A different root cause
Voluntary churn is a verdict on the product. Involuntary churn is a verdict on the card. Counting them together hides both.
Involuntary churn
Voluntary churn is a decision: someone weighed your product and left. Involuntary churn is an accident of payment plumbing - an expired card, a short balance, a bank that refused. The customer still wanted the product. That difference decides whether the fix is your roadmap or your billing.
Measured on your own invoices, not estimated from a benchmark.
Worked example
Illustrative figures, not a benchmark. The point is the shape of the calculation and the gap between the two rates it produces.
| How it ended | Accounts | MRR | Type |
|---|---|---|---|
| Cancelled in the app | 18 | $5,400 | voluntary |
| Downgraded to free | 6 | $1,800 | voluntary |
| Did not renew a term | 4 | $3,200 | voluntary |
| Card expired | 9 | $4,050 | involuntary |
| Insufficient funds | 5 | $1,750 | involuntary |
| Issuer refused | 3 | $1,350 | involuntary |
| Total | 45 | $17,550 |
Involuntary rate, by count
37.8%
17 of 45 departures ended on a failed charge.
Involuntary rate, by revenue
40.7%
$7,150 of $17,550 lost MRR.
The two rates differ because failures do not fall evenly across price points, and the revenue figure is the one a finance team will ask for. Report both. A single blended churn rate hides which of the two problems you actually have, and they have opposite fixes: one is a question about your product, the other is a question about your billing.
Voluntary churn is a verdict on the product. Involuntary churn is a verdict on the card. Counting them together hides both.
No model is required. A subscription that ended after a failed charge is a fact in the billing record, with a date and an amount.
You cannot argue someone back who decided to leave. You can often recover someone whose card expired, because they never decided anything.
The distinction
A cancelled subscription and a failed subscription look identical in a churn rate, and they call for opposite responses. If customers are choosing to leave, the answer is somewhere in onboarding, value or price. If their cards are failing, none of that work moves the number.
Teams that report a single blended churn rate usually cannot say which problem they have, and end up funding the expensive fix for the cheap one.
Cancelled in-app, downgraded, did not renew a term, or asked to stop. A person made a call.
The subscription ended because a charge did not clear: expiry, insufficient funds, an issuer block, a card reissued after fraud.
Involuntary churn is usually the smaller number and by far the cheaper to fix, because the customer is not disputing anything.
How to calculate it
Involuntary churn rate is the share of churn that ended in a payment failure rather than a decision. Take the customers who left in a period, split them by whether the last billing event was a failed charge, and divide.
Both numbers come out of the same billing data, which is why this is one of the few retention figures that needs no model and carries no confidence interval. If your billing system records why a subscription ended, you already have it.
Subscriptions ended by a failed payment, divided by all subscriptions ended in the same window.
A rate by count hides the fact that failures do not fall evenly across price points. The MRR figure is what a finance team will ask for.
A rising share of expired cards points at cohort age. A rising share of insufficient funds points at your customers, or your billing date.
Reducing it
Nearly all of involuntary churn is addressed before the retry schedule runs out. After that the subscription is gone and you are running a win-back, which converts far less often.
A card with an expiry date inside the next renewal window is a known future failure. Requesting an update early is the only fully preventable case.
Fixed retry schedules spend attempts on cards that cannot clear and miss the timing on cards that can.
One link, no login. Most involuntary churn survives a good retry schedule and dies at a password reset.
Keep reading
Where to go next, depending on what you're trying to work out.
FAQ
Involuntary churn is a customer lost because a payment failed rather than because they chose to leave. The subscription ends on an expired card, a short balance or a bank refusal, and in most cases the customer was still using the product.
Voluntary churn is a decision by the customer: they cancel, downgrade or decline to renew. Involuntary churn is a failure of the payment method. The first is feedback about your product, the second is a billing problem, and mixing them makes both harder to fix.
Divide the number of subscriptions that ended after a failed payment by the total number that ended in the same period. Run the same split by MRR as well, because failures do not distribute evenly across price points.
Request a new card before a known expiry lands inside a renewal, retry according to the decline code rather than a fixed schedule, and make updating a card a single link with no login. Work it before the subscription ends, because win-back after cancellation converts far less often.
Stripe retries failed invoices on a schedule you configure and can send its own dunning emails, then applies your chosen end state. It does not differentiate its handling by decline code, which is where most of the recoverable revenue is lost.
Connect Stripe read-only and see which of last quarter's departures ended on a failed charge.
Not ready to switch? Run the free 60-second revenue leak scan and see what staying put costs.