Read a Recharge store's recovery bill the way you would read a story about its month, because that is what it is. Under percentage-of-recovery pricing, the invoice grows in exactly the months the store suffered most: the January when holiday-loaded cards failed in waves, the month a bestseller's renewal cohort came due all at once. The tool did more work in those months, certainly. It also charged its highest fee at the moment the merchant's margin was thinnest, for recoveries that include — unless the counting is unusually honest — saves the platform's own retries and the network's silent card updates would have landed anyway. The invoice is not wrong, exactly. It is just written in a language that flatters the teller.
The alternative language is a flat fee, and its virtue is boredom. The same number in the quiet month and the loud one, judged against a question the merchant can answer from their own records: what is the post-native, genuinely workable failure pool worth, per month, if somebody competent works it? Count last quarter's Recharge failures. Remove the ones the platform's schedule recovered by itself. Remove the endings no sequence could close. What remains is the pool a recovery layer can honestly address, and the flat fee either clears that bar with room to spare or it does not. The arithmetic takes an afternoon and ends most vendor debates before they start.
Counting is where the two pricing models really diverge, because a percentage fee multiplies whatever counting rule it is paired with. If a save credited to the tool includes charges the native schedule recovered, the merchant is paying a commission on their own platform's work. If silent updater saves enter the same column, the merchant is paying for a network service twice. None of this requires bad faith; it requires only a blended report and a busy merchant. The defence is a sentence, agreed before the first invoice: a recovery counts when the tool's own action — its retry, its message — is the mechanism that landed it, and the report shows the other columns beside it, credited to the platform and the network where they belong.
That sentence is the heart of recharge dunning without a revenue-share invoice at the end of the month: the handover set deliberately, the sequence worked by decline family, the saves counted by mechanism, and the price fixed in advance so a bad month for the store is not automatically a record month for its recovery vendor.
There is a quieter benefit to the flat structure that only appears over time. A vendor paid a percentage of recoveries has no particular reason to love prevention — the expiry sweep, the early calm note — because prevented failures never enter the pool the percentage is drawn from. A vendor paid flat is indifferent to where the save comes from, which frees the work to start earlier, in the cheap seats, before a failure exists at all. Merchants rarely choose a pricing model for its effect on the vendor's incentives. They should. They will live inside those incentives every month the tool runs.
A subscription store is a machine for turning small recurring decisions into a business. The recovery invoice is one of the decisions, and like the others, it compounds. Read it as a story — whose work it credits, which months it rewards, what behaviour it quietly pays for — and the right structure usually names itself.
That is the structure Swipe Again sells on Recharge: one flat monthly price, saves credited to the mechanism that earned them, and prevention counted as work rather than treated as a threat to the invoice.
Top comments (0)