GRR vs NRR: Key Differences Every SaaS Team Must Know

Two numbers sit at the center of every serious SaaS investor conversation: gross revenue retention and net revenue retention. They are related, they are frequently confused, and they tell completely different stories about your business. Getting them wrong in a board deck — or in your own operating model — costs credibility and decision quality at exactly the moments when clarity matters most.

This guide covers both metrics in full: exact formulas, worked numeric examples with real dollar figures, the mathematical reason GRR cannot exceed 100% while NRR can, how SaaS Capital, Bessemer Venture Partners, and KeyBanc Capital Markets use each metric in due diligence, and the specific levers that move both numbers up without requiring you to choose one over the other.

What GRR and NRR Actually Measure

Both metrics answer a single question: what happened to the revenue from your existing customer cohort over a given period? The difference is what they include in "what happened."

The relationship between the two metrics is simple: NRR = GRR + expansion revenue as a percentage of starting MRR. If your GRR is 88% and your NRR is 108%, your expansion rate is roughly 20 percentage points on the starting base.

The Exact Formulas

Both formulas use the same starting cohort measured over the same period — typically one month or one year. The period must be consistent; mixing monthly contraction data with annual churn data produces garbage.

A few definitional guardrails that trip up operators:

• New logos never appear in either formula. GRR and NRR are cohort metrics. A customer who signs in month two is not in the month-one cohort and must not be included in that cohort's retention calculation.
• Reactivations are nuanced. A churned customer who returns is typically counted as a new logo for the purposes of the original cohort's calculation.
• Annual vs monthly. Most investors prefer annual cohort calculations (measure January cohort in the following January) because they smooth seasonal patterns. Monthly calculations are valid for operating reviews but show more volatility.

Worked Numeric Examples

Abstract formulas obscure the intuition. The following examples use consistent numbers across both metrics so the relationship between GRR and NRR is immediately apparent.

Example A: Healthy Mid-Market SaaS Company

Starting MRR (January 1): $200,000 from 80 existing customers.

During January:

• 3 customers cancel: $8,000 churned MRR
• 2 customers downgrade: $4,000 contraction MRR
• 12 customers upgrade or add seats: $22,000 expansion MRR

Metric	Calculation	Result
GRR	($200,000 − $8,000 − $4,000) ÷ $200,000 × 100	94.0%
NRR	($200,000 + $22,000 − $4,000 − $8,000) ÷ $200,000 × 100	105.0%

Interpretation: this company retained 94 cents of every dollar from existing customers before any upsells — solid mid-market performance. With upsells included, those same customers now represent 105% of their starting value. The business is growing its existing base modestly.

Example B: World-Class Enterprise SaaS Company

Starting ARR (January 1): $5,000,000 from 40 enterprise accounts.

During the year:

• 1 account cancels: $100,000 churned ARR
• 1 account downgrades: $50,000 contraction ARR
• 15 accounts expand: $750,000 expansion ARR

Metric	Calculation	Result
GRR	($5,000,000 − $100,000 − $50,000) ÷ $5,000,000 × 100	97.0%
NRR	($5,000,000 + $750,000 − $50,000 − $100,000) ÷ $5,000,000 × 100	112.0%

Interpretation: 97% GRR is genuinely excellent — the company lost only 3% of its base ARR to churn and downgrades. With expansion layered on, the 40 accounts now represent $5.6M in ARR, a 12% increase over the starting base. This is a high-quality business.

Example C: High NRR Masking High Churn

Starting MRR: $300,000 from 150 SMB customers.

• 35 customers cancel: $60,000 churned MRR
• 8 customers downgrade: $12,000 contraction MRR
• 18 customers expand (large accounts): $90,000 expansion MRR

Metric	Calculation	Result
GRR	($300,000 − $60,000 − $12,000) ÷ $300,000 × 100	76.0%
NRR	($300,000 + $90,000 − $12,000 − $60,000) ÷ $300,000 × 100	106.0%

Interpretation: on the surface, 106% NRR looks healthy. But 76% GRR reveals that the business lost 24% of its revenue base to churn and downgrades. The expansion from a small number of large accounts is papering over serious retention dysfunction. An investor who only looks at NRR misses the structural problem entirely. This is precisely why GRR matters as a standalone metric.

