Gross revenue retention (GRR) measures how much recurring revenue you keep from existing customers after downgrades and cancellations, ignoring upgrades, so it can't exceed 100%. Net revenue retention (NRR) starts from the same number and adds expansion revenue, so it can go well above 100%. GRR shows how sticky you are; NRR shows whether existing customers grow. The gap between them is your expansion rate.
Key takeaways
- Same inputs, one difference: NRR adds expansion revenue, GRR doesn't.
- NRR minus GRR equals expansion as a percentage of starting revenue.
- SaaS Capital's 2025 medians for private B2B SaaS were 101% NRR and 91% GRR.
- High NRR with low GRR is a warning: a few big upgrades are covering a leak.
- Fix GRR first. Then build expansion to push NRR past 100%.
Picture two subscription businesses. Both proudly report net revenue retention of 105 percent.
On paper, they're twins.
The first keeps 95 percent of its revenue every year and adds 10 percent through upgrades. The second keeps only 75 percent and makes up the difference with 30 percent of expansion from a handful of big customers.
One of those businesses is a fortress. The other is one lost key account away from trouble.
You can't tell which is which from NRR alone. You need its partner: gross revenue retention.
What's the difference between GRR and NRR?
Both start with the same group of existing customers and the same starting revenue. Both subtract downgrades (contraction) and cancellations (churn).
The only difference is that NRR adds back expansion revenue and GRR doesn't.
| Gross revenue retention (GRR) | Net revenue retention (NRR) | |
|---|---|---|
| Also called | Gross dollar retention | Net dollar retention (NDR) |
| Subtracts churn and downgrades | Yes | Yes |
| Adds upgrades, add-ons, price rises | No | Yes |
| Can exceed 100% | Never | Yes |
| Answers | "How sticky are we?" | "Are customers growing with us?" |
| Private B2B SaaS median, 2025 (SaaS Capital) | 91% | 101% |
| What drags it down | Cancellations, failed payments, downgrades | The same, plus a lack of things to upgrade to |
How do you calculate GRR vs NRR?
Use the same four inputs for both.
GRR = (Starting MRR − Contraction − Churn) ÷ Starting MRR × 100 NRR = (Starting MRR + Expansion − Contraction − Churn) ÷ Starting MRR × 100
Subtract one from the other and you get a third, very handy number:
NRR − GRR = Expansion ÷ Starting MRR × 100
That gap is your expansion rate. It tells you how hard your upsells, cross-sells and price rises are working.
A worked example: the twins, side by side
Two businesses, both starting at $100,000 MRR (illustrative)
| Business A | Business B | |
|---|---|---|
| Starting MRR | $100,000 | $100,000 |
| Churned MRR | $3,000 | $18,000 |
| Contraction MRR | $2,000 | $7,000 |
| Expansion MRR | $10,000 | $30,000 |
| GRR | 95% | 75% |
| NRR | 105% | 105% |
| Expansion gap (NRR − GRR) | 10 points | 30 points |
Business A could lose all expansion next year and still keep 95 percent of its revenue. Business B would keep 75 percent. If its biggest expanding customer leaves, NRR could fall below 90 percent in a single year.
Same headline. Completely different risk.
Want to run it on your own business? The NRR and GRR calculator does both at once.
NRR and GRR CalculatorNet and gross revenue retention.Open the free tool →Which matters more, GRR or NRR?
My honest answer: GRR first, NRR second.
GRR is the floor. Expansion can only grow revenue from customers who stay, so a leaky base caps how much NRR you can ever build. SaaS Capital's 2025 report says you need gross retention of at least 90 percent to have "a shot at performance parity with your peers."
NRR is the growth engine. The same SaaS Capital data shows how strongly it lines up with growth among private B2B SaaS companies over $1 million in ARR:
| NRR band (SaaS Capital, 2025) | Median annual growth |
|---|---|
| Below 90% | 15% |
| 90% to 100% | 16% |
| 100% to 110% | 21% |
| 110% to 120% | 30% |
| 120% to 130% | 38% |
| Above 130% | 50% |
So the order of work is simple. Plug the leak, then build the engine.
The retention 2x2: diagnose your business in one minute
Plot your two numbers and see where you land. For B2B, I use 90 percent GRR and 100 percent NRR as the lines.
| NRR under 100% | NRR over 100% | |
|---|---|---|
| GRR 90% or higher | Sticky but flat. Customers stay, but there's nothing to buy next. Build a higher tier or add-on. | Compounding. The goal. Protect it and keep adding value. |
| GRR under 90% | Leaking. Fix onboarding and churn before anything else. | Masked leak. Expansion is hiding churn. Find out who's leaving and why. |
Real companies show what the top-right box looks like. Procore reported GRR of 95 percent and NRR of 106 percent for 2025. Brisbane's TechnologyOne reported NRR of 115 percent for FY25. Both sell software that's painful to switch away from.
Why I look at the gap first
When I see a strong NRR, the first thing I want to know is the gross figure. A 30-point gap between the two isn't automatically bad. Usage-based businesses can have big swings. But it tells me growth depends on a few customers spending more, and that's a concentration risk. In Chapter 5 I suggest no single client should be more than 10 percent of revenue. The same thinking applies to expansion.
See your GRR and NRR side by side
Enter four numbers and the free calculator shows both retention rates and the expansion gap between them.
