New customers cannot repair a retention metric.
NRR follows the same starting customers and includes expansion. GRR follows that cohort while excluding expansion, showing how much of its original revenue survives contraction and churn.
New customer revenue belongs in growth, not retention. Choose a measurement window, usually a month or year, and compare consistent cohorts. Monthly NRR should not be multiplied by twelve to create an annual retention rate.
The formula, made clear.
- Opening cohort
- Customers present at the beginning of the measurement window.
- Gross revenue churn
- (Churn + contraction) ÷ opening; the complement of GRR.
Put the numbers in context.
A 100,000 revenue cohort loses 10,000 to churn and 5,000 to contraction, while expanding by 15,000. NRR is 100%; GRR is 85%.
| Input | Example value |
|---|---|
| Opening cohort recurring revenue | $100,000.00 |
| Cohort expansion | $15,000.00 |
| Cohort contraction | $5,000.00 |
| Cohort churn | $10,000.00 |
| Net revenue retention | 100% |
What this calculation assumes
Same cohort and recurring-revenue basis at both dates. No new customers, FX or acquisitions. Losses cannot exceed starting revenue. Expansion is entered separately and may take NRR above 100%.
Methodology references
What to consider next.
Use GRR to understand underlying leakage, then inspect which accounts drive expansion. Reconcile both with the total MRR bridge.
How we approach financial models →