DEV Community

Swipe Again for Swipe Again

Posted on

Recurly Retries Are the Floor, Not the Ceiling: Finding the Gaps Around a Good Engine

Recurly is one of the more capable billing platforms a subscription business can run on, and its native recovery tooling is genuinely respectable: configurable retry logic, account updater coverage, dunning campaigns with multiple touches. Teams evaluating outside help often start from a fair question — if the platform already does this much, what exactly is left? The honest answer is that what is left is not a replacement engine. It is the ring of work around the engine, where platform defaults quietly stop.

Start with timing intelligence. A platform retries on a schedule shaped by its own rules and by settings a merchant picked, often once, during onboarding. What it generally does not do is reshape that schedule around each subscriber's pay cycle, or hold an attempt because a card update landed an hour ago and the obvious next move is to charge the new card, not the queue position. Those are judgement calls layered on top of mechanism, and they are where a noticeable slice of recoverable revenue sits.

Next, look at the message layer. Platform dunning emails are functional, and functional is the problem: they read like system notices, they often come from a sender the subscriber does not associate with the product they love, and the update path can involve more steps than a phone screen forgives. Recovery communication has one job — get a willing person to a saved card in under a minute — and small frictions there cost more than any retry interval ever will.

Then there is the ending nobody designs. When platform dunning exhausts, the subscription cancels, and from the platform's point of view the story is over. From the merchant's point of view it is not: a subscriber lost to an expired card last month is the cheapest winback candidate there will ever be, and a cancellation flow that offers a genuine save — a pause, a smaller plan, a fix-the-card path — catches people the retry engine never could. Platforms provide the cancellation. The deflection has to be built or bought.

A fair map of which parts Recurly covers natively, which parts it leaves to the merchant, and how the settings interact is set out in this complete walkthrough of Recurly dunning and the work that sits outside it — worth reading before changing a single setting, because a few of the native options change what any outside tool should be asked to do.

One more gap deserves naming because it is invisible until it costs money: what happens when a card is updated somewhere other than the dunning flow. Subscribers fix cards in account settings, through support, sometimes inside the platform's own hosted pages. If the recovery sequence does not learn about that update immediately, it keeps emailing about a problem that no longer exists and may even retry the dead card out of habit. Reconciling card state continuously — not nightly, not at the next campaign step — is unglamorous plumbing, and it is one of the clearest dividers between recovery that feels intelligent to the customer and recovery that feels like a robot reading an old list.

The practical takeaway is not that platforms fall short; it is that their job and the recovery job overlap only partly. Recurly's job is billing correctness at scale. The recovery job is squeezing the human and calendar accidents out of a specific subscriber file, week after week, and being able to prove the count. Teams that understand the boundary get the best of both: the platform as the system of record, and dedicated recovery wrapped around its edges.

That wrapper is what Swipe Again sells for Recurly and the other major billing platforms: payday-aware retry timing, card-update awareness, plain-spoken customer messages, and cancel deflection in the same account — connected in minutes, priced flat, and measured on total recovered revenue rather than activity.

Top comments (0)