Product market fit is overrated. Not because retention does not matter — it does — but because of what people actually mean when they say it. What they mean is: if I can just achieve this thing, I will have something that cannot fail. Something bound to succeed. And if it does not succeed, it must be because the fit was not there. That is false.
Failure to get revenue is a failure to sell. That is it. You can call it market size, channel mismatch, timing — fine. But at the root, it is always a sales problem. If your market is small and your first instinct is to cold call everyone in it, that is not a market size problem. That is a method problem. The acquisition approach was never aligned with the market reality.
What product market fit actually measures
Product market fit does not prove anything before you have customers. It cannot. It reveals something after you have customers — specifically, how long they are willing to stay without being pushed, incentivised, or pressured into it. When people renew on their own volition, when they stay without any extra promotion, that is when you know you have something people genuinely want. That is product market fit. It is a retention metric, not a pre-launch thesis.
Which means chasing it before you have customers is chasing something that does not exist yet. You cannot validate it in advance. You can only earn it in retrospect.
What is actually underrated: problem market fit
You can have problem market fit without a product. Without customers. Without even a way to deliver. But if you have it, you have something worth building toward.
Problem market fit is the thing people should be talking about instead. And the reason it is more powerful is precisely because you can validate it before you build anything.
Here is how to think about it. A problem is a scenario — a workflow, a use case, a job being done in the market — with a component removed. That is all. You identify something people are already doing, find the cost or drag embedded in how they do it, and then ask: if I gave you this same scenario minus that component, would that matter to you? You are not inventing a need. You are finding a need that already exists and proposing a cleaner version of what they are already living.
The beautiful thing about problem market fit is that it is almost the easiest thing to validate. You do not ask people if they would buy. You try to book a meeting on the premise that the problem is solved. If out of a hundred conversations a meaningful number convert to meetings, you have just validated that the problem is worth building on. No product required. No customers yet. Just a premise and a phone.
Validating the unit economics
Once the problem is validated, the next thing to validate is cost — both cost of fulfillment and what the market will pay. And the reason you validate cost is not simply to find out whether someone will buy. Someone will buy at almost any number. The question is how many people will buy at a specific number, and whether that aligns with your cost to acquire and deliver.
Take document automation in immigration as an example. Converting client information into IMM forms is something every firm has to do. It is repetitive, universal within the market, and carries a known drag. You go out and say: we do this without the manual work, here is the mechanism. You find out if they will get on a meeting. Then you price it at your cost of fulfillment and see how quickly people buy. A handful of fast yeses tells you the unit economics work. If no one buys above a certain threshold and that threshold is below your cost to acquire and fulfill, the channel does not work — regardless of how real the problem is.
That is the full validation loop. Problem, then mechanism, then price, then channel. In that order. Product market fit sits at the end of a much longer chain — and it is the last thing you should be thinking about, not the first.