Growth in practice: experiments that move revenue
Growth in practice: how to choose, prioritize and run experiments that move revenue, with a hypothesis template, the ICE method, test ideas and common pitfalls.
How to recognize product-market fit: real signs you have it, false signals that mislead founders, how it varies by business model and what to do if not yet.

Product-market fit is the point at which a product solves an important problem for a specific group of customers so well that they stay, come back and bring others without much pushing. The most reliable signal is retention: the curve of active customers in each cohort stops falling and levels off. Before that, pouring money into acquisition usually just burns cash.
Product-market fit is not a switch that flips on a particular day. It comes in degrees, and it almost always holds for one segment before it holds for the whole market. A piece of software can fit small clinics perfectly and not fit hospitals at all.
It can also be lost. A better competitor, a regulatory change or the end of an acquisition channel can undo a fit that seemed secure.
Day to day, the difference tends to show up like this: before fit, everything is pushed, every sale takes effort and every customer has to be convinced. After fit, demand starts to pull, and the problem becomes keeping up.
No single signal proves fit, but several together form a reliable picture:
A cohort is the group of customers who started in the same period. You track what percentage of that group is still active, or still paying, as the months go by. The numbers below are illustrative:
| Months since joining | Cohort A | Cohort B |
|---|---|---|
| Month 1 | 100% | 100% |
| Month 2 | 55% | 60% |
| Month 3 | 38% | 48% |
| Month 4 | 25% | 45% |
| Month 6 | 12% | 43% |
| Month 9 | 5% | 42% |
Cohort A heads toward zero: people try the product and leave. Cohort B loses customers early, which is normal, but then levels off. The group that stays is the sign that a core of customers has found real value. What counts as a good level varies widely by business model, so the shape of the curve matters more than any fixed number.
A benchmark popularized by Sean Ellis is to ask active users how they would feel if they could no longer use the product, with three answers: very disappointed, somewhat disappointed or not disappointed. The well-known threshold is reaching around 40% very disappointed.
Ask people who used the product recently, not your whole user base. And pay more attention to who answered very disappointed than to the percentage itself: that group shows you your core segment and why it values the product.
Some numbers look like traction but say nothing about fit:
What they share is that they measure attention or acquisition, not value delivered repeatedly. Product-market fit shows up in what happens after the first purchase.
The strongest type of evidence depends on how the business makes money:
| Model | Strongest signal | Common trap |
|---|---|---|
| B2B subscription software | Renewals and account expansion | Pilots that never become contracts |
| Consumer app | Recurring weekly or daily use by cohort | Counting downloads |
| E-commerce or direct-to-consumer brand | Repeat purchase rate | First purchases driven by coupons |
| Marketplace | Retention on both sides and repeat transactions | Subsidizing one side indefinitely |
| Subscription service | Low churn and referrals | Aggressive discounts to hold on to customers |
Not having product-market fit is the normal state of almost every product early on. The mistake is pretending you have it. A sequence for moving forward:
At this stage, spend on acquisition only enough to have new cohorts to measure. Scaling marketing before fit multiplies the cost of learning.
Product-market fit is the line between the adjusting phase and the accelerating phase. Before it, money is better spent on product work and customer conversations. After it, investing in acquisition, technology and operations to grow makes sense.
At Bridgemoon Capital, this question comes up before any conversation about scaling or investment. Understanding retention and where customers come from helps decide whether the next step is to accelerate marketing and technology or to adjust the product, and that is where a partner who builds and sells alongside the founders can make a difference.
There is no single metric. The most useful are cohort retention curves, the share of customers who arrive through referrals or organic search, and the survey asking how disappointed users would be without the product. Use all three together and track how they evolve over the months.
Yes. Market shifts, a better competitor, a regulatory change or the loss of an acquisition channel can undo a fit that once existed. That is why it pays to keep tracking retention and satisfaction even after the company grows.
There is no standard timeline. It depends on the market, the sales cycle, how much the product needs to evolve and how quickly the team learns from customers. What you can control is the speed of your iteration cycles, with clear hypotheses, defined metrics and a deadline for each attempt.