What is product-market fit?
Product-market fit is the moment when buyers actively seek you out instead of being persuaded. It’s pull, not push. If you have to work hard to find every customer, you don’t have it yet, no matter how much they like the product once they show up.
What it actually means
The phrase comes from Marc Andreessen’s 2007 essay defining PMF as being in a good market with a product that can satisfy that market. The definition is precise but easy to misread. Most founders interpret it as “people like the product.” That’s not it. PMF is about demand, not approval. If five customers love what you built but you can’t find a sixth without burning a network favor, the market hasn’t pulled yet.
The clearest signal is direction of motion. Without PMF, every customer is a sales win that took effort: a warm intro, a long email thread, a free trial that needed handholding to get to “yes.” With PMF, customers find you, sign up faster than you can serve them, and tell you about features you haven’t built. The transition is rarely a single moment. It’s a slope change in how hard each new customer is to get.
Sean Ellis popularized a famous heuristic: ask current customers how they would feel if they could no longer use the product. If 40 percent or more say “very disappointed,” you have early PMF. The number is suggestive, not magical. What matters is the underlying signal: a meaningful share of users would be hurt by losing it. Below that threshold you have customers; above it you have a product they depend on.
The wrong metric to chase is revenue alone. Revenue can grow on the back of founder hustle indefinitely, masking the absence of PMF. The right metric is whether customer acquisition cost is going down over time without proportional spend on growth. Pull is making your sales easier. If sales are getting harder as you scale, the market is telling you something specific about who you actually serve.
If you don’t have PMF yet, the worst thing you can do is keep building. Building feels productive but doesn’t address the real question: who is willing to pay, and why. Stop adding features and start having more conversations with the buyers your product is closest to fitting. Find the segment where the product almost works but doesn’t quite, and figure out what would make them go from interested to insistent. PMF is found through buyer dialogue, not through engineering velocity, and the founders who confuse the two often spend an extra year building without progress.
How to know if yours is broken
Are customers showing up faster than you can serve them, or are you working hard for every new logo?
If you stopped all outbound for a month, how many new customers would still arrive?
When new customers tell you how they found you, do they cite each other or only you?
Are existing customers asking for features you haven’t built, or only complaining about ones you have?
Common misconceptions
“Strong product feedback means you have PMF.”
Praise is cheap. PMF is measured by what people pay, recommend, and refuse to give up. Hearing “this is great” from prospects who never buy is the loudest false signal in early-stage companies.
“PMF is binary.”
PMF is segment-specific. You can have it with one buyer type and not another. Generic PMF is rarely real PMF; it’s usually averaged enthusiasm across mismatched buyers.
“PMF means you can scale.”
PMF means the market is pulling. Scaling means you can deliver against the pull without breaking. Different problems, different solutions, often a year apart.
Related concepts
Or browse the full glossary.