When Rahul Vohra first surveyed the users of his email app Superhuman, only 22% said they'd be "very disappointed" if they could no longer use it.
That's well short of the 40% that growth marketer Sean Ellis found separates products with real traction from those still searching. So Rahul didn't celebrate, and he didn't panic. He used the answers to build an engine, and within three quarters the score reached 58%.
This tool scores the survey and points you at the same next steps. I tell the story in the book, and it's still the most practical way I know to measure product/market fit.
How to use the PMF survey scorer
- Copy the questions and send them to people who've genuinely used the product: at least twice in the last two weeks, or in the last billing cycle for less frequent products. The Sean Ellis PMF survey template has the invitation email, segment questions and follow-ups.
- Count the answers to question 1 and enter them: very, somewhat and not disappointed, plus N/A.
- Add segments: role, industry, plan, usage frequency or how they found you. This is where the insight is.
- Read the score, the likely range and the next steps.
Aim for at least 40 responses before you read much into the score.
How the maths works
PMF score = "Very disappointed" ÷ (All answers − "N/A" answers)
People who no longer use the product are excluded, because they're not telling you about the product as it is. The calculator also shows a 95% Wilson interval: the range the true score probably sits in, given how many people answered. With 30 answers that range is wide. With 300 it's narrow.
A worked example
The tool opens with an illustrative bookkeeping subscription. 120 recent users answered: 42 very disappointed, 50 somewhat and 28 not disappointed.
Overall score: 42 ÷ 120 = 35%. Below the line.
Now segment. Of the 50 tradespeople who answered, 26 said very disappointed: 52%. Of the 70 other businesses, only 16 did: 23%.
You don't have a product/market fit problem everywhere. You have fit with tradies and a vague offer for everyone else. That's a far more useful finding than "35%", and it tells you exactly where to focus your marketing and your roadmap.
What good looks like
As our guide to the Sean Ellis 40% test explains, 40% or more suggests product/market fit. Under 40%, the product isn't yet a must-have for enough people. It's a benchmark drawn from comparing nearly a hundred startups, not a law of physics, so watch the trend across quarters as much as the number.
The next steps follow the four-step engine Superhuman used:
- Segment to find your supporters: focusing on users who matched its biggest fans lifted Superhuman from 22% to 33% before it changed the product.
- Find out why they love it: read the fans' "main benefit" answers.
- Find out what holds others back: ask the near-fans whose main benefit matches.
- Split the roadmap: half on what fans love, half on fixing what holds near-fans back.
What to do next
- Read the full story of the Superhuman PMF engine and what happened to the company next.
- See the bigger picture in how to know if you have product/market fit, with examples from Netflix, Looker and GitHub.
- If people like the product but won't pay enough for it, run a willingness-to-pay interview next.
Frequently asked questions
How many responses do I need for the Sean Ellis test?
Aim for at least 40 before you read much into the score, and more if you want to compare segments. With 40 answers, a 35% score could really be anywhere from about 22% to 50%, which is why the calculator shows a likely range. Survey only people who've genuinely used the product recently, because lapsed and brand-new users drag the score around.
Does the 40% benchmark work for B2B and service businesses?
It works as a guide for any product people use repeatedly, including B2B software, memberships and productized services. The benchmark came from startups, so treat it as a reference line, not a pass mark. For a service business with a few dozen clients, the written answers about main benefit and improvements are often more valuable than the percentage itself.
Should I include N/A answers in the score?
No. People who say they no longer use the product can't tell you how they'd feel about losing it, so they're excluded: the score is very disappointed answers divided by all answers minus N/A. Do read their improvement suggestions, though. A high N/A count is a churn signal worth investigating separately.
How often should I run a product/market fit survey?
Quarterly is a good rhythm once you're using it to steer the roadmap. That's often enough to see whether changes are working, without surveying the same people so often that they stop answering. Keep the wording identical each time so the scores are comparable, and always score segments as well as the overall number.
Built on The Subscription Playbook
This tool turns frameworks from Robert Coorey's book into something you can use in minutes. Robert is the author of The Subscription Playbook and a Director at BPAG Accountants. More about Robert.