Exit survey questions: 10 to ask by cancellation reason
10 sample exit survey questions for subscriptions, in the order to ask them, with the follow-up and the offer each answer should trigger, or no offer at all.

Ask the reason before you offer anything to save the customer. A customer one click from cancelling may be leaving because of price, an expired card, an unresolved support ticket or a project that has ended. A discount shown before you know which one can cut the price for someone who only needs fewer seats, or promote a deal to someone who wants an incorrect charge explained. Use the answer to pick a relevant alternative, a support action, or a clean cancellation with no offer.
The 10 questions below work as a branching sequence, not a form every customer must finish. Start with one structured reason, ask only the follow-ups that change your response, and end with optional context and permission to get back in touch.
Practical rule: if an answer won’t change an offer, a follow-up or an operational decision, leave the question out of the cancellation flow.
Keep cancellation available at every step. A useful survey helps you understand the decision without making customers justify their right to leave.
1. What is the main reason you’re cancelling?
Sample question: “What’s the main reason you’re cancelling your subscription?”
Use options that lead to different actions: too expensive, not using it enough, missing features, technical problems, billing trouble, switching to another product, temporary need, and other. For seat-based software, add team size changed when reducing seats is a real alternative. Stripe’s customer portal offers a similar built-in list (too expensive, need more features, found an alternative, no longer need it, customer service, ease of use, quality, other reason), with free text when the customer picks “Other reason”.
A single choice works well for routing because it names the primary issue. It does not prove the customer had no other concerns, so leave room for optional context later.
Give every option a destination
Say a B2B SaaS customer picks team size changed after a reorganisation. Offer a seat adjustment rather than assuming they need a cheaper version of the whole product, and show the new commitment clearly before they accept.
- Price: check whether a lower plan fits what they actually use.
- Temporary need: offer a pause if your billing setup supports it.
- Technical problem: give a support route without blocking cancellation.
- Other: collect optional text and review recurring themes monthly.
Ask this question before any discount, so customers don’t pick a reason in response to an offer. Record the invitation, the answer, the path shown and the final outcome: you need both what customers said and what they did. Our cancellation reasons list covers the taxonomy in detail.
2. Would a pause or a smaller plan meet your needs?
A customer who still likes your product may only need less of it. This question separates a temporary interruption from a permanent mismatch.
Sample question: “Would pausing your subscription or moving to a smaller plan meet your current needs?”
Offer separate answers for a pause, a smaller plan, neither, and not sure. Don’t merge the positive answers into one “interested” bucket: each needs different billing behaviour.
Test the actual alternative
Say a consumer subscription serves customers working on seasonal projects. Someone who just finished a project might accept a pause, while someone using the product steadily but lightly might prefer a downgrade.
Before they accept, explain:
- Billing: what stops, continues or changes. On Stripe, pausing payment collection keeps the subscription active and invoices are still created, but no payment is collected, and invoices created before the pause keep being retried unless you void them.
- Access: which features, seats or saved work stay available.
- Restart: when and how normal billing resumes.
- Exit: where the customer can still finish cancelling.
Whether a pause or a discount fits depends on the underlying need: a discount changes the price, a pause changes the commitment. See pause vs discount retention offers.
Compare stated willingness with actual acceptance, then check whether customers are still subscribed after the alternative ends. Choosing a pause is not the same as returning to paid use. Test one clearly defined alternative rather than a menu of competing offers whose results you cannot read.
3. Was a billing or payment problem part of your decision?
Check payment history before asking customers to explain it. If your billing tool already records a failed payment, use that context instead of showing an unrelated price offer.
Sample question: “Did a payment failure, an outdated payment method or an unexpected charge contribute to your decision?”
Keep those answers separate. A failed payment calls for a recovery path; an unexpected charge may need an explanation, an invoice correction or a human review.
Separate collection problems from cancellation intent
Say a mobile subscriber reports that access stopped after their card expired. Check the subscription’s billing channel and status before offering a way to update the card. An App Store or Google Play subscription is updated through the store, not through a payment form meant for customers you bill directly.
- Outdated details: link to the right hosted card update page or account settings.
- Temporary failure: check the retry schedule already running before adding another attempt.
- Unclear charge: show the invoice and offer billing support.
- Deliberate departure: respect the cancellation, even if a payment issue also exists.
Keeping involuntary churn apart from persuasion matters here. Don’t count a recovered overdue invoice as a saved cancellation. Track payment recovery, subscription status and customer intent separately, or your team may read better collection as happier customers.
4. How likely are you to recommend us?
Sample question: “How likely are you to recommend this product to a friend or colleague?” (0 to 10)
A recommendation score adds a standard sentiment signal, but it does not explain a cancellation on its own. Pair it with the reason and an optional open question so answers stay comparable while keeping their context.
In an exit flow, use it to tell dissatisfaction from changed circumstances. Someone can value your product, recommend it warmly, and still have no reason to pay right now.
Read the score next to the reason
Say a project management customer gives a high score but picks project finished. A pushy save offer is a poor fit: confirm the cancellation and, with permission, get back in touch when a new project may start.
