What Is NRR (Net Revenue Retention) and Why It Beats ARR in SaaS
NRR is the share of revenue kept from existing customers, new sales excluded. Formula, monthly vs annual NRR, a cohort example and good benchmarks by segment.
TL;DR — Summary
- NRR = (Starting MRR + Expansion − Contraction − Churn) / Starting MRR × 100. New customers are excluded.
- ARR shows the size of revenue; NRR shows its quality — only NRR tells you what happens if new sales stop.
- 100% is the neutral line. Below it the base erodes; above it you have negative churn.
- Benchmarks use annual NRR: today's MRR from customers active 12 months ago, divided by their MRR then. To annualise monthly NRR, multiply the 12 months; 101% a month ≈ 112.7% a year.
- NRR is one of the strongest drivers of valuation multiples; 120%+ earns a premium.
What is NRR and how is it calculated?
NRR (Net Revenue Retention) is the revenue produced at the end of a period by the customers you had at the start, divided by their starting revenue. The formula is simple: NRR = (Starting MRR + Expansion MRR − Contraction MRR − Churned MRR) / Starting MRR × 100.
The critical detail is what it leaves out: revenue from newly acquired customers never enters the calculation. That makes NRR the only metric that tells you whether the business would still grow with the marketing and sales engine switched off. Every other growth metric can be masked by new sales; NRR cannot.
Example: a SaaS starts the month at $40,000 MRR and books $6,000 of expansion (upsell, extra seats), $1,500 of contraction (downgrades) and $2,500 of churn. NRR = (40,000 + 6,000 − 1,500 − 2,500) / 40,000 × 100 = 105%. Revenue grew 5% without a single new customer.
Is NRR monthly or annual?
Benchmarks and investor decks use annual NRR: the MRR of customers active 12 months ago is compared with what the same customers pay today. Monthly NRR is an early warning signal; do not compare it directly with annual thresholds like 120%.
Cohort example: customers active 12 months ago paid $40,000 MRR. Those same customers pay $45,200 today, so annual NRR = 45,200 / 40,000 × 100 = 113%. The $9,000 MRR from customers won during those 12 months stays out; include it by mistake and the result reads 135.5%, overstating the strength of the base.
When annualising, multiply rather than add: Annual NRR ≈ Month 1 NRR × Month 2 NRR × … × Month 12 NRR. If every month is the same, Annual NRR ≈ (Monthly NRR)^12, which is why small monthly gaps become large annual ones.
- 101% a month → about 112.7% a year
- 102% a month → about 126.8% a year
- 99% a month → about 88.6% a year
- 97.3% a month → about 72% a year: roughly a quarter of the base gone in 12 months
- A single 105% month → about 179.6%: unrealistic. A month like that usually comes from one large upgrade or renewal, so do not annualise it
- With annual contracts, monthly NRR swings because renewals cluster in certain months; report only the 12-month cohort figure
NRR vs. ARR and MRR
ARR and MRR are volume metrics: they tell you how large the revenue is. NRR is a quality metric: it tells you how durable and expandable that revenue is. Two companies can both sit at $10M ARR, but if one runs 85% NRR and the other 125%, they are fundamentally different businesses.
The company at 85% NRR must sell 15% of its revenue in new business every year just to stand still. To grow, it has to beat that. That creates a permanent dependence on the sales team and the ad budget — the moment CAC rises, growth stalls.
The company at 125% NRR grows 25% a year with zero new customers. Add new sales on top and the effect compounds. That is why two companies with the same ARR can be valued multiples apart.
- ARR: how big is the revenue?
- Growth rate: how fast is it increasing?
- NRR: how much of that increase comes from existing customers — i.e. how much of it is sustainable?
Why the NRR vs. GRR gap matters
GRR (Gross Revenue Retention) counts losses only: (Starting − Contraction − Churn) / Starting. Because it excludes expansion it can never exceed 100%. In the example above, GRR = (40,000 − 1,500 − 2,500) / 40,000 = 90%.
NRR at 105% while GRR sits at 90% means the base is actually eroding and a handful of expanding accounts is covering it. That structure is fragile — when those few accounts leave, the picture deteriorates fast.
This is why a serious investor conversation asks for both. A healthy profile has strong GRR (90%+ in enterprise, 80%+ in SMB) with NRR pushed above it by expansion.
Healthy NRR benchmarks and the investor view
Common references by segment: 90-100% for SMB-focused SaaS, 100-110% for mid-market, and 120%+ for enterprise and usage-based pricing. Usage-based models naturally produce higher NRR because the invoice grows as the customer grows.
Investors care about NRR for two reasons. First, it makes future revenue predictable. Second, it is the most honest proof that the product delivers value — if a customer spends more, they are satisfied. Surveys cannot prove that; invoices can.
How to improve NRR
Improving NRR is mostly product and customer success work, not sales work. Expansion revenue is a natural consequence of delivered value; it cannot be forced.
- Tie pricing to growth: pick an axis that scales with the customer — seats, usage volume or transactions.
- Shorten time-to-value: faster onboarding lowers churn and accelerates expansion at the same time.
- Make expansion frictionless: adding a seat or module should not require a sales call.
- Build a health score: catch usage decline, champion departure and support-ticket spikes before churn happens.
- Track NRR by cohort and segment: knowing which profile expands also fixes your acquisition targeting.
Frequently Asked Questions
What is NRR?
What is a good NRR?
Why is NRR more important than ARR?
What is the difference between NRR and GRR?
Can you show an NRR calculation example?
Is NRR monthly or annual?
What does negative churn mean?
Resources & Related Reading
Digital marketing specialist building websites, performance ad programs and B2B lead generation systems for B2B and e-commerce brands.
Connect on LinkedIn ↗Last updated: