Gross Revenue Retention (GRR): Formula, Benchmarks & Fixes

Your GRR tells investors one thing before any other metric does: whether your customers believe your product is worth keeping.

‍

Gross Revenue Retention, or GRR, measures the percentage of recurring revenue you hold onto from existing customers over a given period, after accounting for churn and contraction. No expansion revenue included. That means it can never exceed 100% and it means a low number has nowhere to hide.

‍

Most content about GRR treats it as a reporting exercise. Calculate the formula, know the benchmark, move on. That's fine if you're studying for a finance exam but it's not enough if you're a CFO preparing for a Series B or C raise in 2026, where investors are triangulating your GRR from three different data sources to see if your numbers hold up.

‍

This article covers the formula, the stage-by-stage benchmarks, and the operational decisions that actually move the number.

What Is Gross Revenue Retention (GRR)?

GRR has one job: it tells you what percentage of your existing recurring revenue survived the period. Churn and contraction reduce it expansion doesn't count toward it. That's what keeps the ceiling at 100%, and that's also what makes it a cleaner signal than NRR for measuring raw retention.

‍

You'll also see it called Gross Dollar Retention (GDR). Same SaaS metric, different label depending on whether the person across the table is a banker or an operator. Don't confuse it with Logo Retention, which measures the count of customers you keep, not the revenue. A company can have 95% logo retention and 80% GRR if it's losing its biggest accounts disproportionately.

‍

The math is closed by design ytou start with a fixed cohort of customers their revenue either holds, shrinks, or disappears over the period. Any upsell or expansion from that same cohort goes into NRR, not here. That separation is why NRR can run above 100% and GRR never will.

The GRR Formula (And Where the Math Goes Wrong)

The math itself isn't complicated.

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

* Fixed cohort, same customers you started the period with, expansion excluded.

‍

Most CFOs have this memorized. Where companies go wrong isn't the formula it's what they feed into it.

‍

These same churn and contraction inputs also feed your ARR bridge with churn and contraction — see how they reconcile in our ARR vs Revenue guide.

Step-by-Step Calculation

Run it through a real set of numbers. Starting MRR: $1,000,000. During the month, $40,000 churns out entirely and another $20,000 contracts through downgrades.

($1,000,000 − $40,000 − $20,000) ÷ $1,000,000 × 100 = 94% GRR

The same logic holds on an ARR basis. Swap in annual figures, measure over 12 months instead of one. Investors will typically ask for both: monthly for operational monitoring, annual for the fundraising conversation.

The Three Most Common GRR Calculation Errors

Most calculation problems aren't formula problems they're input problems. Three come up constantly.

‍

First: new logo revenue bleeding into the starting cohort. GRR runs on a closed set of existing customers the moment new customers enter the denominator, you're inflating starting MRR and making churn look smaller than it is.

‍

Second: calling a contraction a churn. They're not the same thing churn is a cancellation contraction is a downgrade from a customer who's still paying you. For seat-based and usage-based models, this distinction matters every billing cycle. A customer who drops from 50 seats to 20 didn't leave treating that as churn overstates the problem and muddies the signal.

‍

Third: missed partial-period credits on mid-cycle downgrades. When a customer downgrades partway through a billing month, the credit gets complicated fast. Without ASC 606-compliant rev rec,^[1] most billing systems don't attribute it to the right period. The error is small each month over a quarter, it compounds.

‍

If your billing system and your revenue recognition system don't agree on when revenue was earned, your GRR will differ between your finance team and your CSM platform. That discrepancy is a red flag the moment a diligence team shows up.

GRR Benchmarks: What Investors Expect by Stage

90% is the threshold. For most B2B SaaS companies, that's the floor serious investors want to see before a Series B conversation goes anywhere.^[2] Enterprise-focused companies should target 92–95%. Mid-market SaaS sits at 88–92%. Below 85% at Series B is a fundraising obstacle. At Series C, 90%+ is table stakes, not a differentiator.

‍

* (Benchmarks consistent with SaaS Capital 2025 Retention Benchmarks: median GRR is approximately 93% for higher-ACV segments, 90% for lower-ACV segments.^[2])

GRR Benchmarks by ACV

SaaS Capital’s annual survey of more than 1,500 private B2B SaaS companies^[2] benchmarks GRR by Annual Contract Value (ACV) — the primary segmentation lens in their methodology. Higher ACV correlates with higher GRR,^[2] a relationship driven largely by contract structure: enterprise agreements come with annual terms, auto-renew clauses, and contractual minimums that reduce mid-period exit risk before any customer success conversation is needed.^[3]

‍

Top-quartile performers across segments reach 95%+. That spread matters. Investors aren't just checking whether you clear a threshold they're benchmarking you against your cohort.

Source: SaaS Capital 2025 B2B SaaS Retention Benchmarks.^[2] Figures are approximate values from published charts in the downloadable research brief (registration required at saas-capital.com/research). The "~" notation reflects that exact values are from chart data rather than published tabular data.

