Net Revenue Retention (NRR)

NRR tracks the recurring subscription value of customers included at the start of a period through to its end. New customers are excluded. It describes observed changes, rather than predicting what will happen if acquisition stops.

Above 100%, expansion exceeds the included losses. At 100%, the movements balance; below 100%, the cohort’s recurring value has declined. Customer churn measures lost customers, while revenue churn uses a different unit.

The formula for Net Revenue Retention is straightforward:

NRR = (Starting MRR + Expansion MRR - Contraction MRR - Churned MRR) / Starting MRR × 100%

Worked Example

Imagine a company begins a month with $100,000 in Monthly Recurring Revenue (MRR) from existing customers.

• Expansion MRR (from upgrades and cross-sells): $15,000
• Contraction MRR (from downgrades): $5,000
• Churned MRR (from cancellations): $10,000

Applying the formula:

NRR = (100,000 + 15,000 - 5,000 - 10,000) / 100,000 × 100% = 100%

This means the company has maintained its revenue level from existing customers. If NRR had been 110%, it would indicate growth purely from the current base, while 90% would signal a net decline.

NRR separates changes in the existing customer base from new customer acquisition. Price increases or large upgrades from a few accounts can raise NRR even while many other customers leave.

Track GRR, customer churn and concentration alongside it. Define CLV’s time horizon and included costs, and state which acquisition costs enter CAC. NRR does not itself establish customer satisfaction, profit, future cash flow or a valuation multiple.

Compare equivalent periods and segments. State the start and end dates, normalisation method and reactivation policy, keeping the same customer population throughout each measurement.

Investigate whether changes come from prices, more users, greater usage, downgrades or cancellations. A segment at 120% NRR is not automatically more profitable than one at 90%; costs, concentration and the reasons for the difference also matter. Forecasts require separate assumptions about future changes.

While NRR is powerful, it can be misunderstood or misused:

• Ignoring customer count: A high NRR can mask customer churn if revenue growth comes from a few large accounts. Always analyse both NRR and customer retention rates.
• Mixing time periods: Ensure that expansion, contraction, and churn are measured over the same time frame as the starting MRR.
• Overlooking pricing changes: Price increases can raise NRR legitimately under its definition; separate them from usage or seat expansion to explain the cause.
• Comparing across industries: Acceptable NRR benchmarks vary. Enterprise SaaS firms often see higher NRR than consumer subscription services.

NRR should be evaluated alongside other metrics to form a complete view of business health. A company can have strong NRR but still face issues with new customer acquisition or high CAC, which can limit total growth.

Possible measures include better onboarding, relevant support and offers that fit customers’ actual needs. Investigate why customers cancel or downgrade before choosing an intervention.

Track changes in NRR alongside customer counts, GRR, costs and feedback. A price increase can lift NRR but also affect demand and later churn. An observed improvement does not by itself prove that a particular initiative caused it.

NRR summarises the change in recurring subscription value within a specified starting cohort. It is not a direct measure of satisfaction, product-market fit, operating efficiency or profit. Use it with customer churn, GRR, concentration and costs when assessing results and future plans.

Exclude new customers and state how reactivations are treated; tools may have different settings. This example follows the same starting cohort without reactivations. The percentage is undefined if starting MRR is zero.