Mar 14, 2026 · 14 min read
What Is NRR? Net Revenue Retention Explained (Formula, Benchmarks, and What Drives It)
Net Revenue Retention (NRR) is the metric that tells investors more about a SaaS business than almost any other single number. It answers a deceptively important question: if you stopped acquiring new customers entirely, would your revenue grow, shrink, or stay flat?
A company with NRR above 100% would grow revenue even with zero new sales. That's a fundamentally different business than one with NRR of 80%, where churn is quietly eroding the base no matter how fast the sales team moves.
This guide covers the NRR definition, the formula, how to calculate it correctly, what good looks like at different stages, and what actually drives the number up or down.
NRR Definition
Net Revenue Retention is the percentage of recurring revenue retained from an existing cohort of customers over a defined period, including the effect of expansion revenue (upgrades, upsells, cross-sells) as well as losses from downgrades and cancellations.
The "net" in the name is the key word — it nets expansion against contraction and churn, giving you a single number that reflects the total revenue trajectory of your existing customer base.
Because expansion revenue is included and can more than offset losses, NRR can exceed 100%. This is what makes it different from Gross Revenue Retention (GRR), which excludes expansion entirely and therefore caps at 100%.
An NRR of 115% means your existing customers are collectively paying you 15% more than they were at the start of the period — even after accounting for every customer who downgraded or cancelled.
The NRR Formula
NRR (%) = (Starting MRR + Expansion MRR − Contraction MRR − Churned MRR) / Starting MRR × 100
What each variable means
Starting MRR — The total recurring revenue from your customer cohort at the beginning of the measurement period. New customers acquired during the period are excluded — NRR is only about existing customers.
Expansion MRR — Additional recurring revenue generated from those same customers during the period: plan upgrades, seat additions, upsells to higher tiers, cross-sells to additional products. Does not include one-time fees or professional services.
Contraction MRR — Revenue lost from customers who are still active but paying less: plan downgrades, seat reductions, negotiated discounts applied mid-period.
Churned MRR — Revenue lost from customers who cancelled entirely and are no longer paying you.
How to Calculate NRR (Step by Step)
Monthly NRR example
You have four customers at the start of February:
| Customer | Jan MRR | Feb MRR | What happened |
|---|---|---|---|
| Acme Corp | $1,800 | $2,400 | Upgraded plan — $600 expansion |
| Bright Ltd | $900 | $750 | Reduced seats — $150 contraction |
| Comet Inc | $2,200 | $0 | Cancelled — $2,200 churned |
| Drift Co | $600 | $600 | No change |
Step 1: Calculate Starting MRR
$1,800 + $900 + $2,200 + $600 = $5,500
Step 2: Sum each movement
Expansion: $600
Contraction: $150
Churn: $2,200
Step 3: Apply the formula
NRR = ($5,500 + $600 − $150 − $2,200) / $5,500 × 100
NRR = $3,750 / $5,500 × 100
NRR = 68.2%
That's a poor NRR — Comet's churn is doing serious damage that Acme's expansion can't offset. In fact, Acme's $600 expansion barely makes a dent against $2,200 of churn. This is a useful illustration of why NRR above 100% requires your expansion to genuinely exceed all losses combined, not just be positive.
A healthier example
Same starting cohort, different outcomes:
| Customer | Jan MRR | Feb MRR | What happened |
|---|---|---|---|
| Acme Corp | $1,800 | $2,900 | Large upgrade — $1,100 expansion |
| Bright Ltd | $900 | $820 | Small downgrade — $80 contraction |
| Comet Inc | $2,200 | $2,200 | Retained |
| Drift Co | $600 | $600 | No change |
Starting MRR: $5,500
Expansion: $1,100 | Contraction: $80 | Churn: $0
NRR = ($5,500 + $1,100 − $80 − $0) / $5,500 × 100
NRR = $6,520 / $5,500 × 100
NRR = 118.5%
Now your existing customers are collectively paying you 18.5% more than they were at the start of the month. No new customers needed — this cohort alone is growing your revenue.
