The members you are losing to expired cards

Most owners treat churn as a retention problem. Someone stops enjoying the training, drifts, and cancels. The fix is better classes, better follow-up, better culture.

That is real, and it is not the whole picture. A large share of the members you lose never decide to leave at all. Their card expires, the charge fails, the payment quietly stops, and some weeks later they are marked inactive. Nobody had a conversation. Nobody was unhappy.

Payment industry data puts this at up to 30% of membership churn in the fitness sector, with somewhere between 7% and 15% of recurring charges failing on first attempt in any given cycle. If you bill 200 members monthly, that is somewhere between fourteen and thirty charges failing every single month.

The industry calls this involuntary churn. It is worth separating from the voluntary kind, because almost everything you would do about one is useless against the other.

Why it hides

Voluntary churn announces itself. Someone tells you, or stops turning up, and you notice.

Involuntary churn produces no signal at the point it happens. The failure is a line in a billing report, not an empty spot on the mat — and the member usually does not know either. Their card was reissued after a fraud alert, or it expired and the new one never went in. They are still coming to class. They still consider themselves a member. The relationship is intact and the money has simply stopped.

By the time anyone connects the two, weeks have passed, several charges have failed, and the conversation is now about back-payment rather than a card update. That conversation is where members actually leave — not at the failed charge, at the awkward chase that follows it.

The four failures worth telling apart

Decline codes are not all the same, and treating them identically is why retry logic so often fails. Broadly:

Expired card. Entirely predictable, and the only category you can fix before it happens. You know the expiry date the day the card is stored.

Insufficient funds. Timing, not intent. Retrying on the same date next month repeats the same result; retrying a few days later, after payday, frequently does not.

Card reissued or replaced. The account is fine, the number is stale. Fraud reissues are common and the member is often unaware their old number stopped working.

Hard decline. The issuer is refusing outright — closed account, blocked, reported. Retrying achieves nothing and can add fees.

The first three are recoverable. The fourth needs a human. Systems that retry everything on a fixed schedule burn attempts on hard declines and spend nothing extra on the recoverable ones, which is exactly backwards.

What actually recovers it

Published recovery figures cluster around 60–80% for combined retry logic and card-updater services, and managed billing operations report recovery in the mid-eighties on well-run accounts. Three things do most of that work.

Update cards before they expire, not after. You hold the expiry date. A message thirty days out, naming the last four digits, converts far better than a message sent after a failed charge — because it is a favour rather than a demand. This alone removes the most predictable category entirely.

Retry on the failure reason, not the calendar. Ours retries declined cards automatically, before a member ever notices there was a problem — see billing and payments. The principle is the same wherever you run it: insufficient funds wants a few days and a different point in the month. A reissued card wants an updated number, and no number of retries will produce one. Retrying every decline three times over three weeks treats a payday problem and a closed account as the same event.

Ask once, early, in a way that is easy to answer. The first message after a failure decides the outcome. It should say what happened, what it costs, and give a link that takes under a minute. It should not read like a collections notice for a member who has done nothing wrong.

Measure it separately or you will misread everything

Here is the practical consequence, and it is the reason this matters beyond the money.

If a quarter of your departures are billing failures and you count them all as churn, your retention numbers are wrong in a specific and expensive direction. You will conclude the programme has a retention problem. You will spend on curriculum, events and follow-up — all reasonable responses to the problem you think you have, and none of them will move the number, because the actual cause is a card that expired in March.

Split the two. Count a member lost to a failed payment as a billing failure, not a cancellation. That split is exactly what retention analytics is for here — cohort curves, tenure at cancellation and churn as a rate rather than a raw count, read from your own history instead of a benchmark. Then look at what is left: that residue is your real retention picture, and it is usually a good deal healthier than the combined figure suggested.

Our churn calculator will show you the annual cost of your current rate in about thirty seconds — two numbers you already know, nothing leaves your browser. Run it once with your total departures, then again with only the voluntary ones. The gap between those two answers is the money sitting in your billing system rather than in your programme.

The uncomfortable part

Involuntary churn is the cheapest churn to fix and the least interesting to work on. It is card expiry dates and retry timing. There is no version of it that feels like coaching.

It is also, for most businesses, the largest single block of recoverable revenue available without signing one new member — and unlike enrolment, it compounds quietly in the wrong direction for as long as it is ignored.

Look at your last three months of failed charges. Not the cancellations, the failures. If nobody has counted them, that number is usually the surprise.

If your current system cannot show you that number, that is the answer to a different question. Ours retries declined cards on its own, carries aging and collections in the same place rather than as a bolt-on, and reports churn as a rate you can compare month to month instead of a raw count you cannot — on every tier, including the $39 one.

---

Figures on decline and recovery rates in this article are published industry aggregates for the fitness and membership sector, linked at each claim. They are not measurements of our own customer base.