Customer exit interviews: asking the people who left

Almost every cancellation form comes back with price at the top. Almost every team then argues about price, and churn does not move.

Price is the polite answer. It is true of anything nobody wanted enough, it ends the conversation, and it does not require explaining that the product was confusing or that nobody ever got round to using it.

A dropdown cannot ask a second question. That is the whole problem: the real reason is almost always one follow-up past the first answer, and a form has no follow-up.

Ask within a few days. The decision is fresh, the frustration is specific, and the person still remembers what they were trying to do when it went wrong.

Ask a month later and you get a story. People rationalise: a jumble of small irritations becomes one tidy reason, usually price or time.

This is why exit interviews work better as something triggered than something scheduled. The study runs when the cancellation happens, not when the quarterly research plan comes round.

Not 'why did you leave?' as the opener. It asks for a verdict, and you get the rehearsed one.

Start further back: what were you trying to get done when you signed up? You are looking for the job they hired you for, because plenty of churn is people who hired you for the wrong one.

Then: when did you first think about stopping? The gap between that moment and the cancellation is where the fixable problem usually lives.

Then: what did you try before you cancelled? Someone who tried support, docs and a workaround left differently from someone who quietly stopped logging in.

And last: what would have had to be different? Take the answer as a clue, not an instruction. People are better at describing the problem than specifying the fix.

Separate the churn you caused from the churn you never had. Someone who was never going to be a customer is not a bug. Someone who wanted the thing and could not get to it is.

Count, do not weigh. One articulate, furious customer will take over a room. The question is how many people hit the same wall, which is why the volume matters: ten interviews give anecdotes, two hundred give you a count.

Watch for the reason you cannot fix, too. If people leave because they finished — the project shipped, the move is over — that is not churn to prevent. It is a different product, or a lower price for coming back.