What is Net Revenue Retention (NRR)?
Net Revenue Retention (NRR) measures how much recurring revenue a company keeps and grows from its existing customers over a period, expressed as a percentage where anything above 100% means expansion is outpacing churn and downgrades.
Formula
NRR = (Starting ARR + Expansion − Downgrades − Churn) ÷ Starting ARR × 100
Example: A company starting the year with $2,000,000 in ARR that adds $400,000 in expansion, loses $100,000 to downgrades and $200,000 to churn has an NRR of 105%—it grew 5% before winning a single new customer.
NRR looks only at customers you already had. New logos are excluded on purpose, which is what makes it the honest read on whether the business you bought was worth buying. Above 100%, existing accounts fund growth on their own. Below 100%, every quarter starts in a hole that net-new acquisition has to fill before any growth is visible.
That is why NRR and CAC are two halves of one equation rather than separate metrics. CAC tells you what it cost to get a customer in the door; NRR tells you whether that customer stayed long enough to be worth it. When NRR sits below 100%, churned revenue has to be replaced with newly acquired revenue—and replacement revenue is paid for at full CAC, every time. So each point of CAC efficiency you win gets consumed refilling a bucket that leaks, and the blended cost of a dollar of revenue keeps rising no matter how well the top of the funnel performs.
It also caps lifetime value, which is the number CAC has to be judged against. LTV is a function of how long an account stays and how much it grows; churn shortens the term and downgrades shrink the size. A low-NRR business therefore compresses LTV and inflates effective CAC at the same time, which is how an LTV:CAC ratio collapses while marketing dashboards still look healthy. Acquiring customers is expensive work—if you cannot keep the ones you fought for, the economics stop being fixable at the acquisition end.
Reporting growth by logo count hides all of this, because it is a friendlier number to present than a churn conversation. It is also why the GTM data problems that make NRR hard to calculate matter so much: trial-to-paid tracked as a monthly average instead of by cohort smooths over the exact erosion NRR is designed to catch, and handling expansion only at renewal means the conversation happens after the quarter is already decided.
What to do about it
- Calculate NRR by cohort, not blended, so a few large expansions cannot mask broad churn.
- Report NRR next to CAC and payback period—each one is misleading without the other two.
- Trigger expansion conversations from product usage instead of the renewal date.
Frequently asked questions
What is a good NRR?
For B2B SaaS, 100% is the break-even line, 110%+ is strong, and best-in-class enterprise products reach 120% or more. Below 100% means existing customers are shrinking and new logos are funding the gap.
What is the difference between NRR and gross retention?
Gross retention counts only losses—churn and downgrades—and can never exceed 100%. NRR adds expansion back in, so it shows whether growth inside the existing base outpaces those losses.
How does low NRR affect CAC?
It makes CAC impossible to lower at scale. Revenue lost to churn has to be replaced by newly acquired revenue paid for at full CAC, so churn quietly resets every acquisition efficiency gain and caps the LTV that CAC is measured against.