A good net revenue retention rate depends on who you sell to. For private B2B SaaS companies over $1M ARR, SaaS Capital's 2025 survey found a median of 101%, with the upper quartile around 111% in the $25k–$50k ACV band. ChartMogul's 2025 billing data showed lower medians: 82% for B2B SaaS, 49% for consumer SaaS and 48% for AI-native products. Above 100% is the B2B target.
Key takeaways
- Private B2B SaaS median NRR: 101% (SaaS Capital, 2025). Upper quartile in the $25k–$50k ACV band: 111%.
- ChartMogul's broader 2025 data: B2B median 82%, upper quartile 97%; B2C 49%; AI-native 48%.
- Price point is the strongest predictor. AI-native products over $250 a month had 85% NRR; under $50, 32%.
- Annual plans showed median NRR 10 to 20 points higher than monthly plans (ChartMogul, 2025).
- Boom-era figures like Snowflake's 158% are outliers from 2020. Don't benchmark against them.
When I wrote The Subscription Playbook, the net dollar retention numbers were dazzling.
Snowflake: 158 percent. JFrog: 139 percent. DocuSign: 125 percent.
Those figures came from 2020 and 2021, a boom when customers were buying software as fast as it could be shipped. They made a great story for Chapter 4. They make a terrible benchmark for your business in 2026.
So let's look at what retention actually looks like now, segment by segment, with every number sourced and dated. If you need the basics first, start with what net revenue retention is and how to calculate it.
The headline numbers
| Source | Data year | Sample | Median NRR | Upper quartile |
|---|---|---|---|---|
| SaaS Capital | 2025 survey | 1,000+ private B2B SaaS, over $1M ARR | 101% | 111% ($25k–$50k ACV band) |
| ChartMogul | 2025 | ~2,700 B2B SaaS, over $250k ARR | 82% | 97% |
| ChartMogul | 2025 | ~600 B2C (consumer) SaaS | 49% | not reported |
| ChartMogul | 2025 | ~200 AI-native companies | 48% | not reported |
| KeyBanc and Sapphire Ventures | released Nov 2025 | Private SaaS companies | "above 100%" | not reported |
Two surveys, two very different medians. Why?
SaaS Capital surveys established companies (over $1M ARR) that report their own numbers. ChartMogul reads billing data directly, from a much wider spread of businesses, including many small self-serve products with low prices. Both are right about their own population.
The practical answer: if you sell to businesses at a meaningful price, benchmark against SaaS Capital. If you sell self-serve or to consumers, ChartMogul's figures are closer to your world.
NRR by funding and contract length
From SaaS Capital's 2025 survey of private B2B SaaS companies over $1M ARR:
| Segment | Median NRR | Median GRR |
|---|---|---|
| Bootstrapped | 104% | 92% |
| Equity-backed | 101% | 90% |
| Multi-year contracts | 103% | 94% |
| Annual contracts | 101% | 90% |
| Month-to-month | 100% | 89% |
ChartMogul's 2025 analysis found the same direction, more strongly: median NRR for annual plans was 10 to 20 percentage points higher than for monthly plans. Longer commitments mean more time to prove value and fewer chances to cancel. I cover the trade-offs in annual vs monthly plans.
NRR by price point
This is where the gaps get dramatic.
| Source | Segment | NRR |
|---|---|---|
| ChartMogul, 2025 | AI-native, over $250 a month | 85% |
| ChartMogul, 2025 | AI-native, $50–$249 a month | 61% |
| ChartMogul, 2025 | AI-native, under $50 a month | 32% |
| ChartMogul, 2023 report | Top quartile, ARPA over $500 a month | 109.3% |
| ChartMogul, 2023 report | Top quartile, ARPA under $10 a month | 65.1% |
| SaaS Capital, 2025 | $25k–$50k ACV: bottom quartile / median / top quartile | 97% / 102% / 111% |
The pattern holds everywhere. The more a customer pays, the more likely they are to stay and to grow. In ChartMogul's 2023 data, only 2.7 percent of businesses charging under $10 a month had NRR above 100 percent, against 41.1 percent of those charging over $500.
If you're building a cheap AI tool, this table is a warning. Price may be doing more damage to your retention than your product is.
Compare your NRR with these benchmarks
Calculate your net and gross retention in two minutes with the free NRR and GRR Calculator.
NRR by company size
| Source | ARR band | NRR |
|---|---|---|
| ChartMogul, 2023 report (top quartile) | $1M–$3M | 94% |
| ChartMogul, 2023 report (top quartile) | $3M–$15M | 99% |
| ChartMogul, 2023 report (top quartile) | $15M–$30M | above 105% |
| Bessemer, portfolio 2010–1H21 (average) | $1M–$10M | 140% |
| Bessemer, portfolio 2010–1H21 (average) | $100M+ | 120% |
Treat the Bessemer figures as a ceiling, not a target. They come from one top venture firm's portfolio during the best decade software has ever had. Bessemer also notes that SMB-focused companies generally have lower NRR than enterprise ones, comparing Mindbody at 109 percent on roughly $2,000 contracts with Okta at 123 percent on $50,000-plus contracts.