Why GRR Has a 100% Ceiling — and NRR Does Not

The mathematical reason is straightforward. GRR measures retained revenue as a fraction of starting revenue. The numerator is always starting revenue minus losses. You cannot keep more than 100% of what you started with if you exclude additions. The maximum possible GRR is when zero customers churn and zero customers downgrade, producing (Starting MRR − 0 − 0) ÷ Starting MRR = 100%.

NRR's numerator includes expansion revenue added to the same starting base. When expansion is large enough to exceed churn and contraction losses, the numerator exceeds the denominator and the result is greater than 100%. There is no mathematical ceiling. Some usage-based pricing businesses — where every customer is growing their consumption — post NRR above 130%, 140%, or even 150% in strong growth periods.

This asymmetry has an important practical implication: you can never solve a GRR problem by growing NRR. Fixing churn and downgrades is the only path to improving GRR. Expansion only affects NRR.

GRR and NRR Benchmarks

Benchmark data comes from several credible sources. The 2023 SaaS Capital Retention Benchmarks report covers over 1,500 private B2B SaaS companies. Bessemer Venture Partners publishes annual benchmark data through its State of the Cloud series. KeyBanc Capital Markets surveys private SaaS CFOs annually. The figures below synthesize these sources.

GRR Benchmarks by Segment

Customer Segment	Median GRR	Top Quartile GRR	World-Class GRR
SMB (ACV < $5K)	83%–88%	90%–92%	95%+
Mid-Market (ACV $5K–$50K)	88%–91%	92%–94%	95%+
Enterprise (ACV > $50K)	91%–94%	95%–97%	97%+
All Private B2B SaaS	91%	92%–95%	95%+

According to SaaS Capital's 2023 retention benchmark report, the all-segment median GRR for private B2B SaaS is 91%. The 2023 KeyBanc SaaS Survey found that the median GRR across private SaaS companies was 87%, with top-quartile performers exceeding 92%. The gap between reports reflects differences in the respondent pool: KeyBanc skews toward earlier-stage companies with higher SMB exposure and structurally higher churn.

NRR Benchmarks by Segment

Customer Segment	Median NRR	Top Quartile NRR	World-Class NRR
SMB (ACV < $5K)	95%–100%	103%–107%	115%+
Mid-Market (ACV $5K–$50K)	105%–110%	112%–118%	120%+
Enterprise (ACV > $50K)	112%–118%	120%–128%	130%+
All Private B2B SaaS	102%–106%	112%–118%	120%+

Bessemer Venture Partners' canonical NRR framing — Good (100%), Better (110%), Best (120%+) — has become the de facto language of SaaS board reporting. The SaaS Capital 2023 data places the all-segment median NRR at 102%. See also our detailed post on NDR benchmarks for SaaS for a full segment-by-segment breakdown.

How Investors Use GRR and NRR Differently

GRR and NRR answer different due-diligence questions. Experienced investors interrogate both, and the relationship between them often reveals more than either number alone.

How Investors Read GRR

GRR is a product-quality and implementation-quality signal. High GRR means customers renew because the product delivers durable value — not because of contractual lock-in or aggressive account management. Investors use GRR to answer three questions:

1. Is the product sticky? GRR above 90% in SMB or 95% in enterprise suggests customers genuinely depend on the product. GRR below 80% suggests they tolerate it — and will leave at the next contract renewal if a better option appears.
2. Is the revenue base stable? A business with 95% GRR retains 95 cents of every dollar from existing customers before any growth. That makes revenue forecasting reliable and reduces the growth treadmill — the pace of new logo acquisition needed just to offset losses.
3. Is expansion masking churn? When NRR significantly outpaces GRR, investors probe whether upsells are going to the same handful of large accounts that obscure churning small accounts. A 15-point spread between NRR and GRR (e.g., 108% NRR, 92% GRR) is normal. A 30-point spread (e.g., 110% NRR, 80% GRR) requires explanation.

