Cancellation Surveys
A cancellation survey is a structured questionnaire shown to (or sent to) every customer at the moment they cancel, asking why they're leaving. The good ones are short — one required question with 5-7 mutually exclusive options ('Pricing', 'Missing feature', 'Switching to competitor', 'Project ended', 'Onboarding too hard', 'Other'), plus one optional free-text follow-up. The data feeds two things: (1) segmentation for win-back campaigns (different reasons get different sequences), and (2) a closed-loop signal to product/marketing/CS about WHY customers leave. Without exit surveys, churn is a black box — you know the rate but not the cause. With them, churn becomes a diagnosable condition.
The Trap
The trap is making the survey too long. A 12-question exit survey gets a 4% completion rate; a 1-question survey gets a 70% completion rate. Founders who want to learn 'everything' from departing customers learn nothing because nobody fills it out. The other trap: collecting the data and never acting on it. Survey responses pile up in a spreadsheet that nobody reads, while product roadmap and CS playbooks continue based on intuition. The survey only matters if results route to people who can act — and the routing is automated.
What to Do
Build the simplest possible exit survey: ONE required question with 5-7 reason buckets, ONE optional free-text. Show it BEFORE the cancellation completes (not after — completion rates are 5x higher pre-cancel). Tag every cancellation with the reason. Set up automated routing: 'Pricing' → revenue ops sees monthly summary, 'Missing feature' → tagged in product backlog, 'Onboarding' → CS leadership reviews, 'Competitor' → competitive intel team gets battle card update. Review the aggregate monthly: which reason is growing? That's your highest-leverage churn driver to fix this quarter.
Formula
In Practice
Drift built their exit survey as a single question with 6 reason buckets, shown immediately before the cancel-confirmation button. Completion rate was 71%. Aggregating 18 months of data revealed that 'missing feature: account-based filtering' was their #2 cancellation reason — driving ~$1.2M in annual ARR loss. Engineering had been deprioritizing this feature for 14 months. The survey data forced the issue onto the roadmap; the feature shipped 5 months later. Immediately, the win-back team contacted everyone who had cancelled for that reason and reactivated 31% of them.
Pro Tips
- 01
Show the survey BEFORE the customer clicks 'Confirm Cancel'. Once the cancel completes, completion rates drop 5x — they got what they wanted (out) and have no incentive to fill out a form. Pre-cancel surveys often double as a save mechanism: 'You said pricing — would 30% off for 6 months change your mind?' converts 8-15% of pricing-driven churners.
- 02
Always include 'Other (free text)' as an option. The structured buckets capture 80% of reasons, but the 20% in 'Other' often surface NEW failure modes you haven't categorized yet. Read these monthly — they predict the next bucket you need to add.
- 03
Tag survey responses by ACR (annual contract value). 'Pricing' may be 50% of low-ACR cancellations but only 5% of high-ACR cancellations. Aggregate-only reporting hides that high-value churn has different drivers than low-value churn — and the high-value drivers are the ones to fix first.
Myth vs Reality
Myth
“Customers won't tell you the real reason”
Reality
70%+ completion rates on well-designed surveys, and post-survey interviews show ~85% of structured responses match the actual driver. Customers will tell you the truth if you make it easy to. They lie when the survey is too long, requires written responses, or feels like a save attempt instead of a learning exercise.
Myth
“We already know why customers churn”
Reality
CS leaders' intuitive churn reasons are wrong about 40% of the time, by survey vs intuition comparison studies. The CSM team thinks customers leave for X (the loudest customers' reason). The data reveals it's actually Y (the silent majority's reason). Surveys correct the bias toward the noisiest churners.
Try it
Run the numbers.
Pressure-test the concept against your own knowledge — answer the challenge or try the live scenario.
Knowledge Check
Your exit survey shows the top cancellation reasons are: Pricing (40%), Missing feature (28%), Switching to competitor (15%), Project ended (10%), Onboarding (7%). Where should you focus first?
Industry benchmarks
Is your number good?
Calibrate against real-world tiers. Use these ranges as targets — not absolutes.
Exit Survey Completion Rate
B2B SaaS exit survey response rates by designBest-in-Class (1Q pre-cancel)
> 60%
Good (1-3Q in confirmation)
20-40%
Average (5-7Q post-cancel)
8-15%
Useless (>10Q email survey)
< 5%
Source: Hypothetical: KnowMBA composite from CS platform vendors
Most Common Cancellation Reasons (B2B SaaS)
Aggregated B2B SaaS exit survey reasonsPricing / Budget
25-40%
Missing Feature
20-30%
Project Ended / No Longer Needed
10-20%
Switching to Competitor
10-20%
Onboarding / Hard to Use
5-15%
Source: Hypothetical: KnowMBA composite from CS leadership reports
Real-world cases
Companies that lived this.
Verified narratives with the numbers that prove (or break) the concept.
Drift
2022
Drift built their exit survey as a single required question with 6 reason buckets, displayed BEFORE the cancellation confirmation button. Completion rate was 71%. Aggregating 18 months of data revealed 'missing feature: account-based filtering' was their #2 churn driver — accounting for ~$1.2M in annual ARR loss. Engineering had deprioritized this feature for 14 months. The survey data forced it onto the roadmap; the feature shipped 5 months later. The win-back team immediately contacted everyone who had cancelled for that reason; 31% reactivated.
Survey Completion Rate
71%
Top Discovered Reason ARR Impact
$1.2M annually
Feature Win-Back Rate
31%
Time to Fix After Discovery
5 months
Exit surveys aren't passive data collection — they're a forcing function that surfaces controllable churn drivers and prioritizes them with hard ARR numbers. Without the data, the missing feature stayed deprioritized for over a year; with it, the business case became impossible to ignore.
Hypothetical: Mid-Market SaaS
2024
A $15M ARR mid-market SaaS sent a 14-question exit survey via email 2 days after cancellation. Completion rate was 4%. Of the data they DID get, 'pricing' was 60% of responses — driving the team to cut prices by 15% across the board. Margin dropped, but churn barely budged. They later switched to a 1-question pre-cancellation survey: completion went to 68%, and the new data revealed pricing was actually only 32% of churns. Onboarding (38%) was the real top driver. The price cut had been wrong — onboarding investment was the real lever.
Old Survey Completion
4%
New Survey Completion
68%
Old Survey 'Pricing' Reading
60%
New Survey 'Pricing' Reading
32%
Bad survey design produces biased data, and biased data drives wrong strategic decisions. The price cut based on 4% response data cost margin without fixing churn. The right survey would have routed investment to onboarding instead.
Related concepts
Keep connecting.
The concepts that orbit this one — each one sharpens the others.
Beyond the concept
Turn Cancellation Surveys into a live operating decision.
Use this concept as the framing layer, then move into a diagnostic if it maps directly to a current bottleneck.
Typical response time: 24h · No retainer required
Turn Cancellation Surveys into a live operating decision.
Use Cancellation Surveys as the framing layer, then move into diagnostics or advisory if this maps directly to a current business bottleneck.