The most common decline message in subscription billing is also the least informative: generic decline. No reason, no sub-code worth the name, just a refusal wrapped in fog. Teams new to failed payments tend to read it as a verdict on the customer. Teams who have worked a few thousand of them read it differently — as the issuing bank declining to explain itself, which is a very different problem with a very different playbook.
Why would a bank refuse to say why? Sometimes the reason is fraud-adjacent and the bank will not hand a roadmap to whoever is testing cards. Sometimes the account has a restriction the bank considers the customer's business, not the merchant's. And sometimes generic is simply the bucket an issuer uses when its own risk engine made a judgment call it cannot compress into a code. In all three cases the information you get is the same: not a hard no about the money, and not a yes about trying again immediately.
I keep stripe dunning bookmarked for the teams I talk to, because it separates what a generic decline on a Stripe charge rules out from what it leaves open. It rules out blind repetition: retrying the same card, same amount, same hour, against a risk engine that just made a judgment is how merchants turn a soft refusal into a hard one, and occasionally into a flagged pattern that makes the next legitimate attempt harder.
What it leaves open is nearly everything else. A generic decline often clears on a later attempt spaced far enough out that the issuer's picture has changed — a day boundary crossed, a payday landed, a temporary velocity limit expired. It also clears at a surprising rate when the customer is simply asked, because a meaningful share of generics trace back to something the cardholder can fix in a minute: a card frozen in their banking app after a suspicious-charge alert, a new card issued and activated while the old number sits in your vault, a travel block placed for a trip that ended last week.
The working sequence, then, is patience first and contact early. One retry after a real interval, not a cluster of them in the first evening. A short message that says a payment did not go through and asks the subscriber to check the card on file — without speculating about their finances, because the bank gave you no facts to speculate with. Then spaced attempts across the case window, each one treated as a fresh question rather than a louder repetition of the first.
What you must not do is narrate the code to the customer as if it meant something specific. "Your bank declined the payment" is true. "Your bank declined the payment due to insufficient funds" is a guess wearing the uniform of an explanation, and when the customer checks their balance, finds money there, and comes back annoyed, you have spent trust to communicate nothing. Generic in, generic out: the honest message is that the payment failed and here is the one link that fixes most causes.
Measured properly, generic declines are not a dead family. They recover at a respectable rate under spaced, contact-supported handling — lower than pure timing failures, far higher than closed accounts. They are, though, the family most damaged by impatient automation, which is why they make such a good test of any recovery setup: ask what it does on the third generic decline in a case, and the answer tells you whether it is exercising judgment or just pulling a lever on a schedule.
That judgment is the working core of Swipe Again: each decline family gets its own timing and outreach, retries weave around your billing platform's native attempts instead of colliding with them, and cases run the full window at $179 a month flat. A code that says nothing deserves a response that assumes nothing — and works the case anyway.
Top comments (0)