
If you want to know what NRR in SaaS is in one sentence: NRR (Net Revenue Retention) is the percentage of recurring revenue you keep from your existing customers over a year — after subtracting cancellations and downgrades, and adding back any upsells and expansion. Above 100% means your existing customer base spends more this year than last year, even if you never sign a single new customer. Below 100% means it shrinks. That one number is also among the first three or four metrics any acquirer or investor asks for, because it tells them whether your revenue compounds or decays on its own.
Here is why that matters more than most founders realize. A SaaS company with net revenue retention above 100% has, in effect, a growth engine that runs without fuel. I once worked with a founder running a roughly $10M-a-year business he found boring — no chaos, no innovation, nothing to fight. When he finally calculated his NRR, it came back at 140%. I did the math in my head: at 140% retention, with zero new customers, that “boring” $10M business 10x’s to $100M in revenue in well under a decade — roughly seven years of pure compounding off the existing base. He had been ignoring the most valuable asset he owned.
This article answers “what is NRR in SaaS” from the ground up — the plain-English definition, the exact formula, a worked example using realistic numbers at $10M ARR, current benchmarks with the context that makes them useful, the mistakes that quietly corrupt the number, and a short FAQ. If you already know the basics and want the strategic playbook — how NRR drives valuation and how to push it higher — read the deeper guide on net revenue retention. This piece is for getting the definition right first.
What NRR Actually Measures
NRR (Net Revenue Retention) answers a deceptively simple question: take the group of customers you had at the start of a period — usually 12 months ago — and ignore everyone who signed up after that. How much recurring revenue is that same group generating today, compared to what they generated back then?
The word “net” is the important part. NRR nets together everything that happened to that cohort:
- Some customers churned — they cancelled entirely, taking their revenue with them.
- Some customers contracted — they downgraded to a cheaper plan or dropped seats, so they still pay you, just less.
- Some customers expanded — they upgraded, added seats, or bought more, so they pay you more than they did a year ago.
If the expansion from the customers who stayed and grew outweighs the losses from the ones who left or shrank, your NRR is above 100%. If it doesn’t, you’re below 100%.
The reason NRR is so closely watched is what it implies about the future. NRR above 100% means your existing base is a self-funding growth machine — it grows on its own, before you spend a dollar acquiring anyone new. This is the engine behind companies like Slack and Dropbox: once a team is in, they only add more seats and more data over time. There is no realistic scenario where they use Slack less next year. That dynamic is what makes the lifetime value of those accounts effectively infinite, and it’s why those businesses command the multiples they do.
The NRR Formula
Here is the formula, written out explicitly:
NRR = (Starting MRR + Expansion MRR − Contraction MRR − Churned MRR) ÷ Starting MRR × 100%
A quick definition of each term, because the formula is only as good as your understanding of the inputs:
- Starting MRR is the Monthly Recurring Revenue (MRR) — the predictable monthly subscription revenue — from your cohort of existing customers at the start of the period.
- Expansion MRR is the additional monthly revenue those same customers added through upgrades, seat additions, or cross-sells.
- Contraction MRR is the monthly revenue lost when those customers downgraded but did not leave.
- Churned MRR is the monthly revenue lost when those customers cancelled entirely.
You can run the same calculation on Annual Recurring Revenue (ARR) — your yearly subscription revenue — instead of MRR. The result is identical as long as you stay consistent and use the same unit for every term. Whether you use MRR or ARR, the rule that breaks the formula is mixing the two. (Some sources, like Wall Street Prep’s NRR breakdown, fold contraction into the churn term; the math is the same either way, but separating the two makes downgrades easier to see and fix.)
The single most important thing the formula does not include is new customer revenue. NRR measures the existing base only. The moment you let revenue from customers acquired during the period sneak into the numerator, you are no longer measuring retention — you’re measuring growth, and you’ll fool yourself into thinking a leaky business is healthy. More on that mistake below.

