Exit Survey Questions for Cancellations: The SaaS Question Bank

I ran exit surveys for a B2B fintech client and built them for my own e-commerce stores. Most companies set them up so the answers don't sound like how the customer talks, so the cancel flow can't act on the data. Get the wording right and each answer routes to a matched offer.

This is the question bank I built for that work, organized by cancel reason with placement benchmarks included.

Key takeaways

• Write every exit survey question to confirm or rule out one cancel reason.
• Open-text fields boosted one team's response rate from 1.3% to 10.2%.
• Sort questions into the five reasons of price, product-fit, low usage, competitor, involuntary.
• In-context page surveys earn a 55% median response rate versus low single digits for post-cancel email.
• Cap the survey at one screen with no more than 3 questions.
• Route each confirmed reason to a matched offer, not a blanket discount that wastes margin on someone who was never price-sensitive.

What are exit survey questions for cancellations (and why HR templates fail here)?

These questions are what you show a customer when they cancel, written to confirm why they're leaving. Get them right and the cancel flow responds with the right offer. The best sets are short, use the customer's own words, and sort into:

• Price
• Product-fit
• Low usage
• Competitor switch
• Involuntary

Some customers leave for a structural reason, like a company that shut down or a project that ended. No question surfaces a fixable cause there. The survey's job is to tag those exits as unsaveable, so you stop spending offer budget chasing them.

The one rule behind every good exit survey question

Every exit survey question should do exactly one thing: confirm or rule out one cancel reason. Its options use the customer's words, and it caps at one screen.

A question that maps to one reason gives you a signal you can act on. A vague or double-barreled question gives you a flat comment a human has to read later, which is how most exit-survey data dies in a spreadsheet.

The wording matters just as much. Options in marketing-team language ("insufficient value realization") get skipped, while options written the way a customer talks get picked.

I learned this the hard way, watching real customers move through surveys I'd built. They weren't annoyed by the survey itself, but by options that matched nothing they'd actually say, so they clicked the fastest one to escape.

Groove's test swapped a closed dropdown for a plain open-text field and increased exit-survey response from 1.3% to 10.2%. Changing the prompt from "Why did you cancel?" to "What made you cancel?" then pushed it to roughly 19%.

For high-value enterprise accounts, a short follow-up call gets you more than any in-flow form. Above a certain revenue line per account, the survey's job changes. It flags the account for a human instead of being the whole conversation.

Write answer options in the customer's words

The fastest way to kill your response rate is to write answer options in your language, not the customer's. A customer skims the list for one match. They look for the option that sounds like the sentence in their head, then tap it.

If nothing matches, they pick whatever ends the survey quickest. Your data is now noise.

The fix is to pull the wording from real cancellation transcripts. Read what customers actually said when they left, then use those exact phrases as your options.

The customer should feel like you wrote the option after listening to them, because you did. "It got too expensive" beats "cost concerns."

Lead with one open-text question, then the multiple choice

Open with a single open-text question before you show any list. A blank box asks the customer to say it in their own words first. That catches the reason you might not have a checkbox for.

The Groove jump from a closed dropdown to an open field is the same lesson: give people room to talk and more of them will.

Keep it to one box, though. Two open-text questions back to back feels like homework, and the second one goes blank. Ask the open question first, then offer the multiple-choice list as the fast path for anyone who would rather tap than type.

One screen, one decision

Cap the whole survey at one screen, because every question you add costs you responses. The customer is mid-exit and already half gone. A second screen is just a second chance for them to close the tab.

You can’t ask everything, so you ask only what routes to an action you can actually take. If an answer wouldn't change which offer you show next, the question doesn't earn its place on the screen.

Exit survey questions by cancel reason (the question bank)

Build the survey one reason at a time, with a few questions for each of the five cancel reasons. Each question confirms that one reason and nothing else. That is the difference between a question bank and a generic list.

The map below gives one confirming question per reason, in the customer's words:

Building by reason instead of by topic buys you two things a flat list can't. You get coverage, because every reason a customer might leave for has a question that catches it. And you get routability, because each answer points at a cancel-flow response. 

Watch how much one badly merged option hides. "Too expensive?" sounds like a price question, but "I wasn't using it enough" and "the price went up" are two different signals wearing the same coat.

The first is a usage problem and the second a real price objection. Collapse them into one checkbox and you can't tell which customer needs an onboarding nudge and which needs a downgrade.

You don't have to ship all five segments at full depth on day one, and you shouldn't.

Start with the two reasons your customers actually leave for. Most B2C subscription businesses skew toward price and low usage, while most B2B skews toward product-fit and competitor switch. Which two depends on your cohorts, so read your cancellation data before you commit.

Price questions

Price answers are the easiest to misread, so the questions have to separate "too expensive" from "not worth it to me right now." Those are different problems. Each has its own fix, and one blunt price option buries the difference.

Ask the customer to choose between a few specific price sentences instead of a single yes-or-no on cost.

"I wasn't using it enough to justify the cost" tags a value problem. "The price went up and I can't justify the new rate" tags a real price objection. Each one routes somewhere different, which is the whole point.

One thing is worth keeping in mind.

A reflexive "too expensive" is the easiest thing for any customer to click, because it puts the reason on you rather than on them. The open-text question you led with is your check against that, since the customer's own words usually reveal whether price was the real driver.

