Campaign Length and Churn Recognition

This page is about dunning recovery, the campaign that starts when a subscription payment fails. Cancel flows have no campaign length; outcomes settle into definitive sessions within a 24 hour window. See Two Measurement Frames.

Campaign length is how long a recovery campaign runs before you recognize a customer who still hasn't paid as passively churned. It's a business decision, not a fixed number, and the answer depends on what a past-due customer costs you while the campaign runs.

What campaign length decides

Passive Churn is the outcome when the recovery campaign reaches its end with no successful payment and no active cancellation. The customer lapsed silently.

Campaign length is that end, the point at which an unrecovered customer becomes a Passive Churn outcome and the campaign reaches a final state. Until then the campaign is in progress: not recovered, not lost, and not counted in the recovery rate.

A daily cohort completes when its last campaign reaches a final state, and a campaign that ends in Passive Churn takes the full length. Recovery campaign length varies by subscription model. The most common length is around 30 days, but some configurations run shorter or longer. The cohort waits as long as the slowest campaign needs. See Daily Cohorts and Complete Cohorts.

Three business models, three answers

eCommerce. Nothing ships until payment succeeds, so the cost of continuing outreach to a past-due customer is minimal. Recovery can continue until the customer explicitly cancels, and churn recognition often extends to 30 days or beyond. More time means more chances to recover without writing the customer off early.

Narrow fulfillment windows. When shipments go out on a schedule, you have to recover payment before the ship date or the order doesn't ship. A shorter campaign is necessary, and you recognize passive churn sooner.

Paid content and access-based services. A past-due customer who keeps access is consuming the service without paying, a real or perceived cost. Recognizing passive churn earlier, often within days, limits free access and creates urgency to resolve the payment. Where possible, lock access while the customer is past due and restore it automatically on successful payment. Where that is technically hard, a shorter campaign is the fallback.

The tradeoff

Longer campaigns recover more of the long tail. They also delay churn recognition and delay cohort completion, so every number you read arrives later. Shorter campaigns give faster numbers and cost some recoveries.

Campaign length has diminishing returns. Past a point the window stops producing.

A campaign that never ends keeps retrying a dead card and keeps emailing a customer who will never update. Repeated failed attempts drag down clearance on your good payments, and ignored emails degrade inbox placement for the recovery emails that would convert.

What changes when you change campaign length

Cohorts complete on a new schedule. The wait before you can measure performance moves with the campaign length. You still need 30+ days of completed cohorts to measure performance and 60+ to compare a before and an after.

Lengthening campaigns defers passive churn. Subscribers who would have churned early are now in longer campaigns, so their subscription cancellations show up later. For the first few weeks churn looks suppressed. About a month in, the deferred cancellations arrive (initially appearing as a spike) and then the number normalizes. This is a timing shift, not a change in the business.

Changing length or tools creates a transition window. Campaigns already in progress at the change carry over and finish under different logic, and including them contaminates both your before and your after. Exclude the first N days after the change, where N is the previous campaign length. See Comparing Periods, Tools, and Migrations.

Campaign length as a lever

The main levers that move a recovery rate: deliverability, decline-code handling, campaign length, the card-update experience, customer outreach, and the human follow-up that comes out of an escalation to your support team.

Campaign length is the one most operators never deliberately tune. It's also one you can measure: the recoveries a window produces are in your data, so you can see whether it's still producing. Retry timing to the hour or day is not measurable, because the alternate schedule never ran. What Moves a Recovery Rate covers both.

Don't measure it this way

• Comparing two accounts' recovery rates without checking that their campaign lengths match. A 14-day campaign and a 30-day campaign measure different windows.
• Calling the early churn drop after lengthening campaigns a win. Longer campaigns defer the cancellations, and they normalize about a month in.
• Comparing a period under one campaign length against a period under another without excluding the transition. Campaigns in progress at the change carry over and finish under different logic.

The full list of counter-rules is on Practices This Methodology Rejects.

---

Prerequisite: Recovery Campaigns, the anatomy of the campaign this page sets the length of.