Resources

How to Calculate Churn Rate (With Examples)

Churn rate is the most quoted number in subscriptions and one of the easiest to calculate badly. Two operators can look at the same business in the same month, use two reasonable-looking methods, and report rates that differ by a full point — then argue about product strategy based on arithmetic, not reality. The fix is boring and worth it: pick a definition, state the period, count consistently, and separate customers from revenue.

This guide covers the customer churn formula, the revenue churn formula, worked examples, monthly versus annual math, and the cohort notes that keep the number honest.

The customer churn rate formula

Customer churn rate = customers lost during the period ÷ customers at the start of the period × 100.

That is the whole formula. The discipline is in the definitions:

• Customers lost means subscriptions that ended during the period — cancelled, expired, or terminated after a failed payment was never recovered. A subscriber who clicked cancel but whose paid term runs to the end of next month has not churned yet in most reporting setups; count them when the subscription actually ends, and use that rule every month.
• Customers at the start means active subscribers on day one of the period. Do not use the end-of-period count, and do not use an average unless you say so — both quietly change the answer.
• The period must be named every time the number is quoted. A churn rate without a period is not a metric; it is a rumor.

New customers gained during the period do not enter this formula at all. Growth hides churn in a net subscriber count, which is exactly why churn is calculated on the starting base alone: the question is how much of what you started with you kept.

Worked example: customer churn

Example — hypothetical numbers, for illustration only. A subscription business starts March with 2,000 active subscribers. During March, 64 subscriptions end. It also gains 210 new subscribers.

Customer churn rate = 64 ÷ 2,000 × 100 = 3.2% for March.

Note what did not happen: the 210 new subscribers did not lower the rate, and the ending count of 2,146 subscribers did not become the denominator. If you had divided by the ending count, you would have reported roughly 3.0% and understated the loss. Small method choices like that compound across a year of board decks.

The revenue churn formula

Customers are not worth equal amounts, so a second formula tracks the money:

Revenue churn rate = recurring revenue lost from churned and downgraded accounts during the period ÷ recurring revenue at the start of the period × 100.

Example — hypothetical numbers, for illustration only. The same business starts March with $60,000 in monthly recurring revenue. The 64 lost subscriptions were worth $1,540 a month, and two surviving subscribers downgraded, removing another $110 a month. Revenue churn = ($1,540 + $110) ÷ $60,000 × 100 = 2.75% for March.

Customer churn said 3.2%; revenue churn said 2.75%. Both are correct — they answer different questions. The lost subscribers were, on average, smaller accounts than the base. If the mix had run the other way, revenue churn would have been the higher number and the more urgent one. The full comparison — definitions, when each metric misleads, and the idea of negative churn — is covered in our guide to revenue churn vs customer churn.

Monthly vs annual churn

Most subscription businesses calculate churn monthly, because billing is monthly and problems surface faster. Annual churn is useful for annual contracts and for long-range planning. The mistake to avoid is converting between them by simple multiplication.

Example — hypothetical numbers, for illustration only. A steady 3.2% monthly churn does not equal 38.4% annual churn (3.2 × 12). Each month the base shrinks, so the correct compounding is: annual churn = 1 − (1 − 0.032)¹², which works out to about 32.3%. Simple multiplication overstates the loss because it charges the same starting base twelve times.

Run the conversion in the other direction the same way. If you are quoted an annual figure and need a monthly working number, do not divide by twelve and call it exact — state it as an approximation, or better, recalculate from the underlying monthly counts. And never compare one company's monthly rate against another's annual rate in the same table; it happens more often than anyone admits.

For context on how different business types tend to land: Churnkey's published analysis puts average annual churn at about 24% for B2C subscriptions versus about 16% for B2B subscriptions. Treat that as orientation, not a target — your price point, contract length, and customer type move your number more than any average will.

