Product-Market Fit Survey
Run the Sean Ellis test in one click. Name your product, share the link, and read the score with the margin of error it carries at your sample size, plus a breakdown by user segment so you can see which group would put you over 40%.
Free · No response limit · Wilson 95% intervals · Updated September 2026
- 01
Create the survey
Your product name and, if you want the segment cut, up to six options for "which best describes you?". Available in English or Spanish.
- 02
Share one link
Respondents answer the five Sean Ellis questions in about two minutes. They need no account, and each person can answer once.
- 03
Read the results
Score with its 95% interval, the 40% verdict, per-segment scores, the open answers grouped by how much they would miss you, and a CSV export.
The product-market fit survey questions
Five questions. The first one produces the score; the other four produce the reasons. Founders who only tally question 1 throw away the half of the survey that works at any sample size.
- 01
How would you feel if you could no longer use [product]?
The scoring question. Options: Very disappointed · Somewhat disappointed · Not disappointed · I no longer use it. The score is the share who chose "very disappointed" among everyone who still uses the product.
- 02
What type of people do you think would most benefit from [product]?
Answered in the words of your happiest users, this describes the segment to sell to next. Read the answers from the "very disappointed" group separately.
- 03
What is the main benefit you receive from [product]?
The benefit the "very disappointed" group names is your positioning. If it differs from what your homepage says, the homepage is wrong.
- 04
How can we improve [product] for you?
Take this from the "somewhat disappointed" group, who are one fix away from the top box. Ignore requests from people who would not miss the product.
- 05
What would you likely use as an alternative if [product] were no longer available?
Names the real competitor, which is often a spreadsheet, an intern, or nothing. Answers of "nothing" from the top group are the strongest signal on the form.
Product-market fit survey template
Paste this into any form tool. Replace [product] with your product name, keep the answer options in this order, and send it as one email rather than an in-app pop-up, so the people who no longer use the product can answer too.
Subject: 2 minutes to help us decide what to build next 1. How would you feel if you could no longer use [product]? ( ) Very disappointed ( ) Somewhat disappointed ( ) Not disappointed ( ) I no longer use [product] 2. What type of people do you think would most benefit from [product]? 3. What is the main benefit you receive from [product]? 4. How can we improve [product] for you? 5. What would you likely use as an alternative if [product] were no longer available? Send to: people who have used the core feature at least twice, most recently within the last two weeks.
Who to send it to
The score only means something if the people answering have actually experienced the product. Ellis’s guidance, which Superhuman followed when it built its product-market fit engine, is to survey users who have used the core feature at least twice, with the most recent use inside the last two weeks. Everyone else is answering a question about a product they never got to.
Send it to all of them, not to a hand-picked set. A survey sent only to the people who reply to your emails is a survey of your fans, and the score it returns is a compliment rather than a measurement.
How to read the score
A measured 40% is not a true 40%. The table shows where the true value can sit for a measured result of exactly 40%, by sample size, at 95% confidence.
| Responses | True value is somewhere in | Margin |
|---|---|---|
| 10 | 17% – 69% | ±26 |
| 20 | 22% – 61% | ±20 |
| 50 | 28% – 54% | ±13 |
| 100 | 31% – 50% | ±9 |
| 400 | 35% – 45% | ±5 |
To be confident you are above 40%, the whole interval has to sit above it. That depends on both the sample size and how far above the line you land.
| If you measure | You need |
|---|---|
| 45% | 369 responses |
| 50% | 93 responses |
| 55% | 41 responses |
| 60% | 24 responses |
| 70% | 11 responses |
A result close to the threshold is the expensive one. Squeaking over at 45% takes 369 responses to defend; landing at 60% is defensible with 24. If your early sample lands anywhere near 40%, you do not have an answer yet, and collecting a few more responses will not give you one. The full derivation, including why 8 of 20 and 12 of 20 are the same result, is in the 40% rule and sample size.
Already ran the survey somewhere else?
Paste the two counts from Typeform, Google Forms or a spreadsheet and get the same interval and verdict the hosted results page shows.
How to test product-market fit beyond the survey
The survey measures what people say they would feel. Everything below measures what they do, on every user rather than the few who answer, which is why the same population gives a far tighter estimate.