How Investors Read NRR

NRR is a monetization-efficiency and compounding-growth signal. Investors use NRR to model future revenue growth from the existing install base without new logo acquisition assumptions.

A company with 120% NRR grows its existing ARR by 20% annually from the installed base alone. If that company also acquires new customers, total ARR growth is materially higher. This compounding effect is why Bessemer and other leading venture investors treat NRR above 120% as a category-defining threshold — it shifts the business from a "must acquire to grow" model to a "grow from existing customers" model, which is structurally more capital-efficient.

Valuation multiple data reflects this. Per Bessemer Venture Partners' research, public SaaS companies with NRR above 120% have commanded revenue multiples 2–3x higher than companies with below-median NRR. Companies with GRR below 80% face capped valuation multiples regardless of growth rate, because investors discount future revenue streams for the structural churn risk embedded in the business.

NRR Range	GRR Range	Investor Signal	Typical ARR Multiple
Below 90%	Below 80%	Retention crisis — fix before fundraising	1–3x ARR
90%–100%	80%–88%	Marginal — product improvements needed	3–5x ARR
100%–110%	88%–93%	Solid — table stakes for Series A/B	5–8x ARR
110%–120%	93%–97%	Strong — meaningful expansion motion	8–12x ARR
120%+	95%+	World-class — premium multiple justified	12x+ ARR

For a deeper treatment of how retention metrics feed into Series A diligence, see our post on SaaS metrics that Series A investors scrutinize.

When GRR Matters More Than NRR

NRR gets more attention in investor conversations, but there are specific operating contexts where GRR is the more important metric to track and optimize.

Early-stage companies (pre-$3M ARR). At this stage, the customer base is small enough that one or two large churns dramatically distort NRR. GRR gives a cleaner signal of whether the core product experience is working. Optimizing for NRR before establishing strong GRR means you are building an expansion motion on top of a leaking bucket.

SMB-focused businesses. SMB customers have structurally higher churn rates driven by factors outside your control: company failures, budget cuts, headcount reductions. In this context, GRR is the primary indicator of whether your product is genuinely essential — not just nice-to-have. An SMB SaaS business with 90%+ GRR is performing well; that same business boasting 115% NRR may be papering over churn with upsells to surviving customers.

Post-acquisition integration. When a SaaS company acquires a complementary product, the first 12 months are dominated by churn risk as customers evaluate whether the combined product meets their needs. GRR is the right operational dashboard metric during this period — NRR can look artificially inflated by cross-sell opportunities between the two customer bases.

Turnaround situations. If NRR has been declining for two or more consecutive quarters, GRR is the diagnostic tool. Decomposing NRR into GRR and expansion rate isolates whether the problem is churn (a GRR problem requiring product/CS investment) or declining expansion (an upsell/pricing problem requiring GTM changes).

When NRR Matters More Than GRR

Fundraising and valuation conversations. NRR is the single most cited retention metric in term sheets and data rooms. Investors model revenue growth from existing customers using NRR, not GRR. If your NRR is strong, lead with it — but always have GRR ready for the follow-up question.

Usage-based pricing models. When customers pay for what they consume, NRR is the primary operational metric because revenue changes continuously within accounts. GRR still matters, but contraction is often revenue-level contraction rather than full churn events, and expansion is driven by usage growth rather than explicit upsell motions.

Platform and multi-product businesses. Companies with multiple product lines use NRR to measure the health of the cross-sell motion. The gap between GRR and NRR quantifies how much revenue the platform strategy generates beyond base retention.

Competitive positioning. In conversations with prospective enterprise customers, quoting a 120%+ NRR from existing accounts is a credible signal of customer success and product depth that GRR alone cannot communicate as directly.