What NRR does to growth
From SaaS Capital's 2025 survey (private B2B SaaS over $1M ARR, median annual growth 24%):
| NRR | Median growth |
|---|---|
| Below 90% | 15% |
| 90%–100% | 16% |
| 100%–110% | 21% |
| 110%–120% | 30% |
| 120%–130% | 38% |
| Above 130% | 50% |
That's the real reason investors obsess over this number, and it feeds straight into valuation. Try your figures in the valuation estimator.
Subscription Business Valuation EstimatorEstimate a revenue multiple range from growth, retention and margin.Open the free tool →Public company NRR, then and now
| Company | NRR | Period | Notes |
|---|---|---|---|
| Snowflake | 158% | 2020 IPO | From the book, Ch 4 |
| JFrog | 139% | 2020 IPO | From the book, Ch 4 |
| DocuSign | 125% | 2021 | From the book, Ch 7 |
| SendGrid | 112%, 111%, 117% | 2015, 2016, 2017 | From the book, Ch 9 |
| Linear | 177% | 2026 | Company-reported via ARR Club |
| TechnologyOne | 115% | FY25 | Australian enterprise SaaS |
| Procore | 106% (GRR 95%) | FY2025 | Usage-based pricing |
| Zoom (enterprise) | 98% net dollar expansion | FY2026 | Slight shrinkage on balance |
| Udemy Business | 93% (97% large customers) | FY2025 | Before merging with Coursera |
Procore is the clearest example of the reset. Its volume-based pricing grew with builders in the boom, and contracted with them as construction slowed. The Procore case study shows what changed.
My reading: rough bands by segment
These are my own rules of thumb from the sources above, not a published benchmark.
| Who you sell to | Concerning | Solid | Excellent |
|---|---|---|---|
| B2B, $25k+ annual contracts | under 95% | 100–110% | over 115% |
| B2B, small business self-serve | under 80% | 90–100% | over 100% |
| Consumer subscriptions | falling year on year | stable | over 80% |
| AI-native, under $50 a month | under 30% | 40–60% | over 70% |
Whatever your band, the trend matters more than the level. A business moving from 88 percent to 96 percent is in better shape than one sliding from 112 percent to 104 percent.
What to do next
- Calculate your own NRR and GRR over the last 12 months with the NRR and GRR calculator.
- Find the row above that matches your price point and billing period.
- If you're below it, check gross revenue retention first. Most NRR problems are churn problems.
- If you're above it, protect it. The Chapter 9 playbook on reducing churn still applies.
- For benchmarks across every metric, see our compiled subscription benchmarks for 2026.
Frequently asked questions
What NRR do investors expect at seed or Series A?
Few published sources split NRR by funding round, so be wary of anyone quoting a precise figure. What the data does show is that NRR and growth move together. In SaaS Capital's 2025 survey, companies with NRR of 110% to 120% grew at a median of 30%, against 16% for those at 90% to 100%. Investors look for that link, so show both.
Is 90% NRR bad?
For B2B SaaS over $1M ARR, it's below the 2025 median of 101% (SaaS Capital), and SaaS Capital found companies under 90% grew at a median of only 15%. For a consumer subscription it can be perfectly respectable, since ChartMogul's 2025 median for B2C SaaS was 49%. Compare yourself with your own segment, price point and billing period.
Why is ChartMogul's median NRR so much lower than SaaS Capital's?
They measure different populations in different ways. SaaS Capital surveys more than 1,000 private B2B companies over $1M ARR, which report their own figures. ChartMogul reads billing data directly from about 3,500 software companies over $250k ARR, including many smaller, self-serve businesses with lower prices. Smaller and cheaper products churn more, which pulls the median down.
How often are NRR benchmarks updated?
SaaS Capital publishes its retention benchmarks annually from a survey run early in the year; the 2025 edition came out in September 2025. ChartMogul publishes retention reports and updates its metric pages periodically; its latest analysis was December 2025. KeyBanc and Sapphire Ventures release their private SaaS survey each year. We update this page when new editions appear.
Does company size change what good NRR looks like?
Yes. ChartMogul's 2023 report found top-quartile NRR of 94% for companies with $1M to $3M ARR, 99% for $3M to $15M, and above 105% for $15M to $30M. Bessemer's older portfolio data (2010 to 2021) showed the opposite pattern among venture-backed leaders, with averages of 140% at $1M to $10M ARR. Sample matters as much as size.
Get both books free
The Subscription Playbook (ebook + audiobook) and Feed a Starving Crowd, instantly. In Australia? Claim the free hardcover.
Robert Coorey, MBA
Robert is the author of The Subscription Playbook and Feed a Starving Crowd and a Director at BPAG Accountants in Sydney. More about Robert.
Sources
- SaaS Capital, 2025 B2B SaaS Retention Benchmarks (Sept 2025)
- SaaS Capital, What is a good retention rate for a private SaaS company in 2025? (18 Sept 2025)
- ChartMogul, SaaS retention: the AI churn wave (10 Dec 2025)
- ChartMogul, SaaS Retention Report (Nov 2023)
- KeyBanc Capital Markets and Sapphire Ventures, Private SaaS Company Survey press release (13 Nov 2025)
- Bessemer Venture Partners, Scaling to $100 Million (portfolio data 2010–1H21)
- Procore, Q4 and full-year 2025 results (12 Feb 2026)
- Udemy, Q4 and full-year 2025 results (4 Feb 2026)
Facts and prices checked 27 October 2026. Spotted something out of date? Tell us.