DEV Community

Swipe Again
Swipe Again

Posted on

Counting Retries Is Easy. Knowing When to Stop Is the Product

Every retry tool on the market can tell you how many times it tried. Almost none of them lead with the more interesting number: how many times it decided not to. That asymmetry is worth sitting with if you are evaluating this category, because the economics of recovery are made as much by the attempts a system declines to make as by the ones it places well.

A retry is not free, even when the per-attempt price rounds to zero. Each attempt is a signal to an issuer about this card, this merchant, and this amount. Spend those signals carelessly — four charges in forty-eight hours against a card whose owner gets paid on Friday — and you have not merely failed to recover the money; you have made Friday's well-timed attempt look like the fifth knock from a merchant the bank is learning to distrust. The tools worth paying for are the ones that treat restraint as a feature, which is why any serious reading of smart retry tools compared on timing evidence, not retry counts should start with their stopping rules, not their schedules.

Three questions separate the field quickly. First: what does the system actually sense? A tool that knows a card was replaced last Tuesday — through the processor, through a customer update on any surface, through the network's own updater — is playing a different game from one that only knows its own attempt history. Retrying a replaced card's old number is the category's most common unforced error, and it is invisible in a dashboard that only counts attempts and successes.

Second: whose calendar does it use? Payday timing is the oldest fact in consumer billing and the one most retry ladders still ignore, marching attempts out at fixed intervals from the failure date as if incomes arrived on the merchant's schedule. A system that holds a soft decline for the customer's likely payday, or for the morning rather than the small hours, will beat a louder system on the same card file almost every month. Ask a vendor to show you the distribution of their successful retries across the month. If it is flat, their timing is flat, whatever the brochure says about intelligence.

Third: what does it count as a win? A replacement subscription started by a customer who gave up on the old one is not a recovery of the failed dollars, and a pause that later resumes is not a loss — teams get gamed, sometimes by their own dashboards, when these get blended. The evaluation question is boring and decisive: show me a recovered payment, traced to the original failed charge, with the failed amount collected. Tools that can answer plainly are usually the ones whose other numbers can also be trusted.

None of this makes the loud tools useless; a small merchant with no retry logic at all gains from almost any structure. But the gap between a fixed ladder and a sensing, calendar-aware system compounds quietly, month after month, in the population of subscribers who wanted to stay and whose banks were ready to say yes on the right day. That population is the business. Evaluate for it: stopping rules first, sensing second, calendar third, counting always — and let the attempt totals look after themselves.

That is the standard Swipe Again publishes for itself: retries woven around the platform's own schedule rather than stacked on top of it, card changes sensed on every route they can arrive by, and a guarantee judged on total recovery over the first 30 days — recover $179 or the next month is free at $179 a month flat. Bring the three questions above to every demo in this category, including ours. The vendor who answers them without reaching for an attempt count is the one who understands the job.

Top comments (0)