Gross Revenue Retention: Formula and How It Differs From NRR

Gross revenue retention measures how much of your starting recurring revenue you keep from the same customers once downgrades and cancellations are taken out, with no credit for upgrades. It cannot exceed 100 percent, and that ceiling is exactly what makes it useful: it shows how leaky the base is, without the expansion that net revenue retention allows to cover for losses. Two businesses can report the same net figure and be in very different positions because one keeps its customers and the other relies on a few large accounts growing fast. This guide gives the formula, shows the customer-level calculation that is easy to get wrong, compares two companies with identical net retention and different gross retention, and explains how to read and improve the metric. It belongs to unit economics and budgeting and forecasting, builds on net revenue retention and annual recurring revenue, and leads to the Rule of 40 and the SaaS magic number.

The formula

GRR = (Starting recurring revenue - Contraction - Churn) / Starting recurring revenue

Measure it for the customers who existed at the start of the period, usually over twelve months. Expansion is left out entirely, and so are new customers. Using the cohort from the net retention example, a starting base of $1,000,000 with $30,000 of downgrades and $70,000 of cancellations gives GRR = (1,000,000 - 30,000 - 70,000) / 1,000,000 = 900,000 / 1,000,000 = 90.0 percent. The same cohort had $150,000 of expansion, giving a net retention of 105.0 percent. The 15-point gap is the expansion.

The customer-level cap

The most common error in the calculation is to net each customer's expansion against other customers' losses. GRR should be calculated customer by customer, with each customer's ending revenue capped at the starting level. Consider three customers:

The correct GRR is 160 / 300 = 53.3 percent. If you simply added ending revenue, you would get 210 / 300 = 70.0 percent, which is the net retention figure, and you would overstate how well the business is keeping revenue. The cap treats customer A's growth as irrelevant to the question "how much of what we had did we keep?" In a spreadsheet, the formula for each customer is =MIN(ending, starting), summed and divided by the total starting revenue.

Same net retention, different risk

Compare two companies, each starting with $1,000,000 of recurring revenue.

Check: X: 1,000,000 + 100,000 - 20,000 - 30,000 = 1,050,000, and GRR = (1,000,000 - 50,000) / 1,000,000 = 95 percent. Y: 1,000,000 + 300,000 - 100,000 - 150,000 = 1,050,000, and GRR = (1,000,000 - 250,000) / 1,000,000 = 75 percent. Both report 105 percent net retention. Company X loses 5 percent of its base and grows slightly through expansion. Company Y loses a quarter of its base and relies on survivors to grow by a larger amount. Y depends on a small number of accounts expanding, which is riskier: if the expansion slows, its net retention will fall quickly. Y also has a larger churn problem to solve, which means its sales and marketing effort goes partly toward replacing lost revenue.

What gross retention tells you

GRR answers three questions that net retention cannot.

• How good is the product's core value? A high GRR means customers stay and keep what they bought.
• How much new business do you need just to stand still? At 75 percent GRR, you must replace a quarter of your base each year. At 95 percent, 5 percent.
• How reliable is your forecast? Losses are more predictable when they are small.

A rough guide for the arithmetic of standing still: with starting recurring revenue R and GRR g, the new business needed to hold revenue flat, ignoring expansion, is R × (1 - g). For $5,000,000 of revenue, a 90 percent GRR requires $500,000 of new business, and 80 percent requires $1,000,000. Each ten-point improvement in GRR here is worth $500,000 of annual new business that does not have to be won.

Monthly and annual views

Monthly GRR compounds. A monthly GRR of 99 percent is 0.99 to the twelfth power, or about 88.6 percent a year. A monthly GRR of 98 percent is about 78.5 percent a year (0.98 to the twelfth power is 0.7847). Small monthly differences matter a great deal over a year. Report the annual figure on the cohort from twelve months ago, and show monthly trends alongside to catch changes early. Do not annualize by multiplying the monthly loss by 12, since that overstates the loss slightly (1 percent a month is not exactly 12 percent a year).

Segment it

As with net retention, aggregate GRR hides variation. Break it out by customer size, plan, acquisition channel, industry, contract length and customer age. Typical findings: smaller customers on monthly plans churn more than larger customers on annual contracts, customers acquired through discounts churn more than customers acquired through referrals, and the first year is the riskiest. Each finding suggests an action. If monthly-plan customers churn at twice the rate of annual-plan customers, test an annual-plan incentive and measure the effect on GRR after allowing for the discount cost. See revenue churn vs. logo churn for how counting customers and counting revenue can tell different stories.

How to improve gross retention

1. Fix onboarding. Early churn usually reflects customers who never reached value. Measure time to first value.
2. Track health signals. Usage, logins, support tickets, billing failures and stakeholder changes warn of trouble.
3. Reduce involuntary churn. Failed payments, expired cards and billing errors cause cancellations that customers did not choose. Retry logic and reminders recover a share.
4. Offer the right plan. A downgrade is better than a cancellation. Make downgrades easy and keep customers who need less.
5. Use contract terms. Annual and multi-year contracts reduce churn at the cost of some upfront discount. Compare the discount with the retention gain.
6. Run exit interviews and record reasons for every cancellation and downgrade.
7. Invest in the product where customers say it falls short.

Common calculation errors

• Letting expansion offset losses by not capping each customer.
• Including new customers acquired during the period.
• Excluding customers you consider poor fits.
• Treating pauses and reactivations inconsistently.
• Mixing customer counts and revenue.
• Changing the period after seeing the result.

FAQs

What is gross revenue retention?

The share of starting recurring revenue retained from existing customers after downgrades and cancellations, excluding any expansion.

How is GRR different from NRR?

NRR includes expansion and can exceed 100 percent. GRR counts only losses and cannot exceed 100 percent.

Why cap each customer at their starting revenue?

So that growth from one customer does not hide losses from others. Without the cap, the calculation becomes net retention.

What is a good gross revenue retention?

It depends on your customer size, contract length and market. Compare it with your own history and with similar businesses, and look at segments.

How do I improve GRR?

Improve onboarding, monitor health signals, reduce failed payments, offer downgrades as an alternative to cancellation and fix the product problems customers cite.