What Is a Good Gross Revenue Retention Rate?

What Is a Good Gross Revenue Retention Rate?

GRR is capped at 100%, so the only useful target is the smallest gap below 100% your contract mix can sustain. Unlike net retention, gross retention has no upside. It counts the ways revenue leaves and nothing else, which makes it the cleanest read on whether customers keep paying for what they already bought.

That cap changes how you should read the number. A 12-point gap below 100% means you replace 12% of your base every year before growth starts. On a $20 million ARR base, that is $2.4 million of new revenue spent standing still, and it competes for the same pipeline as your growth number.

Published GRR benchmarks fail for a structural reason. A company selling annual enterprise contracts measures losses at renewal dates clustered in specific months. A company selling monthly SMB subscriptions measures losses continuously. Identical customer satisfaction produces different GRR curves. Compare yourself to your own trailing twelve months before you compare yourself to anyone.

How Should You Read the Gap Between GRR and NRR?

The gap tells you how much expansion is covering for retention losses, and a wide gap is a warning even when NRR looks strong. The third row is the dangerous one. A team reporting 118% net retention feels safe until you learn gross retention sits at 82%, which means expansion worth 36% of beginning ARR is covering an 18-point loss in the base. Expansion concentrates in a few accounts, and when one of those accounts leaves, both numbers move at once. Reporting net revenue retention without gross retention beside it is how boards get surprised.

Which ARR Movements Belong in the Calculation?

Only the contraction movements, measured against beginning ARR, with expansion excluded entirely. At ORM we run gross and net retention off the same monthly waterfall, where beginning ARR always equals the prior month's ending ARR.

The waterfall has six movements between the two endpoints:

- Churned customer ARR. The customer left completely. Contraction. - Churned product ARR. The customer stayed but dropped a product. Contraction. - Product decrease ARR. The customer kept the product at lower value, usually fewer seats. Contraction. - New customer ARR. Expansion. - New product ARR. Expansion. - Increased product ARR. Expansion.

Gross retention uses the first three. Net retention uses all six. The reconciling requirement is what makes the waterfall trustworthy, because every dollar of movement has to land in exactly one bucket and the endpoints have to tie.

Most retention disputes inside a company are really classification disputes. A customer who drops two products and adds one has produced churned product ARR and new product ARR in the same month. Netting them into a single line erases the churn and inflates GRR.

What Does a Single GRR Number Hide?

Mix. A blended GRR moves whenever the composition of your base moves, with no change in any customer's behavior. Three cuts expose it.

Cut by ACV band first. Small accounts churn at higher rates almost everywhere, so a quarter that added many small logos will drag blended GRR down while enterprise retention held. Cut by cohort second, because a customer who has been with you three years behaves differently from one who signed last quarter, and a growth spurt loads your base with young accounts. Cut by product third, since a single underperforming product can carry the entire gap.

The fourth thing a blended number hides is concentration. If your top ten accounts represent 40% of ARR, your GRR is a bet on ten renewal conversations. The rate reads as a statistical property of a large base when it is actually the outcome of a small number of decisions. Report GRR alongside the share of ARR concentrated in the top ten accounts, or the number tells the board something it does not mean.

What Is the Earliest Signal That GRR Will Fall?

Support ticket volume at both extremes, checked before the renewal date enters the window. The pattern in ORM customer data runs against intuition.

A customer filing no support cases is at risk. Silence usually means low usage, and low usage means nobody inside the account will defend the line item when budgets get reviewed. A customer filing seven or more cases in a year is also at risk, for the obvious reason. The healthy middle sits at three to five cases a year, typically tier two or tier three severity rather than severe outages. Those customers are engaged, getting help, and generally happy.

Build the signal into a report rather than a dashboard nobody opens. Rank every account by ticket count over the trailing twelve months, flag the zero-ticket accounts and the seven-plus accounts, and route both lists to customer success before the renewal is 90 days out. Usage telemetry sharpens it further, but ticket count works with data every company already has.

How Do You Forecast GRR Instead of Reporting It?

Forecast the renewal base by month, apply retention rates by segment and cohort, then reconcile the result against the waterfall. Reporting GRR tells you what already happened. Forecasting it tells you which quarter has a problem while there is still time.

Start with the contract base. Every renewal has a date, so the renewal exposure for each of the next twelve months is a known quantity today. Apply segment-level retention rates to each month's exposure rather than one blended rate, since a month heavy with SMB renewals carries different risk from a month heavy with enterprise.

Then treat the at-risk accounts individually. Below a certain number of renewals, statistical rates stop being useful and account-level judgment takes over, the same way large deals get handled in a new business forecast. Feed the output into the same model that produces your revenue forecast, because renewals and new business compete for the same coverage and the board sees one number.