Skip to content

Retention

Gross Revenue Retention

What is Gross Revenue Retention?

Gross revenue retention (GRR, also called gross dollar retention) is the share of a cohort's starting recurring revenue still present at the end of a period, counting cancellations and downgrades but excluding all expansion. Because expansion is excluded, GRR can never exceed 100%, which makes it the only retention figure a strong upsell quarter cannot flatter.

Formula

GRR = (Starting MRR − Contraction − Churn) ÷ Starting MRR × 100

Starting MRR
MRR of the customer cohort active at the start of the period
Contraction
MRR lost to downgrades and seat reductions on customers who remained
Churn
MRR lost to full cancellations within the cohort

Worked example

The same April cohort used in the net revenue retention example, with expansion removed from the calculation.

Step-by-step calculation of Gross Revenue Retention
StepValue
1Starting MRR$42,500
2Contraction MRR$900
3Churned MRR$1,250
4Retained MRR$40,350
5Gross revenue retention94.9%
6Net revenue retention (for contrast)102.2%
7Annualised GRR (0.949 ^ 12)53.0%

Result

94.9% monthly gross retention sounds close to intact. Compounded across a year it means barely half of today's revenue base survives untouched, and the 102.2% NRR that looked reassuring is entirely a function of expansion outrunning that decay.

The metric expansion cannot rescue

GRR has one structural property that makes it valuable: it is capped at 100%. No upsell, no seat expansion, no price increase can push it higher. Whatever it reports is pure leakage from the base you started with, which is exactly what NRR is capable of concealing.

Read together, the two decompose the base into its parts. GRR is how much of the bucket stays full on its own; NRR is what happens after you pour expansion in. A company at 90% gross and 115% net is running a very effective expansion motion over a badly leaking base — a real business, but one whose growth depends on continuing to outrun the leak. A company at 98% gross and 105% net is far more durable at a lower headline number.

Compounding is brutal at monthly cadence

Monthly gross retention percentages read reassuringly and compound viciously. 95% monthly is 54% annually. 97% is 69%. 99% is 89%. The difference between a 95% and a 99% month — four points that would barely register in a board discussion — is the difference between keeping half your revenue base for a year and keeping nearly all of it.

This is why GRR is more commonly reported on a trailing-twelve-month basis, where the compounding has already happened and the number means what it appears to mean. If you report it monthly, state the annualised equivalent alongside, or the figure will be read as far healthier than it is.

Contraction is the half people forget

GRR counts downgrades, not just cancellations, and a business can post excellent logo retention while quietly bleeding gross revenue through seat reductions. That pattern — customers staying but shrinking — is characteristic of a product that survived a budget review without winning it, and it is invisible in any metric that only counts cancellations.

Because contraction and churn are the only two inputs, the boundary between them does not affect GRR at all: a customer who drops to a free tier reduces retained MRR by the same amount whichever bucket it lands in. This makes GRR unusually robust to the classification arguments that distort churn rate, and it is a good reason to reach for it when two teams cannot agree on definitions.

Where GRR is the wrong tool

GRR is deliberately blind to upside, so it says nothing about whether a product deepens in an account. For a business whose entire model is land-and-expand, judging it on GRR alone would condemn a strategy that is working as designed. And in a self-serve business with a small, fixed price point, GRR and logo churn converge to nearly the same number, so reporting both adds a row without adding information.

The pairing that carries the most signal is GRR with NRR, on the same cohort and the same window, with the gap between them stated explicitly. That gap is the size of your expansion motion, and it is a number very few companies report directly.

Where GRR goes wrong

  • Reading monthly GRR without annualising it. 95% a month is 54% a year; four points of monthly retention separate a durable base from one that halves annually.
  • Allowing any expansion into the calculation. GRR above 100% is not a strong result, it is a definitional error — usually reactivations or upgrades leaking into the retained figure.
  • Counting only cancellations and ignoring downgrades. Customers who stay but shrink are the pattern GRR exists to catch, and omitting contraction removes the metric's main advantage.
  • Judging a land-and-expand business on GRR alone. The metric is blind to upside by design, so it will report a working expansion strategy as mediocre retention.
  • Comparing your GRR to a competitor's without matching the window. Monthly and trailing-twelve-month figures for the same business differ by tens of points.

