Every shortlist starts the same way: a spreadsheet, a column per vendor, rows for the features somebody remembered to ask about. The dunning rows are always the vaguest — "smart retries: yes," "email sequences: yes," "reporting: yes" — because the demo showed each one working on a tidy example. Six months later the spreadsheet is forgotten and the tool is being judged by entirely different things: whether anyone on the team actually opens it, whether its numbers can be trusted in a finance review, and whether its attempts behave well toward real customers on real paydays. Score those things at shortlist time and the spreadsheet earns its keep.
Start with the queue test. Ask each vendor to show you not a dashboard but a single failed payment's life: the moment it failed, how it was classified, what happened next and why, what the customer was told, and how the case ended. Tools that manage dunning as a system answer this fluently, because a case is their native unit. Tools that are really a messaging feature with retries attached go vague exactly where the classification and the ending should be. That vagueness is the most useful signal on your shortlist, and it costs you one question to collect.
Next, score the counting. Recovered revenue is the number the whole purchase will be justified by, so ask how it is defined: do platform-native retries count toward the tool's total, does a customer who updates a card after a nudge count the same as a silent retry success, and can the definitions be read somewhere public before you sign? In our own reading of the category, the best dunning management software is the one whose numbers survive a Monday audit — definitions published, platform recoveries separated from the tool's own work, and a total you can reconcile against your processor in minutes rather than take on faith.
Then score the calendar, because timing is where recoveries are quietly won. Does the tool retry on fixed intervals regardless of the failure in front of it, or does it read the decline — waiting for payday on a funds failure, asking immediately on an expired card, standing down on a hard decline? Does it respect quiet hours where the customer lives, not where your servers sit? These questions sound operational because they are; dunning management is an operation, and software that cannot describe its calendar behaviour cannot manage one.
Score the exit as well. What does the tool charge once it is working — a fixed fee you can budget, or a percentage that taxes every recovered dollar and grows with your success? What happens if you leave: do your cases, histories, and customer-facing links survive the transition, or does cancellation strand subscribers mid-sequence? Vendors confident in their retention answer plainly. The spreadsheet should have a row for leaving, filled in before arriving feels urgent.
Finally, weight the scores by your actual failure mix. A business whose losses are mostly expired cards needs card-update craft above all; a business bleeding on insufficient funds needs timing; a business with angry support tickets needs tone. There is no universal winner hiding in the category — there is a best fit for the queue you actually have, and the shortlist that scores the work instead of the demo is the one that finds it.
Swipe Again is built to be scored this way: cases classified by decline from the first minute, payday-aimed timing inside recipient-local hours, published counting rules, and one flat monthly fee with no percentage taken from recovered revenue.
Top comments (0)