How to Improve Both GRR and NRR Simultaneously

Many operators treat GRR and NRR improvement as separate programs — reduce churn for GRR, drive upsells for NRR. The most effective operators treat them as a single retention and expansion system where the same inputs drive both metrics.

1. Build a customer health score that predicts churn and expansion

A well-constructed customer health score uses product usage signals, support ticket frequency, stakeholder engagement, and contract utilization to identify accounts at churn risk and accounts ready for expansion — simultaneously. Acting on health score data improves GRR by surfacing at-risk accounts before they churn; it improves NRR by surfacing expansion-ready accounts before they seek competitive alternatives.

The same investment in customer data infrastructure lifts both metrics. For more on the metrics that underpin this, see our post on customer success metrics that matter.

2. Fix the onboarding and time-to-value problem

The majority of churn is decided in the first 90 days of a customer relationship. Customers who do not reach a meaningful outcome within the initial contract period — what product teams call the "aha moment" — never develop the product dependency that sustains renewal. Compressing time-to-value through better onboarding, implementation support, and early success reviews directly improves GRR by reducing first-renewal churn.

The same outcome — faster, deeper adoption — also accelerates expansion by creating internal champions who push for additional seats or modules. Fast time-to-value improves both metrics through the same mechanism.

3. Introduce usage-based pricing components

Pure seat-based subscriptions create a natural churn trigger at every renewal: the customer can always downgrade to a lower seat tier. Usage-based components — API calls, data volume, transaction count — decouple revenue from headcount and tie it to business activity. As customers grow, usage grows, and NRR expands naturally. At the same time, usage-based pricing removes one of the most common downgrade triggers, improving GRR.

The data supports this. Usage-based pricing models structurally produce higher NRR — typically 115–130% — versus flat-rate subscription models at 95–105%, according to Bessemer Venture Partners' cloud benchmarks.

4. Build a formal renewal motion 90 days out

Renewals handled by account managers in the final 30 days of a contract cycle are reactive and lose to inertia. Renewals managed 90 days out — with a structured success review, documented outcomes, and a clear renewal conversation — close at higher rates and at higher contract values. The 90-day runway gives the CS team time to address any product gaps, escalate unhappy stakeholders, and position an upsell conversation alongside the renewal rather than after it.

5. Segment and tier your customer base by expansion potential

Not all customers have the same expansion ceiling. A 10-seat SMB customer may grow to 15 seats; a 50-seat enterprise customer may grow to 500 seats if the product proves its value. Tiering customers by expansion potential — and allocating CS resources accordingly — concentrates effort where both NRR and GRR returns are highest. High-value accounts with complex deployments need dedicated CSMs. Low-value accounts with simple use cases benefit from automated health-score-driven interventions.

Common Calculation Mistakes That Distort Both Metrics

Bad GRR and NRR data is worse than no data — it creates false confidence or false alarm. These are the most frequent calculation errors that produce inaccurate numbers.

Including new logo revenue in the cohort. Revenue from customers acquired after the cohort start date must be excluded from both formulas. Including it inflates the denominator (if included in starting MRR) or inflates the numerator (if treated as expansion) and produces numbers that cannot be compared to industry benchmarks.

Netting contraction against expansion. Some operators calculate a single "net expansion" number (expansion minus contraction) and add it to the NRR formula. This is mathematically equivalent but obscures the churn components needed to calculate GRR independently. Always track expansion, contraction, and churn as separate line items.

Using inconsistent time periods. Monthly GRR and annual NRR cannot be directly compared. Both metrics should be calculated over the same period — 12 months is the standard for investor reporting.

Ignoring multi-year contract timing. Customers on two- or three-year contracts may show no churn in year one but a significant churn event at year two. GRR and NRR should be tracked on a contract-renewal basis, not calendar-year basis, for businesses with long average contract durations.

How Fairview Tracks GRR and NRR

Frequently Asked Questions