Gross retention vs net retention comes down to whether expansion counts. Gross retention (GRR) keeps only what customers hold on to, and net retention (NRR) adds what they grow. GRR tops out at 100% because nothing is added, so for the same cohort NRR is never lower than GRR.
Comparing NRR with GRR matters because a high NRR can hide a leaking base. For the same $100,000 cohort, losing $12,000 to churn and downgrades and adding $17,000 in upsells gives 88% gross retention and 105% net retention.
Below, that one cohort runs through every section. You get a terms table, a list of calculation mistakes, benchmarks from 2023 or later by contract size, and a decision table.
Gross vs Net Revenue Retention at a Glance
Revenue retention is the share of recurring revenue from a starting group of customers that is still there after a period. Gross revenue retention stops at what the cohort keeps, while NRR also counts its expansion.
The table compares the two metrics on the six questions teams ask most.
| Question | Gross revenue retention (GRR) | Net revenue retention (NRR) |
|---|---|---|
| What counts | Starting revenue minus churn and contraction | Starting revenue minus churn and contraction, plus expansion from the same customers |
| What it ignores | Expansion and new customers | New customers and new logos |
| Range | 0 to 100% | 0% and up, so it can pass 100%; NRR is never lower than GRR for the same cohort |
| Answers the question | How much of the base did we keep? | Did existing customers grow revenue? |
| Who reads it first | CS and finance, on churn risk (our reading) | Board and investors, on growth efficiency (our reading) |
| What a drop means | More churn or downgrades, so the base leaks | Less expansion or more leakage; check GRR to tell which |
What Is Gross Retention (GRR) and How Do You Calculate It?
Gross revenue retention (GRR) is the share of starting recurring revenue a cohort keeps after churn and contraction, ignoring expansion and new customers. Investors call it gross dollar retention.
The gross retention formula is:
GRR = (starting recurring revenue minus churned revenue minus contraction) ÷ starting recurring revenue × 100
- Starting recurring revenue: what the cohort paid on day one of the period.
- Churned revenue: revenue from customers who left.
- Contraction: revenue lost from customers who stayed but pay less.
On the shared cohort, $100,000 minus $8,000 of churn minus $4,000 of contraction leaves $88,000, so GRR is 88%.
Contraction covers downgrades, seat cuts and discounts given to keep a customer, and it lowers GRR the same way churn does.
GRR tops out at 100%: GRR revenue is only what the starting cohort already paid, so the ceiling is keeping every dollar.
What Is Net Retention (NRR) and How Do You Calculate It?
Net revenue retention (NRR) is gross retention plus expansion from the same customers, such as upgrades, add-ons, extra seats and price increases. It is also called net dollar retention.
The net retention formula is:
NRR = (starting recurring revenue minus churn minus contraction plus expansion) ÷ starting recurring revenue × 100
- Expansion: extra recurring revenue from customers already in the starting cohort.
On the shared cohort, $88,000 plus $17,000 of expansion gives $105,000, so NRR is 105% against a GRR of 88%.
A 120% NRR means $120 is left for every $100 the cohort started with, or $120,000 on a $100,000 cohort, even if some customers left.
New customers and new logos are never expansion, because NRR counts only revenue from the starting cohort.
Terms That Get Mixed up with GRR and NRR
Five terms sit near GRR and NRR: two are the same ratios under investor names, and three measure something else.
| Term | What it measures | Includes expansion? | How it relates to GRR and NRR |
|---|---|---|---|
| Net dollar retention (NDR) | Revenue kept plus expansion from the starting cohort | Yes | Another name for NRR |
| Gross dollar retention (GDR) | Revenue kept from the starting cohort, before expansion | No | Another name for GRR |
| Logo or customer retention | The share of customers kept, counted by customers, not dollars | No | Ignores contract size, so it can differ from GRR |
| MRR | One month’s total recurring revenue | It is a total, so it holds whatever revenue exists | An input for both ratios, not a rate |
| ARR | The annual total of recurring revenue | It is a total, so it holds whatever revenue exists | The base the ratios start from, not a rate |
MRR vs NRR is a total against a ratio: MRR is one month’s recurring revenue, while NRR compares the same customers’ revenue at two dates.
A 90% customer retention rate and a 90% GRR are different claims, because logo retention counts customers and GRR counts dollars. Whether the lost customers were large or small decides how far apart the two sit.
Mistakes That Distort Both Numbers
SaaS retention rate calculation goes wrong in six places once a real spreadsheet replaces a clean example.
- New-customer revenue in NRR: adding $12,000 from new logos turns the cohort’s 105% into 117%. Fix: freeze the customer list on day one.
- Expansion in GRR: counting the $17,000 of expansion lifts GRR from 88% to 105%, and a GRR above 100% is always this error. Fix: GRR subtracts only.
- Price rises booked as expansion: if $6,000 of the $17,000 is a price rise, NRR is 105%, but 99% without it. Fix: report the price share separately.
