Product-Market Fit (PMF)
Product-market fit (PMF) is the state in which a product satisfies a real, sizeable market well enough that customers keep using it, pay for it and recommend it. Marc Andreessen, who popularized the term, described it as being in a good market with a product that can satisfy that market. Reaching it marks the shift from searching for a working product to scaling one.
How Product-Market Fit Works
In his 2007 essay "The only thing that matters," Andreessen argued that the market matters most: a strong market pulls a product out of a startup, while a weak market defeats even a great team. He described PMF as something you can feel. Without it, customers are not getting quite enough value, word of mouth is not spreading and deals stall. With it, customers buy as fast as the company can deliver and usage grows faster than the team can keep up.
Because "feel" is hard to manage, teams use proxy signals:
- Retention: usage cohorts that flatten into a stable line instead of decaying to zero. See retention.
- Organic demand: referrals and word of mouth that bring customers without paid acquisition.
- The Sean Ellis test: a survey asking users how they would feel if they could no longer use the product. Ellis found that companies that struggled to grow almost always had fewer than 40% of users answer "very disappointed," while companies with strong traction usually exceeded it.
- Commercial pull: shorter sales cycles, renewals and expansion without heavy discounting.
No single signal is proof. The survey reflects stated feelings, while retention and revenue reflect behavior, so the strongest case combines both.
PMF is also not permanent. Markets move, competitors arrive and a product can lose fit as its customer base changes.
Why Product-Market Fit Matters
Before PMF, the main job is learning: talking to customers, running a minimum viable product and changing direction when the evidence says so. Spending heavily on growth at this stage tends to buy users who leave. After PMF, the job shifts to scaling acquisition, onboarding and operations. Knowing which side of the line a product is on decides how a founder spends money and time.
Rahul Vohra of Superhuman showed that the Ellis score can also guide the work. His team scored 22% at first, then focused on the users who were most disappointed, learned what they loved, and addressed what held back the next most promising group. The score rose to 58% within three quarters.
Product-Market Fit Example
A team sells shift-planning software to restaurants. After six months it has 120 paying venues, but most stop logging in after the second month and new sign-ups come almost entirely from ads. A survey shows 18% would be very disappointed without the product. Segmenting the answers reveals that multi-site restaurant groups score above 45% and retain well, while single cafés churn. The team narrows its focus to multi-site groups, rebuilds onboarding around their head-office workflow and watches retention in that segment before spending more on growth.
For the metrics that make this measurable, see how to measure product-market fit with AARRR metrics and the Sean Ellis survey.
Product-Market Fit vs. Problem-Solution Fit
Problem-solution fit comes first: evidence that a real problem exists and that the proposed solution addresses it, usually gathered through interviews and small tests. Product-market fit comes later and is proven by market behavior at scale: retention, payment and referral from a meaningful number of customers.