Interventions matched to the actual cancellation reason, including the accounts you should let go.
Design our cancellation flow. Right now everyone gets 20% off.
Price accounts for 22% of cancellations, so the blanket 20% discount is being spent on the other 78%. Two reasons — wrong fit and a capability you do not plan to build — get a clean exit instead of an offer. The pre-churn trigger fires about five weeks before the median cancellation, which is where the recoverable revenue is.
A churn email template that offers everyone a discount saves the wrong customers. This one maps cancellation reasons to interventions — price, unrealized value, missing capability, or wrong fit — and writes the sequence for each, including the reasons where the honest answer is to let the account cancel cleanly.
A set of interventions triggered by cancellation intent, mapped to reason. Each path has its own message, its own offer or absence of one, and a defined outcome — save, downgrade, pause, or clean exit.
Ask one question about why, then respond to the answer. A cancellation flow that fires the same save offer regardless of reason converts a small number of price-sensitive customers and annoys everyone whose problem was something else.
When price is the actual constraint and the account otherwise gets value. Discounting a customer who never activated buys a month of delay, and pausing beats discounting when the reason is seasonal or a budget freeze.
Well before cancellation — at the usage decline that precedes it by weeks. The template includes the pre-churn trigger, because the email you send at the cancel button is already the low-leverage one.
The agent reads usage trends and cancellation reasons from your product analytics and CRM to build the reason map, and can weight interventions by account value. The flow exports to Customer.io, HubSpot, or Klaviyo, with the pre-churn trigger defined as a usage condition rather than a date.
From your cancel survey, CRM notes, or support tickets.
Each reason gets an intervention, or an explicit decision not to intervene.
A usage condition that fires weeks earlier, where the leverage is.
Deploy branches in your ESP and log reasons back to the CRM.
Interventions are matched to reason, not fired blindly
Names the accounts that should be allowed to cancel
Includes the pre-churn trigger where leverage actually is
Offers pause and downgrade instead of defaulting to discount
Trigger it on declining usage rather than on cancellation, and lead with the specific thing they stopped doing. By the time someone reaches the cancel button, the decision is usually already made.
Only when price is the real reason. Otherwise a discount delays churn by a billing cycle and lowers the revenue you eventually lose, which is why reason-matching matters more than the size of the offer.
No. Friction produces chargebacks, support load, and public complaints, and in several jurisdictions it is now a regulatory issue. Make it easy and put the effort into the weeks beforehand.
One required question with a short list of reasons, plus an optional free-text box. Long exit surveys go unanswered precisely when you most need the data.