DEV Community

Swipe Again for Swipe Again

Posted on

Plan the Quarter From the Failure Numbers, Not From Hope

Every subscription business plans its quarter from a handful of numbers: expected signups, expected churn, expected revenue. The number that rarely makes the planning sheet is the one that quietly edits all the others — the share of renewals that will fail at the card before anyone has decided anything. That share is not a mood. It is measurable, it is published in benchmark form, and planning without it is planning with a hole in the middle of the model.

Start with the shape of the problem. Across subscription businesses, a meaningful single-digit-to-low-double-digit percentage of renewal attempts fail on first presentation. Some of those failures are hard — a closed account, a stolen card — and no process recovers them at the point of sale. Most are soft: insufficient funds on the wrong day, an expired card nobody replaced, a generic decline the issuer will reverse on a better-timed second attempt. The planning mistake is treating the whole failure rate as weather. Only the hard slice is weather. The soft slice is operations, and operations can be budgeted.

That distinction changes the forecast. If ten percent of renewals fail and roughly two-thirds of the failures are soft, then the recoverable pool is not ten percent of revenue at risk — it is closer to six or seven percent of revenue waiting on process. A recovery effort that collects half of that pool adds three points of retained revenue that the signup forecast never had to produce. Put differently, recovering failed payments is the cheapest growth line in the plan, because the customers are already acquired, already onboarded, and already willing. They simply have not paid yet.

The second planning use of the numbers is cadence honesty. Benchmarks for recovery timelines show the same pattern everywhere: most recoverable payments land in the first two weeks, and the tail after day thirty is thin. A quarter plan that assumes a sixty-day dunning cycle will rescue the month is assuming a tail the data does not support. Plan the rescue window at two to three weeks, staff the work accordingly, and let anything beyond that be upside rather than forecast.

The third use is comparative. A business whose failure rate sits well above the published band does not have a recovery problem first — it has an input problem: stale cards collected at signup, no pre-expiry prompts, billing dates that ignore paydays. The statistics tell you which problem you own before you buy a solution for the other one. Read failed payment statistics worth planning a quarter around, find your own rate in the band, and only then decide whether the quarter's investment belongs in prevention, in recovery, or in both.

None of this requires false precision. The benchmarks are ranges, honest ranges, and a range is enough to plan with: assume the low end for revenue you promise, the high end for work you schedule. What the numbers forbid is the two common failure modes of planning — ignoring the leak entirely, or pricing it as unrecoverable loss. It is neither zero nor total. It is a pool with a known shape, and a quarter planned around its real shape starts closer to the truth.

So put the failure rate on the planning sheet, between signups and churn, where it belongs. Give it an owner, a recovery target stated as a share of the soft pool, and a review date. The businesses that do this stop being surprised by a number that was published all along.


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

Top comments (0)