Revenue retention curves: GRR vs NRR over time (how to model)

A practical guide to revenue retention curves: how GRR and NRR compound, how to interpret expansion vs churn, and how to avoid common mistakes.

Updated 2026-01-28

Try it in a calculator

Why retention curves matter

NRR and GRR are often reported as a snapshot, but the compounding effect over 12-24 months is what drives growth quality. A retention curve makes that compounding visible and helps you see which lever matters: expansion vs churn/contraction.

GRR vs NRR (quick recap)

  • GRR excludes expansion; it answers how leaky the bucket is after churn and downgrades.
  • NRR includes expansion; it can exceed 100% when upgrades outweigh churn/contraction.
  • Both can be true: strong NRR can hide weak GRR (expansion masking churn).

Modeling approach (simple monthly compounding)

  • Start with cohort MRR.
  • Apply churn and contraction to get GRR.
  • Apply expansion (and contraction/churn) to get NRR.
  • Compound monthly to see 12-24 month outcomes.

Common mistakes

  • Using blended averages across segments (plan/channel) and hiding weak cohorts.
  • Mixing time units (annual NRR used as monthly rates).
  • Confusing logo churn with revenue churn (different denominators).

FAQ

What's a good GRR-
It depends on segment and stage, but GRR is a 'leakiness' metric: higher is better. Track GRR over time and by segment; improvement is often driven by product quality and customer success.
Can I forecast growth using NRR alone-
NRR is only the existing base. Overall growth also depends on new customer MRR. Use an MRR forecast model that combines new MRR with retention and expansion.

More in saas metrics

NRR/GRR targets: how to translate targets into expansion and churn goals
ROAS: What it is and how to use it