How to Achieve Product-Market Fit (Without Guessing)
Product-market fit is a number you can measure, not a feeling. The Sean Ellis 40% test, Superhuman's PMF engine, and how to deliberately increase your score before you scale.
Product-Market Fit Is a Number, Not a Feeling
Most founders describe product-market fit as something you just know when you feel it. That instinct is exactly why so many teams pour money into growth before the product is ready. The breakthrough insight from the last decade of startup practice is that PMF can be measured, tracked, and deliberately increased.
The 40% Test
The most durable measure comes from Sean Ellis: survey users who have genuinely experienced your product and ask one question, how would you feel if you could no longer use this product. Count the percentage who answer very disappointed. Across hundreds of startups, Ellis found that 40% is the threshold, teams above it tend to scale sustainably, teams well below it tend to struggle no matter how hard they push. Below 25%, you have a positioning or product problem. Between 25 and 40%, you are close, and your high-intent users will tell you what to fix.
Turn the Score Into an Engine
A single survey is a snapshot. Rahul Vohra turned it into a system at Superhuman. As First Round documented, Superhuman's PMF score climbed from 22% to 58% over a few quarters. The first jump, 22% to 32%, came in five minutes, not by changing the product, but by segmenting the survey down to the high-expectation customers who represent the real market. The engine then runs in a loop: double down on what those users love, and systematically remove the objections that hold back the almost-converted.
Listen to the Right Users, Not the Loudest
PMF is found at the intersection of a specific user and a specific problem. The fastest way to lose it is to average feedback across everyone. Identify the segment that would be most disappointed to lose you, learn what they value, and build for them before you broaden. Every other request is noise until the core is undeniable.