How to calculate churn rate for B2B SaaS
The monthly customer count in your dashboard can't give you a churn rate. Going from 1,005 to 1,100 customers could mean 30 left, or 130: it depends on how many new customers came in.
Churn in customers
Add the customers who joined or came back during the month, and the ones who left fall out of the difference.
Churn in revenue
Revenue needs two more lines, because customers who stay also upgrade or downgrade. Billing tools usually list these movements separately.
- MRR on 1 Oct
- $200k
- New
- + $14k
- Expansion (upgrades)
- + $6k
- Contraction (downgrades)
- − $3k
- Churned
- − $9k
- MRR on 1 Nov
- $208k
Revenue churn = $9k ÷ $200k = 4.5%, against 3% of customers: two of the 30 were large accounts. Net of contraction and expansion, the flip side of net revenue retention: ($9k + $3k − $6k) ÷ $200k = 3%.
Finance will look at the revenue figure first. Count only paying customers: a free user who never pays isn't churn, while a customer who drops back to a free plan is lost revenue.
Expect self-serve churn to run higher
Self-serve customers are smaller, pay month to month and can leave in a few clicks. Sales-led customers sign a contract, usually annual, and can only leave at renewal. Higher self-serve churn is the normal state of things, so judge each against its own history rather than against the other.
For contracts, the renewal rate reads more clearly than a monthly rate: of the contracts that came up for renewal, how many renewed, and at what value.
What is a good churn rate?
It depends on what a customer costs to win. Say a new customer costs $3,600 to win, through ads, events and sales time, and brings in $200 of margin a month. It takes 18 months to earn that $3,600 back. Churn decides whether customers stay that long. If 3% of customers leave each month, the average customer stays about 33 months (1 ÷ 3%). If 6% leave, about 17 months.
The line marks month 18, when a customer has paid back what it cost to win. At 6%, the average customer leaves just before.
At 3%, customers keep paying long after they've covered their cost. At 6%, the average customer leaves before you've earned back what you spent to win them. That's why self-serve, where customers cost far less to win, can live with a higher churn rate. The LTV:CAC ratio builds on the same idea.
Before a price increase, agree on the churn you can afford
Take 100 customers paying $100 a month: $10,000 in total. Raise the price by 10%, and if nobody leaves, you make $11,000. If 9 customers leave, the 91 who stay pay $10,010, about what you made before. So with a 10% increase, you can lose up to about 9% of the customers concerned and still come out even. With a 20% increase, about 17%.
Agree on that margin before the change, then compare the departures that follow with it. On annual contracts, wait for the renewals to see the effect. Changing how you charge, from a price per user to a price based on usage for example, raises some bills and lowers others. The customers whose bill goes up are the ones likely to leave, so look at them account by account before the switch.
What to check first in a churn analysis
Before concluding that something is wrong, check whether one of these explains the rise.
- MixYou won more new customers than usual. New customers leave more often in their first months, so churn rises with them.
- PlanYou stopped targeting a segment, like small accounts after a move upmarket. Their departures were expected.
- PriceYou raised prices, and the departures stay below the margin you agreed on.
- Real changeThe same kind of customer, after the same time with you, leaves more often than a year ago.
The acquisition source matters as well. A channel whose customers leave within a few months costs more than its acquisition cost suggests. Funnelsight reads churn by source and by cohort in the same view as revenue and payback, so you can tell the channels that bring customers who stay from those that mostly bring sign-ups.
See churn by source and by cohort, next to the revenue each channel brings.
Start free trialHow do you calculate an annual churn rate?
Annual churn rate = customers lost during the year ÷ customers at the start of the year. Count only the customers you had on day one, not those who joined during the year. If you only have a monthly rate, don't multiply it by twelve: each month, churn applies to fewer customers, because those who have left can't leave again. Compound it instead: annual churn = 1 − (1 − monthly churn)^12. A 3% monthly rate gives about 31% a year, not 36%.
What's the average churn rate in B2B SaaS?
Recent surveys measure it in revenue. SaaS Capital's 2025 survey of more than 1,000 private B2B SaaS companies puts median gross revenue retention at 91%, so about 9% of revenue lost in a year, and 95% for companies whose contracts exceed $250k a year. Customer churn usually runs higher than revenue churn, since small customers leave more often. Compare yourself with companies selling at a similar contract size, and above all with your own trend.
What's the difference between churn rate and retention rate?
They are two sides of the same measure. Over the same period and on the same base, churn rate is 100% minus retention rate. Net revenue retention is different: it adds expansion, so it can go above 100%.