DEV Community

Swipe Again for Swipe Again

Posted on

The Expiry Date Printed on the Card Is a Renewal Forecast

Every card in a subscription file carries its own failure date, printed on its face months or years in advance. Almost no billing operation treats it that way. Expiry is handled as an event — the renewal declines, the code comes back, the dunning sequence starts — when it could have been handled as a forecast: a known date, attached to a known customer, arriving on a schedule the business can read today. The gap between those two postures is where a surprising share of involuntary churn quietly lives.

Start with the audit, because it takes an afternoon and most teams have never run it. Pull the active card file and bucket it by expiry month. The shape is always instructive: a smooth business expects a smooth distribution, and instead finds cliffs — the month a popular signup cohort's cards all expire together, the quarter a bank reissued a whole portfolio onto new dates. Those cliffs are future failure spikes with dates on them. A team that can see a cliff coming in ninety days can staff for it, message ahead of it, and measure whether the preparation worked. A team that cannot see it experiences the same spike as a mysterious bad month.

The message ahead of expiry deserves more care than it usually gets, because it is the easiest billing email a customer will ever receive. It asks for nothing urgent, threatens nothing, and arrives while the relationship is healthy: your card ends next month, here is the thirty-second update. Response rates on that email dwarf dunning response rates, for an obvious reason — the customer is being helped before a problem exists rather than chased after one does. That asymmetry is the whole argument in expired cards failing on a schedule the file already predicts: the cheapest recovery in the entire system is the one that happens before the failure.

Two complications keep the forecast honest. First, account updater services silently replace many expiring cards, which is wonderful for revenue and terrible for measurement: the file you audited in January is not the file that bills in March, and expiry cliffs shrink without anyone knowing why. Re-audit quarterly and treat a shrinking cliff as a question, not a victory lap. Second, the cards the updater cannot save — prepaid, some debit portfolios, customers who changed banks entirely — concentrate in exactly the segment least likely to answer an email quickly. The forecast tells you how many of those are coming too.

None of this requires new tooling to start. It requires a calendar habit: a monthly look at next month's expiries, a standing pre-expiry message, and a line in the billing review where the cliff chart is shown next to the failure chart so the two can be compared. Businesses that build the habit stop describing expired cards as bad luck. They were never luck. They were appointments the business forgot it had made.

One more discipline separates the forecasters from the merely organized: the pre-expiry message itself gets measured like a campaign, not filed as a courtesy. Track its update rate by cohort month, and watch what happens to the following month's expiry-decline count when the message improves. That closed loop — message sent, cards updated, cliff shrunk, decline count down — is the proof other teams ask for when they wonder whether the afternoon audit was worth it. In files where the loop has been run for a year, expiry quietly stops appearing in post-mortems altogether. It does not become exciting. It becomes handled, which is the highest compliment a billing process can earn.


Published by Swipe Again — failed payment recovery at a flat $179/mo.

Top comments (0)