NRR vs GRR: The Difference That Actually Matters
NRR and GRR are both retention metrics, and they're often confused or used interchangeably. They measure different things and tell different stories.
GRR (Gross Revenue Retention) strips out all expansion revenue. It only counts what you lose — churn and contraction. It can never exceed 100% because you can only retain up to what you started with. It answers: how well are you holding on to existing revenue, ignoring any upsells?
NRR (Net Revenue Retention) includes expansion revenue. It can exceed 100% when expansion outpaces losses. It answers: what is the net revenue trajectory of your existing customer base?
Here's the same cohort, both metrics side by side:
| Customer | Start MRR | End MRR | Change |
|---|---|---|---|
| Customer A | $3,000 | $4,200 | +$1,200 expansion |
| Customer B | $1,500 | $1,100 | −$400 contraction |
| Customer C | $2,000 | $0 | −$2,000 churned |
GRR (expansion excluded):
Retained: $3,000 + $1,100 + $0 = $4,100
GRR = $4,100 / $6,500 × 100 = 63.1%
NRR (expansion included):
Net retained: $4,200 + $1,100 + $0 = $5,300
NRR = $5,300 / $6,500 × 100 = 81.5%
NRR looks meaningfully better — but that 18-point gap is entirely explained by Customer A's expansion. If expansion slows, NRR collapses toward the GRR floor. This is why investors look at both: NRR tells you the current trajectory, GRR tells you how durable it is.
A business with 115% NRR and 65% GRR is growing through upsells while simultaneously losing a lot of customers. It's a more fragile position than one with 108% NRR and 91% GRR — even though the first number looks better on the headline metric.
What Is a Good NRR in SaaS?
NRR benchmarks vary by stage, market, and business model, but these are the reference points most commonly cited:
| NRR Range | What it signals |
|---|---|
| 130%+ | Exceptional. Typical of top-tier PLG or enterprise companies with strong expansion motions. |
| 110–130% | Strong. Existing customers are growing meaningfully. Solid foundation for efficient scaling. |
| 100–110% | Healthy. Expansion roughly offsets losses. Positive but not compounding. |
| 90–100% | Concerning. More revenue is leaving than expanding. Growth depends entirely on new customers. |
| Below 90% | Significant retention problem. Hard to build a durable business at this level. |
Notable benchmarks from public SaaS companies:
- Snowflake has reported NRR above 160% — driven by a usage-based model where customers naturally expand as data volumes grow
- Twilio and other usage-based companies often exceed 130% when growth is strong
- Most Series B SaaS companies targeting institutional funding aim for NRR above 110–120%
- The median NRR across public SaaS companies has historically sat around 106–112%
These numbers shift with market conditions. During periods of budget scrutiny (like 2022–2023 for many SaaS businesses), NRR compressed across the industry as customers downsized contracts. Benchmarking against your own cohort trend is more actionable than chasing an industry average.
Cohort NRR vs. Blended NRR: An Important Distinction
Most discussions of NRR treat it as a single company-wide number, but there are actually two ways to calculate it that tell different stories.
Blended NRR
Blended NRR takes your total starting MRR (all existing customers) and compares it to ending MRR from that same group. It's the easiest number to produce and the most commonly reported.
The limitation: blended NRR is an average that can mask significant variation. If your enterprise cohort has 130% NRR and your SMB cohort has 70% NRR, your blended number might be 100% — appearing flat when you actually have a serious SMB retention problem and a strong enterprise expansion story happening simultaneously.
Cohort NRR
Cohort NRR tracks the revenue from a specific group of customers — typically those who joined in the same month or quarter — and measures how their revenue changes over subsequent months.
