NRR and GRR are built from the same billing data and measure the same customers. The difference is one line: NRR counts expansion revenue, GRR doesn't. That one line is why a business can report 110% NRR while quietly losing a fifth of its revenue base every year.
This post compares net revenue retention vs gross revenue retention side by side: what each includes, how both come out of the same customer base, what a good result looks like for each, and why investors ask for the pair rather than either number alone.
Table of Contents
- NRR vs GRR: The Short Answer
- What NRR and GRR Each Include
- The Formulas: Expansion, Contraction and Churn
- NRR vs GRR: A Worked Example From the Same Customers
- Why You Need Both: Expansion Can Hide Churn
- How to Read NRR and GRR Together
- What Good NRR and GRR Look Like
- Which One Investors Look At
- Common Mistakes When Comparing NRR and GRR
- How to Track NRR and GRR
- FAQ
NRR vs GRR: The Short Answer
GRR (gross revenue retention) tells you how much of your starting recurring revenue you kept after cancellations and downgrades. It ignores upgrades completely, so it can never go above 100%.
NRR (net revenue retention) starts from the same place, then adds back expansion - upgrades, extra seats, add-ons and usage growth from those same customers. It can go well above 100%.
Put simply, GRR is the floor: what your existing customers are worth if nobody upgrades. NRR is the full picture: what they're worth once upgrades are counted. The gap between them is your expansion revenue, expressed as a share of the starting base.
If you want both numbers for your own business right now, the NRR calculator returns NRR and GRR from the same four inputs.
What NRR and GRR Each Include
Both metrics use a fixed cohort: the customers who were paying at the start of the period. New customers who join during the period are excluded from both. That's what makes them retention metrics rather than growth metrics.
| GRR | NRR | |
|---|---|---|
| Starting MRR from existing customers | Yes | Yes |
| Churned MRR (cancellations) | Subtracted | Subtracted |
| Contraction MRR (downgrades, fewer seats) | Subtracted | Subtracted |
| Expansion MRR (upgrades, seats, add-ons) | Excluded | Added |
| New customer revenue | Excluded | Excluded |
| Can exceed 100%? | No | Yes |
| Question it answers | How much did we keep? | Is the existing base growing? |
You'll also see NRR called net dollar retention (NDR). It's the same metric under a different name. GRR is sometimes called gross dollar retention. If your dashboard, board deck and investor update use different labels, write the formula next to each one.
The Formulas: Expansion, Contraction and Churn
Both formulas start from the same four revenue movements. Here they are side by side:
NRR = (Starting MRR + Expansion MRR - Contraction MRR - Churned MRR) ÷ Starting MRR × 100
GRR = (Starting MRR - Contraction MRR - Churned MRR) ÷ Starting MRR × 100
Where each input comes from in your billing data:
- Starting MRR: recurring revenue from customers active on the first day of the period. In Stripe or Paddle, that's the normalised monthly value of their active subscriptions, with annual plans divided by 12.
- Expansion MRR: increases from those same customers - plan upgrades, added seats, add-ons, usage tiers, and price increases if you choose to count them.
- Contraction MRR: decreases from customers who stayed - downgrades, removed seats, discounts applied mid-term.
- Churned MRR: the full MRR of customers who cancelled, including subscriptions that ended after failed payments.
Two rules keep both numbers honest. Use the same time unit for every input (all monthly or all annual). And keep one-off charges, refunds and taxes out - both metrics are about recurring revenue only.
For the full definitions, see what net revenue retention is and what gross revenue retention is.
NRR vs GRR: A Worked Example From the Same Customers
Here's one SaaS business over twelve months. It started the year with 200 paying customers and $50,000 MRR. During the year, those same 200 customers produced these movements:
- Churned MRR: $6,000 (customers who cancelled)
- Contraction MRR: $2,000 (customers who downgraded)
- Expansion MRR: $9,000 (customers who upgraded or added seats)
Now calculate both from the same inputs:
| Step | GRR | NRR |
|---|---|---|
| Starting MRR | $50,000 | $50,000 |
| Minus churn | - $6,000 | - $6,000 |
| Minus contraction | - $2,000 | - $2,000 |
| Plus expansion | not counted | + $9,000 |
| Ending MRR from the cohort | $42,000 | $51,000 |
| Result | 84% | 102% |
Same customers, same year, two very different stories. NRR says the base grew 2% without any new sales. GRR says 16% of the revenue you started with walked out the door, and expansion from the rest covered the hole.
Neither number is wrong. They answer different questions, and you need both answers.
Why You Need Both: Expansion Can Hide Churn
Take a second business with the identical $50,000 starting base. Over the same year it loses $2,500 to churn and $1,000 to contraction, and gains $3,500 in expansion.
| Business A | Business B | |
|---|---|---|
| Churn + contraction | $8,000 | $3,500 |
| Expansion | $9,000 | $3,500 |
| GRR | 84% | 93% |
| NRR | 102% | 100% |
On NRR alone, Business A looks slightly better. On GRR, Business B is far healthier: it keeps almost all of its revenue, while A has to win back 16% of its base in upgrades every year just to stand still.
That matters because expansion is usually less reliable than retention. In many small SaaS businesses, a large share of expansion comes from a handful of accounts that are growing fast. If two or three of them stop adding seats, or churn themselves, NRR drops sharply and the underlying churn becomes visible all at once. GRR shows that risk a year early.
