DEV Community

Swipe Again for Swipe Again

Posted on

Stripe Tells You More About a Failed Payment Than the Dashboard Shows

A failed Stripe payment leaves a paper trail most teams only half read. The dashboard shows a red row and a decline code; the underlying event carries the issuer's actual reasoning, the attempt history, and the signals that decide whether a second attempt is sensible or theatre. Working Stripe failures well is mostly a reading discipline.

Start with the code, but do not stop there. Two charges can share a generic decline and have opposite prospects: one was a momentary issuer risk decision on an otherwise healthy card, the other the fourth refusal in a week on a card the customer has already replaced. The attempt count and the spacing between attempts are part of the diagnosis. Stripe's own automatic retries are useful background pressure, but they run on a platform-wide logic that knows nothing about this subscriber's payday, time zone, or support history.

Next, read the card's life events. A payment method that was updated — by the customer, or silently by the network's account updater — changes what a retry means. Retrying a superseded card teaches you nothing except that dead cards decline. The operational habit worth building is simple: when card details change, that change should wake the case up and put a fresh attempt in motion while the customer's attention is still on it, rather than leaving the case asleep until the next scheduled sweep.

Then read the customer's side of the timeline. Did the failure notice reach them, and could they act on it in one sitting on a phone? A remarkable share of Stripe recovery is lost in that gap: the email lands, the update page asks for something the customer does not have to hand, the session expires, and a willing payer becomes an involuntary cancellation. Instrument that path like a checkout, because for this customer, in this moment, it is one.

There is also the question of what not to retry. A lost or stolen card, a closed account, a customer who has already cancelled — these are endings wearing the costume of failures, and every attempt spent on them is taken from a case that could have recovered. Stripe surfaces the distinction in its codes and network advice; the discipline is believing it the first time, closing the case, and letting the winback conversation, if there ever is one, start fresh months later rather than as the ninth retry.

And measure by cohort, not by anecdote. The loudest failed payment in a support inbox is rarely the typical one. Group a month of Stripe failures by code family, by card age, and by whether the customer ever saw a notice, and the queue stops being weather and becomes a short list of fixable causes — usually an update path that drops mobile users, a retry pattern that fires before payday, or a descriptor a bank's app truncates into nonsense. Fix one cause and the whole family shrinks the following month, which no amount of per-case heroics can match.

Finally, close cases honestly. A subscription cancelled by the customer mid-recovery is an ending, not a pending success. A charge that later succeeds through a completely separate purchase is a new sale, not proof the sequence worked. Clean endings are what make the remaining numbers — attempts, recoveries, time to recovery — worth steering by.

For the complete Stripe-specific walkthrough — decline-code handling, retry placement alongside Stripe's own schedule, and the card-update path — stripe dunning works through it in order.

That read-first, retry-second discipline is how Swipe Again approaches Stripe recovery: attempts woven into the gaps of the native schedule, at one flat price, with cancellation stopping everything at once.

Top comments (0)