This is more work to calculate but far more revealing. It shows:
- Whether early cohorts are expanding or eroding over time
- How NRR changes as customer maturity increases (new customers often have lower NRR; customers who've been with you 18 months often expand more)
- Whether product or pricing changes had a measurable impact on specific cohort behaviour
When investors ask to see NRR cohort data, they're looking for this view — not the blended number. A chart showing cohorts that expand over time is a much more compelling story than a single NRR percentage.
What Actually Drives NRR Above 100%
Getting to and sustaining NRR above 100% requires two things working together: expansion revenue that consistently exceeds churn and contraction, and a customer base where both are predictable. Here's where each comes from.
Expansion drivers
Natural usage growth — In usage-based or seat-based models, customers who find value tend to use more over time. More users get added, more data gets processed, more transactions get run. This is the most durable form of expansion because it doesn't require your sales team to do anything — the product's value does the work.
Tiered pricing with clear upgrade paths — When you have distinct tiers that offer meaningfully more value, customers who outgrow their current plan have a natural next step. The key word is "meaningfully" — feature-gating things customers expect to be included creates resentment, not expansion.
Land and expand sales motion — Closing an initial deal with a team or department and then systematically expanding to adjacent teams or use cases. This works when the initial deployment is successful enough that internal champions advocate for broader rollout.
Cross-sell to complementary products — Customers who already trust you are easier to sell additional products to than new prospects. Companies with multi-product portfolios often have structurally higher NRR than single-product ones.
Retention drivers (reducing the denominator losses)
Customer success investment — Proactive engagement with at-risk accounts before they decide to churn. Customers who feel supported are less likely to downgrade or leave, and more likely to expand. The return on CS investment shows up directly in NRR.
Onboarding that drives activation — Customers who fully activate — who experience the core value of the product deeply, not just superficially — have dramatically better retention. Poor onboarding is one of the most common root causes of early churn that shows up in NRR 3–6 months later.
Annual contracts over monthly — Annual contracts don't prevent churn, but they delay it and create natural renewal conversations. They also make the churn signal more deliberate — a customer on an annual plan who wants to leave has to actively decide not to renew, rather than just not logging in until their card fails.
Product-led stickiness — When your product is embedded in workflows, holds data customers care about, or connects to other tools in their stack, switching costs are real. This kind of stickiness is reflected in GRR as much as NRR — and as discussed above, a high GRR floor is what makes strong NRR durable.
NRR as a Fundraising Signal
Of all the retention metrics, NRR is the one most investors focus on during due diligence, and understanding why helps you present it well.
NRR above 100% means the business compounds on itself. If your existing customer base grows revenue by 15% per year without any new sales, the value of every new customer you acquire is higher — because once acquired, they're likely to expand too. This has significant implications for LTV calculations, payback period, and ultimately, valuation multiples.
NRR reveals the quality of growth. A company growing 100% YoY with 80% NRR is growing because sales is working hard to replace revenue that's leaving out the back door. A company growing 80% YoY with 120% NRR has a fundamentally different unit economics story — new customers layer on top of a base that's already growing. The second company is often worth more despite the lower growth rate.
NRR predicts capital efficiency. High NRR reduces the amount of new ARR you need to acquire just to maintain your current revenue level. If your NRR is 80%, you need to replace 20% of your ARR every year before you can show any growth. If your NRR is 115%, your existing base is already growing — every dollar of sales spend goes toward net new growth, not replacement.
For Series A and B conversations specifically, investors will typically want to see 12 months of monthly NRR data, cohort NRR by join date, and a breakdown of what's driving expansion vs. what's driving churn. Having that analysis prepared before the conversation is a meaningful signal of operational maturity.
Common NRR Calculation Mistakes
Including new customer revenue. NRR is exclusively about existing customers — those who were already paying you at the start of the measurement period. Revenue from customers acquired during the period is new ARR, not NRR. Mixing them inflates the metric and obscures what's actually happening with retention.
Counting one-time fees as expansion. Implementation fees, onboarding charges, and professional services are not recurring revenue and should never be counted in expansion MRR. NRR should reflect only changes to the recurring subscription component.