A Worked Example at $10M ARR
Numbers make this concrete. Consider a B2B SaaS company at $10M ARR — squarely in the range where this question actually matters. To keep the arithmetic clean, that’s roughly $833,333 in MRR at the start of the year ($10M ÷ 12).
Over the next 12 months, here is what happens to that starting cohort — and only that cohort, ignoring any new logos signed during the year:
| Component | Monthly Amount | What Happened |
|---|---|---|
| Starting MRR | $833,333 | The existing base, 12 months ago |
| Expansion MRR | +$125,000 | Upgrades, added seats, cross-sells |
| Contraction MRR | −$33,333 | Downgrades from customers who stayed |
| Churned MRR | −$58,333 | Customers who cancelled entirely |
| Ending MRR (same cohort) | $866,667 | What the original base pays now |
Plug those into the formula:
NRR = ($833,333 + $125,000 − $33,333 − $58,333) ÷ $833,333 × 100% = $866,667 ÷ $833,333 × 100% = 104%
So this company has 104% net revenue retention. The existing base grew about 4% on its own over the year, before any new sales. That’s a healthy, self-sustaining number — modest, but it means the business is not running uphill.
It’s worth noting what Gross Revenue Retention (GRR) would be for the same company. GRR measures the same cohort but excludes expansion — it only counts what you kept before any upsells:
GRR = ($833,333 − $33,333 − $58,333) ÷ $833,333 × 100% = $741,667 ÷ $833,333 × 100% = 89%
The 15-point gap between 89% GRR and 104% NRR is the contribution of expansion. That gap is one of the most diagnostic things you can look at — it’s covered in depth in the comparison of gross revenue retention versus net.

NRR Benchmarks (and the Context That Makes Them Useful)
A raw NRR number means little without a benchmark — and benchmarks mean little without context. Here is the general interpretation scale:
| NRR | Interpretation |
|---|---|
| Below 90% | Leaky bucket — the base is in net contraction |
| 90–100% | Stable, but the base isn't growing on its own |
| 100–110% | Healthy — the existing base grows without new sales |
| 110–130% | Strong — expansion-driven growth |
| Above 130% | Elite — a powerful upsell and cross-sell engine |
For 2026, the median NRR for private B2B SaaS sits in the low-to-mid 100s. SaaS Capital’s 2026 benchmarking survey of more than 1,000 private B2B SaaS companies puts the median NRR for bootstrapped firms in the $3M–$20M ARR range at 103%, with 90th-percentile performers reaching about 118%. But the median hides enormous variation by customer segment, and this is where most founders misread their own number:
| Segment (by deal size) | Typical Median NRR |
|---|---|
| Enterprise (ACV above $100K) | ~118%, top performers 130–135% |
| Mid-market ($25K–$100K ACV) | ~108% |
| SMB (under $25K ACV) | ~97% |
The same NRR means different things at different scales. Holding 100%+ at SMB scale is genuinely strong — small customers churn more and expand less, so staying flat is an accomplishment. At enterprise scale, 100% would be a warning sign, and even 120% is merely “good.” Before you celebrate or panic over your NRR, compare it to companies with a similar customer size, pricing model, and contract structure. A usage-based business will naturally post higher NRR than a flat-seat business; comparing the two tells you nothing.
A note on the numbers: Benchmark figures shift year to year and vary by source and methodology. The ranges above reflect 2025–2026 conditions and are included to show relative differences across segments, not as fixed targets. Pull the latest data for your specific segment before setting goals against it.
The Mistakes That Corrupt Your NRR
NRR is simple to state and surprisingly easy to calculate wrong. These are the errors I see most often, and each one makes a struggling business look healthier than it is.
- Counting new customers in the numerator. This is the cardinal sin. NRR tracks the existing cohort only. If you fold in revenue from customers acquired during the period, you’ve turned a retention metric into a growth metric and lost the entire point. The whole value of NRR is that it isolates whether your base grows on its own.
