The churn you never see: failed payments between your CRM and Stripe

A customer decides to leave, and you get a signal: a cancellation, a downgrade, an unhappy call. A customer’s card fails, and you get nothing. The revenue stops just the same, but there’s no decision, no conversation, no event your CRM ever hears about. This is involuntary churn — the churn you never see — and the gap between your CRM and Stripe is where it does its quietest damage.

Most churn conversations are about the voluntary kind: the customer who chose to go, the feature that was missing, the competitor who won. That’s the churn you can fight in the product and the renewal. Involuntary churn is different. Nobody chose it. A card expired, a bank declined a charge, a billing detail went stale — and a paying customer silently stopped paying. It’s an operational failure dressed up as a loss, and it’s a bigger slice of the total than most teams assume.

How big it actually is

ProfitWell’s widely-cited estimate is that 20–40% of subscription churn is involuntary — customers lost to failed, expired, or declined cards rather than a decision to leave. Paddle, which now owns ProfitWell’s research, frames delinquent churn as accounting for up to 40% of a SaaS company’s overall churn. Recurly’s benchmark data lands in the same neighborhood, putting involuntary churn at roughly a quarter of total churn across its base. The exact share varies by business, but the direction is consistent: a large, recoverable fraction of your logo loss is a payments problem, not a product problem.

The reason it stays high is mundane and unavoidable: payment details rot. Cards expire on a roughly three-year cycle, and Paddle notes that on average about a third of your customers will need to update their card information with you each year. Layer in banks declining otherwise-valid charges — industry decline rates run in the low teens as a share of attempts — and a steady stream of perfectly happy customers lapse for reasons that have nothing to do with how they feel about you.

The blind spot that doubles the cost

Here’s the part that makes involuntary churn worse than it has to be. A failed payment is a billing event. It happens inside Stripe. Your CRM has no wiring to hear about it, so the account stays exactly as it was: active, green, forecast to renew. The CSM doesn’t intervene, because from where they sit nothing is wrong. Finance keeps counting the ARR. And the customer — who might have fixed a card in thirty seconds if anyone had asked — drifts out of the window where a save was easy.

A failed payment is a churn event your CRM never hears about. The account stays active, the forecast stays wrong, and the one window where a human could have saved the account closes while everyone believes it’s healthy.

The CRM↔Stripe blind spot

So the cost lands twice. Once as the lost revenue itself, and again as the wasted CSM cycles and inflated forecast that follow — because the system your team actually looks at still shows the customer as present. Dunning tools work on the billing side to recover the charge, but they don’t reach into the CRM to tell a human that a real account is quietly lapsing. That’s the seam, again, and it’s the exact gap we mapped in where your CRM and billing quietly disagree about money.

Recovery is real, but it’s time-boxed

The good news is that involuntary churn is the most recoverable churn there is. A large share of failures are “soft” declines — insufficient funds, temporary bank holds — that clear on their own within a few days, especially if a retry is timed near a customer’s payday. The catch is that recovery falls off a cliff as time passes and as the customer forgets they were ever your customer. Speed is everything, and speed requires knowing.

• No automated recovery: businesses relying on a processor’s default retries recover only ~20–31% of failed payments.
• Basic dunning (scheduled retries plus email) recovers roughly 45–55%.
• Comprehensive recovery (smart, issuer-aware retries plus multi-channel outreach) pushes recovery to 70–85%.

The spread between the bottom and the top of that range is enormous, and it’s all in the execution — retry timing, the dunning cadence (industry data points to 4–6 messages over about two weeks), and, critically, whether a human on the account side ever gets pulled in. Recovery tooling is worth it: Recurly reports recovering $1.6 billion in revenue for its customers annually through exactly this kind of retry-and-outreach work.

Two fixes, not one

Beating involuntary churn takes both halves of the seam, because it lives in both:

• Recover the payment (billing side). Smart retries, card-updater services, and a real dunning sequence turn a decline into a re-collected charge. This is table stakes and it works.
• Close the CRM blind spot (the seam). Surface every account that’s failed-but-still-active, so a CSM can reach out while the relationship is warm and finance stops forecasting revenue that already stopped. This is the half dunning tools don’t touch — and it’s pure read-only comparison.

The second fix is what a Junction scan gives you for this specific break: connect the CRM and Stripe read-only, and get the list of accounts where billing shows a lapse or decline but the CRM still shows active — ranked by the ARR at stake. No writes, no migration; just the accounts your dunning tool is fighting for and your CRM is quietly hiding. It’s one of the four silent revenue breaks — and the one where seeing it a week sooner most directly turns into money kept.