Form questions
Churn survey questions
Ask six to ten, after the cancellation is confirmed rather than as a step in confirming it. Three are never cut: one primary reason as a single choice, a free text box for the reason your list missed, and what would have had to be true for them to stay. The free text box is where the useful answers live.
A cancellation survey gets one honest answer or none at all, and which one you get is usually decided before the first question. Run it after the cancellation has already gone through, keep every retention offer out of it, and the answers change completely.
The questions
The reason
What is the main reason you are cancelling?Single choiceRequired
One reason, the biggest one. A multiple choice version produces four ticks and no ranking, which is how a churn report ends up saying everything matters.
If none of those is right, what is the real reason?Single choice
Ask it as a short text field rather than relying on an Other box, so the answer lands in a column somebody reads.
How long did you use it before deciding?Single choice
When did you first start thinking about leaving?Single choice
Usually months earlier than the cancellation date, which is the finding.
Was this your decision or somebody else's?Single choice
What would have kept them
What would have had to be true for you to stay?Long textRequired
At what monthly price would you have kept the plan you are on?Single choice
Bands, not a free number, and no offer follows from the answer. Say that in the question or the answers are negotiating positions.
Which of these did you run into?Multiple choice
Slow or unreliable, needed something on a higher plan, did not fit how we work, nobody else adopted it, support too slow, no integration.
Did you raise any of this with us before deciding?Single choice
If you did, what happened?Long text
Would you come back if the thing you named were fixed?Single choice
Where they went
What are you moving to?DropdownRequired
A paid tool, a free tool, spreadsheets, something built in house, nothing at all, not decided. Six answers that each mean something different.
What does the new one do that this did not?Long text
How long did it take you to choose the replacement?Single choice
Who made the final call?Single choice
Closing
What did it do well while you had it?Long text
Almost always answered, and it is the only question in a churn survey that protects a feature from being cut for the wrong reason.
How likely are you to recommend it to someone else anyway?Single choice
0 to 10. A customer can leave and still recommend it to a different sort of team.
How was the cancellation itself?Single choice
Do you want your data exported before the account closes?Single choice
May somebody from the product team reply to this?Single choiceRequired
Anything else you want to say?Long text
Which questions to make required
Four are required. The primary reason and what they moved to are the two fields any churn report is built from, so a response missing either cannot be counted. What would have kept them is required because it is the question the whole survey exists to ask. The permission question is required because it decides whether a reply is welcome, and the safe default when it is blank is silence.
- What is the main reason you are cancelling?
- What would have had to be true for you to stay?
- What are you moving to?
- May somebody from the product team reply to this?
The form built from these questions
A static copy of a form built from part of the bank, exactly as it renders.
Three ways to change this bank
For a membership or a gym, replace the moving to question with the circumstance: moved house, changed job, injury, cost, stopped going. Most memberships end because of a life event rather than a competitor, and a list built for software finds none of them. For a shorter version, ask three: the main reason, the free text box for the reason not listed, and what would have kept you. Three questions after a confirmed cancellation get answered at a much better rate than eight inside the cancellation flow. For a follow up wave, write six months later to the people who said they would come back if a specific thing were fixed, once it is fixed, and say what changed. That is a genuine reactivation message rather than a discount email, and the response is worth reading on its own.
Questions people ask
Should the survey come before or after the cancellation?
After. A survey that stands between a customer and the cancel button is read as an obstacle, and the answers are chosen to get past it. Cancel first, confirm it plainly, then ask. Say on the first screen that the plan is already ended, or people assume the questions are a delaying tactic.
How many questions will people answer?
Fewer than you would like. Six to ten if the cancellation is already done, three or four if you are asking inside the flow. Put the primary reason first so that even an abandoned response carries the one field your report needs, since a partial submission is not stored.
Why ask for one reason instead of several?
Because a multiple choice reason question produces four ticks and no weighting, and a churn report built on it concludes that price, features and support all matter equally. One primary reason plus a free text box gives you a ranked list and the thing your list missed, which is usually the interesting part.
Should I offer a discount in the survey?
No. An offer changes what people write, because the answer becomes a bargaining position rather than a description. If you want to run save offers, run them somewhere else in the flow and keep this form clean. Saying plainly that no discount follows from any answer is worth a sentence in the question itself.
What response rate should I expect?
Lower than any other survey you run, and that is normal. People who are leaving have no ongoing relationship to maintain. A short form shown immediately after the confirmation does best. Read the results as directional rather than representative, because the customers who leave quietly are the ones least likely to answer.
Is a churn survey anonymous?
It is attached to the account that cancelled, which is what lets you read the answer next to how the account was actually used. Say so. If you promise anonymity you cannot deliver, the first customer who gets a reply referring to their answer will tell everybody, and the next wave is worthless.
Build this form in about a minute
The button opens the form generator with this form already described. Change the wording to match your own, generate, then edit any field.
Create a form with OneCraftOther question banks
Subscription pause survey questions
A pause is the last moment a subscription business gets a straight answer, and most of them waste it by building a retention screen instead of a survey. Apply the pause first, ask afterwards, and the answers are worth reading.
Customer onboarding survey questions
An onboarding survey is worth sending at the point where a customer either has the thing working or has quietly stopped trying. These questions separate what somebody signed up to do from what they have actually managed, because the gap between the two is the whole finding.
Brand perception survey questions
Brand perception is measured in a fixed order, because the moment a survey names a brand it has destroyed its own unaided recall question. Open questions first, then aided awareness, then attribute by attribute, and the sequence matters more than the wording.
On writing the questions themselves, read How to write survey questions. For everything the builder can do, see the form maker. Step by step: Create a conversational form with AI, then View, export and manage your responses.
Written and checked by the OneCraft team. Last checked .