Churn Rate
Churn rate is the percentage of customers or recurring revenue a business loses over a given period, such as a month or a year.
Key takeaways
- Voluntary and involuntary churn have separate causes and require entirely separate fixes.
- Annualising monthly churn compounds survival; it is not the monthly figure times twelve.
- The honest denominator is accounts actually up for renewal, not the whole customer base.
- Annual contracts relocate churn to a renewal cliff rather than removing it.
- On a small base one cancellation swings the rate by several points of noise.
In depth
Two mechanical details decide what the number means. The first is the denominator: counting every customer at the start of a period is standard, but on annual contracts only a slice was ever able to leave, so the honest base is accounts actually up for renewal. The second is the split between voluntary churn, where someone decides to go, and involuntary churn, where a card expires and the payment simply fails. They arrive in the same report and have nothing else in common.
Billing frequency is one of the strongest levers, because an annual contract removes eleven monthly opportunities to reconsider. Payment failure rates, depth of integration into a customer's workflow, and the number of colleagues using the account all push in the same direction. The trade-off is that longer contracts relocate churn rather than remove it: dissatisfaction accumulates invisibly and arrives as a renewal cliff, at which point the account is expensive to save and there is no early signal to act on.
Practical work starts by separating the two kinds. Involuntary churn is a billing problem handled with card updaters, retry schedules and pre-expiry notices. Voluntary churn is then grouped by cohort age and reason. A scorecard quiz taken at acquisition earns its place at this stage in an unusual way: it preserves what the buyer said their problem was, so a cancellation can be compared against that stated need to see whether the account was mis-sold or simply never got the help it needed.
The rate is unstable on small bases. With forty customers, one cancellation swings it by more than two points, so month-to-month movement is mostly noise. Monthly churn is close to meaningless on annual contracts, where most months structurally contain no renewal decisions. And the measure flattens every exit into one category: a customer who was acquired by a competitor, one who went out of business, and one who left angry all count identically, which is exactly the distinction a fix depends on.
Example in practice
How to measure it
Report three cuts rather than one: customer churn, revenue churn, and churn among accounts that actually reached a renewal decision. Split each into voluntary and involuntary. Then read the numbers by cohort age, because churn is nearly always highest in the first period after purchase, which means a blended rate moves whenever the proportion of recently acquired customers changes.
Convert between periods by compounding survival rather than multiplying rates. Annual retention is the monthly survival share raised to the twelfth power, and treating monthly churn as one twelfth of annual churn understates the loss. Track the average revenue of churned accounts against the average of retained ones as well; when the leavers are larger than the stayers, the customer-count rate is understating the damage.
Common mistakes
The most costly error is reporting failed payments and cancellations as one number. Involuntary churn is a billing failure, fixed with automatic card updating, sensible retry schedules and notices before expiry. Cancellation is a fit or value failure that needs product and onboarding work. Reported together, teams launch save campaigns aimed at customers who never intended to leave, and never notice that a meaningful share of their losses required no persuasion at all.
The second is watching churn only at the moment of cancellation. By the time the request is recorded, the decision is usually weeks old and the account has been drifting for longer. Define a disengagement threshold in usage terms, such as no meaningful activity for a fixed number of days, and monitor the share of accounts crossing it. That leading indicator is where intervention still has some chance of working.
Frequently asked questions
Can lead qualification lower churn rate?
Yes. Much churn originates from poor-fit customers who never should have converted, so qualifying leads on budget, intent, and fit before sign-up reduces early cancellations. A scorecard quiz that scores respondents helps you route only strong-fit prospects to paid plans.
How do you calculate churn rate?
Divide the customers lost during a period by the number at the start of that period. For revenue churn, divide the recurring revenue lost by the recurring revenue at the start. On annual contracts, use the accounts that actually came up for renewal as the denominator instead, since customers who could not leave should not dilute the rate.
What is the difference between voluntary and involuntary churn?
Voluntary churn is a decision to leave; involuntary churn is a payment that failed, usually from an expired or declined card. The distinction matters because the fixes share nothing. Involuntary losses respond to card updating services, retry logic and advance notices, while voluntary losses respond only to changes in fit, onboarding or product value. Report the two separately.
How do you convert monthly churn to annual churn?
Compound the survival rate rather than multiplying the churn rate. Take one minus the monthly churn rate, raise it to the twelfth power, and subtract the result from one. Multiplying monthly churn by twelve overstates annual loss, because each month's churn applies to a base that has already shrunk. The gap widens as the monthly rate rises.
What counts as a good churn rate?
The figure is not portable across business models, because contract length, price point and buyer type dominate it. A monthly self-serve product and an annual enterprise agreement are not comparable in any useful way. Judge your own trend by cohort at the same age, and pay more attention to which segments churn than to the blended headline number.
How do you actually reduce churn?
In order of leverage: fix failed payments, tighten who you sell to, and improve the first weeks of use. Payment recovery is cheap and immediate. Qualification prevents losses that no retention work could have avoided. Onboarding decides whether the product becomes part of a routine. Save offers aimed at people already cancelling are the least effective and most expensive option.
Should churn be measured by customers or by revenue?
Both, because they answer different questions. Customer churn tells you how well the product suits the people who buy it, and revenue churn tells you what the losses cost. When the two diverge, that gap is the diagnosis: stable logos with falling revenue means downgrades, while stable revenue with falling logos means the small accounts are leaving.