The reverse also happens. A product with no upgrade path can have excellent GRR and an NRR stuck near the same number. That isn't a retention problem - it's a pricing and packaging question.
How to Read NRR and GRR Together
Look at where each number sits and the size of the gap between them:
- High GRR, high NRR: customers stay and grow. Protect what's working and find out which plans and segments drive the expansion.
- Low GRR, high NRR: expansion is covering a leak. Find the segment that churns - often the smallest plan or one acquisition channel - before you plan around the headline NRR.
- High GRR, low NRR (near 100% or just under): customers stay but rarely pay more. Look at upgrade paths, seat-based pricing, usage limits or add-ons.
- Low GRR, low NRR: the base is shrinking and nothing offsets it. Fix churn and failed payments before spending more on acquisition.
The gap itself (NRR minus GRR) is your expansion rate. A wide gap with a low GRR is the pattern to watch, because the business depends on a few accounts continuing to grow.
What Good NRR and GRR Look Like
These ranges are rough rules of thumb for annual figures, not results from a specific study. They move with contract size, customer type and pricing model, so compare against businesses that sell the way you do.
| Customer type | GRR (rough rule of thumb) | NRR (rough rule of thumb) |
|---|---|---|
| Self-serve SMB / prosumer | 70-85% | 90-105% |
| Mid-market B2B | 85-90% | 100-115% |
| Enterprise B2B | 90%+ | 110%+ |
A few practical reads:
- GRR is the harder number to move and the more honest signal of product fit. For most B2B products, GRR in the 80s or higher means the base holds up.
- NRR above 100% means the existing base grows on its own. For small self-serve SaaS with low prices and few upgrade options, landing near 100% can be a solid result.
- The two numbers should be judged together. 105% NRR on 92% GRR is a stronger position than 110% NRR on 75% GRR.
For a deeper look at NRR targets by stage and pricing model, see what counts as a good NRR.
Which One Investors Look At
NRR is usually the headline. It's the number that ends up in decks and growth-stage comparisons, because above 100% it shows the business can grow from its existing customers.
GRR tends to come up in diligence. Once someone sees a strong NRR, the natural follow-up is "how much of that is covering churn?" GRR answers it, and a healthy GRR next to the NRR makes the story much more believable than NRR alone.
For lower-priced, self-serve products, expect more attention on GRR and logo churn, since expansion is often limited. For larger contracts with seat or usage pricing, NRR carries more weight because expansion is built into the model.
Either way, present both, with the formula and period next to each. Investors are more likely to trust two well-defined numbers than one flattering one.
Common Mistakes When Comparing NRR and GRR
Using different cohorts. Both numbers must come from the same starting customers and the same period. If NRR is annual and GRR is a monthly average, the gap between them means nothing.
Multiplying monthly figures by 12. Retention compounds. A monthly GRR of 98% isn't an annual GRR of 76% exactly (100% minus 12 × 2%) - it's closer to 78% if the rate holds steady, because each month's loss applies to a slightly smaller base. Better still, measure the annual cohort directly.
Counting new customers in the ending number. Total MRR going up doesn't mean NRR went up. New revenue stays out of both metrics.
Counting reactivations as expansion. If a churned customer comes back, decide once whether that's expansion in the original cohort or a new customer, and keep the rule fixed.
Leaving failed payments out of churn. A subscription that ends because the card failed reduces GRR exactly like a deliberate cancellation.
How to Track NRR and GRR
You can get a quick read with the NRR and GRR calculator: enter starting MRR, expansion, contraction and churn, and it returns both numbers.
For an ongoing view, Chartsy pulls the movements straight from Stripe and Paddle, as well as Polar.sh, Lemon Squeezy, Creem.io and Freemius. Expansion, contraction and churned MRR are tracked separately, so you can see NRR and GRR for the same period and break the movements down by plan or country to find which segment is behind the gap.
FAQ
Can NRR ever be lower than GRR?
No. NRR is GRR plus expansion, and expansion can't be negative. At worst, with no upgrades at all in the period, NRR equals GRR. If your report shows NRR below GRR, the two numbers were calculated from different cohorts, periods or data sources.
Do price increases count as expansion in NRR?
They can, but say so. A price increase on existing customers raises NRR without any change in how customers use the product. Many teams report NRR with and without price-driven expansion so the board can see how much came from real upgrades.
Should failed payments count as churn in GRR?
Yes, once the subscription actually ends. A customer whose card fails and never recovers is lost revenue, and GRR should reflect it. Customers who are past due but still active usually stay in the base until the retry window closes.
How big a gap between NRR and GRR is normal?
There's no fixed number. The gap equals your expansion rate, so it depends on whether your pricing lets customers pay more over time. Seat or usage-based products often have a wide gap, flat-priced products a narrow one. What matters is whether GRR holds up on its own.
Do reactivated customers count toward NRR or GRR?
Neither formula has a single standard rule. Some teams count a returning customer as expansion within the original cohort, which lifts NRR only. Others treat them as new customers after an inactive period. Pick one rule, write it down, and keep it consistent across periods.
Should I report NRR and GRR monthly or annually?
Report the annual figure to investors and the board, since that's how most benchmarks are framed. Review monthly movements internally to spot changes early. Just don't mix the two when comparing NRR against GRR.
Written by
Chartsy TeamAnalytics team at Chartsy
The Chartsy Team writes guides, product updates, and resources to help SaaS and eCommerce founders make sense of their metrics, without SQL or spreadsheets.
Chartsy