Typical ranges

Investor commentary on B2B SaaS commonly treats gross revenue retention in the high 80s to low 90s (trailing twelve months) as acceptable for SMB-focused businesses and above roughly 90% as the expectation for enterprise, where contracts are longer and switching costs are higher. Self-serve products typically sit lower and are not usefully compared to either. These are circulated conventions rather than study findings — the internal comparison, GRR against your own prior cohorts on a fixed window, is the one that carries information.

Source: SaaS investor convention, not an empirical study

Related

Metrics that move with this one

No metric explains a business on its own. These are the figures that qualify, offset or explain GRR.

Net Revenue Retention (NRR)

Net revenue retention (NRR, also called net dollar retention) is the recurring revenue a fixed group of existing customers generates at the end of a period, expressed as a percentage of what the same group generated at the start, including expansion and after churn and contraction. New customers are excluded entirely. Above 100% means the existing base grew on its own; it does not mean customers are staying, because heavy expansion from a few accounts can cover substantial churn among the rest.

Learn more

Revenue Churn

Revenue churn is the share of recurring revenue lost from existing customers over a period. Gross revenue churn counts cancellations and downgrades against starting MRR and can never be negative; net revenue churn subtracts expansion from those losses and can go below zero when upgrades from surviving customers outweigh everything lost. Neither version includes revenue from new customers.

Learn more

Retention Rate

Retention rate is the share of customers or revenue from the start of a period that is still present at the end, calculated as 100% minus the churn rate over the same period and definition. Customer retention rate is bounded at 100%, while net revenue retention can exceed it, so the two are not interchangeable despite both being described as retention.

Learn more

Contraction MRR

Contraction MRR is recurring revenue lost from customers who stayed but now pay less — downgrades to a cheaper plan, removed seats, dropped add-ons and newly applied discounts. It is distinct from churned MRR, which comes from customers who left entirely, and keeping the two separate matters because a shrinking account is a retained relationship with a product or pricing problem, while a churned one is gone.

Learn more

Logo Churn

Logo churn is the share of customer accounts lost over a period, counting each account once regardless of what it paid. It is the customer-count view of churn, and comparing it to revenue churn reveals whether the accounts leaving are larger or smaller than average — the two rates diverging is usually more informative than either level on its own.

Learn more

Cohort Analysis

Cohort analysis groups customers by when they started and tracks each group separately over elapsed time, producing a triangular table where rows are signup periods and columns are months since signup. It exposes what an aggregate churn rate cannot: whether retention is improving for newer customers, where in the lifecycle customers leave, and whether a flat headline number is hiding a deteriorating base propped up by durable older cohorts.

Learn more

GRR: frequently asked questions

What is the difference between GRR and NRR?

Gross revenue retention excludes expansion, so it caps at 100% and measures only how much of the starting revenue base survived. Net revenue retention includes expansion and can exceed 100%. The gap between them is the size of your expansion motion — and because NRR nets expansion against churn, it is the only one of the two that can hide a leak.

Can gross revenue retention be above 100%?

No. GRR excludes expansion by definition, so the highest possible value is 100%, reached only when a cohort loses nothing at all to cancellations or downgrades. A GRR above 100% always indicates an implementation error, most often reactivations or upgrades being counted as retained revenue.

What is a good gross revenue retention rate?

For enterprise SaaS, above roughly 90% on a trailing-twelve-month basis is the common expectation; SMB-focused businesses are generally accepted in the high 80s to low 90s, and self-serve products typically sit lower still. These are conventions rather than measured benchmarks. State the window whenever you quote one, because monthly and annual figures differ enormously.

Does gross revenue retention include downgrades?

Yes. Both full cancellations and downgrades reduce GRR, which is a large part of its value: a business can retain nearly every logo while losing significant revenue to seat reductions and plan downgrades, and a metric counting only cancellations would show nothing wrong.

Stop recalculating GRR by hand.

Connect Stripe and Bastle computes this metric — and the rest of the glossary — against your own billing history, with the definitions under your control. Free while in beta, no card required.