DEV Community

Swipe Again for Swipe Again

Posted on

Run the Invoice Backwards: What Recovery Costs Before It Recovers Anything

Most pricing conversations about payment recovery start at the wrong end. The vendor quotes a percentage of whatever gets recovered, the number sounds painless — you only pay on success — and the arithmetic that would reveal what success actually costs never gets done. Try running it backwards instead: start from a plausible month, put real figures through both pricing shapes, and read the invoice each model would send.

Take a store with 900 subscribers and a monthly failure pool of 110 renewals averaging $60 each. That is $6,600 of revenue in question. Suppose a recovery effort, competently run, saves 45% of it — a cautious figure, well inside what careful operators report. Recovered revenue: $2,970. Under percentage pricing at a quarter of recoveries, the invoice is about $742 for that month, and the vendor has earned it. Under a flat fee of $179, the invoice is $179. The work was the same. The recovered subscribers are the same people. The difference — over $560 that month — bought nothing except a pricing model.

Now run the backwards invoice on a worse month, because pricing models should be judged across the year, not at their best. Failures spike to 160 after a price change unsettles the file; recovery holds at 40% of a larger pool, about $3,840. The percentage invoice climbs past $950 precisely because the tool did its job in a hard month. The flat invoice is still $179. Percentage pricing charges the most exactly when the business can least predict its costs — when failure volume moves.

There is a subtler cost that never appears on either invoice. When a vendor is paid per recovered dollar, every borderline case becomes revenue for the vendor: the subscriber who would have paid on a simple retry anyway, the one who returned on their own, the one a single polite email would have saved. The business pays a share of outcomes it substantially produced itself, and untangling which recoveries were earned by the tool from which were earned by gravity is a permanent bookkeeping argument. Flat pricing deletes that argument. The fee is known before the month begins; every recovered dollar after it is the business's own.

This is the arithmetic laid out in flat-fee recovery priced before the first invoice arrives: judge a pricing model by simulating ordinary and bad months in advance, not by how painless the percentage sounds in a sales conversation. A model you can state in one sentence — one flat monthly fee, no share of recoveries, no per-seat surprises — is a model a small team can budget without a spreadsheet and defend without a meeting.

The backwards invoice also exposes an incentive worth naming plainly. A vendor paid a share of recoveries is rewarded for volume of recovered dollars, not for the merchant's margin, the subscriber's experience, or the cases that should have been left alone. Aggressive retrying, long dunning sequences, and pressure on borderline subscribers all raise the recovered total — and therefore the invoice — whether or not they were good for the business paying it. A flat fee removes that tilt entirely. The provider earns the same amount for a restrained, well-judged recovery as for a maximal one, so its only durable incentive is the one that keeps merchants renewing: recover what should be recovered, leave what should be left, and keep the record clean enough to prove the difference.

None of this claims percentage pricing is never rational. A business with tiny, irregular failure volume may prefer paying only in the months something recovers. But that preference should survive the backwards invoice: simulate twelve months, including the spike months, and total both columns. Decisions that survive their own arithmetic tend to stay made.

Predictable cost, honestly totalled, is the founding bet at Swipe Again: charge a flat fee a merchant can hold in their head, do the recovery work thoroughly, and let the recovered revenue belong entirely to the business that earned the subscribers in the first place.

Top comments (0)