The Four Outcomes of a Recovery Campaign

This page is about dunning recovery, the campaign that starts when a subscription payment fails. Cancel flows have their own outcomes and use Save Rate.

Complete cohorts. Every outcome counted. The four states below are what "every outcome" means.

Every non-voided campaign reaches exactly one of four final states: Card Update, Successful Retry, Cancellation, or Passive Churn. The first two are recoveries. The last two are losses. All four sit in the same denominator.

A non-voided campaign is active from the moment the payment fails until it reaches one of these four states. Active (or in progress) is not a fifth outcome. It means the outcome is unknown, so the campaign counts as neither recovered nor lost, and the cohort it belongs to stays incomplete until it resolves. Voided campaigns are the exception, covered below. What that does to the rate is on Daily Cohorts and Complete Cohorts.

Card Update (recovered)

Recovery via the customer updating their payment method, where the new method succeeded. Counts as Recovered.

A card update needs the customer to act, so this outcome reflects how well your outreach reaches them and how easy the update itself is. It counts when the new method pays, not when the customer saves the card.

Successful Retry (recovered)

Recovery via the original payment method succeeding on a retry, whether we ran the retry or the platform did, without a card update. Counts as Recovered.

Many declines are temporary, and retrying the same card recovers the payment with no action from the customer. A successful payment is a successful payment, whether it came from a card update, a retry we ran, or a retry the platform ran.

Cancellation (lost)

The customer actively cancelled the subscription while a recovery campaign was running. Counts as Lost in the recovery rate denominator.

This is the outcome most tempting to leave out, and the one that most needs to stay in.

First, the recovery experience itself can cause the cancel. Billing confusion, friction in the card update flow, emails that sound off-brand, badly timed retries, a support rep closing the account mid-campaign. Any of these can turn a recoverable failure into a lost customer, and there's no clean way to know afterward which cancellations the campaign caused and which would have happened anyway. So a cancellation is a measure of dunning effectiveness, and it counts.

Second, at the moment a payment fails you can't know which of the four outcomes will occur. Removing one of them after the fact skews the result and hides spikes in active churn.

Pulling cancellations out of the math can inflate a reported recovery rate with no change in performance.

Cancellation stays separate from Passive Churn so it can also feed active churn analysis.

Passive Churn (lost)

A campaign that reached the end of its recovery window without a successful payment and without an active cancellation. The customer lapsed silently. Counts as Lost.

When a campaign ends in Passive Churn depends on the campaign length. The most common length is around 30 days, but it varies by business model. Campaign length is a lever with tradeoffs of its own; see Campaign Length and Churn Recognition.

Voided campaigns, the one exit that isn't an outcome

Sometimes the customer or merchant cancels the underlying order (a skip, a pause, a delay) rather than the subscription. The charge that triggered the campaign no longer exists, so there's no payment left to recover. That voids the campaign, which leaves its cohort. If the next renewal fails, a fresh campaign starts in a fresh cohort.

A voided campaign is not a recovery, not a loss, and not in the recovery rate. A voided campaign is the only state that legitimately leaves a cohort.

This is mostly an eCommerce subscription pattern, where a customer can skip or pause a shipment without ending the relationship. Some SaaS billing models offer similar functionality.

Source of recovery doesn't determine credit

A successful payment is a successful payment, whether it came from a card update, a retry we ran, or a retry the platform ran.

The alternative, counting only the recoveries one tool triggered, sounds rigorous and isn't. It rewards retrying early and aggressively to beat the platform to the payment. The recovery rate measures what happened to the customer, not which system got there first. You can still identify which retry or channel produced a given recovery from the underlying campaign data, which is useful when troubleshooting.

Don't measure it this way

Each rule above has a mirror image that distorts the recovery number.

• Excluding cancellations from the denominator. They're a dunning outcome and they stay in.
• Attributing recoveries by which system got there first. Count outcomes, not credit.
• Treating in-progress as a fifth outcome. It's the absence of an outcome, and the cohort waits.
• Counting a voided campaign as recovered or lost. The charge is gone, so the campaign leaves the cohort.

The full list is on Practices This Methodology Rejects.

---

Prerequisite: none. This is the first teaching page. Next: Daily Cohorts and Complete Cohorts, where every campaign belongs to the day it started and a rate waits for the whole cohort to finish. The formula: Recovery Rate: the Formula.