Churn rate formula: customer, revenue, and cohort methods
The basic churn rate formula, plus the revenue and cohort variants that catch what logo churn hides. Worked examples you can copy into a spreadsheet.
THE SHORT ANSWER
The basic churn rate formula is: customers lost during a period ÷ customers at the start of the period × 100. If you start January with 200 customers and lose 8, your monthly churn rate is 4 percent. But customer churn alone hides revenue concentration and signup-timing distortions, so pair it with revenue churn and cohort analysis.
What is the customer churn rate formula?
Customer churn (also called logo churn) counts accounts, not dollars. The formula:
Customer churn rate = customers lost in period ÷ customers at start of period × 100
A worked example. You run a B2B SaaS with 240 customers on January 1. During January, 12 customers cancel and 15 new customers sign up.
| Input | Value | |---|---| | Customers at start of January | 240 | | Customers lost during January | 12 | | New customers added during January | 15 | | Customers at end of January | 243 |
Customer churn rate = 12 ÷ 240 × 100 = 5 percent monthly.
Two things to notice. First, the new signups do not appear in the formula. Churn measures how well you keep the customers you already had, and mixing in acquisition hides retention problems behind growth. Second, the denominator is the start-of-period count, not the end-of-period count. Using the end count understates churn whenever you are growing.
What is the revenue churn rate formula?
Customer churn treats a $50 per month account and a $5,000 per month account as equal. Revenue churn does not. The standard version is gross MRR churn:
Gross revenue churn rate = MRR lost to cancellations and downgrades ÷ MRR at start of period × 100
Same company, same January. Your 240 customers represent $120,000 in monthly recurring revenue. The 12 customers who cancelled were mostly larger accounts, worth $9,000 in combined MRR. Two more customers downgraded, cutting another $600.
| Input | Value | |---|---| | MRR at start of January | $120,000 | | MRR lost to cancellations | $9,000 | | MRR lost to downgrades | $600 | | Total gross MRR churn | $9,600 |
Gross revenue churn rate = 9,600 ÷ 120,000 × 100 = 8 percent monthly.
That is the number logo churn hid. You lost 5 percent of customers but 8 percent of revenue, because the accounts that left were bigger than average. If your revenue churn consistently runs above your customer churn, your best customers are leaving faster than your small ones. That is the pattern worth losing sleep over.
This week: calculate both numbers for last month. If revenue churn exceeds customer churn, pull the list of churned accounts and sort by MRR. The pattern in the top five usually tells you why they left.
What is net revenue churn, and why can it go negative?
Gross revenue churn only counts losses. Net revenue churn subtracts the expansion revenue you gained from existing customers who upgraded, added seats, or moved to a higher tier:
Net revenue churn rate = (churned MRR + downgrade MRR − expansion MRR) ÷ MRR at start of period × 100
Continuing the example: during January, 10 existing customers expanded, adding $4,200 in MRR.
Net revenue churn = (9,600 − 4,200) ÷ 120,000 × 100 = 4.5 percent monthly.
When expansion revenue exceeds churned and downgraded revenue, this number goes negative. Negative net churn means your existing customer base grows in revenue even if you sign zero new customers. It is the single most attractive property a SaaS business can have, which is why investors ask about it before almost anything else.
Report gross and net together. Net churn alone can mask a leaky base: a company losing 10 percent gross and expanding 11 percent looks healthy at negative 1 percent net, but it is replacing unhappy customers with upsells, and that treadmill eventually stops.
How do mid-period signups distort churn rate?
The basic formula has a blind spot: customers who sign up and cancel within the same period never appear in the start-of-period denominator, and depending on how you count, may not appear in the numerator either.
Say you start the month with 100 customers, add 40, and 10 of those 40 cancel within the same month. Your basic formula reads 0 ÷ 100 = 0 percent churn if you only count losses from the starting base. Meanwhile a quarter of your new signups evaporated. Fast-growing companies systematically understate churn this way, because the flood of new customers keeps diluting whatever measurement convention they pick.
How does cohort analysis fix the distortion?
Cohort analysis groups customers by their signup month and tracks each group separately over time. Every customer belongs to exactly one cohort, so nobody gets lost between denominators, and you can see whether retention is improving for newer signups.
You can build this in a spreadsheet in an afternoon. Rows are signup cohorts, columns are months since signup, and each cell is the percentage of the original cohort still active:
| Cohort | Month 0 | Month 1 | Month 2 | Month 3 | |---|---|---|---|---| | January signups (40) | 100% | 85% | 78% | 74% | | February signups (35) | 100% | 88% | 82% | | | March signups (50) | 100% | 91% | | |
Read it two ways. Across a row: January's cohort lost 15 percent in its first month, then the losses slowed. Down a column: month 1 retention improved from 85 percent to 91 percent across three cohorts, so whatever you changed in onboarding is working. A blended churn rate would have shown you neither.
To build this from your billing data: export every subscription with its start date and cancel date, group by start month, and count how many remain active at each monthly anniversary. Stripe, Chargebee, and Recurly exports all include both dates.
How do you annualize a monthly churn rate?
The tempting shortcut is monthly churn × 12. It is wrong, because churn compounds: each month's losses come out of an already-shrunken base.
The correct formula:
Annual churn rate = 1 − (1 − monthly churn rate)^12
For 3 percent monthly churn:
1 − (1 − 0.03)^12 = 1 − 0.97^12 = 1 − 0.694 = 30.6 percent annual, not 36 percent.
30.6%the real annual rate behind 3% monthly churn, not 36%
The gap widens as churn rises. Multiplying by 12 overstates annual churn, which sounds conservative until you run the same mistake in reverse: dividing an annual target by 12 sets a monthly goal that is easier than the real one, and you miss the annual number while hitting every monthly one.
| Monthly churn | Naive ×12 | Correct compounded annual | |---|---|---| | 1% | 12% | 11.4% | | 3% | 36% | 30.6% | | 5% | 60% | 46.0% | | 7% | 84% | 58.1% |
Which churn number should you report where?
Different audiences need different numbers. Reporting the wrong one wastes the meeting.
| Audience | Report this | Why | |---|---|---| | Board | Gross and net revenue churn, annualized | They think in ARR and compounding, not logos | | Investors | Net revenue churn (or NRR) plus logo churn | Net churn drives valuation; logo churn shows base health | | Internal ops and CS | Monthly cohort retention plus logo churn | Cohorts show whether this quarter's fixes work | | Pricing and packaging decisions | Gross revenue churn split by plan tier | Reveals which tiers leak and which hold |
One convention to fix now if you have not: write down your definitions. Which date counts as churned, the cancellation request or the end of the paid term? Do downgrades count in gross churn? Does a failed payment count immediately or after the dunning window? None of the answers matter as much as answering them once and never changing mid-year, because a definition change looks exactly like a trend.
Related resources
Churn rate tells you what you lost. Its mirror metric, net revenue retention, tells you what you kept and grew, and it is the number investors will quote back to you.
RELATED
01 ·Net revenue retention explained for founders
What NRR actually measures, how to calculate it from a cohort, and why investors treat one number as a proxy for product-market fit.