Systematically categorize why subscribers are canceling, spot the patterns driving churn, and leave with a prioritized list of actions to reduce it.
The template
Response Snapshot
Record the raw response, the cancellation reason selected, and the respondent's plan tier so you have context before coding.
Example: Respondent on Pro plan selected "Too expensive" and wrote: "We loved the product but after our funding slowed down we just couldn't justify $299/month for a tool only two people used."
Primary Churn Theme
Assign one theme label that best captures the core reason for leaving, chosen from a consistent set of categories you define at the start.
Example: Theme: Price-to-value mismatch — the user perceives the outcome as worthwhile but the cost is not justified by their current usage volume or company stage.
Supporting Evidence & Secondary Signals
Note any secondary issues mentioned in the response that compound the primary reason or hint at a fixable product or onboarding gap.
Example: Secondary signal: low seat utilization (only 2 of a possible 10 seats used), suggesting the team never fully adopted the product — possible onboarding or activation failure underneath the price objection.
Recommended Action
Write one specific, ownable next step tied directly to this response — something a real team member could act on this sprint.
Example: Action: Flag to Sales for a downgrade-to-Starter offer within 7 days; flag to Product that the Pro plan seat minimum may be misaligned with early-stage startup budgets.
How to use it
Export and normalize your responses Pull your exit survey data into a spreadsheet and add four columns matching the template sections — do this before reading a single response so your coding stays consistent.
Define your theme taxonomy first Before coding any responses, agree on 5–8 churn theme labels (e.g. Price, Missing Feature, Poor Onboarding, Competitor Switch) so every analyst uses the same language.
Code each response using the four sections Work through responses one at a time, filling in all four sections for each — resist the urge to skip the "Supporting Evidence" column, as that's where hidden patterns live.
Tally themes and rank by frequency and revenue impact Once all responses are coded, count theme occurrences, weight them by the MRR of churned accounts, and present the top three themes with their recommended actions to stakeholders.
What it looks like filled in
Price-to-Value Mismatch
"It's not that the product doesn't work — it does. We just couldn't justify the cost when half our team stopped logging in after month two."
→ Introduce a usage-based or seat-flexible pricing tier and trigger a check-in email when active seat usage drops below 50% in any 30-day window.
Competitor Feature Gap Closed
"We switched to [Competitor] because they launched native Salesforce sync last quarter — that was a blocker for us for over a year and we finally gave up waiting."
→ Escalate Salesforce native integration to the roadmap committee with this churn cohort's MRR as the business case; add an interim Zapier workaround to the help docs immediately.
Failed Onboarding and Low Adoption
"Honestly we never really got it set up properly. We had one kickoff call and then were kind of left on our own — by month three we'd forgotten it existed."
→ Redesign the 30-day onboarding sequence to include a human check-in at day 14 for accounts that have not completed their first core workflow milestone.
Why teams skip the template
Manual coding takes hours per batch Reading, categorizing, and tagging even 50 exit responses by hand typically takes a full analyst day — and that's before you've written a single recommendation.
Theme labels drift across reviewers When more than one person codes responses, "price" to one analyst is "value" to another, quietly corrupting your frequency counts and making trends impossible to trust.
Insights arrive too late to save the account By the time a manual analysis is complete, reviewed, and presented to stakeholders, the churned customers are long gone and the window to act on retention has closed.
Analyze your exit survey responses automatically — no template needed
