When to send it
The product-market fit survey is meant for people who know your product well enough to imagine losing it. Send it to customers who have used its core at least twice and have been active recently, typically within the last couple of weeks. Leave out people who signed up and never came back, and anyone in their first days: their answers describe your onboarding, not your fit. Paste the link into an email from your usual email tool, and run it again after major changes to see whether things have moved.
Before sending, replace "our product" in the questions with your product's name.
Why these questions
The first question is the one Sean Ellis popularised: how would you feel if you could no longer use the product? There are four answers: very disappointed, somewhat disappointed, not disappointed, and N/A for people who have already stopped. Keep the wording exactly as it is, so your results stay comparable with other companies running the same test.
Ellis's rule of thumb is that when around 40% or more of customers say they'd be very disappointed, a product has usually found a market it can grow in. Treat that as a benchmark drawn from experience, not a law. With a small number of answers, a couple of people can swing the share by several points, and the line means little for products people naturally use once a year.
Customers who choose N/A skip ahead to one question: what made them stop. Everyone else answers two open questions from the original method. "What type of people do you think would benefit most?" shows who your keen customers think you're for, often in sharper terms than your own positioning. "What is the main benefit you get?" tells you what to lead with. The survey then asks what they would use instead, which names your real competition, and how you could improve the product for them.
Why there is no coupon
A discount would draw in people who barely use the product and nudge them to answer anyway, which waters down the very disappointed share. The test works best when replies come only from people who care enough to give them.
What to do with the answers
Hearsay's analytics show the share of each answer. Many teams leave N/A out when they work out the percentage, since those customers have already gone; whichever way you choose, do it the same way every round.
Then read the benefit and "who benefits" answers from the very disappointed group only. Those are the customers to build for: find more people like them, and strengthen the main benefit they describe. Use the improvement answers from the somewhat disappointed group to learn what would win them over, as long as it doesn't blunt what the first group loves.
Teams still before launch should start with early customer problem discovery. To learn who your buyers are, see who are our customers?.