Gross Revenue Retention (GRR) is the percentage of recurring revenue kept from an existing customer base over a set period, counting only losses from cancellations and downgrades, with no expansion revenue included. Because expansions are excluded, GRR can never exceed 100%.
At a glance
- GRR measures revenue retained from existing customers before any upsell or expansion activity.
- Formula: (Starting ARR minus churn minus downgrades) divided by Starting ARR.
- 90%+ is a common Series B benchmark; enterprise-focused SaaS often runs 93-95%.
- GRR below 85% in mid-market SaaS typically triggers scrutiny in investor due diligence.
- Blended GRR can mask poor retention in specific segments, especially SMB.
How is GRR actually calculated?
Start with your MRR or ARR at the beginning of a period. At the end, subtract revenue lost to churn and downgrades. Divide that result by the starting figure. No expansion revenue enters the calculation at any point.
Example: a quarter begins with $500K ARR from existing accounts. Three customers cancel ($20K lost) and two downgrade ($10K lost). Retained ARR from that cohort is $470K. GRR = 94%. Net Revenue Retention (NRR) adds expansion back in, which is why a company can show 120% NRR while GRR sits at 88%. Both numbers tell a different story, and both are necessary.
Why does GRR matter for B2B revenue teams?
GRR strips away the flattering effect of upsells. A company posting 115% NRR looks healthy until GRR comes in at 82%, which means the business is covering serious churn with aggressive expansion activity. If expansion slows, the revenue floor drops fast.
For CEOs and heads of revenue, GRR is a direct read on whether customers find enough value to stay at their current spend level. It is not a growth metric. It is a signal about whether the base is holding.
When does GRR mislead you?
Blended averages hiding segment problems
A 91% blended GRR can conceal a 78% GRR in an SMB segment and a 96% GRR in enterprise. Tracking segments separately is necessary to diagnose where intervention is actually needed.
Confusing GRR with account churn rate
Churn rate counts lost accounts. GRR counts lost dollars. Losing ten small accounts while retaining two large ones can produce a high account churn rate alongside a healthy GRR. Revenue churn is what matters for the income statement.
Ignoring measurement timing
GRR measured monthly versus annually tells different stories because contraction timing varies. Picking a consistent measurement window and holding to it is necessary for any meaningful trend analysis.
Common mistakes teams make with GRR
- Declaring victory at 90%. Ten percent annual revenue erosion on $5M ARR is $500K gone before a single new customer is acquired.
- Using NRR to mask GRR problems. Strong expansion can inflate NRR while a deteriorating retention base goes unaddressed.
- Not tying CS initiatives to GRR movement. Playbooks, QBR cadences, and early warning scoring should all connect back to measurable GRR improvement.
- Skipping cohort analysis. Aggregate GRR hides whether newer cohorts retain worse than older ones, which matters for forecasting.
How does GRR connect to adjacent metrics?
GRR functions as the floor of a recurring revenue model. CLV calculations break down when GRR is unstable, because the assumed retention window shrinks. CAC Payback Period also gets distorted: modeling a 24-month payback while GRR implies customers churn at month 18 means the unit economics do not hold.
A 2-point GRR improvement on $10M ARR preserves $200K in annual revenue, and that figure compounds. Retention initiatives only show their true return when measured against GRR movement over time.