- Ignoring contraction. A customer who downgrades from $1,000 to $600 a month is not “retained” — they represent $400 of monthly contraction. Treating partial losses as zero quietly inflates the number.
- Letting NRR hide a churn problem. A company can post 120% NRR while its GRR sits at 75% — meaning it loses a quarter of its base revenue every year and papers over the hole with expansion from a handful of accounts. Always read NRR and GRR together. NRR alone can mask a leaky bucket. If churn is the real issue, the fix lives in the work on reducing SaaS churn, not in more upsells.
- Mixing time periods or units. Compare a monthly figure against an annual one, or MRR against ARR, and the calculation is meaningless. Pick one period (12 months is standard) and one unit, and hold them constant.
- Blending all segments into one number. A company-wide NRR averages your healthy enterprise cohort with your leaky SMB cohort and hides both. Segment it — by deal size, vertical, and acquisition channel. In my experience, 100% of the time there are significant variances across segments, and the blended number conceals the ones that matter.

Why Acquirers Care So Much About NRR
NRR is not just an operating metric — it’s a valuation lever. When an investor or acquirer evaluates a SaaS business, NRR is among the first numbers they request, because combined with your growth rate and gross margin they can ballpark what the business is worth in about ten minutes.
The logic is straightforward. A business with NRR above 100% is de-risked — it will be larger next year even if sales stalls completely, because the existing base keeps expanding. A business below 100% is the opposite: it has to win new customers just to stand still, and any disruption to its sales engine means decline. Acquirers pay a premium for the first kind and a discount for the second. The spread is real — companies in the 100–110% NRR band tend to command meaningfully lower revenue multiples than those above 120%.
This connects directly to how you should think about building toward an exit. NRR is one of the cleanest signals of the kind of recurring-revenue quality that drives a high multiple — a theme that runs through how NRR compares to ARR as a measure of business health. If you’re building to sell, NRR isn’t a metric you monitor; it’s an asset you cultivate.
Frequently Asked Questions
What does NRR stand for in SaaS? NRR stands for Net Revenue Retention. It measures the percentage of recurring revenue retained from your existing customer base over a period (typically 12 months), after accounting for churn and downgrades and including upsells and expansion.
What is a good NRR in SaaS? Above 100% is the threshold for “healthy” — it means your existing base grows on its own. The 2026 median for private bootstrapped B2B SaaS at $3M–$20M ARR is about 103%. But “good” depends on segment: 100% is strong for SMB-focused businesses, while enterprise-focused companies should target 110%+ and elite performers exceed 130%.
What’s the difference between NRR and GRR? GRR (Gross Revenue Retention) excludes expansion revenue — it only measures how much of your existing revenue you keep before any upsells, so it can never exceed 100%. NRR includes expansion, so it can rise above 100%. Reading them together reveals whether healthy NRR is hiding a churn problem underneath.
Does NRR include new customers? No. This is the most common mistake. NRR measures only the cohort of customers you had at the start of the period. Revenue from customers acquired during the period is excluded entirely — including it turns NRR into a growth metric and defeats its purpose.
How is NRR calculated? NRR = (Starting MRR + Expansion MRR − Contraction MRR − Churned MRR) ÷ Starting MRR × 100%. You can use ARR instead of MRR as long as you use the same unit for every term and measure over a consistent period.
What time period should NRR cover? Twelve months is the standard, because it’s the most intuitive to benchmark and it smooths out seasonal swings. The critical rule is consistency: use the same cohort for the starting and ending measurement, and keep the period and unit fixed.
The Bottom Line
NRR in SaaS is the percentage of recurring revenue you keep and grow from your existing customers over a year. Above 100% and the base compounds on its own; below 100% and it decays, forcing you to acquire new customers just to stay flat. Calculate it on the existing cohort only, read it alongside GRR, and segment it before you trust the blended number. Get those three things right and you’ll have one of the most honest signals available about whether your business is the kind that compounds — the kind acquirers pay the most for.

