Net revenue retention, usually shortened to NRR and sometimes called net dollar retention, measures how much revenue you keep and grow from your existing customers over a year. It's one of the metrics investors look at first in a subscription business, because it captures two things that are hard to fake: whether customers stay, and whether they spend more over time.

Quick answer

The NRR formula is (starting revenue + expansion − contraction − churn) ÷ starting revenue, measured on customers you had at the start of the period. That's how to calculate NRR. Net revenue retention vs gross revenue retention: NRR includes expansion and can exceed 100%, while GRR excludes it and can't. Investors watch both closely in subscription businesses.

An NRR above 100% means your existing customers, as a group, are paying you more than they did a year ago, even after accounting for those who left or downgraded. A business with strong NRR can grow even if it signs no new customers at all.

The formula

NRR is calculated on a cohort: the customers you had at the start of the period.

NRR = (Starting revenue + expansion − contraction − churn) ÷ Starting revenue

  • Starting revenue: recurring revenue from the cohort at the beginning of the period, often expressed as ARR or MRR.
  • Expansion: extra revenue from those same customers through upgrades, more seats, more usage, add-on products or price increases.
  • Contraction: revenue lost from those customers downgrading but staying.
  • Churn: revenue lost from customers who cancelled entirely.

Crucially, revenue from new customers acquired during the period is excluded. NRR is only about the customers you started with.

A worked example

Illustrative SaaS company. On January 1, it has 200 customers paying a combined $2,000,000 in ARR.

Over the next 12 months, among those same 200 customers:

  • Expansion (more seats, upgrades): +$450,000
  • Contraction (downgrades): −$80,000
  • Churn (cancellations): −$170,000

Ending ARR from the cohort: $2,000,000 + $450,000 − $80,000 − $170,000 = $2,200,000.

NRR = $2,200,000 ÷ $2,000,000 = 110%.

New customers signed during the year don't count here, however many there were.

NRR vs GRR

Gross revenue retention (GRR) is NRR's stricter sibling. It ignores expansion and looks only at what you lose.

GRR = (Starting revenue − contraction − churn) ÷ Starting revenue

In the example above: ($2,000,000 − $80,000 − $170,000) ÷ $2,000,000 = 87.5%.

NRRGRR
Includes expansion?YesNo
Can exceed 100%?YesNo (maximum 100%)
What it showsOverall growth from existing customersHow well you keep the revenue you have
Why investors look at itGrowth efficiency and upsell potentialProduct stickiness and downside risk

Look at both. A company with 120% NRR but 75% GRR is growing its best customers fast while losing a lot of others. That's a different business from one with 110% NRR and 95% GRR, which keeps almost everyone and grows them steadily.

Why investors care so much

  • Cheaper growth. Expanding existing customers usually costs far less than acquiring new ones.
  • Evidence of value. Customers don't spend more on products that don't work for them.
  • Compounding. With NRR above 100%, each year's cohort keeps growing, so new customer acquisition adds on top of an expanding base.
  • Durability. High retention makes revenue more predictable, which supports higher valuations.

That's why NRR comes up so often in Series B and growth-stage diligence, and why it pairs naturally with efficiency measures like the Rule of 40.

What's a good NRR?

It depends heavily on who you sell to.

  • Selling to small businesses: small customers churn more often because they close, switch, or cut costs. NRR around or slightly below 100% can still be respectable.
  • Selling to mid-market companies: above 100% is a common expectation for healthy companies.
  • Selling to large enterprises: the best enterprise software companies often report NRR well above 110% or 120%, because large customers expand across departments.

Business model matters too. Usage-based pricing can produce high NRR when customers grow, and sharp drops when they cut back. Seat-based pricing tends to be steadier.

Benchmarks shift with economic conditions. During downturns, many companies see NRR fall as customers trim seats and budgets.

Common calculation mistakes

  1. Including new customers. The most common error. It inflates NRR and makes the metric meaningless.
  2. Using a monthly snapshot and annualising. One strong month multiplied by twelve misleads. Use a full twelve-month cohort.
  3. Counting one-off fees. Implementation or services revenue isn't recurring and shouldn't be in the base or the expansion.
  4. Treating paused customers as retained. Be consistent about how you classify customers who stop paying temporarily.
  5. Cherry-picking cohorts. Reporting only your best year's cohort without saying so damages trust when investors check.
  6. Mixing currency effects. If you bill in several currencies, exchange rate movements can make NRR look better or worse than the underlying business.
Be precise in your data room. Write down exactly how you calculate NRR, show the cohort table behind it, and keep the definition consistent over time. Inconsistent metric definitions are one of the most common issues in due diligence.

