Resources

Recurly Dunning: The Complete Guide

If you bill on Recurly, you already have dunning — the question is whether it's a system or a pair of settings. Recurly's native Revenue Recovery stack pairs its Intelligent Retries engine with configurable dunning campaign emails, and it's a genuinely solid baseline. What it doesn't run is the full recovery playbook: a decline-code-specific schedule as the organizing system, follow-up beyond email under your brand, cancel deflection in the same product, and one recovery view across every platform you bill on. This guide explains exactly where Recurly's native dunning stops, what a dedicated recovery layer adds, and how to decide whether the gap is worth paying for in your account. Its sibling guide covers the same ground for Chargebee dunning.

What "Recurly dunning" actually is, out of the box

Dunning is everything that happens after a subscription payment fails: retrying the charge, telling the customer, getting a working card on file, and deciding when to stop. Recurly organizes that work into two tools that activate on a failure and run independently of each other:

• Payment Retries — Recurly re-attempts the failed charge on a schedule. You can run Intelligent Retries, Recurly's machine-learning model, which times each attempt per failure; Recurly describes the model as trained on billions of transactions, capped at 20 attempts per invoice or 60 days, whichever comes first. Or you can set a custom retry schedule of your own. Note the fine print in Recurly's own documentation: Intelligent Retries requires an eligible plan tier — the custom schedule is the fallback where it isn't available.
• Dunning Campaigns — emails sent at day offsets you configure after a failure. You can run multiple campaigns for different cohorts or plans, with templates for different decline situations, and the campaign length effectively sets how long retries continue.

Recurly's own recommended starting windows are 27 days for monthly subscriptions, 45 for bimonthly, and 60 for longer cycles. Around the two tools sit prevention and reporting: an account updater that refreshes card details before renewals, expired-card management, and a Dunning Recovery dashboard that breaks recovered revenue down by which method won it — updater, customer update, intelligent retry, or manual collection. That's a better-instrumented native stack than most platforms ship.

Where native Recurly dunning stops

The edges are structural, not quality judgments:

• Two independent tools, one uncoordinated sequence. Recurly is explicit that retries and dunning emails don't wait for each other — an email can land on day one while a retry fires on day two, or both at once, depending on your settings. The merchant is the coordinator. A recovery system makes the retry and the message one plan: the email goes when the decline says it should, not when an independent offset says it might.
• The play doesn't reorganize around the decline code. Intelligent Retries times attempts well. But a replaced card needs a card-update ask, not a better-timed retry; an insufficient-funds failure needs a payday-aware pause, not a generic cadence. Native tooling gives you a smarter clock; a decline-code playbook gives each failure the treatment its cause requires — including not retrying at all when the decline is final.
• Email is the channel. Dunning campaigns are email campaigns. There is no native SMS or voice step for the subscriber who never opens email — and for consumer subscriptions especially, that subscriber is a large share of the failures that matter.
• No cancel deflection in the recovery stack. The subscriber who moves to cancel during a payment problem meets Recurly's subscription management, not a save flow with reason-matched offers executed on the subscription. Voluntary and involuntary churn are handled by different machinery, and only one of them is in this stack.
• It only sees Recurly. The Dunning Recovery dashboard is a good dashboard — for Recurly. Bill anywhere else as well, and your recovery picture splits in two.

What a dedicated recovery layer adds on Recurly

Swipe Again connects to Recurly and runs recovery as one coordinated sequence, layered with Recurly's native tooling so no customer is ever worked by two uncoordinated schedules:

• A decline-code playbook as the organizing system. Up to six attempts over roughly sixteen days for declines that clear with timing — payday-aware for insufficient funds, a dedicated day 1 / 3 / 7 cadence for expired cards — and no blind retries for hard declines like lost or stolen cards, which go straight to a card-update ask. The retry and the message are scheduled together, from the same decline signal.
• Outreach beyond email. Branded email you control per day of the sequence, SMS on day 1 and day 7, and voice calls reserved for high-value accounts — every message carrying a secure, hosted card-update page under your brand.
• Cancel deflection included. Reason-matched save offers — discount, pause, downgrade — at the cancel moment, executed on the subscription automatically, in the same product and the same price.
• Backfill on connect. Your last 90 days of failed payments are pulled in automatically, so recovery starts on the backlog already sitting in your account — not just on tomorrow's failures.
• Attribution you can audit. Total recovered revenue and the portion directly recovered by Swipe Again's own actions, reported separately — the split that lets you judge any recovery tool, including ours, on evidence. The definitions are contractual, not cosmetic; they're in our Terms, and the counting rules are public: how Swipe Again counts recovered revenue.

The coordination rule matters more than any single feature: decide explicitly which system owns which part of the sequence before layering anything. That discipline — one schedule, one voice to the customer — is the core of our dunning management guide.

What it costs, and how the models compare

Recurly's dunning is bundled into your Recurly plan (with Intelligent Retries gated to eligible tiers). Swipe Again is one flat price — $179/month, or $1,799/year — dunning and cancel deflection together, no percentage of recovered revenue, no per-session math. Full details on the pricing page. If you're comparing the whole field rather than one platform's stack, our best dunning software comparison runs the models side by side.

And the flat fee carries the guarantee: recover $179 in your first 30 days, guaranteed — or Swipe Again is free the next 30 days. Recovered revenue means every failed payment collected while Swipe Again is active, from any recovery source, as defined in Section 3 of our Terms. How we count: how Swipe Again counts recovered revenue

Frequently asked questions

Should I turn off Recurly's Intelligent Retries?

Not blindly. Swipe Again coordinates with Recurly's native retries so the same failure is never chased by two uncoordinated schedules. The thing to avoid is the default accident: two systems retrying and emailing on independent clocks. Whatever you run, run one plan.

Does Swipe Again replace Recurly?

No. Recurly stays your billing system of record. Swipe Again is the recovery layer on top: decline-code retries, multi-channel outreach, card updates, and cancel deflection, with attribution across every platform you bill on — not just Recurly.

How is this different from Recurly's Dunning Recovery dashboard?

The dashboard reports what Recurly's native methods recovered, by method. Swipe Again adds the recovery work itself — the channels and plays Recurly doesn't run — and reports total recovered versus directly recovered in one view across platforms. Measurement and machinery are different jobs; you want both, and they should agree.

My failures are mostly expired cards. Does the playbook change?

Yes — that's exactly what decline-code routing means. Expired cards get the dedicated 1 / 3 / 7 cadence and an immediate card-update ask instead of repeated retries against a dead date, and the account-updater layer is checked first. Retrying an expired card on a smarter clock is still retrying an expired card.

The short version

Recurly's native dunning is a well-built pair of tools. It is not a recovery system: retries and emails run on independent clocks, the play doesn't reorganize around the decline code, outreach stops at email, cancel deflection lives elsewhere, and the view ends at Recurly's border. If your unrecovered gap is bigger than $179 a month, a dedicated layer closes it — and a 14-day trial on your own backlog proves it one way or the other. For the full playbook behind all of this, see our complete guide to recovering failed payments.

Related reading: Recurly alternatives