Net Revenue Retention: How to Calculate It, With an Example

Follow these steps for a 12-month window:

1. Pick the cohort. List every customer with recurring revenue on the start date, such as January 1 last year.
2. Add up their starting recurring revenue. Use monthly recurring revenue times 12, or annual recurring revenue, consistently.
3. Look at the same customers today. Find each one's recurring revenue on the end date.
4. Break down the change: expansion (upgrades, added seats, cross-sell), contraction (downgrades) and churn (customers who left).
5. Apply the formula: (starting revenue + expansion − contraction − churn) ÷ starting revenue.
6. Exclude anyone who became a customer after the start date, even if they've grown since.

For example, say your starting cohort paid $1,000,000 in annual recurring revenue. Suppose that over the year, existing customers expanded by $180,000, downgraded by $40,000 and churned $110,000. As an example, that gives ($1,000,000 + $180,000 − $40,000 − $110,000) ÷ $1,000,000, so NRR is 103 percent.

Gross revenue retention (GRR) ignores expansion. It's (starting revenue − contraction − churn) ÷ starting revenue, and it can't exceed one hundred percent. For example, with the numbers above, GRR would be 85 percent, since expansion isn't counted.

The two numbers tell you different things. NRR shows whether the customer base as a whole grows without new sales. GRR shows how much revenue you keep from what you had. A company can post a healthy NRR because a few big accounts expanded, while GRR reveals that many smaller customers left. Track both, and if they diverge, look at which segments are churning and which are expanding.

Logo churn, the share of customers that left, is a third view. It can move opposite to revenue churn if the customers leaving are your smallest.

Retention and expansion compound. In one 2025 SaaS Capital survey, median ARR growth was 15 percent for companies with NRR below 90 percent and 44 percent for those with NRR above 130 percent¹. The direction is what's useful here: companies whose existing customers grow tend to grow faster overall, because new sales sit on top of a base that isn't shrinking.

That doesn't mean NRR causes growth in every business. A high NRR on a tiny cohort says little, and it can be driven by a single large customer. Use the number alongside the size of the cohort and the share of revenue from your top accounts. For benchmarks by stage, see NRR benchmarks by growth stage.

Check your calculation for these errors:

• Including new customers. Anyone who joined after the start date belongs to the next cohort.
• Mixing time windows. Don't divide a 12-month change by a monthly starting figure. Keep the window and the base consistent.
• Counting price increases as expansion without labeling them. They're valid revenue, but they hide whether customers are actually buying more. Show them separately.
• Including one-time or non-recurring revenue. Implementation fees and one-off services don't belong in recurring revenue.
• Ignoring usage-based volatility. For usage pricing, decide whether you use committed or actual revenue and stay consistent.
• Leaving out reactivated customers or treating them inconsistently. Decide the rule once.

Write your definition down and apply it to history, so trends are comparable.

Split the number into its parts and work on the weakest:

• Churn is the biggest drag: interview customers who left, then fix the causes that repeat, such as poor onboarding or weak early results.
• Contraction is high: review pricing and packaging, and check whether customers are downgrading because features go unused.
• Expansion is low: identify accounts ready for more seats or products and assign an owner for expansion, as laid out in the CRO playbook for NRR expansion.

Customer success platforms such as ChurnZero and Gainsight combine health scores and renewal tracking for larger customer bases. ChurnZero fits teams that want to monitor account health and act on expansion signals. Gainsight fits larger teams with more complex renewal and success workflows. See Gainsight vs ChurnZero vs Salesforce to compare. Related calculations include sales velocity and pipeline coverage.