MRR, Churn, NRR and GRR: SaaS Retention Examples

Give every amount the same basis

Before calculating subscription metrics, choose a reporting window and normalize recurring amounts consistently. An annual invoice is not directly comparable with a monthly plan price. Keep one-off fees, collected cash and accounting revenue separate from recurring run rate. Write down how you treat delinquency, pauses and reactivation, then use that policy throughout the review. A clean definition matters more than adding decimal places to a percentage built from amounts that describe different things.

Start with the whole-business MRR bridge

The MRR tool separates starting revenue, first-time new customers, expansion, reactivation, contraction and cancellation losses. Its example starts at 10,000, adds 2,000 new, 500 expansion and 250 reactivation, then subtracts 300 contraction and 1,000 churn. Ending MRR is 11,450. Multiplying by twelve produces a 137,400 annualized run rate, not a record of actual annual revenue or a guaranteed forecast. Enter an observed closing balance to expose a reconciliation difference before analyzing growth.

Freeze the opening cohort for retention

Retention asks what happened to customers already present at the start. New customers belong in the growth bridge, but not in the opening-cohort retention numerator. Keep a customer-level reconciliation so an upgrade and its later reversal do not inflate both expansion and contraction. A business can add enough new customers to grow while losing a substantial part of its original cohort. Looking at total ending revenue alone cannot distinguish those two processes.

Read gross and net revenue retention together

With opening cohort revenue 100,000, cancellation losses 5,000, downgrades 3,000 and expansion 15,000, the retained base before expansion is 92,000. GRR is therefore 92%. Adding expansion gives 107,000 and NRR of 107%. Strong expansion offsets losses in this example, but the losses still occurred. Inspect both percentages and the underlying account changes rather than deciding that the cohort has no retention problem because the net figure is above 100%.

Keep customer churn distinct from revenue churn

The churn tool’s example begins with 200 customers and loses ten, giving 5% customer churn. Opening recurring revenue is 20,000, with 1,500 cancellation losses, 500 downgrade losses and 2,500 expansion. Gross revenue churn is 10% and net revenue churn is −2.5%. Negative net revenue churn means expansion outweighs revenue losses; it does not mean a negative number of customers left. Rates must also carry their reporting period, since a monthly percentage is not an annual percentage.

Use customer churn in a bounded LTV scenario

The LTV tool models survival using monthly customer churn, not net revenue churn. With monthly revenue 60, contribution margin 50%, customer churn 10%, CAC 100 and a twelve-month horizon, expected contribution is 215.27 and LTV:CAC is about 2.1527 to one. The long-run constant-churn calculation gives 300 and three to one. The gap is the effect of the chosen horizon. Both scenarios assume month-one contribution and fixed subsequent churn, excluding expansion, discounting and changing costs.

Make a review explain the changes

Bring the reconciled MRR bridge, cohort retention and customer churn to the same review, but keep their denominators visible. Investigate concentration: a few large upsells can offset many smaller losses. Compare segments only when their definitions and periods match. Then vary churn, margin and CAC in the finite model to see which assumption drives the planning conclusion. A calculation organizes the evidence; customer-level behavior and collection timing still determine whether the proposed acquisition budget is affordable.

Sources and further reading

• ChartMogul: opening-cohort net and gross retention definitions
• Stripe Atlas: SaaS economics and modeled customer lifetime value

Our editorial standards explain how we check sources, examples and assumptions. Suggest a correction to this article.