Product-fit and missing-feature questions

Product-fit questions confirm the customer needed something the product doesn't do, so they should ask what was missing.

Don't ask whether the customer was happy. A satisfaction question returns a vague rating, while a missing-capability question returns a roadmap signal.

Give them options that name concrete gaps drawn from your support and sales notes. "It was missing a feature I needed," with a short open field, turns a cancellation into product feedback you can route to the right team.

"It didn't work the way I expected" is a different tag, closer to an onboarding problem than a true gap.

This is also the segment where one question earns its keep more than the rest. When a customer leaves for something specific your product lacks, that answer surfaces the exact gap a competitor is filling.

It's the most useful sentence in the whole survey, so make sure a question is positioned to catch it.

Low-usage and activation questions

Low-usage questions confirm the customer never got far enough to see the value, which is an activation problem hiding inside a cancellation. Ask about value, not logins. 

Find out whether they ever reached the point where the product pays off, not just whether they signed in.

Activation is the first moment a customer hits real value in your product, like saving their first report or connecting their first integration, so ask in those terms.

"I never got it fully set up" and "I set it up but didn't use it" are two different activation failures, and they need two different follow-ups. The first wants a setup assist, while the second wants a reason to come back.

This segment matters so much because a low-usage canceler is often your most saveable customer.

Nothing was wrong with the product or the price. They just didn't get over the first hill.

Catch that reason and the cancel flow can offer help instead of a discount the customer didn't need.

Competitor-switch questions

Competitor-switch questions confirm the customer is moving to a named alternative, the most valuable answer your survey can capture. A competitor win names your weak point. It tells you exactly which gap or price point beat you.

Ask directly and give room to name names. "I'm switching to another tool," with an open field for which one, turns a lost customer into competitive intelligence.

The follow-up that matters is why, since a missing feature, a lower price, or a better experience each send you somewhere different.

In my experience this is the one cancel reason where the open-text answer is worth more than the multiple choice, because the specifics are the prize.

Hold this segment a little looser than the others, though. A customer who has already signed up elsewhere is usually past saving in this flow. The job is to record why you lost, so the next customers in that segment don't leave for the same reason.

Involuntary and "I'll be back" questions

Involuntary churn isn't a choice, so the survey's job here is to separate a failed payment from a real cancellation.

A failed payment routes to a card update, not a save offer, and a temporary leaver routes to a pause. If cancel flow vs. dunning is new territory, that article covers why the two tools handle different triggers.

There's a limit on what the in-flow survey can do here. Most involuntary churners never reach the cancel flow at all, because a failed payment ends the subscription quietly in the background.

The survey catches only the ones who show up to fix it themselves. The rest need a different surface, which is where placement comes in.

Where and when to ask: survey placement in the cancel flow

Ask the survey inside the cancel flow at cancellation, not in a post-cancellation email, because in-flow placement earns more responses. Placement sets your response-rate cap.

A survey shown in context catches the customer while they're still on the screen, mid-decision. An email reaches an already-gone customer and asks them to come back and do unpaid work, which almost none will.

Response rate is the share of surveyed customers who actually answer.

Survicate's 2025 report found page surveys in context earn a 55.07% median response rate. A widget tucked in the corner earns 7.64%.

Groove's post-cancel email started near the bottom of that range, which is where an after-the-fact email lands. That gap is why in-flow placement is a core cancel flow best practice.

The email still has one job the in-flow survey can't do, which is reaching involuntary churners whose payment failed and never saw a cancel flow. So run in-flow as the main surface, with a short email as the backstop for the payments that fail quietly.

The part most teams skip is capturing the reason before the offer shows, which is what our Churn.io exit surveys handle. The reason a customer picks routes to the matched offer on the same screen.

Turning answers into action (route each reason to an offer)

An exit survey only pays off when each confirmed reason routes to a real cancel-flow response, so design it backward from your offers. No offer, no question. A reason with no action waiting for it is just a comment.

Every question you keep should have a response on the other side. A price answer needs a downgrade or a pause ready to show, and a low-usage answer needs an onboarding nudge or a setup call.

If you can't act on a reason, asking about it only lowers your response rate, since it's one more question between the customer and the exit.

A customer who selects "I wasn't using it enough" should never see a discount. Cutting the price doesn't fix the problem, and it spends margin on someone who was never price-sensitive. Routed correctly, that same answer triggers a pause or a re-onboarding push, which protects the revenue a blanket discount would have given away.

Weighting reasons by how often they happen and the revenue each represents, plus the full reason-to-offer map, is its own discipline. Our churn reasons analysis walks through that routing table in full.

Design the survey so every answer is routable, then let the Churn.io cancel flow tool handle the match automatically.

FAQ

How many questions should a cancellation exit survey have?

Keep it to one screen, which in practice means one open-text question plus one or two multiple-choice questions. Every question past that drops your response rate more than it adds in data.

Should a cancellation survey be required before the customer can cancel?

No, never block the cancellation on the survey. A forced survey produces fast junk answers and the kind of cancel-flow friction that kills the customer's odds of ever coming back.

Are customer exit surveys different from employee exit surveys?

Yes, completely. An employee exit interview gathers open reflection after the fact. A cancellation survey runs live in the cancel flow, with one screen to capture a reason precise enough to trigger an offer.