What moves NRR

Things that raise it

  • pricing that grows with customer success (seats, usage, modules),
  • strong onboarding so customers get value quickly,
  • a customer success team focused on adoption and expansion,
  • additional products customers can add,
  • regular, well-communicated price increases tied to added value,
  • focusing sales on customer segments that tend to expand.

Things that lower it

  • selling to customers who are a poor fit,
  • weak onboarding and slow time to value,
  • pricing that caps how much customers can spend,
  • product gaps that push growing customers to competitors,
  • economic downturns that hit your customers' budgets.

How to present NRR to investors

  • Show a cohort table. Each row is a starting cohort (for example, customers acquired in a given quarter), and columns show their revenue over time. Investors can see whether newer cohorts behave like older ones.
  • Show NRR and GRR together.
  • Break it down by segment. SMB and enterprise customers often behave very differently.
  • Explain movements. If NRR dipped, say why and what you're doing about it.

Real-world context

Companies that publish NRR are usually public. Most private companies don't disclose it, but some share related signals. For example, 1Password has said more than three-quarters of its revenue now comes from businesses, and Clio's ARR roughly doubled in about 18 months, partly through acquisition. Neither figure is NRR, but both hint at how revenue is growing. When you research private companies, look for clues like these and treat any third-party NRR estimate with caution. Our guide on finding private company revenue covers how to read these signals.

Building an NRR cohort table step by step

If you've never built one, here's a simple way to do it in a spreadsheet using your billing data.

  1. Export monthly recurring revenue by customer for at least the last 24 months. One row per customer, one column per month.
  2. Tag each customer with a start month, the first month they paid.
  3. Group customers into cohorts by start quarter. Quarterly cohorts are usually easier to read than monthly ones.
  4. Sum each cohort's revenue for every month since it started.
  5. Divide each cell by the cohort's starting revenue to get a percentage. A cohort that started at $100,000 MRR and is now at $118,000 shows 118%.
  6. Read across and down. Across a row shows how one cohort evolves. Down a column shows whether newer cohorts behave better or worse than older ones at the same age.

Illustrative cohort table (revenue as % of starting revenue)

CohortStart6 months12 months18 months
Q1 2025100%104%109%115%
Q2 2025100%103%111%—
Q3 2025100%106%——

Newer cohorts expanding slightly faster than older ones at the same age is a good sign: the product, onboarding or targeting is improving.

NRR for usage-based and hybrid pricing

Usage-based pricing, where customers pay for what they consume, makes NRR more volatile. A customer can expand sharply in a busy quarter and contract the next. A few practical adjustments help:

  • Use a trailing average of usage revenue rather than a single month's figure for the start and end points.
  • Report committed contract revenue and variable usage revenue separately.
  • Explain seasonality to investors upfront so a quiet quarter isn't misread as churn.

Hybrid models, with a platform fee plus usage, often produce the steadiest NRR: the platform fee anchors gross retention while usage drives expansion.

NRR and valuation

Investors care about NRR partly because it affects how much a company is worth. Higher NRR means each dollar of new revenue lasts longer and grows, which makes future revenue more valuable today. Analysts of public software companies have long observed that businesses with higher NRR tend to command higher revenue multiples, although NRR is only one of many factors. For private companies raising growth rounds, a strong, well-documented NRR is one of the most persuasive things you can show.

Frequently asked questions

What does NRR stand for?

Net revenue retention. It measures revenue retained and expanded from existing customers over a period, usually a year.

Is net dollar retention the same as net revenue retention?

Yes. The two terms are used interchangeably.

Can NRR be more than 100%?

Yes. If expansion from existing customers exceeds what you lose to churn and downgrades, NRR is above 100%.

What is the difference between NRR and GRR?

NRR includes expansion revenue. GRR excludes it and only measures how much of the starting revenue you keep, so it can't exceed 100%.

Do new customers count in NRR?

No. NRR only measures the customers you had at the start of the period.