- Monthly rate against an annual benchmark: a 99% monthly GRR compounds to 88.6% a year, and an 88% annual GRR is about 98.94% a month. Fix: annualize before comparing.
- A period that does not match the cohort: a mid-year signup measured on a calendar year lands in different totals than finance uses. Fix: one cohort, one 12-month window.
- Reactivations netted out of churn: netting $5,000 of returning customers against churn shows GRR at 93% instead of 88%. Fix: in our approach, count a returning customer as new revenue in the year it returns.
Every figure comes from the shared $100,000 cohort in our examples script, and the mistakes are the ones the ranking pages’ clean examples leave out.
What Good Looks Like: GRR and NRR Benchmarks
Revenue retention benchmarks differ by method and sample, so read each row with its source. The rows come from SaaS Capital, Sep 2025, Benchmarkit, 2025, KeyBanc Capital Markets and Sapphire Ventures, Nov 2025 and ChartMogul, Dec 2025.
| Segment | GRR | NRR | Source | Data year | What it measured |
|---|---|---|---|---|---|
| ACV under $12k | 90% | 98% | SaaS Capital | 2024 | Median, private B2B SaaS survey |
| ACV $12k to $25k | 91% | 103% | SaaS Capital | 2024 | Same survey, median |
| ACV $25k to $50k | 91% | 102% | SaaS Capital | 2024 | Same survey, median |
| ACV $50k to $100k | 90% | 104% | SaaS Capital | 2024 | Same survey, median |
| ACV $100k to $250k | 91% | 102% | SaaS Capital | 2024 | Same survey, median |
| ACV over $250k | 95% | 106% | SaaS Capital | 2024 | Same survey, median |
| All respondents | 88% | 101% | Benchmarkit and Pavilion | 2024 | Median of survey respondents |
| Private SaaS | 86% | Above 100% | KeyBanc Capital Markets and Sapphire Ventures | 2023 | Private company survey |
| B2B SaaS on ChartMogul | Not reported | 82% | ChartMogul (vendor) | 2025 | Billing data, about 2,700 B2B SaaS companies, many small self-serve |
SMB (ACV under $12k) posts a lower NRR, 98%, than enterprise at 106%, in our reading.
What is a good gross retention rate? Below $250k ACV, about 90%, because SaaS Capital’s private B2B SaaS medians are 90% to 91%, while the overall panels sit at 86% to 88%. Above $250k ACV the median is 95%, so 90% falls short. These panels measure dollars, so convert a 90% logo retention rate to GRR first.
What is a good net revenue retention rate? For private B2B SaaS above $12k ACV, SaaS Capital’s medians run 102% to 106%, and 98% below $12k.
Reading the Gap Between GRR and NRR
The gap between GRR and NRR equals expansion as a share of starting revenue, so a wide gap means expansion is covering for leakage. On the shared cohort the gap is 17 points, 105% against 88%.
| Gap size | What it usually means | What to check first |
|---|---|---|
| Small, with high GRR | A sticky base with little expansion | Packaging and add-ons |
| Medium | Expansion and leakage are both active | Which customers expand and which shrink |
| Wide, with low GRR | Expansion is masking churn | Churn and contraction by cohort |
These rows are our reading and carry no benchmark thresholds. A few large expanding accounts can carry NRR while many others shrink, so check expansion by customer, not only in total.
Three Companies, Same NRR
Three companies can report the same NRR on a $100,000 starting cohort and still be three different businesses.
| Company | Churn | Contraction | Expansion | GRR | NRR |
|---|---|---|---|---|---|
| A | $3,000 | $1,000 | $9,000 | 96% | 105% |
| B | $8,000 | $4,000 | $17,000 | 88% | 105% |
| C | $15,000 | $7,000 | $27,000 | 78% | 105% |
Company A’s 9-point gap says its base is sticky, and company C’s 27-point gap says expansion is hiding heavy leakage.
The three companies share the $100,000 starting cohort and are computed in our examples script; only churn, contraction and expansion change.
Takeaway: Company C needs a churn fix before any expansion plan, because its NRR depends on finding $27,000 of new upsell every year.
Gross Retention vs Net Retention: Which One Do You Read First?
Read gross retention first, because it shows whether the base leaks, then read net retention to see whether expansion makes up for it. In any net vs gross retention review, that order stops a healthy NRR from hiding a shrinking base.
| Situation | Read first | First action |
|---|---|---|
| GRR below 85% | GRR | Find churn and contraction by cohort before any upsell push |
| GRR 90% or higher with NRR below 100% | NRR | Build an expansion motion with packaging and add-ons |
| Raising money | Both, NRR leads | Tell the growth story with NRR and show GRR holding it up (our reading) |
| Planning CS and sales headcount | GRR, then NRR | GRR sizes CS capacity, NRR sets the expansion quota |
| A board question | Both | Show both on the same cohort, with the gap (our reading) |
| A pricing change | GRR, then NRR | Watch GRR for downgrades in the next two renewals, and show price rises separately in NRR |
McKinsey & Company, Nov 2025 found top-quartile-valued B2B SaaS companies at 113% NRR against 98% for bottom-quartile peers. That is a correlation across more than 100 companies, not proof that NRR causes the valuation.