Inconsistent cohort definitions. If you define the starting cohort differently each month — sometimes including customers from the last 30 days, sometimes not — your NRR trend becomes meaningless. Fix the cohort at the start of each measurement period and don't adjust it.
Too small a sample size. NRR on 10 customers is extremely volatile — a single churn event moves it by 10 percentage points. Monthly NRR is only a reliable signal once you have enough customers that individual events don't dominate. For very early-stage companies, quarterly NRR is often more stable and more informative.
Not separating contraction from churn. Both reduce NRR, but they represent completely different problems. Contraction means customers see partial value and are paying for less. Churn means they see insufficient value to continue at all. Treating them as a single "losses" line loses the diagnostic signal.
Calculating NRR from total revenue instead of MRR. If your revenue reporting mixes recurring and non-recurring revenue, your NRR will be distorted. Always start with a clean MRR figure — subscription revenue only, normalized to monthly amounts — before applying the NRR formula.
Frequently Asked Questions (FAQs)
What does NRR stand for?
NRR stands for Net Revenue Retention. It's sometimes written as Net Dollar Retention (NDR) — same metric, different name. Both measure the percentage of recurring revenue retained from an existing customer cohort including expansion, contraction, and churn.
What is a good NRR for SaaS?
Above 100% is the key threshold — it means your existing customer base is generating more revenue over time without any new customers. Top SaaS companies aim for 110–130%+. For most growth-stage SaaS businesses, 100–115% is a healthy operating range. Below 100% means revenue is leaking faster than it's expanding.
How is NRR calculated?
NRR is calculated as: (Starting MRR + Expansion MRR − Contraction MRR − Churned MRR) ÷ Starting MRR × 100. Starting MRR is your existing customer revenue at the beginning of the period. Expansion, contraction, and churn are the movements from that cohort during the period. New customer revenue is excluded entirely.
What is the difference between NRR and net revenue retention?
They're the same thing — NRR is just the abbreviation for net revenue retention. You'll also see it called Net Dollar Retention (NDR) or Net Dollar Retention Rate, particularly in US-based investor contexts. The formula and interpretation are identical regardless of which name is used.
Can NRR be above 100%?
Yes — and this is one of the key differences between NRR and GRR. NRR can exceed 100% when expansion revenue from upgrades and upsells is larger than the combined losses from churn and contraction. Companies with strong product-led growth or usage-based pricing often achieve NRR of 120–140% or higher.
How often should I calculate NRR?
Monthly NRR provides the fastest feedback loop and is the most common cadence for internal tracking. Quarterly NRR smooths out month-to-month volatility and is more useful for board reporting. Annual NRR is a useful long-term benchmark but too slow for operational decisions — problems visible in annual NRR were typically detectable in monthly NRR six to twelve months earlier.
What is the NRR formula?
The NRR formula is: NRR (%) = (Starting MRR + Expansion − Contraction − Churn) / Starting MRR × 100. All variables should use only recurring subscription revenue from the existing customer cohort. Non-recurring fees and new customer revenue are excluded.
What's the difference between NRR and revenue retention formula?
The revenue retention formula typically refers to the same NRR calculation. Some contexts use 'revenue retention' more loosely to mean any measure of retained revenue — which could include GRR (gross revenue retention, expansion excluded) or NRR (net revenue retention, expansion included). When precision matters, specify whether you mean gross or net.
Summary
NRR is one of the most consequential metrics in SaaS because it tells you whether your business model is fundamentally sound — whether customers find enough value to stay and pay more over time, or whether you're on a treadmill of constantly replacing revenue that leaves.
Getting the calculation right requires discipline: same cohort at the start and end of the period, expansion limited to recurring revenue only, new customers excluded entirely. When those conditions are met, NRR becomes one of the clearest signals available about the health of your customer relationships and the long-term trajectory of the business.
A number above 100% means the engine is working. Below 100% means you're growing despite the engine, not because of it — and at some point, that catches up.