Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

Metric

Churn Rate

Share of customers at the start of a period who left during it.
BusinessmetricUpdated

Formula

Customers lost in the period ÷ Customers at the start of the period × 100

Unit: % · lower is better

Definition

Customer churn rate counts relationships that ended. Take the customers who were active on the first day of a period, count how many of them were no longer customers on the last day, and divide. The result is usually expressed as a percentage per month or per year, and the period has to travel with the number: a monthly rate and an annual rate describe the same business at very different scales and cannot be compared by eye.

Customers lost in the period ÷ Customers at the start of the period × 100(%)

  • Customers at the start of the period is the fixed base. People who joined during the period are left out of it, even if some of them also left before the period closed; including them inflates the base and makes the rate look lower than it is.
  • Customers lost in the period are members of that starting base who cancelled, failed to renew, or were closed for non-payment. Whether failed payments count is a choice to state, not to leave implicit.
  • The period is a calendar month, quarter or year, chosen to match the billing cycle.

Customer churn and revenue churn

The word churn covers two different questions. Customer churn, sometimes called logo churn in business software, asks how many relationships ended. Revenue churn asks how much recurring revenue the existing base lost through cancellations and downgrades, divided by the recurring revenue it started with. The two move apart whenever the customers who leave are larger or smaller than average, so a business can lose many small accounts and little revenue, or a handful of large accounts and a great deal. Reporting both prevents either story from going unnoticed.

Businesses without subscriptions

A retailer has no cancellation event, so nobody formally leaves. Churn then has to be inferred from silence: a customer is treated as lapsed once the gap since their last order is long relative to their normal buying cycle. Because that threshold is a judgement, many transactional businesses prefer to report the positive measure, Repeat Purchase Rate, and keep churn for subscription lines.

How to interpret it

Lower is better, but a single churn figure says little until it is split. The useful readings are:

  • By cohort. Customers who joined in the same month, followed together, show whether newer customers leave faster than older ones. A blended rate is dominated by long-standing loyal customers and can sit still while each new Cohort is worse than the last.
  • By age. Most subscription businesses lose customers fastest in the first few billing cycles, so the same rate means something different at month two and at month twenty.
  • By reason. Voluntary churn (the customer chose to go) points to value, product and service. Involuntary churn (a card expired or a payment failed) points to billing operations and is usually the cheapest share to recover.
  • By acquisition source. A channel that brings customers cheaply and loses them quickly may cost more per retained customer than a dearer one.

The churn and retention metrics guide sets out the full family of retention measures and the leading indicators that move before customers leave; this page is the definition those measures rest on.

Illustrative example — not a real result

A gym chain recounts its base

Glintspire Fitness Studios reports monthly churn by dividing members who left in a month by all members on its books that month, including the month's new joiners. In March that gives 180 leavers over 4,600 members, or 3.9%. Recounted on the start-of-month base of 4,000 members, and counting only leavers from that base (150 of the 180 left), churn is 3.75%; the 30 joiners who quit within weeks are tracked separately as early cancellations. Split by joining cohort, members who signed up in January on a no-fee offer leave at 9% a month, while members who joined in the previous autumn leave at 2%. The blended figure had been hiding a promotion that recruited people who were never going to stay.

What moves it

  • Onboarding and early value. Customers who reach a first useful outcome quickly stay longer; most early churn is a failure to get there.
  • Fit of the customers acquired. Discounts and broad targeting bring in people with weaker intent, and the churn they cause arrives a cycle or two after the acquisition report looked good.
  • Price changes and renewals. Rates step up around price rises and the first annual renewal.
  • Service failures. Unresolved problems and repeat contacts tend to precede cancellations.
  • Payment operations. Card updaters, retry schedules and reminder emails reduce involuntary churn without changing anything the customer experiences.

What good looks like

There is no reliable universal benchmark, and the site does not cite one. An acceptable rate depends on the price point, the contract length, the category and what each customer is worth, so a figure that would sink one business is comfortable for another. The useful comparison is internal: this cohort against earlier cohorts at the same age, and the current rate against the rate the business's economics can sustain. That second test is arithmetic, not convention: given the margin each customer earns and the cost of winning them, how long must they stay for acquisition to pay back?

Churn sets expected lifespan, and lifespan is one of the multipliers of LTV: a lower rate lengthens the time a customer pays and raises their value. It also stretches or shortens Payback Period, the months of margin needed to recover acquisition cost, because customers who leave early never repay it. For transactional businesses Repeat Purchase Rate carries the same information from the positive side, and Cohort tables are the format in which churn is most honestly read. The LTV and CAC guide shows how churn feeds the lifetime value calculation, and the break-even calculator for subscription businesses turns an assumed churn rate into the most a subscriber can cost to acquire. Where retention sits in a measurement plan, and which drivers hang beneath it, is worked through in the KPI tree for retention and the wider customer experience measurement section. How churn behaves differently in self-serve and sales-led software models is compared in product-led versus sales-led growth.

Common mistakes

  • Diluting the base with customers acquired during the period, which makes churn look lower.
  • Quoting a rate without its period, so a monthly figure is compared with an annual one.
  • Reporting customer churn alone when revenue churn tells a different story about large accounts.
  • Mixing voluntary and involuntary churn, which hides the cheapest losses to fix.
  • Reading only the blended rate, so worsening new cohorts stay invisible until they dominate the base.
  • Treating annual plans as retained for the months they cannot cancel, then discovering the losses all at once at renewal.

More terms to look up

Calculators

Where this term is used

Related