A customer who reports a technical failure with a low score needs something else: access to someone who can investigate, not a referral request.
- Positive score, temporary need: consider a pause or a reminder they agreed to.
- Positive score, low value: look at packaging and usage fit.
- Negative score, unresolved problem: route the issue to its owner.
Keep the wording and the scale direction the same across cohorts. Segment by plan, tenure and acquisition source, but don’t label every unhappy respondent a product problem: billing, onboarding, support and expectations all move the answer.
5. What would have made you stay?
Closed choices tell you where to route a customer. Optional text can show why that route exists.
Sample question: “What would have needed to change for you to keep your subscription?”
This beats a general comments box because it asks for a condition you can assess. Keep it optional, with no minimum length.
Code the underlying need
Say a B2B customer writes that they needed a sandbox environment. Tagging it only as missing feature loses the detail. Keep the specific need, the workflow it serves and the customer segment.
Use one shared set of themes (pricing, features, onboarding, reliability, support, outside circumstances) and review new answers weekly. Don’t promise roadmap changes just to reverse a cancellation.
A specific request is evidence of a need, not proof that building the feature will keep the customer.
Say who can see the feedback and how you will use it. Don’t promise anonymity when answers are attached to subscription records, and ask before quoting identifiable feedback publicly.
6. Are you switching to another product?
A competitor’s name helps only if you know what the customer expects to gain. A lower price, a simpler setup, a required integration and a company-wide consolidation call for different decisions.
Sample question: “Are you moving to another product? If so, which one, and what made it a better fit?”
Use a maintained list of competitors with other, not decided yet and not switching. Keep the explanation optional.
Respond to the advantage, not the brand
Say a customer of an AI writing tool names a cheaper alternative. Check whether they also report lower usage: the problem may be a plan that no longer fits their workload, not a verdict that the whole product costs too much.
- Cheaper fit: compare a smaller plan with what they need.
- Required capability: record the gap and don’t promise an uncommitted release.
- Company consolidation: offer a clean exit, unless you can address the actual buying constraint.
- Simpler workflow: ask which task became easier.
Use these answers for pricing and product decisions as well as retention messages. Don’t trigger a discount every time a competitor is named: it won’t solve a mandated vendor change, and it can hide the workflow problem your team needs to see.
7. How would you rate the value you received?
A customer can like a product and still decide it no longer earns its price. Separate satisfaction from perceived value so the answer points to a useful path.
Sample question: “Considering what you paid, how would you rate the value you received?”
Use a 1 to 5 scale, label both ends clearly, and keep the direction consistent with your other rating questions.
Match the answer to the context
A subscriber who rates value low and uses only a narrow part of the top plan may need a smaller package, not a temporary discount on features they don’t use. Compare the rating with plan, tenure, billing cycle and usage you already have. Don’t ask customers to re-enter data you hold.
- Low value, low usage: a smaller plan, a pause or a clean cancellation.
- Low value, heavy usage: ask what outcome or price they expected before proposing anything.
- Good value, temporary need: price is not the obstacle; ask what changed and whether a pause fits.
For annual renewals, say whether the question covers the year that ends or the renewal ahead. Treat ratings from small groups as directional. If a customer accepts an offer, check retained revenue and later status rather than counting the first click as a save.
8. How often did you use the product recently?
Usage logs show what the system recorded. A short question shows whether that activity still means something to the customer.
Sample question: “How would you describe your recent use of the product?”
Use plain options: regularly, occasionally, rarely, not at all. Add an optional explanation when the answer would change your response.
Measure the workflow, not just logins
For an API product, a customer may stop opening the dashboard while production usage continues. For seat-based software, the admin who cancels may never have used the core workflow. Compare the answer with the events that matter: completed tasks, active seats, API calls, or use of the workflow they bought.
- Need disappeared: a pause or a reactivation path they agreed to.
- Setup never worked: focused onboarding help, if they want it.
- Core workflow stopped: find the blocker before suggesting a discount.
- Activity without outcomes: check whether your engagement metric measures value.
Say users stop completing a review workflow after an integration breaks. A reminder to log in won’t fix that. Route the feedback to the integration owner, and keep it apart from customers whose projects simply ended.
9. How would you rate your support experience?
Don’t ask customers to rate support they never contacted. Use ticket history to decide whether the question applies, and always include not applicable.
Sample question: “How satisfied were you with the help you received before deciding to cancel?”
If the answer is negative, ask one focused follow-up: unresolved issue, slow reply, unclear answer or repeated handoffs. Don’t make the customer retell a conversation your support tool already holds.
Separate a bad conversation from an unresolved cause
Say a SaaS customer leaves after an invoice question went unanswered. Read the ticket next to the subscription record: the right response may be a billing explanation, not troubleshooting or a discount.
- Unresolved billing issue: route to someone who can investigate the charge.
- Technical blocker: link the case to the product or engineering owner.
- Repeated explanations: carry the existing context into any follow-up.
- No more contact wanted: confirm the cancellation without another pitch.
Boundary: offer human help when it is useful, but never make talking to a person a condition of cancelling.
