I'mBoardDocs
Board OntologyCustomers

Gross Revenue Retention (GRR)

Recurring revenue retained from the cohort of customers present at the start of the period, excluding expansion — so the metric captures only churn and contraction. Per the SaaS Metrics Standards Board (SMSB) GRR standard. GRR is bounded at 100% (cannot exceed it) and reads as the "no-defense-against-churn" floor on retention. The board reads GRR alongside NRR (`customers.net_revenue_retention`) — the gap between them is the expansion contribution. Common pitfall: treating GRR and NRR as substitutes — they answer fundamentally different questions, and a healthy NRR with sliding GRR signals churn masked by upsell. — Customers KPI anchored to SaaS Metrics Standards Board.

Rogue ID: customers.gross_revenue_retention Type: Percentage (%) Domain: Customers

Definition

Recurring revenue retained from the cohort of customers present at the start of the period, excluding expansion — so the metric captures only churn and contraction. Per the SaaS Metrics Standards Board (SMSB) GRR standard. GRR is bounded at 100% (cannot exceed it) and reads as the "no-defense-against-churn" floor on retention. The board reads GRR alongside NRR (customers.net_revenue_retention) — the gap between them is the expansion contribution. Common pitfall: treating GRR and NRR as substitutes — they answer fundamentally different questions, and a healthy NRR with sliding GRR signals churn masked by upsell.

Formula

GRR = (Starting ARR − Contraction − Churn) ÷ Starting ARR, on the cohort active at period start. Excludes expansion. Capped at 100% by definition. Per SMSB GRR standard.

Why it matters

Isolates the "do customers stay and not shrink" signal from expansion noise. GRR is the true downside floor on retention — boards use it to spot product or onboarding deterioration that NRR can hide.

How to interpret

Per KBCM/Sapphire Private SaaS Company Survey 2024 (§ "Gross Dollar Retention"), private SaaS GRR medians typically sit in the high-80s to low-90s, with top-quartile in the mid-90s — distributions vary by ACV cohort and vintage, so pull the current edition. The NRR − GRR gap quantifies expansion contribution; a widening gap with declining GRR is a yellow flag (expansion masking churn). Trend it quarterly — a single-period drop with steady NRR usually means a one-off contraction; persistent decline with flat NRR is a product issue.

Calculation policy

How an AI agent should compute this KPI from messy company data. Free-text rules consumed at reasoning time — not a deterministic DSL. The most common ways to get this wrong are listed under Common miscomputations.

Inclusion rules

  • Cohort = customers active at the start of the period (same cohort as NRR).
  • Numerator: Starting ARR − Contraction − Churn. Expansion is excluded by definition.
  • Capped at 100% — if a calculation produces > 100%, expansion is leaking in.

Exclusion rules

  • Expansion ARR, cross-sell ARR, price increases — none of these enter GRR (that is exactly what distinguishes GRR from NRR).
  • New-logo ARR (same exclusion as NRR).

Required inputs

  • Customer-level ARR snapshot at period start (cohort identity).
  • Customer-level ARR snapshot at period end for the same cohort.
  • Classification of each per-customer change as contraction vs churn vs (excluded) expansion.

Data-source priority

  • Same cohort-tracking source as NRR; the two metrics share inputs.

Edge cases

  • Customer with both expansion and contraction in the period: zero out the expansion, keep the contraction.
  • Customer renaming or re-keying: hold cohort identity by account, not by name.

Validation checks

  • GRR ≤ 100% always; > 100% means expansion is being counted (recompute with strict exclusion).
  • GRR ≤ NRR for the same cohort; the gap is the expansion contribution.

Common miscomputations

  • Letting expansion in — produces GRR > 100%, which is definitionally impossible; the most common GRR error.
  • Reporting GRR alone without NRR — boards always want the pair to see whether expansion is masking churn.
  • Conflating GRR with logo retention — GRR is dollar-based; logo retention is count-based; they answer different questions.
  • Using a different cohort base than NRR — the two must use the same starting cohort to make the gap meaningful.
  • customers.net_revenue_retention
  • customers.logo_retention_rate
  • customers.logo_churn_rate
  • customers.churn_risks
  • sales.arr

Source

SaaS Metrics Standards Board · section: GRR — published 2023-01-01.

Why does this cite SaaS Metrics Standards Board? Read the ontology methodology for the published vs editorial tier system, attribution rules, and dispute process.

Metric definitions reference standards published by the SaaS Metrics Standards Board (saasmetricsboard.com). imboard is not affiliated with, endorsed by, or a member of SMSB.

Industry benchmark

A reference distribution sourced from KBCM/Sapphire SaaS Survey 2024 (15th Annual) (2024):

PercentileValue
25th82%
Median91%
75th95%

Higher is better.

Stage relevance

Company stagePriority
Series ACore
Series BCore
Series C+Core
PublicCore

Suggested for stages: Series A, Series B, Series C+, Public.

Default owning functions

  • Finance
  • Sales

Machine-readable

On this page