DEV Community

Swipe Again for Swipe Again

Posted on

Three Numbers In, One Argument Out: Using a Leak Estimate Honestly

Every subscription business has a leak estimate floating around it — a number somebody produced once, in a meeting, that now anchors decisions it was never built to carry. The number is usually either theatrical (every failed payment priced as lost forever, as if no one ever updates a card) or timid (only the failures somebody happened to notice). A calculator's real job is not to produce the number. It is to force the three arguments that make a number honest, in public, where they can be disagreed with.

The first argument is about the pool. Not every failed payment is workable: some belong to customers who cancelled in spirit months ago, some are fraud blocks that should never be retried, some are duplicates of the same underlying case. The workable pool — failures attached to subscribers who want the product and have a fixable payment problem — is always smaller than the raw failure count, and starting from the raw count inflates everything downstream. Any estimate that leads with raw exposure is selling something.

The second argument is about the rate. Of the workable pool, what share can a competent recovery system actually save? The credible band, anchored in published benchmarks rather than vendor decks, runs around half to a little over half over a full cycle — higher on credential problems, lower on genuine funds problems. Rates of eighty or ninety percent exist in marketing copy and nowhere else. If your estimate needs an eighty percent rate to justify action, the honest conclusion is that the action is not justified at your current volume, and that is useful information too.

The third argument is about value: what is a saved subscription worth — one month, or the months it goes on to live? Conservative estimates use a few months of margin; aggressive ones use full lifetime value and thereby smuggle a retention claim into a payments calculation. Pick the conservative one for decisions and let the upside surprise you.

That discipline is built into a failed payment calculator for sizing the leak before fixing it, which asks for the small set of inputs, applies a benchmark-anchored rate to a workable pool rather than to raw failures, and returns a range instead of a prophecy. Run it with your real trailing-twelve-month failure count, not the month you remember most clearly. Then argue with it — change the rate, shrink the pool, and see whether the conclusion survives. An estimate that collapses when you touch an assumption was never an estimate; it was a mood.

Used this way, the calculator earns its place before any purchase: it sizes whether the leak justifies a flat monthly fee, a part-time manual effort, or nothing yet. Businesses under the threshold should hear that answer plainly. The leak does not care whether you measure it flatteringly. It just keeps dripping at the rate your own numbers describe.

Finally, keep the worksheet. An estimate written down, with its three assumptions visible, becomes a baseline the next quarter can be judged against: did the workable pool shrink, did the realised rate land inside the band, did saved subscriptions live as long as the conservative value assumed? Teams that keep the worksheet discover their leak estimate getting more accurate every quarter, because each assumption gets replaced, one by one, by a measured fact. Teams that re-run the calculator from memory each time get a fresh mood instead. The tool is the same. The habit is the difference.


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

Top comments (0)