What's a good save rate? Benchmarks for cancel flows
25–40% is the honest range. Here's how that breaks down by reason, what drags it down, and which number you should actually optimise.
Posted by
CallumRelated reading
Cancel flows without dark patterns
Click-to-cancel rules are coming for retention flows that trap people. Here's how to save customers and stay honest — they're the same thing.
Win-back emails that actually bring people back
One well-timed email beats a drip campaign. When to send it, what to offer, and when to leave cancelled customers alone.
Failed payments are your quietest churn
Up to 40% of SaaS churn is a card that expired, not a decision. The dunning playbook: retry windows, three emails, and an in-app banner.
The honest range is 25–40%
Vendors love quoting their best customer's best month. Across real deployments, a well-targeted cancel flow saves 25–40% of the customers who enter it. Below 20%, something specific is broken. Above 45%, check whether you're training customers to cancel for discounts — a suspiciously high save rate on the too expensive reason is usually repeat negotiators, not rescued churn.
Save rate varies wildly by reason
Blended averages hide everything useful. Typical per-reason rates:
- Too expensive: 35–55%. The most saveable reason — price objections respond to discounts and downgrades.
- Not using it / temporary need: 30–45%. Pauses work startlingly well here; the customer wants an off-ramp that isn't goodbye.
- Missing features / bugs: 15–30%. Only a human conversation plus a free extension moves this; discounts read as tone-deaf.
- Switching to a competitor: 5–15%. By the time they tell you, they've already migrated their data. Don't overspend here.
The action item: look at your per-reason save rates monthly. A reason at near-zero means the offer is mismatched — change the offer, not the copy.
Measure it like a funnel
Sessions → reason selected → offer shown → offer accepted. Losses at each step mean different things: many abandons before reason selection means the flow loads too slowly or looks like a trap; reasons selected but offers declined means targeting is off; offers accepted but churn a month later means the offer bought time without fixing anything — watch your second cancellation rate on saved customers.
The number that actually matters
Save rate is the diagnostic; net MRR saved per month is the business result — MRR saved minus the discount cost of saving it. A 50% discount that saves a customer for two months and then loses them anyway is a loss dressed as a win. Churntide's dashboard tracks both, per offer, so you can kill offers with good accept rates and bad economics. Want to see what the range means in money? Run your numbers.