointed" to lose you

This is the Sean Ellis test, and it's the closest thing the industry has to a standard. Ask your engaged users one question — "How would you feel if you could no longer use this product?" — with three options: very disappointed, somewhat disappointed, not disappointed.

If 40% or more say "very disappointed," you have product-market fit. Below that, you don't yet, and the size of the gap tells you how far you have to go. The number isn't arbitrary — Ellis calibrated it across roughly a hundred startups and found 40% was the line that separated the ones that took off from the ones that stalled.

Why it works when other surveys don't: "very disappointed" measures genuine dependency, not polite approval. Plenty of people will call your product "nice." Far fewer will say their week gets worse without it. That second group is the one that drives word of mouth and retention.

3. Growth shows up without you pushing it

Before fit, every new user is a user you dragged in — through ads, outreach, founder hustle. After fit, growth starts arriving on its own. People sign up because a friend told them to. A waitlist forms that you didn't engineer.

Lenny Rachitsky describes two flavors of this market pull: sudden and significant (the hockey stick that catches you off guard) and gradual but compounding (a line that bends quietly upward, month after month, without a clear single cause). Both count. The tell is the same: demand outpacing your effort to create it.

4. Usage gets deeper, not just wider

New signups are a vanity number. The signal that matters is whether the people who show up keep coming back and use you more, not less, over time.

Are daily-active-to-monthly-active ratios climbing? Are people reaching for you for the core job, habitually, without a nudge? Meaningful, repeated usage is the difference between a product people tried and a product people rely on. You want the second one — and it shows up in engagement long before it shows up in revenue.

5. Revenue compounds

For anything with a price, money is a brutally clear signal — if you read it correctly. Brad Feld's MRR ladder is the most useful map I've seen of how fit reveals itself in revenue:

• $0 MRR — no fit. Obviously.
• $1k–$10k MRR — the illusion of fit. Someone's paying you, but don't pour on the gas. You've found a buyer, not a market.
• $10k–$100k MRR — a semblance of fit. If you're compounding ~10% month over month here, you've found something real. If you're not, you haven't.
• $100k–$500k MRR — the sweet spot. Customers start coming to you.
• $500k–$1M MRR — you've found it. This is "initial scale." You have a business.

The number that matters in that ladder isn't the total — it's the compounding rate. Flat revenue at $50k MRR is the illusion dressed up as success.

6. You can't keep up with demand

This is Andreessen's felt signal, and it's last on the list because it's the most obvious and the least actionable. Support tickets pile up. You're hiring frantically. Usage is outrunning your roadmap. If you're genuinely drowning in demand, congratulations — you almost certainly have fit, and you didn't need this article to tell you.

But notice the problem with this signal, and with most of the five above it.

The problem with five of these six

They're either lagging, fuzzy, or both. The retention curve needs months. Revenue compounding needs revenue. "Drowning in demand" only fires once you're already there. Organic pull and deepening usage are real, but they're judgment calls — you can argue yourself into seeing them, or miss them entirely, depending on the week.

There's exactly one signal on this list you can turn into a single number, measure on a schedule, and watch move before the lagging signals catch up: the 40% survey.

That's why operators who take this seriously — Superhuman, Slack, Dropbox among them — don't wait for the feeling. They run the survey, watch the percentage, and treat it as a tracked metric. Superhuman's product team made the score their primary OKR and walked it from 22% to 58% in three quarters. They didn't guess whether they had fit. They measured it, every cycle, and watched the number climb.

How to use these signals this week

1. Run the survey first. It's the fastest read and the only one that gives you a number today. Send the "very disappointed" question to your engaged users and calculate the percentage. Here's the step-by-step.
2. Pull your retention curve. If you have a few months of data, plot cohort retention and look for the plateau. No flattening means the survey number is the one to trust for now.
3. Sanity-check growth and usage. Is any growth organic? Is usage deepening? These confirm what the survey suggests.
4. Track it, don't snapshot it. One reading is a data point. The trend is the signal. Re-run the survey every month and watch which way the number moves.

Product-market fit isn't a lightning bolt you wait to feel. It's a set of signals you can read — and the most useful one is a number you can measure this afternoon and again next month. Get the number, then watch it move.