Skip to content
Kuza Ndoto Academy

Product-Market Fit

All levelsChapter 4 · Page 17 of 294 min readUpdated 15 Nov 2026

Product–market fit is the point where a startup has built something a clearly defined group of customers truly wants, so much that they keep using it, pay for it and tell others. Reaching it is the most important milestone for an early startup, because scaling before it usually wastes money.

What product–market fit means

Investor Marc Andreessen popularised the term in a 2007 essay. He described product–market fit as being in a good market with a product that can satisfy that market. He argued that the market matters most: a strong market "pulls" a good-enough product out of a startup, while even a brilliant team struggles in a market that does not care.

Andreessen also described how it feels. Before fit, customers do not get much value, word of mouth is weak, usage grows slowly and sales take forever. After fit, customers buy as fast as you can deliver, usage grows, and the team struggles to keep up with demand.

Signs you have (or do not have) fit

Signs of fitSigns of no fit yet
Customers keep coming back without remindersMost users try once and disappear
Growth comes partly through word of mouthEvery customer must be pushed with discounts or adverts
Customers complain loudly when the product is unavailableNobody notices when the service is down
Sales cycles get shorterEvery sale takes long persuasion
Customers pay without heavy negotiationPeople only use it when it is free

Measuring retention

A powerful way to check fit is a retention curve. Take a group of customers who started in the same week (a cohort) and track what percentage are still active each week afterwards. If the curve flattens at a healthy level, a core group finds lasting value. If it keeps falling towards zero, you have not yet found fit.

Week012345678
Product A, % active1006045383534333333
Product B, % active100503018106321
0%25%50%75%100% 012345678 Weeks since first use % of cohort still active A: flattens at 33% B: falls towards 0%
Figure 1. Cohort retention for two fictional products (data from the table). Product A keeps a stable third of its users, a sign of fit with that group; product B loses almost everyone.

What counts as a "healthy" level differs by product. A daily messaging tool should keep many more users than a once-a-year tax service. Compare against similar products and your own earlier cohorts.

The Sean Ellis "very disappointed" test

Growth marketer Sean Ellis suggested a survey question for users who have recently used the core product: "How would you feel if you could no longer use this product?" with the options very disappointed, somewhat disappointed and not disappointed. From his experience across many startups, he suggested that when 40% or more answer "very disappointed", the product often has fit.

Treat 40% as a heuristic, not a law. Survey enough real users (not just friends), and use it alongside retention and revenue. The follow-up questions are just as useful: "What is the main benefit you get?" and "What type of person would benefit most?" Those answers tell you which segment you fit best.

Example

Joseph, 26, runs a bookkeeping app for small shops in Mwanza. After six months he has 900 sign-ups, but only 8% of each cohort is still active after two months. He surveys 120 recent active users: 24 of them (20%) say "very disappointed". When he splits responses by type of shop, he notices that 18 of the 35 pharmacies (about 51%) answer "very disappointed", because the app helps them track stock expiry dates. Joseph narrows his focus to pharmacies, improves the expiry features and sees two-month retention for new pharmacy cohorts rise to 35%. He found fit with a narrower segment than he first planned.

Common mistake

Spending heavily on marketing to "grow out of" weak retention. If most customers leave, more advertising just fills a leaky bucket faster. Fix retention with a focused segment first; then scale.

Key points

  • Marc Andreessen described product–market fit as being in a good market with a product that satisfies it.
  • Retention curves that flatten at a healthy level are strong evidence of fit.
  • The Sean Ellis test (40% or more "very disappointed") is a useful heuristic, best combined with behaviour data.
  • Fit is often found by narrowing to the segment that values you most, before scaling.
Up next · New chapter: Go to MarketMarket Sizing: TAM, SAM, SOMEstimate how big your opportunity is.
→

Tip: use the ← and → keys to move between pages.