Gross Revenue Retention (GRR)

Gross revenue retention measures how much of your existing recurring revenue you keep over a given period, after customers cancel or cut back, and before any expansion is added in. For a subscription business, it’s one of the clearest signals of whether the revenue you’ve already won is staying with you.

The number itself is easy to calculate. Reading it well is harder, and that’s where the value is: knowing what your GRR is telling you, what drives it, and how to move it as your Customer Success team matures.

What Gross Revenue Retention Really Tells You

The value of gross revenue retention comes from what we exclude from the calculation.

By excluding upsells, cross-sells, seat growth, and price increases, it shows the health of your existing revenue base with the effects of growth stripped away. Two things can pull it down: churn, where a customer leaves entirely, and contraction, where a customer stays but pays less. Nothing on the growth side gets to compensate for either.

That exclusion is why GRR can never exceed 100%. The best you can do is keep everything you started with. A result of 92% means 92 cents of every dollar stayed and 8 cents left through churn or contraction.

It also makes GRR more revealing than the metrics sitting next to it. Net revenue retention (NRR) runs the same calculation but adds expansion back in, so it can climb above 100% even while customers are leaving, because growth from your strongest accounts hides losses among the rest. Logo retention counts customers instead of dollars, so a lost $5,000 account and a lost $500,000 account weigh the same. GRR weights by revenue, which means a few large departures show up clearly even when your customer count looks stable.

Read together, the three answer different questions:

• Logo retention: how many customers stayed.
• Gross revenue retention: how much recurring revenue stayed.
• Net revenue retention: whether your existing base grew overall.

When NRR looks strong but GRR is slipping, expansion is masking a retention problem that surfaces eventually.

How to Read Your GRR

There is no single figure that counts as a good GRR. The right target depends almost entirely on who you sell to.

Smaller customers churn more by nature: budgets shift quickly, more of them go out of business, and switching costs are low. A GRR in the low-to-mid 80s can be healthy for an SMB-focused product. Enterprise works differently, with multi-year contracts and deeper implementations that push retention into the mid-90s, where anything below 90% is worth investigating. The benchmarks table on this page sets the targets for your segment, so read your result against that row rather than a universal ideal.

Two habits matter more than the raw number.

Watch the trend, not the snapshot. A steady 88% across six quarters describes a predictable base you understand. The same 88% sliding down from 93% is an early warning the rest of the business hasn’t felt yet.

Distrust the aggregate. An overall 94% can hide a recent cohort retaining at 82%, with mature accounts covering the gap. The headline tells you the base is healthy on average. It says nothing about where the soft spots are, which is why cohort and segment views belong next to any retention number you report.

What Moves GRR

Only two events lower the number directly, and they call for different responses.

Churn is a full departure. Contraction is a customer who stays but spends less, through a downgrade, fewer seats, or a dropped module. Teams often track the first closely and let the second slip, which quietly overstates retention. A customer who halves their contract hasn’t churned, but they have taken a real bite out of GRR.

The causes underneath those outcomes are where the leverage is.

Value realization is the big one. Customers who reach a clear, repeated payoff from the product rarely leave; those who never quite get there are the first to cut when budgets tighten. Onboarding feeds straight into this, since accounts that stall in their first ninety days tend to resurface as churn at renewal. Adoption depth matters too: software woven into daily workflows, with integrations and several users depending on it, is far harder to remove than a tool sitting at the edge of the business. Relationship coverage carries weight in larger accounts, where a single champion leaving can put the whole contract at risk.

One driver gets overlooked more than the rest: failed payments. A share of churn is never a decision at all. Expired cards, declined charges, and billing errors cancel customers who fully intended to stay. This involuntary churn is among the most fixable causes of a sagging GRR, because the customer still wants the product. The payment just needs to go through.

How to Improve Gross Revenue Retention

The right moves depend on the maturity of your CS team. Work them in stage order instead of attempting everything at once.

Crawl: Build a Baseline You Trust

The first job is a number you believe.

Many early teams overstate retention by leaving out downgrades, misclassifying late renewals, or working from incomplete billing data. Get the measurement right before you try to move it. The Establish a Retention Baseline playbook walks through building that figure from your billing records.

This is also the moment to split churn into voluntary and involuntary. The involuntary slice, the failed payments and expired cards, is often the cheapest retention you will ever recover. Basic payment hygiene, automated retries and card-update reminders, can lift GRR with no change to product or pricing.

Walk: Catch Risk Before the Renewal

Once the reporting is solid, the work shifts to spotting trouble early enough to act.

Customer health scoring gives you a leading signal of which accounts are drifting, so you can step in around month three rather than discover the problem in month eleven. The Implement Health Scoring playbook covers standing that up. Pair it with a deliberate renewal and save motion, so at-risk accounts get worked on a schedule instead of rescued in a scramble; the Build a Renewal and At-Risk Save Motion playbook lays that out. Segmentation begins to matter here too, since a renewal play built for enterprise rarely fits a self-serve SMB book.

Run: Find the Weak Spots Under a Healthy Average

By this stage GRR should be stable and high, and improvement comes from understanding the detail rather than chasing the headline.

Break the number down by signup quarter, plan, industry, acquisition channel, and customer size, and look for the pockets dragging quietly beneath a healthy average. Predictive customer intelligence, covered in the Build a Predictive Customer Intelligence Model playbook, extends this by flagging churn risk across a large base long before a manual review would catch it. The goal shifts from raising GRR to protecting it and knowing exactly what holds it up.

An Example Scenario

Picture a company reporting 91% GRR for the year. Comfortable territory, and easy to feel fine about.

Then they break it down by signup cohort. Customers who joined eighteen or more months ago are retaining at 96%. The cohort from the last two quarters is retaining at 82%. The blended 91% looked stable, but it was hiding a real shift: something about how recent customers are sold to, onboarded, or activated has gotten worse, and the strong older cohorts were covering for it.

The calculator on this page would have told you the company is at 91%. The judgment is in noticing that one healthy-looking number held two very different stories, and that the newer, more fixable one deserves attention now. That is the difference between calculating GRR and reading it.

Nuances and Edge Cases

Cohort versus aggregate trips up more teams than anything else here. An aggregate GRR pools new and mature customers together and smooths over the exact trends you want to see. A cohorted GRR follows a defined group from a fixed starting point, so this spring’s signups can be compared against last spring’s. For spotting trends, cohort wins almost every time.

Voluntary versus involuntary churn deserves its own line in reporting. The two look identical in the formula, since a lost dollar is a lost dollar, but they need opposite responses. Voluntary churn is a value and fit problem for product and CS. Involuntary churn is an operational problem for billing. Combine them and you bury the cheapest win you have.

Cadence and seasonality shape how the number should be read. An annual GRR is a rear-view mirror; by the time it moves, the cohort that caused it churned months ago. Monthly or quarterly cohort GRR surfaces the same trend while you can still act on it. Renewal-heavy quarters are worth watching too, since they can make a single period look alarming out of context.

The expansion boundary is worth guarding. The moment any upsell or seat growth slips into the calculation, you are measuring NRR, not GRR. Keeping that line clean is what makes GRR worth tracking as its own number.