Product-market fit

What product-market fit means

Marc Andreessen described product-market fit in a 2007 essay as being in a good market with a product that can satisfy that market. He credited the idea to the investor Andy Rachleff. Andreessen's point was that, for a startup, it's the only thing that matters: with it, many mistakes are survivable; without it, nothing else saves you.

In practice you can feel the difference. Before product-market fit, every sale is a push, users drift away and growth comes only from your own effort. After it, customers come back without reminders, recommend you to others and complain loudly when something breaks.

The Sean Ellis test

Sean Ellis, who led early growth at Dropbox and LogMeIn and coined the term growth hacking, created the most widely used survey for product-market fit. It asks one question: how would you feel if you could no longer use this product?

Users pick one of four answers, listed below. Ellis compared results across about 100 startups and found that those where 40% or more answered very disappointed tended to grow well, and those below it usually struggled.

• Very disappointed
• Somewhat disappointed
• Not disappointed (it isn't really that useful)
• N/A, I no longer use it

How to run the survey

1. Survey only people who've used the core of your product, at least twice, in the last two weeks. Newcomers and long-gone users skew the result.
2. Aim for at least 40 answers. With fewer, a handful of people changes the score.
3. Add three follow-up questions: what type of person would benefit most from the product, what's the main benefit you get from it, and how could we improve it for you?
4. Calculate the share of very disappointed answers, leaving out N/A.
5. Look closely at who answered very disappointed. That group, and what they say the main benefit is, is your market.

How Superhuman used it

Rahul Vohra, the founder of the email app Superhuman, wrote about using the Ellis test to steer the product. When they first ran it, 22% of users said they'd be very disappointed. Instead of trying to please everyone, the team studied the users who loved the product, built more of what those people valued, and fixed what held back the users who were nearly there. The score rose with each round, and Vohra reported reaching 58%.

The lesson travels: product-market fit is found in one group first. Find the people who would be very disappointed without your product, then build for more people like them.

Other signals of product-market fit

Retention that flattens

Plot what share of each month's new users are still active over time. If the curve drops and then levels off, a core group keeps using the product. If it keeps sliding toward zero, you're not there yet.

Growth you didn't pay for

New users arriving from word of mouth, referrals and search, not only from ads and outreach.

Customers who pay and stay

Renewals and few cancellations, at a price that supports the business.

Pull, not push

Sales get shorter and easier. People ask when a feature will ship instead of whether they need the product at all.

Signs you don't have it yet

The most common sign is sign-ups without use: people try the product once and disappear. Founders often respond with more marketing, but more traffic into a product people don't keep using just makes the problem bigger and more expensive.

These posts from our ideas library show what it looks like from the inside:

Before you have a product

You can't measure product-market fit before people use a product, but you can look for its early signs: people describing the problem without being asked, a high share of visitors leaving an email on a landing page, and clicks on a paid plan. These don't prove fit. They tell you which market is worth building for.

Then use the product's first users to find fit, with the Ellis survey and retention as your guide. Expect to change the product, the audience or both. Fit is rarely exactly where you started.

How to validate a business idea