Cohort notes: where the simple formula hides things

The headline formula treats every subscriber as identical. Cohort analysis — grouping subscribers by the month they started — shows you what the blended number hides:

• Early-life churn is usually the story. A cohort's first one or two billing cycles behave differently from its tenth. If new cohorts churn faster than old ones, you have an onboarding, expectation, or acquisition-quality problem that a blended rate will dilute for months.
• Plan and price cohorts split the cause. Calculate churn separately for monthly versus annual plans and for each price tier. A single tier driving the losses points at that tier's value or audience, not at the whole product.
• Channel cohorts keep marketing honest. Subscribers from different acquisition sources can retain very differently. Blended churn can look stable while one channel quietly deteriorates.
• Keep cohorts and the headline rate on the same definitions. If the headline counts churn at subscription end, cohorts must too, or you will spend meetings reconciling two numbers that were never measuring the same event.

You do not need a cohort table to calculate churn correctly. You need one the moment you ask why the rate moved.

Split the number by cause before you act on it

A single churn rate mixes two different problems. Voluntary churn is a decision: the subscriber chose to leave. Involuntary churn is a payment failure: the card expired, funds were short, or the bank declined the charge, and the subscription ended through billing plumbing. Recurly research estimates that 20–40% of subscriber churn is involuntary — a large slice of the number most teams hand to their product team.

The responses have nothing in common. Voluntary churn is worked with product, pricing, onboarding, and save offers at the cancellation moment. Involuntary churn is worked with decline-aware retries, card-update requests, and dunning messages, on a clock of days. Our explainer on subscription churn: voluntary vs involuntary walks through the split in detail, and our failed payment recovery benchmarks page collects the published figures behind the involuntary side.

Practically: tag every ended subscription with its cause at the moment it ends, then calculate two churn rates alongside the headline one. If you only fix the formula and not the tagging, you will have a precise number that still points your team at the wrong work.

Common calculation mistakes

• Changing the denominator mid-year. Start-of-period, end-of-period, and average bases each produce different rates. Pick start-of-period and stay with it.
• Counting a cancel click as churn and an ended subscription as churn, in different months. One event definition, applied always.
• Letting new sales mask the rate. Net growth is a fine number to report separately; it is not a churn rate.
• Mixing trial conversions into the base inconsistently. Decide whether trials count as customers, write it down, and apply it to every period you compare.
• Annualizing by multiplying by twelve. Compound it, as shown above, or label the figure a rough approximation.
• Rounding away movement. At a few thousand subscribers, a tenth of a point is real money. Report at least one decimal place, and keep the raw counts next to the percentage so anyone can recheck the arithmetic.

The bottom line

Customer churn rate = customers lost ÷ customers at the start of the period. Revenue churn applies the same shape to dollars. Name the period, hold the definitions steady, compound — never multiply — when you annualize, and split the result by cause before assigning anyone to fix it. A churn rate calculated this way will not flatter you, but it will tell you the truth early enough to act on.

Frequently asked questions

What is the formula for churn rate?

Customer churn rate equals customers lost during the period divided by customers at the start of the period, multiplied by 100. Revenue churn uses the same shape with recurring revenue lost divided by recurring revenue at the start of the period.

Should I calculate churn monthly or annually?

Monthly for most subscription businesses, because billing cycles are monthly and problems show up sooner. Use annual figures for annual contracts and long-range planning, and compound monthly rates instead of multiplying by twelve when you convert.

Do new customers count in the churn calculation?

No. New customers gained during the period are excluded from both the numerator and the denominator. Churn measures how much of your starting base you lost; growth is reported separately.

What is a good churn rate?

There is no universal good rate — it depends on price, customer type, and contract length. Published analysis puts average annual churn higher for B2C subscriptions than for B2B, so compare against your own trend and segment first, and external figures only for rough orientation.

Why track revenue churn as well as customer churn?

Because subscribers are not worth equal amounts. Losing a small number of high-value accounts can hurt more than losing many small ones. Revenue churn weights the loss in dollars, and the gap between the two rates tells you which end of your base is leaving.

Prefer to work the math by hand? The churn rate calculator worksheet walks through the same calculation step by step, with cohort and revenue splits.

Related reading: Churn seasonality · Retention curves