NRR and GRR: Measuring revenue retention in subscription businesses

Net Revenue Retention (NRR) and Gross Revenue Retention (GRR) measure what happens to recurring revenue from an existing customer base over time. NRR includes expansion as well as losses, while GRR excludes expansion and focuses on how much of the starting revenue base remains.

For venture investors evaluating a subscription business, the pair answers two different questions. GRR shows how well the company protects the revenue it already has. NRR shows whether expansion within retained customers is strong enough to offset churn and contraction.

How NRR and GRR are calculated

NRR = (Starting recurring revenue + Expansion − Contraction − Churn) ÷ Starting recurring revenue × 100

GRR = (Starting recurring revenue − Contraction − Churn) ÷ Starting recurring revenue × 100

Both calculations should use the same starting customer cohort. Revenue from customers acquired during the measurement period stays outside the numerator because retention is meant to describe what happened to the customers the company already had.

GRR cannot exceed 100% because expansion is excluded. NRR can exceed 100% when expansion from retained customers is greater than revenue lost through churn and downgrades.

Worked example

Assume a SaaS company starts the year with $2.0 million in ARR from its existing customers. During the year, those customers generate $300,000 of expansion ARR, while $120,000 is lost to downgrades and $180,000 is lost to churn.

NRR = ($2.0M + $0.30M − $0.12M − $0.18M) ÷ $2.0M = 100%

GRR = ($2.0M − $0.12M − $0.18M) ÷ $2.0M = 85%

The company ends with the same amount of recurring revenue from the starting cohort that it began with, so NRR is 100%. The 85% GRR reveals that 15% of the starting revenue base was lost and had to be replaced by expansion within retained customers.

How venture investors interpret retention

NRR is a compact view of how the existing customer base contributes to growth. A business with NRR above 100% can grow recurring revenue from its installed base even before adding new customers. That can make growth more durable and improve the economics of customer acquisition.

GRR exposes the retention quality underneath that expansion. Strong expansion can produce attractive NRR while masking meaningful churn or contraction. Reading GRR alongside NRR helps an investor determine whether growth comes from a durable retained base or from repeatedly replacing lost revenue with larger contracts and upsells.

The cohort and segment matter. Enterprise customers, SMB customers, geographies, products, and contract types can have materially different retention behavior. A blended company-wide percentage can hide where the actual strengths and weaknesses sit. Cohort retention analysis provides a deeper view when enough customer history exists.

What does a good NRR or GRR look like?

For annual NRR, 100% means expansion exactly offsets churn and contraction in the starting cohort. Above 100%, expansion exceeds these losses. Annual NRR scenarios of 110% or 120% can illustrate expansion-led growth, but are not sourced market benchmark cutoffs. The right benchmark still depends on customer size, pricing model, and how much expansion is structurally available.

GRR should be read as the durability of the starting revenue base before expansion. Illustrative annual GRR of 90% means 10% of starting revenue was lost before expansion; annual GRR of 85% means 15% was lost. Evaluate these examples in the context of segment, renewal frequency and customer mix.

The two metrics are most useful together. Strong NRR with weak GRR can indicate that a smaller set of expanding accounts is compensating for substantial churn elsewhere.

Common mistakes

Including new customers

Retention should follow the customer base that existed at the start of the period. Adding revenue from newly acquired customers turns the metric into a broader growth measure.

Confusing customer retention with revenue retention

Logo retention measures how many customers remain. Revenue retention measures how much recurring revenue remains. Losing one large account can have limited impact on logo retention and a large impact on GRR.

Using inconsistent periods or cohorts

Monthly, quarterly, and annual retention can tell different stories. The same cohort definition and measurement window should be used when comparing periods.

Looking only at NRR

NRR can look healthy because a subset of customers expands rapidly. GRR provides the complementary view of how much recurring revenue the business loses before expansion.

Limitations

NRR and GRR work best for recurring-revenue models with a clearly defined starting customer base and consistent revenue measurement. Usage-based businesses can still use retention analysis, but revenue can move because customer usage changes rather than because a subscription expands or contracts in a conventional sense.

Both metrics are backward-looking summaries. They do not explain why customers churned, which segments are deteriorating, whether expansion is sustainable, or how much it cost to retain and grow the accounts. Venture investors usually need cohort-level and segment-level context before treating a headline retention figure as representative of the whole business.

For each review, record the starting cohort, annual or monthly measurement period, recurring-revenue policy and dated source file. Treto's portfolio workspace can preserve these inputs across company updates so the team can review changes on a consistent basis.