Every subscription business eventually faces a payment that will not go through, and the language it reaches for in that moment reveals how it thinks about its customers. Two vocabularies compete. Collections language talks about debtors, delinquency, and escalation. Dunning language talks about subscribers, failed renewals, and recovery. They are not synonyms with different tones. They describe different relationships, different legal postures, and different futures for the account — and using the wrong one on a subscriber who simply had a card replaced is how a billing hiccup becomes a cancellation.
Start with the simplest distinction. Collections assumes the money is owed by someone who is choosing not to pay, and its tools — demand letters, agency referrals, credit reporting where it applies — are built for adversarial recovery. Dunning assumes the money failed to move for a mechanical reason, and its tools — timed retries, card update links, plain-language notices — are built for cooperative repair. Most failed subscription payments belong firmly in the second world. The subscriber never decided anything. Their bank replaced a card, their balance dipped on the wrong day, or their issuer shrugged with a generic decline.
The boundary cases are where teams get into trouble. A subscriber sixty days past a failed renewal, ignoring every notice, starts to look like a debtor. A subscriber who disputes a charge they plainly authorised starts to look adversarial. The temptation is to escalate the language as the balance ages. Resist sorting by age alone. Sort by evidence: did this account ever signal intent to leave, dispute a charge, or refuse contact? If none of those happened, the account is still a dunning case wearing an old timestamp, and collections framing will manufacture the hostility it presumes.
This is the line a clear comparison of dunning and collections and where each one belongs draws in operational terms: dunning ends when the evidence says the relationship ended, not when a calendar page turns. Until then, every message should assume the subscriber wants the service and the payment is the only problem — because in the overwhelming majority of failed renewals, that assumption is simply true.
There is also a practical cost to crossing the line early. Collections language in a dunning email — words like delinquent, outstanding balance, final notice — measurably changes how the message is received. Subscribers who would have updated a card in thirty seconds instead forward the email to a partner, a manager, or a lawyer, or they cancel defensively to stop the tone. The recovery rate falls, and the team concludes the account was never recoverable, when in fact the vocabulary did the damage.
A useful test for any message in the queue is to read it as the subscriber will: at a kitchen table, on a phone, with no context except the subject line. If the message would make sense — and make sense calmly — to someone whose only mistake was receiving a replacement card in the post, it belongs in the dunning flow. If it only makes sense addressed to someone who has decided not to pay, it belongs in a flow most subscription businesses should reach rarely, late, and on purpose.
None of this argues for softness about real non-payment. Services should pause when grace periods end, access should follow payment, and genuinely abandoned accounts should close. The argument is narrower: keep the two toolkits separate, escalate on evidence rather than age, and let the subscriber's own behaviour — not your frustration with the queue — decide which relationship you are in.
Swipe Again is built entirely on the dunning side of that line: retries timed around decline reasons, notices written as service messages, and a card update path that treats the subscriber as a customer with a broken payment method. Draw your own line in the same place. The subscriber is not a debtor until they make themselves one — and most never do.
Top comments (0)