NRR vs GRR: What's the Difference?
NRR and GRR both measure revenue retention, but they answer different questions. Here's how each works, when to use them, and what good looks like.
The core difference
Net revenue retention (NRR) includes expansion revenue. Gross revenue retention (GRR) doesn’t. That single distinction changes what each metric tells you about your business.
GRR answers: how much of my existing revenue am I keeping? NRR answers: is my existing customer base generating more or less revenue than before?
The formulas
NRR = (Starting MRR + Expansion − Contraction − Churn) / Starting MRR × 100
GRR = (Starting MRR − Contraction − Churn) / Starting MRR × 100
The only difference is that NRR adds expansion MRR to the numerator. GRR ignores it entirely.
A worked example
Say you start the quarter with $200,000 in MRR. Over three months:
| Movement | Amount |
|---|---|
| Expansion (upgrades, add-ons) | +$18,000 |
| Contraction (downgrades) | −$4,000 |
| Churned (cancellations + failed payments) | −$10,000 |
NRR = ($200,000 + $18,000 − $4,000 − $10,000) / $200,000 = 102%
GRR = ($200,000 − $4,000 − $10,000) / $200,000 = 93%
Same business, two very different readings. NRR says you’re growing. GRR says you’re losing 7% of your base revenue. Both are true.
Benchmarks
| Metric | Median B2B SaaS | Strong | Best-in-class |
|---|---|---|---|
| NRR | 100-105% | 110-120% | 120%+ |
| GRR | ~90% | 93-95% | 95%+ |
GRR varies significantly by segment. Enterprise SaaS with annual contracts and high switching costs tends to sit above 95%. SMB products with monthly billing typically land at 85-92%.
When to use each
Use GRR when you want the unvarnished truth about retention. GRR can never exceed 100%, so there’s nowhere to hide. A declining GRR trend means your product is getting less sticky, regardless of how well your sales team upsells.
Use NRR when you’re evaluating overall revenue efficiency. NRR captures the full picture of how your installed base contributes to growth. It’s the metric investors reference most because it signals whether customers are becoming more valuable over time.
In practice, you want both on your dashboard. The gap between them is just as telling as either number alone. A wide gap (e.g. 87% GRR, 115% NRR) means you’re heavily dependent on expansion to compensate for churn. A narrow gap (e.g. 96% GRR, 105% NRR) means your foundation is solid and expansion is a bonus.
How to improve NRR
NRR has three levers: grow expansion, shrink contraction, and reduce churn. Most teams fixate on the first one, but the other two often move faster.
Expansion revenue is the obvious play. Upsells, cross-sells, seat growth, and usage-based pricing all push NRR above 100%. The best SaaS companies design their pricing so customers naturally spend more as they get value. Slack’s per-seat model is the textbook example: as teams grow, revenue grows without a single sales conversation.
Contraction is harder to spot but easier to fix than you’d think. Customers downgrading usually signals a mismatch between their plan and their actual usage. Right-sizing plans proactively, before a customer goes looking for a cheaper option, keeps contraction low. We’ve seen this play out across ChurnWard customers: the ones who monitor feature usage and reach out before renewal tend to hold contraction under 1% of MRR.
Then there’s churn itself. Voluntary churn needs product and customer success work. Involuntary churn, the kind caused by failed payments, needs dunning. The difference matters because one requires months of iteration, and the other can be fixed in a week.
How to improve GRR
GRR improvement is purely defensive. You can’t upsell your way to a better number here.
Start with involuntary churn. Failed payments account for 20-40% of total churn in most subscription businesses, and they sit in the churned MRR line of the GRR formula. Recovering those payments through automated dunning is the single fastest way to improve GRR because it requires zero product changes. A business recovering even 50% of failed payments can move GRR by 2-4 percentage points.
Voluntary cancellations are the harder problem. The fixes are structural: better onboarding so customers reach value faster, proactive engagement when usage drops, and pricing that matches the value customers actually receive. None of these are quick wins, but they compound.
Reducing contraction helps GRR too. If customers are regularly downgrading, it’s worth investigating whether your tier structure makes sense. Sometimes a simpler pricing model with fewer tiers reduces the temptation to drop down.
The involuntary churn lever
A portion of the churn hitting both metrics comes from failed payments, not deliberate cancellations. We’ve seen involuntary churn account for 20-40% of total churn across SaaS businesses. Recovering those payments through dunning improves both NRR and GRR simultaneously, and it’s usually the fastest retention win available because it requires no product changes.
If your GRR is below 90%, it’s worth checking how much of that loss is involuntary before assuming you have a product problem. Sometimes you just have a payments problem.