Involuntary Churn and Dunning
Date: 2026-08-16
Subscriptions ending because a card failed, not because anyone decided to leave. It’s an engineering problem with a revenue number attached, it’s usually a substantial share of total churn, and it’s the cheapest churn there is to fix.
What it is
Involuntary churn is a subscription lost to a payment failure. Dunning is the process of retrying and recovering it.
VOLUNTARY INVOLUNTARY
customer decided to leave card expired
→ a product or value problem → a systems problem
→ expensive to fix → cheap to fix
→ they may not come back → they never wanted to leave
These are not the same event and should never share a metric. Blending them means running win-back campaigns at people whose card simply expired, and reading a systems failure as a product failure.
Why the cards fail
Roughly in order of frequency:
- Expiry. Every card expires, so a proportion of your base hits this annually by arithmetic alone
- Insufficient funds. Timing-related, and highly recoverable on a retry
- Reissue after fraud. New number, subscription unaware
- Bank declines on recurring transactions, particularly cross-border
- Authentication failures where strong customer authentication is required and the flow can’t complete
Most of these are temporary or fixable, which is what makes recovery rates high compared with voluntary churn.
Dunning that works
The mechanics matter more than the messaging:
Retry intelligently, not mechanically. Retrying the same declined card five times in five days recovers little and can trigger bank blocks. Space retries, and vary the timing — a retry shortly after payday recovers meaningfully more than one on the same day each time.
Update card details before failure. Card account updater services, offered by most processors, refresh expired or reissued cards automatically. This is the single highest-return intervention because it prevents the failure rather than recovering from it — [CHECK: availability and coverage varies by processor and card scheme].
Pre-emptive contact. Message customers whose card expires next month, before it fails.
Make updating trivial. A one-click link to update payment details, no login required, mobile-friendly. Most dunning emails fail here rather than at persuasion.
Cap the sequence. Three or four attempts over a couple of weeks, then stop. Beyond that you’re damaging deliverability for nothing.
The economics
Recovery is worth a subscriber’s entire remaining lifetime value, not one payment:
monthly contribution £15.00
monthly churn rate 5%
subscriber LTV £300
recovering one failed payment recovers £300 of expected
value, for the cost of an email
Which makes dunning improvements one of the highest-return pieces of work available in a subscription business — and it usually sits with nobody, being too technical for marketing and too commercial for engineering.
Measuring it
- Split churn into voluntary and involuntary, always. This is the first step and most reporting doesn’t do it — Subscription Metrics
- Recovery rate — proportion of failed payments eventually collected
- Time to recovery, since a subscription recovered on day 2 keeps its cycle and one recovered on day 20 has skipped a period
- Failure rate by card type, issuer and country, which frequently reveals one issuer or one flow causing a disproportionate share
- Never count a recovered payment as a winback. It was never lost
Where it applies beyond subscriptions
The same mechanism affects any stored-card flow — saved cards for one-click reorder, and replenishment subscriptions on a retail site. If you hold cards for repeat purchase, you have involuntary churn whether or not you call it that — Replenishment Timing.