How it's calculated
Leave new customers and their revenue out of every figure: these numbers are about the customers you had at the start. For the full method, with a worked example, see how to calculate churn rate.
- Customer churn rate = customers lost ÷ customers at the start × 100
- Revenue churn rate = (cancellations + downgrades) ÷ recurring revenue at the start × 100
- GRR = (starting revenue − downgrades − cancellations) ÷ starting revenue × 100, which is 100% minus revenue churn
- NRR = (starting revenue + upgrades − downgrades − cancellations) ÷ starting revenue × 100
- Lifetime value (estimate) = average revenue per customer per month × gross margin ÷ monthly customer churn
Monthly to yearly: rates compound, so the calculator raises a monthly figure to the 12th power rather than multiplying by 12. A monthly NRR of 101% is about 112.7% a year. The guide to net revenue retention covers NRR step by step, and gross revenue retention sets GRR and NRR side by side.
What good looks like
| Figure (yearly) | B2B SaaS median, 2025 | Source |
|---|---|---|
| Revenue churn | 12.5% | Lighter Capital |
| Gross revenue retention | 91% | SaaS Capital |
| Net revenue retention | 101% | SaaS Capital |
Compare with your own past periods first, then with a benchmark. A product sold to small businesses usually loses more customers each year than one sold to large companies on annual contracts. There's no customer churn or lifetime value benchmark here: we haven't found one we can trace to its original source.
Questions
Should I calculate churn monthly or yearly?
Benchmarks are yearly, so compare a yearly figure. If you track churn monthly, don't multiply it by 12: the months compound. 3% a month is about 30.6% a year, not 36%. Pick Month in the calculator and it works out the yearly figure for you.
Why can NRR go above 100% when GRR can't?
Gross revenue retention counts only what you lose to cancellations and downgrades, so it can never go above 100%. Net revenue retention adds upgrades back in, so when existing customers grow by more than you lose, it goes above 100%.
How accurate is the lifetime value estimate?
It's a rough guide. The formula assumes churn stays the same every month, so a very low churn rate gives a very large number. An early-stage company rarely has enough history for a steady churn rate, so treat it as a range and lean on churn and NRR instead.