Read NRR with GRR

Net revenue retention compresses base retention, contraction, and expansion into one figure. That summary is useful, but it cannot show whether expansion is broad or concentrated, or why customers remained, contracted, or left.

The same NRR, two different companies

Consider two hypothetical cohorts. Company A retains 97% of starting revenue and adds expansion equal to 11%; Company B retains 84% and adds 24%, concentrated in three accounts. Both report 108% NRR under the same simplified formula. Their base erosion and concentration risk differ, but these retention measures alone do not establish which business is healthier or what caused either result.

GRR makes the base-retention difference visible because it excludes expansion and is capped at 100%. It answers a narrower measurement question: how much starting recurring revenue remained before expansion. It does not identify whether product value, customer mix, service, pricing, budget pressure, or contract timing drove the change.

Same cohort, or the pair is meaningless

The pair decomposes cleanly only when both measures use the same cohort, entity grain, revenue basis, currency treatment, and recognition points. A common mismatch is customer success computing from CRM accounts while finance computes from billing contracts. The output is then two valid measures that should not be subtracted from each other.

• —Fix the cohort basis: customer or contract, defined once.
• —Fix the revenue basis: ARR at a point in time, or recognized revenue over the period.
• —Fix the recognition point for churn: notice, term end, or final service date.
• —Fix the treatment of downgrade-then-expand within the period.

What to report

For operating review, report starting cohort value, churn, contraction, expansion, ending value, GRR, and NRR on one bridge. The bridge localizes movement; it does not explain causality. Use account evidence, product usage, service records, pricing changes, and customer research to investigate why the components moved.