The only place a leaving customer will tell you the truth. For most solo products it is a single button, which loses you both the revenue and the reason.
Your cancellation flow is the only place a leaving customer will tell you the truth, and for most solo products it is a single button that ends the conversation. That is a wasted opportunity in both directions — you lose the revenue and the reason.
A good cancellation flow saves some subscriptions and diagnoses the rest. The second part is worth more.
A single screen with five specific options and a free-text box. Specific matters — "other" as a dominant answer means your options are wrong.
Options that map to the five churn types work best, because the answer routes directly to a fix:
Those five map onto onboarding, value, product, price and positioning churn respectively — the churn framework covers what each one implies. Collecting this at the moment of cancellation beats a survey a week later, because the reason is still concrete.
A blanket discount offered to everyone is the standard approach and the weakest. It converts people who were leaving over something else at a permanently lower price, and it does nothing for the reasons that are not about money.
| They said | Offer | Not |
|---|---|---|
| Never got it set up | 15 minutes of your time, personally | A discount |
| Do not need it now | Pause for 3 months | A discount |
| Missing a feature | Tell them if it is coming; otherwise let them go | A vague roadmap promise |
| Too expensive | A lower tier, if one genuinely exists | An indefinite discount |
| Found something better | Let them go. Ask what won. | A counter-offer |
The pause option is the most underused and often the highest-converting. Seasonal businesses, consultants between projects and anyone with a quiet quarter frequently want to return, and pausing keeps the relationship without the awkwardness of a cancellation.
Hiding the cancel button, requiring an email to support, or adding four confirmation screens all reduce measured churn and increase the real cost. You get chargebacks, which threaten your payment processing, and public complaints, which cost more than the saved subscription.
It also poisons the win-back. Someone who had to fight their way out will not return and will tell other people — the win-back sequence depends on people leaving without resentment.
Regulatory direction is moving the same way, with several jurisdictions requiring cancellation to be as easy as signup. Designing friction in now creates work later.
Make the return path obvious on the confirmation screen. Keeping data for 90 days and saying so removes the biggest practical barrier to coming back, and it costs almost nothing to store.
Confirm clearly: when access ends, what happens to their data, and how to come back. Ambiguity here produces support emails from people who have already left, which is the worst use of your time.
Send one personal message shortly afterwards asking about their answer. The cancellation form gives you a category; the reply gives you the specifics, and the two together are what make the diagnosis actionable — getting churned customers to reply covers the wording that gets responses.
Review the answers monthly and count buckets rather than reading individually. One reason usually dominates, and that is the only one worth working on this quarter.
If "never got it set up" leads, your problem is upstream of churn entirely — see the onboarding framework. If "too expensive" leads among people who were actively using the product, look at your value metric before your price. And if the rate itself is what worries you, churn benchmarks at solo scale covers what is normal.
One screen with a single question and five specific options mapped to the five churn types, a targeted alternative matched to the answer rather than a blanket discount, and a clear confirmation covering when access ends and what happens to their data.
Only when the stated reason is price, and only if a genuine lower tier exists. A blanket discount converts people who were leaving over something else at a permanently lower price and does nothing for reasons that are not about money.
Yes. Hidden cancel buttons and support-only cancellation reduce measured churn while increasing chargebacks, which threaten your payment processing, and public complaints, which cost more than the saved subscription. Several jurisdictions now require cancellation to be as easy as signup.
A pause is the most underused and often the highest-converting. Seasonal businesses, consultants between projects and anyone with a quiet quarter frequently want to return, and pausing preserves the relationship without a cancellation.
Review monthly and count buckets rather than reading individually. One reason usually dominates and that is the only one worth working on that quarter. If 'never got it set up' leads, the problem is onboarding rather than churn.
Bring a month of cancellation reasons. Marcus names the dominant type and the one fix worth making this quarter.
Try GhostCoach free →14-day free trial · cancel anytime · 30-day money-back on Lifetime