The 85% and 90% marks sit around the 86% to 91% panel medians.
What a Few GRR Points Cost over Time
With no expansion, a 90% GRR leaves 59% of starting revenue after five years, against 77.4% at 95%.
| GRR | After 1 year | After 3 years | After 5 years |
|---|---|---|---|
| 80% | 80.0% | 51.2% | 32.8% |
| 90% | 90.0% | 72.9% | 59.0% |
| 95% | 95.0% | 85.7% | 77.4% |
Each row raises the annual GRR to the power of the year count in our examples script, assuming no expansion and no new customers.
Takeaway: Every GRR point lost must be won back with new or expansion revenue each year, so fix leakage before raising the expansion target.
How to Raise Each Metric
Protect GRR first, then expand, because expansion revenue gets used up replacing lost revenue.
| Lever | Moves GRR or NRR | What to measure |
|---|---|---|
| Onboarding | GRR | Time to first value and 90-day churn |
| Support response | GRR | First response time against renewals |
| Health-score outreach | GRR | Save rate on flagged accounts |
| Downgrade saves and pricing | GRR | Contraction dollars per quarter |
| Upsell and add-ons | NRR | Expansion revenue per customer |
| Renewal timing | Both | Renewals closed before the end date |
ChurnZero’s 2025 Customer Revenue Leadership Study with 6sense (n = 793) found NRR of 98% for teams with customer success managers against 90% without. That is vendor data and an association, not proof of cause.
Which Gap Is Yours?
Gross retention vs net retention is a reading order, GRR first and NRR second, and never NRR alone. Run your MRR through the churn rate calculator to get gross and net revenue churn from one set of counts. Which row of the gap table does your own GRR-NRR gap fall in today?
Frequently Asked Questions
What Is the Difference Between NRR and ARR?
NRR is a ratio on one starting cohort, while ARR is a total that also grows from new customers. Take a $1,200,000 cohort at 95% NRR: it is worth $1,140,000 a year later. Add $300,000 from new logos and total ARR reaches $1,440,000, up 20%, while NRR is still 95%. A growing ARR line can therefore sit on a shrinking base.
Gross Dollar Retention vs Net Dollar Retention: What Is the Difference?
Gross dollar retention (GDR) and net dollar retention (NDR) are the same ratios as GRR and NRR under investor names. GDR excludes expansion and NDR includes it. A 110% NDR in a deck and a 110% NRR from finance compare only on the same cohort. Both also need the same rules for price rises and reactivations. Ask which rules were used first.
What Is the Difference Between MRR and NRR?
NRR is calculated from MRR, but only from the starting cohort’s MRR, then and now, so one month’s total MRR cannot give it. If those customers paid $100,000 in MRR a year ago and pay $105,000 now, NRR is 105%. Total MRR today may be $140,000 once new customers join, but that figure is not NRR, and dividing it by $100,000 would wrongly give 140%.
What Is the Difference Between Net Logo Retention and Gross Logo Retention?
Gross logo retention counts only the customers you lost, while net logo retention also adds back reactivated or regained customers. Both count customers, not dollars, so they can look healthy while GRR falls. Take 1,000 customers with 80 lost: logo retention is 92%. If the lost customers averaged $25,000 against a $10,000 base average, GRR is 80%.
Our Gross Retention Is High but Our Net Retention Is Low. What Does It Mean?
Gross retention that is high while net retention is low means the base is not leaking, but expansion is weak. Check the upsell and add-on rate per customer, and whether price rises are counted as expansion. Price-led expansion can flatter NRR while product expansion stays flat, so compare NRR with and without price rises.
Is Gross Retention the Same as 1 Minus the Churn Rate?
No. Gross retention is not 1 minus the churn rate, because it also subtracts contraction. An 8% churn rate alone suggests 92%, but adding 4% contraction from downgrades and seat cuts gives a GRR of 88%. The two match only when no customer pays less without leaving, so check contraction before you quote 1 minus churn as GRR.
How Often Should You Calculate Gross and Net Retention?
Track gross and net retention monthly for direction, review them quarterly, and quote the headline on a trailing 12-month cohort. The 12-month cohort lets you compare your figure with annual benchmarks, which a monthly number cannot do. Use the monthly series to spot a leak early, and the quarterly review to decide whether to act on it.
Does Usage-Based Pricing Change How You Calculate Gross and Net Retention?
Usage-based pricing does not change the formulas for gross and net retention. Fix the cohort on day one of the period, and count a usage drop as contraction when the customer’s bill falls. For context, Benchmarkit and Pavilion’s 2025 report puts hybrid subscription-plus-usage companies at a 110% median NRR.