Check the order of events too. A ticket sent after the decision may describe friction in cancelling, not the original cause. Fix both, but don’t merge them into one churn reason.
10. Would you consider coming back, and under what conditions?
Interest in returning is not permission to send marketing. Ask about the condition first, then offer a separate, optional contact choice.
Sample question: “Would you consider subscribing again, and what would need to change?”
Useful answers include a new project, a smaller plan, a missing feature becoming available, a technical issue fixed, not sure, and no interest.
Store the condition with the permission
Say a customer would return once a given integration exists. Record the integration and their contact preference together, and keep them out of unrelated discount campaigns.
- Temporary interruption: time the follow-up to what they told you.
- Feature dependency: write when the feature ships.
- Price or packaging: share a relevant plan change or offer.
- No permission: exclude them from win-back.
Win-back campaigns should follow the reason for leaving, not just the time elapsed. Keep a record of the wording shown, the customer’s choice and later opt-outs, and check eligibility again before sending, in case they have already come back. Measure reactivations by stated condition and look at retention afterwards: a customer who returns briefly for an offer has not necessarily found a lasting reason to stay.
How do these exit survey questions differ?
| Question | Format | When to ask | What it reveals | What to do with it |
|---|---|---|---|---|
| 1. Main reason | Single choice with “Other” | Before any offer | The main reason for leaving | Route to the right offer, support path or clean exit |
| 2. Pause or smaller plan | Multiple choice | After the reason | Which lighter commitment they would consider | Offer a pause or downgrade; compare with real uptake |
| 3. Billing or payment problem | Yes or no, optional detail | When payment history shows an issue | Whether the churn is involuntary | Offer a card update; track involuntary churn apart |
| 4. Likelihood to recommend | 0 to 10 | Next to the reason | Sentiment of departing customers | Segment sentiment; guide product or win-back decisions |
| 5. What would have made you stay | Optional free text | Near the end | Unexpected needs or gaps | Code themes and share them with the owners |
| 6. Switching product | Yes or no, optional competitor and reason | During the survey | Competitors and what draws customers to them | Inform pricing, packaging and win-back |
| 7. Value received | 1 to 5 | Next to usage | Whether the price felt worth it | Choose between a discount, a pause or a downgrade |
| 8. Recent usage | Frequency options | During the survey | Engagement and adoption gaps | Trigger onboarding, a pause or reactivation |
| 9. Support experience | Rating, optional detail | Only for customers who contacted support | Whether service played a part | Escalate open cases; fix recurring problems |
| 10. Coming back | Yes, no or later, optional condition | At the end | Openness to returning and its condition | Build win-back based on consent and reason |
How do you turn each answer into a next step?
The best exit survey is not the longest. It is the shortest one that lets your team make a better decision while leaving the customer free to cancel. Ask the main reason before any offer, show only the follow-ups that fit that reason, keep free text optional, and pull plan, billing, usage and support data from your systems instead of asking for it.
- 01 Payment problem The right card update route or a billing investigation
- 02 Temporary need A pause with clear restart terms
- 03 Plan mismatch A relevant downgrade or seat change
- 04 Unresolved service issue A support follow-up with the context attached
- 05 No suitable alternative No offer, and immediate cancellation
Assign a path and an owner
- Payment problem: the right card update route or a billing investigation.
- Temporary need: a pause with clear restart terms.
- Plan mismatch: a relevant downgrade or seat change.
- Unresolved service issue: a support follow-up with the context attached.
- No suitable alternative: no offer, and immediate cancellation.
Review other answers monthly to find gaps in your list. Compare stated willingness with actual acceptance, then look at later subscription status and retained revenue. Keep completed cancellations, accepted alternatives and recovered payments apart, so your reporting does not overstate saves.
Treat answers as evidence, not as everyone
Customers who answer may differ from those who skip. Report how many were asked and how many answered, compare respondents with all cancellations, and resist turning a few answers into broad product conclusions. A recurring complaint deserves investigation, but it does not prove every silent customer had the same problem. Combine survey answers with usage, support, billing and renewal records before a major change.
How NoChurn helps
NoChurn’s cancel flows ask the reason first, then show an offer that fits it (a pause, a downgrade or a discount) or no offer, with one-click cancel always visible. Win-back emails go out at 30, 60 and 90 days according to the reason given. It works with Stripe, Chargebee, Recurly, Paddle and other billing tools, or with any billing through the API.
FAQ
How many questions should an exit survey have?
As few as change your response. One structured reason is enough for routing; add only the follow-ups that fit that reason, and keep free text optional. If an answer would not change an offer, a follow-up or a decision, drop the question.
Should you show a retention offer before or after asking the reason?
After. A reason chosen in front of an offer is biased by the offer, and an offer chosen without the reason is often the wrong one: a discount for someone who only needs fewer seats, or a price deal for someone with a billing error.
Can an exit survey be mandatory?
Make it skippable. Cancellation should stay available at every step, and talking to a person should never be a condition of leaving. Answers forced out of customers are also less reliable.
Sources
All accessed 22 November 2026.