GRR and NRR benchmarks side by side
| Source and year | Segment | Median GRR | Median NRR |
|---|---|---|---|
| SaaS Capital, 2025 | Private B2B SaaS over $1M ARR | 91% | 101% |
| SaaS Capital, 2025 | Bootstrapped | 92% | 104% |
| SaaS Capital, 2025 | Equity-backed | 90% | 101% |
| SaaS Capital, 2025 | Month-to-month contracts | 89% | 100% |
| SaaS Capital, 2025 | Multi-year contracts | 94% | 103% |
| ChartMogul, 2025 data | AI-native products | 40% | 48% |
| ChartMogul, 2025 data | AI-native over $250 a month | 70% | 85% |
| ChartMogul, 2025 data | AI-native under $50 a month | 23% | 32% |
Two things jump out.
First, bootstrapped companies beat equity-backed ones on both measures in SaaS Capital's data. Growing on your own cash seems to force a focus on customers who stay.
Second, cheap AI tools have a retention problem, not an expansion problem. When GRR is 23 percent, no upsell strategy will save you. For more segment detail see the NRR benchmarks.
How your pricing model shapes the gap
Your pricing model decides how far apart these two numbers can ever get.
One flat price. If everyone pays the same and there's nothing to upgrade to, NRR and GRR will sit almost on top of each other. The only expansion is a price rise. Most local subscriptions start here.
Good-better-best tiers. Tiers create room to expand, so a gap opens up. How wide depends on whether customers actually need the next tier, which is why tier design matters so much. See packaging: what goes in each tier.
Usage-based pricing. This produces the widest swings in both directions. In Chapter 4 I explained one of its hidden benefits: when a customer's business shrinks, they scale down instead of cancelling. That shows up as contraction rather than churn. You keep the customer, but GRR takes the hit. Then, when they grow again, NRR jumps. Procore lived through exactly this as construction slowed.
Seats. Seat-based pricing expands when teams grow and contracts when they shrink. In 2026 that's a real risk, because AI lets customers do the same work with fewer people. WiseTech removed seat fees from CargoWise in December 2025 for precisely that reason.
So before you judge your gap, ask whether your pricing gives customers anywhere to grow.
Mistakes people make comparing the two
- Different starting groups. Both must use the same customers who were there at the start of the period.
- Different periods. Don't compare a monthly GRR with an annual NRR.
- Counting the full upgrade price as expansion. Only the increase counts. A $20 customer moving to $50 is $30 of expansion.
- Reporting only NRR. If you show an investor or buyer one number, show both. A strong gross figure is what makes the net figure believable.
How the book frames it
Chapter 9 is about net negative churn, which is NRR above 100 percent. But I opened the chapter by saying "churn is one of the biggest killers of subscription businesses." That's GRR.
Chapter 7, land and expand, is the playbook for the gap. Land with a small, attractive offer, then expand. The gap between GRR and NRR is how you measure the "expand" part. And Chapter 9's tips on upselling, cross-selling and expanding your resources are how you widen it. More on that in expansion revenue.
What this looks like for a gym or a cafe
A local gym rarely gets NRR above 100 percent. Members don't add seats. But the two numbers still tell you where to focus.
If your gym keeps 80 percent of its membership revenue over a year and adds 5 percent through personal training packs and family add-ons, you're at 80 percent GRR and 85 percent NRR (illustrative figures). The 5-point gap says your upsell exists but is small. The 20 points of leakage says retention is the bigger prize.
Move ten members a month from "leaving" to "staying" and you'll beat anything a new add-on can do.
What to do next
- Calculate GRR and NRR for the last 12 months from the same customer list.
- Place yourself in the 2x2 above.
- If you're leaking, start with how to reduce churn. If you're flat, design your next tier with the pricing table builder.
- If you're thinking about selling or raising money, run both numbers through the valuation estimator.
Frequently asked questions
Which do investors care about more, GRR or NRR?
Both, for different reasons. NRR is the headline because it drives growth: SaaS Capital's 2025 data shows companies above 130% NRR grew at a median of 50% a year. GRR is the due-diligence check because it shows how durable revenue is if expansion slows. A strong NRR on weak GRR tends to get marked down once a buyer looks closely.
Can NRR be lower than GRR?
No. NRR is GRR plus expansion, and expansion can't be negative. If your NRR comes out lower than your GRR, something in the calculation is wrong, usually new customers in one figure and not the other, or contraction counted twice. The lowest NRR can ever be is equal to GRR, in a year with no upgrades at all.
Should GRR and NRR use the same time period?
Always. Calculate both from the same starting group of customers over the same 12 months. If you compare a monthly GRR with an annual NRR, the numbers are meaningless together. Most benchmarks, including SaaS Capital's and ChartMogul's, are annual, so make your headline figures annual too.
Where do price increases go?
Into NRR as expansion, never into GRR. A price rise on existing customers lifts NRR even if nobody chooses to buy more. That's legitimate revenue, but it's worth reporting separately so you know how much of your NRR is pricing and how much is customers choosing to expand.
Is there a quick way to estimate one from the other?
Only if you know your expansion. NRR equals GRR plus expansion revenue as a percentage of starting revenue. So if your GRR is 90% and existing customers added upgrades worth 12% of last year's revenue, your NRR is 102%. Without the expansion figure you can't convert between them.
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)
- ChartMogul, SaaS retention: the AI churn wave (10 Dec 2025)
- Procore, Q4 and full-year 2025 results (12 Feb 2026)
- TechnologyOne, FY25 full-year results presentation (Nov 2025)
- The Subscription Playbook, Chapters 4, 7 and 9 (Robert Coorey, 2022)
Facts and prices checked 27 October 2026. Spotted something out of date? Tell us.