Product market fit survey calculator
By Arun Agrahri, founder of Gavel · Published
A product market fit survey asks users how they would feel if they could no longer use your product. Your score is the share who answer very disappointed. Sean Ellis, who made the survey, calls 40% or more a leading sign of product-market fit (Lenny's Podcast, 6:29).
Score your survey here. Paste the answers or type the counts, then press Get my score. Each person's answer is very disappointed, somewhat disappointed or not disappointed.
Copy the answer column from your survey tool. Any line that is not one of the three answers is left out and shown to you.
Example result: Lookout's first survey
7%
Sean Ellis says on Lenny's Podcast that only 7% of Lookout's users would have been very disappointed without it (9:20). That is far under the 40% line.
Your own result shows your score, the share of each answer and how many more very disappointed answers would reach 40%.
Have your score? Ask Gavel what to do next.
Gavel answers startup questions with what a named founder did, and links to the minute they said it.
Ask Gavel freeA real score: Lookout went from 7% to 40% in two weeks
Sean Ellis ran this survey at Lookout, a phone security app, and tells the story on Lenny's Podcast. Only 7% of users said they would be very disappointed without it. Most of that 7% used one feature: antivirus. So the team led its pitch with antivirus and made it the first thing a new user set up. The next group of users scored 40%.
| When | Very disappointed | Source |
|---|---|---|
| First survey | 7% | Ellis at 9:20 |
| Two weeks later | 40% | Ellis at 11:45 |
| Six months later | 60% | Ellis at 11:53 |
Your score is very disappointed answers divided by all answers
Count the very disappointed answers. Divide by the number of people who gave one of the three answers. Show that share as a percent, and you have your score. Blank lines and other answers are left out, and the calculator tells you how many it skipped.
The survey to send: one question, three answers
Send this question with your product's name in it. Offer these three answers and no others.
How would you feel if you could no longer use [your product]?
- Very disappointed
- Somewhat disappointed
- Not disappointed
Ask people who have really used the product
Ellis sends the survey to a random sample of people who signed up, reached the main use and came back at least twice (Lenny's Podcast, 22:38). Brand new sign-ups have not felt the value yet, so they pull the score down.
a random sample of people who've really used your product
Rahul Vohra did the same at Superhuman. He asked only users who had felt the main benefit, so a slow start would not look like a weak product (Vohra interview, 33:23).
Ellis does not trust a score until at least 30 people have replied. He says he made up that floor, then heard from others that it is the usual minimum (Lenny's Podcast, 33:11). The calculator warns you when fewer people replied.
Where experts disagree: ignore the somewhat disappointed, or win them over?
Ellis tells a founder stuck near 10% to ignore the somewhat disappointed group (Lenny's Podcast, 0:20). He says those users see the product as a nice-to-have. Change it for them and you may dilute it for the users who need it.
Just ignore the people who say they'd be somewhat disappointed.
Vohra splits that group in two. The somewhat disappointed users who like the same main benefit as his fans get half of each planning cycle. He sets the rest aside (Lenny's Podcast, 56:35).
spending half your time doubling down on what people really love
Our call: follow Ellis while your score is low and you cannot yet name what your fans love. Once the very disappointed group points to one main benefit, use Vohra's split to pick who to win next.
Not sure which call fits your product? Ask the question with your own numbers.
Gavel answers startup questions with what a named founder did, and links to the minute they said it.
Ask Gavel freeQuestions about the product market fit survey
How firm is the 40% line?
Not very firm, Sean Ellis says. Asked if 39% and 41% differ much, he calls the line a target. The team agrees not to push growth hard until it gets there. Hear him at 17:45.
My score is around 10%. What should I do?
Read only the very disappointed answers and ignore the somewhat disappointed ones, Sean Ellis says at 0:20. Find what that small group values and lead with it. At Lookout he did this with antivirus, and the score went from 7% to 40% in two weeks, he says on Lenny's Podcast at 9:20.
Do I count people who say they no longer use the product?
No. The calculator counts only the three answers Ellis names. It leaves out any other line, such as a header row or N/A, and tells you how many it skipped.