NRR vs GRR: Find the Churn Expansion Can Hide

Net revenue retention includes expansion from existing customers; gross revenue retention excludes that expansion. Use both to understand whether a SaaS business is retaining its base and growing the customers who remain. A strong NRR can coexist with meaningful churn, so it should not replace the gross view.

These are revenue metrics, not customer counts or cash receipts. They become useful when the cohort, revenue basis, period, and movement categories are defined consistently. Without those definitions, two dashboards can report different numbers while both appear precise.

Start with one fixed cohort

Choose customers with recurring revenue at the beginning of the period. Keep that cohort fixed through the calculation. New customers acquired during the period belong in a separate growth measure, not in retention of the starting base.

Choose a consistent recurring-revenue unit, such as normalized monthly recurring revenue. Do not mix annual contract value for one customer with monthly revenue for another. Decide how usage-based charges, discounts, paused subscriptions, delinquent accounts, and reactivations are treated.

Stripe's comparison of NRR and GRR explains the distinction between expansion-inclusive and expansion-excluding retention. Your implementation still needs a reconciled movement table rather than blindly copying a formula whose terms may use different definitions.

Use a clear movement bridge

Define starting recurring revenue as S, expansion as E, contraction as C, and churn as L. Under a simple cohort movement model:

NRR = (S + E - C - L) / S
GRR = (S - C - L) / S

Multiply by 100 for percentages. Expansion is growth from members of the starting cohort. Contraction is a reduction among those who remain, while churn removes the relevant recurring revenue under the chosen policy.

If starting revenue is zero, the percentage is not defined. Report that case explicitly instead of substituting 0% or 100%. Also check that contraction and churn do not double-count the same lost revenue.

Work through an example

Suppose an illustrative cohort starts with 100,000 currency units in normalized monthly revenue. Existing customers expand by 20,000, others contract by 5,000, and departed customers remove 10,000.

NRR is 105%: (100,000 + 20,000 - 5,000 - 10,000) / 100,000. GRR is 85%: (100,000 - 5,000 - 10,000) / 100,000.

The cohort produces more revenue overall, yet 15% of its starting revenue has been lost through contraction and churn. If all expansion comes from one large account, the headline also depends on that customer's continued growth. The two measures reveal different operating questions.

Do not add new-customer revenue to make retention look better. That changes the question from retaining an existing base to total business growth.

Investigate the causes, not only the percentage

Split losses into price pressure, product fit, implementation failure, budget cuts, business closure, service quality, and other evidence-supported causes. Keep unknown causes visible. An account owner guessing “budget issue” does not establish the customer's actual reason.

For expansion, distinguish more seats, higher usage, a price change, and adoption of another product. Those sources have different durability and service costs. A price increase can raise NRR even if customer satisfaction or activity declines.

Use the customer concentration guide to consider dependence on large accounts. Retention percentages do not eliminate concentration risk or prove that expanded revenue is profitable.

Keep revenue retention separate from cash

An annual prepayment can improve cash without changing the normalized recurring-revenue movement in the same way. Overdue invoices can leave a good revenue metric alongside a collection problem. Reconcile retention with the receivables and cash views rather than treating one as a substitute.

For usage-based products, a customer's lower activity may count as contraction under the chosen policy even when the account remains active. State the treatment clearly and avoid comparing it directly with a fixed-seat product without explanation.

Currency changes can also affect reporting. Use a documented currency policy and show a constant-currency view where appropriate. Do not attribute exchange movement to customer expansion or churn without separating it.

Compare cohorts and periods carefully

Report customer count, starting revenue, age, segment, and sample size alongside the percentages. A newly acquired cohort and a mature enterprise cohort face different dynamics. A small cohort can swing sharply from one account.

Version changes in metric definitions and, where possible, restate comparable historical periods. A dashboard improvement that reclassifies reactivations can create a discontinuity even when the business did not change.

Turn the pair into actions

High NRR with weak GRR suggests investigating the leaking base, not simply celebrating expansion. Stable GRR with low expansion may point toward adoption, packaging, or product opportunity. Neither pattern determines the answer without customer and margin evidence.

NRR shows how much the existing cohort grows after gains and losses. GRR shows how well its original revenue is preserved. Read them together, reconcile the movement bridge, and use the underlying reasons to decide what to improve.