Subscription analytics
Net Revenue Retention: The Formula, a Worked Example and the Usage-Billing Catch
By the Orvlin team · · 3 min read
Start with the customers, not the percentage
A company can add customers while losing value from the accounts it already has. Net revenue retention, or NRR, separates those stories. Choose a starting cohort and compare revenue from those same customers at the end of the measurement period.
A common formula is (starting recurring revenue + expansion − contraction − churn) ÷ starting recurring revenue × 100. “Recurring revenue” needs a written definition, particularly when contracts include variable usage. NRR is an operating metric, not a substitute for recognized revenue or cash reporting.
Walk through the arithmetic
Imagine an existing cohort starts with $100,000 of monthly recurring revenue. By the comparison date, those accounts add $20,000 of expansion, lose $5,000 to contraction and lose $10,000 to churn. Ending cohort revenue is $105,000. NRR is $105,000 ÷ $100,000 × 100 = 105%.
If new customers contribute another $30,000, leave that money out of this NRR calculation. Including it would mix acquisition with retention. Under a common gross-retention definition, expansion is excluded: ($100,000 − $5,000 − $10,000) ÷ $100,000 = 85%. The two metrics answer different questions.
Why usage billing needs extra care
For a flat subscription, recurring value may be relatively easy to identify. With consumption pricing, last month’s usage can rise because the customer grew, ran a temporary job or bought capacity earlier than usual. An invoice increase is not automatically durable expansion.
Choose and disclose whether you compare actual consumption revenue, a normalized run rate, contracted commitments or another consistent measure. Do not silently switch between them. An annual commitment, a prepaid balance and recognized revenue are not interchangeable inputs.
A reporting policy worth keeping
Record the cohort start date, measurement window, currency treatment, reactivation policy and excluded charges. Explain how acquisitions, customer-account merges, one-time services and price increases affect the calculation. Use the same definitions when showing a trend.
Public-company NRR methods vary, as Ordway’s research documents. A peer’s published percentage can be useful context, but not a directly comparable target until you understand its definitions. Avoid an unsupported universal “good NRR” threshold.
Use the result to ask better questions
Break the cohort into understandable segments: product, contract size, customer age or pricing model. A healthy combined percentage can conceal heavy contraction in one segment and strong expansion in another.
When a number changes, trace it back to account-level movements. Did customers leave, reduce seats, use less, receive temporary credits or move to a new contract? That explanation is more actionable than a large percentage on a dashboard.
Frequently asked questions
- Can NRR exceed 100%?
- Yes. When expansion from the starting cohort exceeds contraction and churn, ending cohort revenue is higher than starting revenue.
- Is new business included in NRR?
- Not in the common existing-cohort calculation. New-customer revenue should be reported separately.
- Is NRR the same as revenue growth?
- No. Overall growth can include new customers; NRR follows the existing cohort under a disclosed revenue definition.