Why ACV Determines Your GRR Exposure

A $150K ACV account is much less likely to churn mid-year than a $5K one — partly because the contract won't let it. If your GRR is lagging, your contract architecture is worth examining before your product roadmap.

GRR vs. NRR

Think of them as floor and ceiling. GRR tells you how stable your revenue base is, NRR tells you whether that base is growing. A company with 92% GRR and 115% NRR is losing some revenue to churn but replacing it and then some through expansion. Investors read both numbers together. They're not interchangeable, and a strong NRR doesn't mask a weak GRR.

Why Your Revenue Recognition Process Determines Your GRR

GRR is only as accurate as the revenue data underneath it. Companies that misclassify mid-period downgrades, contract modifications, or usage true-ups will produce GRR figures that don't hold up when someone looks closely. And someone always looks closely.

The Rev Rec → GRR Accuracy Chain

Accurate GRR requires clean separation of two things: revenue that churned and revenue that contracted. Making that separation correctly depends on your billing system passing properly classified revenue events to your accounting system. Most mid-market SaaS companies don't have that handoff working cleanly.

‍

The failure mode looks like this. Your CS platform reports 92% GRR and your finance spreadsheet shows 88%, both teams are confident in their numbers. The difference isn't a math error it's that the CS platform and the finance team are using different definitions of contraction. One treats a seat reduction as contraction the other treats it as partial churn. Neither is necessarily wrong by their own logic but they can't both go into a board deck. This judgment is governed by your revenue recognition method for subscription contracts.

‍

ASC 606^[1] exists, in part, to resolve exactly this. It creates a clear framework for when a revenue modification is a contract modification versus a termination.^[1] Companies with proper ASC 606 workflows don't need a separate GRR reconciliation process. They get clean revenue event classification as a byproduct of compliant accounting.

‍

DualEntry was built for this layer specifically. When your revenue recognition is automated and ASC 606-compliant,^[1] the classification that drives your GRR calculation is the same classification your GL is using. One number. No reconciliation meeting before every board call.

The Investor Due Diligence Red Flag

At Series B and C, investors don't just take your GRR figure at face value. They triangulate it from three sources: the CRM, the billing system, and the general ledger. If all three agree, the conversation moves on. If they don't, it opens a full audit of your revenue recognition practices.

‍

In our experience working with growth-stage SaaS companies preparing for fundraising, GRR reconciliation appears with increasing frequency as a line item in investor due diligence checklists — a shift that reflects how much more sophisticated revenue quality scrutiny has become at Series B and C over the past several years.

‍

GRR is a management KPI, not a GAAP line item. Define your methodology once, document it in your board reporting template, and don't change it without a disclosure note. Customer downgrades that reduce deferred revenue balances are contraction events.^[1] If your rev rec system doesn't classify them that way, your GRR and your deferred revenue schedule will disagree. At Series C and beyond, that disagreement is a diligence event, not an accounting footnote.

‍

f your system is the gap, see our comparison of choosing revenue recognition software for growth-stage SaaS

How to Improve GRR: The Two Levers CFOs Control

CFOs improve GRR through two levers: contract architecture and early warning signals from billing data. Not all GRR loss is equal contraction from a renewing $200K account is a different problem than full churn from a $20K account. Both hurt, but they call for different responses.

Lever 1: Contract Architecture

This is the highest-leverage structural decision available to a CFO, and most companies underuse it. Annual contracts create a structural barrier to mid-period cancellation and downgrade — a customer on an annual term with auto-renew has to actively decide to change course at renewal rather than quietly lapsing.^[3] SaaS Capital's research on contract length found the relationship between term length and overall retention is more nuanced than it might appear, but annual structures do reduce the frequency of mid-cycle contraction events that hit GRR directly.^[3]

‍

For usage-based models, the equivalent lever is contractual minimums. No floor means full contraction exposure every billing period. A customer whose usage drops 40% in Q3 takes your GRR with it, a $10K annual minimum doesn't. The commercial terms are doing the retention work that customer success conversations can't always win. Both levers are automated in DualEntry's automated subscription renewals and contract minimums feature

Lever 2: Early Warning from Billing Data

Contraction signals show up in the billing system 60 to 90 days before renewal. Seat reduction requests. Invoice disputes. Downgrade tickets. These are finance system events before they're CRM events. CFOs with ERP-level visibility into billing data catch them early enough to intervene. Those on spreadsheets, or relying entirely on the CS platform, often don't see them until the renewal comes in short.

‍

The billing system is an early warning system. Most finance teams aren't using it that way.

GRR Decision Logic: A Quick Diagnostic

Where you sit on the GRR spectrum tells you what to fix first.^[2] Below 85%, churn is the primary problem and expansion investment is premature. Between 85–90%, contraction is the leak and contract architecture is the fix. Above 90%, GRR becomes a monitoring metric and the growth work shifts to NRR.

GRR Diagnostic Framework

One thing that gets missed: a company with 88% GRR and 120% NRR still has a problem. The expansion motion is papering over a retention issue. At some point, the expansion ceiling drops and the churn floor becomes visible. Investors at Series C know this. Fix the floor first.

Gross Revenue Retention FAQ