Involuntary Churn

Counting departed customers as one pile hid the cause. Three percent leaving every month read as a product problem, so features got fixed, and six months later the number sat where it had been.

Pagedeck went through that. Grace opened last month's churn list and worked down it one account at a time. Close to half had been editing documents until two days before the account closed. That isn't the shape of someone who stopped finding the product useful. Theo lined the list up against payment records and the answer appeared: the card had expired, or a reissue had changed the number, the charge failed, and the account shut on the spot. The customers didn't know their subscription had ended.

It starts by splitting closed accounts by cause, and the test is usually one line. The customer pressed cancel, or a charge failed and the account closed.

Charges fail for a short list of reasons.

1. An expired card. Past its valid date, or reissued with a new number. This is the largest group.
2. A limit or balance problem. Often true for one month and fine the next.
3. Fraud screening at the issuer. A charge flagged as foreign and blocked automatically.
4. Mismatched billing details. A name or address that doesn't match what the card has on file.
5. A change of owner at a company account. When whoever registered the corporate card leaves, the card gets suspended with them.

Twenty to forty percent of all churn in a subscription product sits in this group. The good news is that this half shrinks without touching the product. Add a grace period, which keeps service open for a few days after a charge fails, and attach a dunning process that retries the charge several times and emails the customer while it does.

Which is why churn gets recorded split by cause. Mixed together, an improvement earned by fixing the product disappears into payment failures, and a win from rewriting a billing email gets read as a product win.

"Churn was 3.2 percent this month, and 1.1 of that was involuntary."

Only 2.1 percent has a cause the product team can go looking for.

"Let's add a notice 30 days before card expiry. Half of involuntary churn was expiry."

Attacking the single largest cause with a billing-side fix.

"This one is involuntary, so there's nothing for the product meeting."

The reason for leaving sits in payment handling rather than in the experience.

After adding the split, Theo gave failed-charge accounts a seven-day grace period with three retries across three days. Grace now lists both numbers side by side in the monthly report. When both rise at once the company has two separate problems, and the teams that have to fix them are different.