Every operator eventually goes looking for the chart. What is a normal churn rate for a business like mine? The search is understandable: without a reference point, your own number floats in space, impressive or alarming depending on your mood. But the chart people find is usually an average of businesses that share a category label and almost nothing else — different prices, different contract lengths, different mixes of customers who chose the product last week and customers who have paid for six years.
Used well, churn benchmarks broken down by industry do one honest job: they tell you whether you are in the same neighbourhood as your peers or in a different county entirely. If your voluntary churn looks ordinary but your involuntary churn — the failed-payment kind — is double the reference range, that is not a verdict on your product. It is a signpost pointing at your billing operation. Benchmarks are best at splitting a problem into its kinds, not at grading the whole business.
The traps are worth naming. First, most published benchmarks blend voluntary and involuntary churn into one number, which hides the only part you can fix without changing the product. A company can sit exactly on the industry average while leaking twice as much to failed cards as its peers, compensated by unusually loyal customers — until the loyalty runs out and the leak looks sudden. Second, benchmarks rarely separate new cohorts from old ones. Early-tenure subscribers churn at multiples of the rate of settled ones; a fast-growing company will look worse than a stagnant peer for reasons that have nothing to do with quality.
Third, and most quietly: the inputs are self-reported, survivorship-flavoured, and often gathered to sell something. That does not make them useless. It makes them a starting line — a reason to pull your own numbers apart, not a substitute for doing it.
Your own numbers, pulled apart, usually give you three figures that matter more than any chart: churn split by cause (asked to leave versus payment failed), churn split by tenure (first ninety days versus after), and recovery rate on the failures (of the payments that failed, how many did you actually get back?). Compare those three against the benchmark ranges and you have a work list. Compare a single blended rate and you have a mood.
There is also a time dimension benchmarks flatten away. Churn is seasonal in many categories — January resolutions, summer pauses, budget cycles in business markets. A month that looks alarming against an annual average may be the most ordinary month of your year, and a calm month may be the anomaly. This is why serious operators trend their own split numbers monthly and visit the industry charts quarterly, for orientation, rather than the other way around.
If the split points at involuntary churn — and for most subscription businesses it eventually does — the follow-up is mechanical: fix the credential layer, time retries to real paydays, write recovery emails people act on, and count each recovered dollar once. That is the work Swipe Again's homepage describes, priced flat so the recovered revenue is not shared with the tool that recovered it. But start with the split, not the chart. The chart can tell you where you stand. Only your own numbers can tell you what to do on Monday.
Top comments (0)