Get FP&A best practices, research reports, and more delivered to your inbox.
Last updated: July 2026. Benchmarks reflect full-year 2025 data
The median B2B SaaS company has a net revenue retention (NRR) rate of 102% — just above the 100% break-even line. Top-quartile companies reach 110%. The sharpest split is by pricing model: usage-based companies post a median NRR of 108% versus 98% for seat-based, a 10-point structural gap. These figures come from full-year 2025 data across 342 SaaS and AI-native companies in the 2026 Aleph × Benchmarkit SaaS & AI Performance Benchmarks.
A 102% median means the typical SaaS company now grows its existing customer base by 2% a year before adding a single new logo. That is positive, but thin — any slip in retention or expansion pushes the cohort below break-even and back onto the new-logo treadmill.
{callout}
Bottom line: 100% is the floor, 110%+ is top-quartile, and 120%+ is best-in-class. The 2025 median is 102%. The single biggest lever is pricing model — usage-based companies clear the bar by 8 points while seat-based companies sit below it at 98%.
{/callout}
What's a good NRR rate in 2026?
A good net revenue retention rate is at or above 100%, with clear tiers above it:
- 120%+ — best-in-class. Strong organic expansion, common in usage-based models.
- 110% — top quartile. Meaningful growth from the installed base alone.
- 102% — median. Above break-even, but with little margin for error.
- 92% — bottom quartile. Below 100%: these companies must replace 8% of ARR through new logos every year before they grow at all.
NRR is the most direct measure of whether your existing customers fund your growth or quietly drain it — one of the core SaaS metrics that matter.
What's the difference between NRR and GRR?
These two retention metrics answer different questions, and you need both:
- Net Revenue Retention (NRR) includes expansion — upsell, cross-sell, and usage growth. It can exceed 100%. 2025 median: 102%.
- Gross Revenue Retention (GRR) excludes expansion; it measures only what you kept, so it caps at 100%. 2025 median: 84%.
The gap between them is your expansion engine. A 102% NRR sitting on top of an 84% GRR means expansion is doing heavy lifting to cover real churn underneath.
{callout}
Definition: NRR = (starting ARR + expansion − contraction − churn) ÷ starting ARR. GRR uses the same formula without the expansion term, so it can never exceed 100%.
{/callout}
How do you calculate net revenue retention?
NRR = (Starting ARR + Expansion ARR − Contraction ARR − Churned ARR) ÷ Starting ARR × 100
Measured for a fixed cohort of customers over a set period (usually 12 months), and it excludes any new logos acquired during the period. The most common error is letting new-logo revenue leak into the expansion term — which inflates NRR and hides a retention problem. In usage-based models especially, how you draw the line between “new ARR” and “expansion ARR” materially changes the number.
Why does pricing model drive NRR?
Pricing architecture is the strongest structural determinant of NRR in the 2025 data — and the gap is widening. Usage-based companies post a 108% median NRR; seat-based companies sit at 98%, below the break-even line. Usage-based pricing creates organic expansion: as customers use more, revenue grows automatically, without a renegotiation. The 75th percentile for usage-based models reaches 155%.
Seat-based pricing faces the opposite pressure. In an era of AI-driven headcount efficiency, seat counts are under active threat — customers consolidating roles or deploying AI agents shrink their own seat count, dragging NRR down. The report's framing is blunt: seats are under attack, and seat-based NRR needs active management or a pricing-model evolution to hold the line.
This is the same dynamic that gives hybrid pricing the strongest Rule of 40 profile, and it compounds: every year, the usage-based cohort grows its installed base about 10 points faster than the seat-based cohort. For a related SaaS benchmark, see Magic Number.
How does NRR vary by growth rate?
Expansion and growth reinforce each other — fast growers do not rely on new logos alone:
Companies growing above 50% are not doing it on new logos alone — product adoption and an effective post-sale motion fuel expansion at scale. Low-growth companies at 92% NRR face the opposite: a shrinking installed base compounds the new-logo challenge into a deeply unfavorable growth equation.
How does NRR vary by company size and deal size?
Scale and deal size both help expansion economics:
- >$100M ARR: 103% median (75th percentile 115%). Scale brings more product surface area and stronger customer-success infrastructure.
- $20M–$50M ARR: 101% median. The inflection point where expansion infrastructure is being built but has not fully paid off.
- <$5M ARR: 94% median. Too little installed base and CS capacity to drive systematic expansion.
- By deal size: the $25K–$50K ACV band leads at 105%, while sub-$5K ACV sits at 98%. Notably, 2025 was the first year the $10K–$25K segment dropped below 100% — a new warning signal.
Is NRR getting harder to hold?
Yes, at the bottom and the top. The 25th percentile slipped from 95% in 2024 to 92% in 2025, worse than the 2023 trough — the weaker half of the market has not recovered. Even the 75th percentile eased from 110–111% to 108%. The report attributes the pressure to buyer budget scrutiny, slower upsell cycles, and deferred commitments as buyers evaluate AI-native alternatives in parallel. Expansion is harder to come by than it was two years ago, which makes pricing model and a deliberate expansion motion more decisive, not less.
How should finance teams benchmark and improve their own NRR?
- Report NRR and GRR together. NRR alone can mask churn that expansion is papering over. The spread between them is the metric that tells the real story.
- Benchmark by pricing model first. A seat-based company at 100% is outperforming its cohort; a usage-based company at 100% is underperforming its. The all-in 102% median hides this.
- Separate new ARR from expansion ARR cleanly. Especially under usage-based pricing, a sloppy definition inflates NRR and hides a retention problem you need to see.
- Treat expansion as a revenue function. The report's broader finding is that expanding a customer costs about half what acquiring one does ($0.80 vs $1.63 per dollar of new ARR) — so the cohorts winning on NRR resource expansion deliberately, not as a CS afterthought.
Getting this right depends on a clean cohort view of starting ARR, expansion, contraction, and churn — which is exactly what most teams rebuild by hand each quarter. Aleph ties the billing and CRM data into a live model so NRR, GRR, and the expansion-versus-churn breakdown stay current and board-ready, segmented by the cuts that actually move the number.
See how finance teams track NRR, GRR, and expansion in Aleph → Book a demo.
Methodology and sources
These benchmarks come from the 2026 SaaS & AI Performance Benchmarks report, published jointly by Aleph and Benchmarkit on June 1, 2026. The report draws on 342 B2B SaaS and AI-native software companies; NRR figures are based on the 230 participants that reported the metric. Figures reflect full-year 2025 (CY-2025) actuals. The underlying metrics are explorable in Benchmarkit's interactive benchmarks.
{callout}
A note on the year: This report was published in 2026, but the benchmarks reflect full-year 2025 results — the latest complete data. Where this page says “2025,” it means the data year. “2026” refers to the report edition and the current planning year.
{/callout}
This page is reviewed against each new edition of the benchmark data.
Get FP&A best practices, research reports, and more delivered to your inbox.
