Cancellation and save flows

When a customer clicks cancel, most subscription businesses just cancel them. That’s a missed conversion. The cancellation flow is one of the highest-intent moments in the whole relationship - the customer is actively making a decision about your product - and it’s a conversion surface in exactly the way a checkout is, except the conversion you’re trying to win is “stay” rather than “buy”. Unlike dunning, this is voluntary churn: the customer has decided to leave, and the question is whether the right offer or option changes their mind.

What a save flow does

The flow sits between the cancel button and the actual cancellation, and it works by understanding why the person is leaving and responding to that reason rather than throwing a blanket discount at everyone:

• Ask the reason. A short “why are you leaving?” isn’t just research, it routes the save. Someone leaving over price needs a different response to someone who’s finished their project or found it too complex.
• Offer the matched save. Too expensive - a discount or a cheaper tier. Going away for a while - a pause instead of a cancel. Not using it - a downgrade or a how-to. The matched offer converts far better than a generic “here’s 20% off”.
• Pause over cancel. A pause keeps the relationship alive when the alternative is losing it entirely, and a paused customer is far easier to bring back than a churned one.

The save offer is just the offer applied at the exit, and a retained customer’s lifetime value usually dwarfs the cost of the discount that saved them - though not always, which is the catch.

The ethics line, which is also a CRO line

There’s a strong temptation to make cancelling hard - hiding the button, forcing a phone call, burying it in menus. This is a dark pattern, and beyond the regulatory risk now attached to it, it backfires on its own terms. Friction-based “saves” trigger reactance: the customer who had to fight to cancel doesn’t come back, leaves a bad review, and tells people. A genuine save flow makes cancelling easy and makes staying attractive. A dark pattern makes leaving hard and makes the brand resented. They produce opposite long-term results even when the short-term save rate looks similar.

Where it goes wrong

• One discount for everyone. A blanket save offer trains customers to threaten cancellation to get a discount, and gives margin away to people who’d have stayed anyway. The save has to be matched to the reason and ideally targeted, not universal.
• Saving customers you shouldn’t. Discounting hard to retain a chronically unprofitable or unhappy customer can lower margin and inflate retention metrics with people who churn next cycle anyway. Not every cancellation is worth fighting.
• Optimising save rate in isolation. A flow can lift the immediate save rate with aggressive friction or steep discounts and lose on second-order churn and brand. The metric is retained value over time, not saves at the door.
• Measuring the discount as free. Every save discount is real margin. A save flow that looks like a win on retention can be a loss on contribution if the offers are too generous.