Retention that flattens. Plot the share of each signup cohort still active by week. A curve that keeps falling toward zero has no fit, whatever the survey says. A curve that flattens at any level has found a group of people who keep coming back, and that group is the one to survey.
Repeat purchase and expansion. Second payments are harder to fake than first ones. A customer who renews, upgrades, or adds a seat has run the survey on themselves with their own money.
Unprompted word of mouth. Signups that arrive with no campaign attached, and support tickets that describe what the user is trying to do rather than asking whether the product does it.
Before you have users at all. The survey needs a product. The question it stands in for, whether people would miss this enough to pay for it, does not. What your target user already pays for, and which workarounds they tolerate today, is available before a line of code exists, and it is the check Scoutr runs on an idea.
Where the 40% rule comes from
Sean Ellis published the question in 2009 after running it across roughly a hundred startups he had worked with, and noticed that the ones growing well cleared 40% while the ones stalling did not. Superhuman later made the method famous by surveying users until the score rose from 22% to 58%, and by segmenting on question 2 so that the score was computed only for the users the product was built for.
Ellis never published the underlying data. Researchers at MeasuringU reviewed the item and found no peer-reviewed validation of it, advising against giving the threshold undue weight. That is a fair criticism of the number. The practical problem is separate and applies even if 40% is exactly right: at the sample sizes founders use, a measured result cannot tell clearing the bar from missing it.
Sources
- How Superhuman Built an Engine to Find Product-Market Fit
First Round Review. The survey questions, the sampling rule, and the segmentation method described above.
- MeasuringU: The Product-Market Fit Item
A research review of the Sean Ellis question, tracing the 40% threshold to its originator’s experience rather than to published data.
- The 40% Product-Market Fit Rule Needs 369 Responses, Not 20
The sample-size arithmetic this calculator applies: Wilson score intervals at 95% confidence, reproducible from the count and the proportion alone.
Cite this tool
scoutr, "Product-Market Fit Survey" (September 2026), hosted Sean Ellis test with Wilson 95% intervals. https://www.scoutr.dev/product-market-fit-survey
Frequently asked questions
What is a product-market fit survey?
A short survey, designed by Sean Ellis, whose key question asks users how they would feel if they could no longer use the product. If at least 40% answer "very disappointed", the product is read as having product-market fit. The remaining questions ask who would benefit most, what the main benefit is, how to improve, and what the user would switch to.
How do I run a product-market fit survey?
Create one above with your product name, send the link to users who have used the core feature at least twice in the last two weeks, and read the results page. It shows the share who would be very disappointed, the 95% interval around it, whether the interval clears 40%, and the score per user segment if you asked for one. The survey is free with no response limit, and respondents do not need an account.
What is the 40% rule for product-market fit?
Sean Ellis benchmarked roughly a hundred startups and observed that the ones growing well had over 40% of users answer "very disappointed" to the key question, while the ones that stalled did not. The threshold is a rule of thumb from that experience. The underlying data was never published, and a MeasuringU review found no peer-reviewed validation of the item.
How many responses does a product-market fit survey need?
It depends on how far from 40% the result lands. Confirming a measured 45% takes 369 responses for the whole 95% interval to sit above the threshold. A measured 50% takes 93, 60% takes 24, and 70% takes 11. Results close to the line are the expensive ones to defend, and at 20 responses the margin of error is about ±20 points.
Who should receive the survey?
Users who have experienced the core of the product: used the main feature at least twice, most recently within the last two weeks. Surveying everyone who ever signed up mixes in people who never got to the value and drags the score down for reasons that have nothing to do with fit.
What questions are in the Sean Ellis test?
The scoring question about how the user would feel without the product, plus four open questions: what type of people would benefit most, what the main benefit is, how the product could be improved, and what the user would use as an alternative. The full template is on this page.
Is the Sean Ellis test reliable?
As a threshold at small sample sizes, no. With 20 responses, 8 "very disappointed" answers read as 40% with a true value anywhere from 22% to 61%. As a source of qualitative signal it is useful at any size: the free-text answers from the "very disappointed" group describe your best segment and your positioning in the users' own words.
Get the demand answer without waiting for 369 responses
Scoutr checks what your target user already pays for, which alternatives exist, and which of your assumptions is unproven. It works before you have a product to survey.
Validate my idea free →