GRR vs Churn Rate: Are They the Same Number?

Is Gross Revenue Retention Just One Minus Churn Rate?

Only when the churn rate measures revenue, includes contraction, and covers the same period and the same starting base. Meet those three conditions and GRR plus revenue churn equals 100%. Miss any one of them and the identity breaks.

The conditions get missed constantly, because most companies keep several churn numbers and label them all churn. Logo churn counts customers. Revenue churn counts dollars. Some revenue churn definitions include downgrades and some count only full cancellations. Each produces a different complement to GRR, and only one of them is the right one.

So the honest answer is that GRR and churn rate are two views of the same event, expressed from opposite ends. GRR states what stayed. Churn states what left. The arithmetic ties only when both are measuring the same thing.

How Do You Calculate Each One?

GRR is starting ARR minus cancellations and downgrades, divided by starting ARR, with expansion excluded. Revenue churn is those losses divided by the same starting ARR.

Work an example. A company begins the year with $12 million in ARR. Over the year, $900,000 of it cancels outright and $300,000 contracts through downgrades and seat reductions. Expansion inside the base adds $1.8 million.

GRR is $12 million minus $1.2 million, over $12 million, which is 90%. Revenue churn is $1.2 million over $12 million, which is 10%. They sum to 100% because both used dollars, both included contraction, and both used the same denominator. Net revenue retention on the same base is 105%, since expansion counts there and nowhere else.

Now swap in logo churn. If 40 of 500 customers left, logo churn is 8%, and 100% minus 8% is 92%, which is not the 90% GRR. Neither number is wrong. They just are not complements.

Where Does the Equivalence Break?

Four mismatches account for nearly every case where the two numbers refuse to reconcile. The last row is the one that gets past reviews. A denominator built from average ARR over the period instead of ARR on day one quietly includes revenue that was never at risk of churning. Retention improves, churn drops, and nothing about the business changed. Fix the denominator to the opening balance and both numbers become comparable across periods.

Which One Should You Report?

Report gross revenue retention externally and break out churn by cause internally. They serve different readers.

GRR is the number investors and boards ask for, because it is bounded, it is comparable across companies, and it reconciles to an ARR waterfall. At ORM the retention view is a monthly reconciling waterfall that runs from beginning ARR through churned customer ARR, churned product ARR, product decrease ARR, and the expansion lines, to ending ARR, with gross and net retention sitting on the same chart. Every point of GRR traces back to a line item, which is what makes it defensible when someone asks where the drop came from.

Churn is the operating number, and it is only useful when it is decomposed. Churn caused by a failed implementation, churn caused by a champion leaving, and churn caused by a budget cut are three different problems with three different plays. A single company-wide churn percentage tells a customer success team nothing about which one they are facing this quarter.

How Does Each Feed the Forecast?

Churn rates drive the cohort math inside a forecast, and GRR is the output you check that forecast against. You model the loss, then you validate the retention.

Cohort decay is where churn rates do the work. Apply a monthly churn rate to each cohort of the base, let it compound, layer expansion on top, and you have a base revenue forecast that responds to how long a customer has been with you. Retention almost always improves with tenure, so a single blended rate applied to the whole base overstates losses in the mature cohorts and understates them in the newest.

The forecast gets sharper when the churn rate itself is predicted rather than assumed. One of the strongest early signals we see at ORM comes from support cases. Customers with no support cases at all are at risk, and customers with seven or more in the last year are at risk. The safe zone is three to five cases, usually tier two or three, because those customers are engaged and getting help. That pattern turns a lagging churn number into an input you can act on months before a renewal date, and it belongs in the same model that produces your revenue forecast rather than in a separate customer success spreadsheet. Once churn is predicted at the account level, net revenue retention becomes a forward number instead of a report on a quarter that already happened.