How to Reduce Subscription Churn

Subscription churn concentrates in three places: the gap between the first and second delivery, failed payments, and the cancel flow. Work them in that order. Brands that start with retention offers at the cancel screen are treating the last stage of a problem created two stages earlier, which is why those offers buy one cycle at best.

The first-to-second delivery gap

More subscribers leave between delivery one and delivery two than at any later point, and the reason is rarely the product. It is that the subscriber has not yet built the habit the subscription assumes, and the second charge arrives before the value of the first delivery has landed.

Three fixes, in order of effect. First, match the default interval to actual consumption rather than to your preferred revenue cadence. A 30-day default on a product most people use in 45 days guarantees a pile-up, and a pile-up is the most common stated reason for cancelling. Let the subscriber set the interval at checkout and make changing it obvious afterwards.

Second, send a pre-charge notification with a one-tap route to delay. Framing it as a reminder with control rather than a charge warning converts a would-be cancellation into a skip. A skipped cycle keeps the subscription; a surprise charge frequently ends it.

Third, use the first cycle to teach usage. A short onboarding sequence covering how to use the product, what to expect and when, does more for cycle-two retention than any discount, because the churn cause it addresses is confusion rather than price.

Failed payments are churn, and mostly recoverable

A meaningful share of what brands record as churn is an expired card or a declined authorisation, not a decision. This is the cheapest churn to recover because the subscriber never intended to leave.

Retry on a schedule that respects issuer behaviour rather than retrying immediately and repeatedly, which raises the chance of a hard decline. Pair the retries with an email and SMS sequence that gives a direct card-update link, and keep the link short-lived for security but long enough to be usable. Add card-account-updater support through your payment processor if your platform offers it, since it resolves a share of expiries with no customer action at all.

Time the first retry against payday patterns where your data supports it. A decline for insufficient funds on the 28th and a retry on the 2nd recovers at a materially better rate than four retries inside 48 hours, and it costs nothing to schedule.

Set a defined dunning window and a clear end to it. An indefinite retry cycle keeps dead subscriptions in your active count and corrupts every churn number you report.

One measurement note before moving on. Report failed-payment churn separately from voluntary churn in every summary you produce, and report both against the same denominator. A blended number moves for reasons nobody can explain afterwards, and it hides the fact that the two have completely different remedies.

The cancel flow, done honestly

By the time someone reaches the cancel screen, the useful job is to capture the reason and offer the alternative that matches it. Too much product is a skip or an interval change. Too expensive is a smaller size or a lower-frequency plan. Do not want it any more is a cancellation, and obstructing it produces chargebacks and complaints rather than retention.

Rank the alternatives by what the reason actually implies, and show one, not four. A screen offering a skip, a smaller size, a lower frequency and a discount all at once reads as a brand bargaining, and it converts worse than a single relevant option.

Offer pause prominently. A pause keeps the relationship, the payment method and the data, and a meaningful share of paused subscribers resume without any further intervention. Make the reason field structured rather than free text so the answers aggregate into something you can act on, and read them monthly.

Then act on the aggregate. If the top reason for three months running is pile-up, the fix is the default interval, not a better offer at the exit.

When we take over a subscription programme, [BMO NUMBER: insert the share of reported churn that turned out to be failed payments, with the account and window] is usually the finding that reorders the client’s priorities. Our dtc subscriptions  [link to: /services/subscriptions] team works the three stages in sequence, starting with the dunning sequence  [link to: /blog/dunning-email-sequence] because it is the fastest recovery available.