FinanceIntermediate6 min read

Gross vs Net Retention

Gross Retention (GRR) measures how much revenue you keep from existing customers BEFORE expansion. Net Retention (NRR) measures revenue from existing customers AFTER expansion (upsell + cross-sell). Formula: GRR = (Starting ARR − Churned ARR − Downgrade ARR) / Starting ARR. NRR = (Starting ARR − Churn − Downgrade + Expansion) / Starting ARR. GRR is capped at 100%; NRR can exceed 100% (which means existing customers grow your business even if you stop acquiring new ones). Best-in-class public SaaS posts GRR > 90% and NRR > 120%. Snowflake hit NRR of 178% at peak — a single 1.78x compounding force on the existing book.

Also known asGRR vs NRRGross Dollar RetentionNet Dollar RetentionNDRGDR

Challenge a friend Browse library

The Trap

The trap is celebrating high NRR while ignoring weak GRR. NRR of 130% sounds amazing — but if your GRR is 75%, you're losing 25% of customers per year and just papering over it with massive upsells from a few whales. This creates customer concentration risk and mistakes a few accounts' growth for product-market fit. The opposite trap exists too: founders proudly cite GRR of 95% but their NRR is 100%, meaning they have NO expansion motion and zero pricing power. Both numbers tell the truth only TOGETHER.

What to Do

Report BOTH metrics every board meeting. Decompose NRR into three components: GRR, Upsell rate (more seats/usage at same plan), Cross-sell rate (different products). Calculate by COHORT (new customers from Q1 2024 vs Q1 2023) — newer cohorts often retain better as you improve product. Calculate by SEGMENT (enterprise NRR is usually 130%+, SMB NRR 90-100%). Set targets: GRR > 90% (if SMB) or > 95% (if enterprise), NRR > 110% (good) or > 120% (great).

Formula

GRR = (Starting ARR − Churn − Downgrades) / Starting ARR NRR = (Starting ARR − Churn − Downgrades + Expansion) / Starting ARR

In Practice

Snowflake at IPO (2020) reported the highest NRR in SaaS history: 158%, peaking at 178% in 2021. The driver: consumption-based pricing meant customers automatically spent more as their data grew. Even with GRR around 95%, the structural expansion created an organic 1.7x compounding force on the existing book. By 2024, NRR moderated to ~127% as consumption optimization tools (FinOps) emerged and customers right-sized usage. Stock collapsed 60% from peak partly because the market re-rated the durability of the NRR story.

Pro Tips

01
NRR > 100% means your existing customer base is a growing business by itself — even if you acquired ZERO new customers, revenue would grow. This is the single most powerful predictor of valuation multiple. Companies with NRR > 130% trade at 2x the multiple of peers with NRR < 110%.
02
GRR ceiling reveals product stickiness. If GRR caps at 80%, fundamentally something about your product or market is causing customer death. No amount of CSM hiring fixes a 80% GRR — you need product changes.
03
Watch the GAP between GRR and NRR. NRR − GRR = expansion contribution. If gap is < 10 points, you have weak expansion motion. Build account-based expansion plays, usage-based pricing tiers, or new product modules.

Myth vs Reality

Myth

“NRR over 120% is always sustainable”

Reality

Consumption-based NRR (Snowflake, Twilio, Datadog) can revert quickly when customers optimize. Seat-based NRR (Slack, Zoom) reverts when headcount slows. Product-add NRR (HubSpot multi-hub) is most durable but requires constant new product launches. Know which type you have.

Myth

“GRR doesn't matter if NRR is high”

Reality

Two companies with NRR 125%: Company A has 95% GRR + 30% expansion; Company B has 75% GRR + 50% expansion. Company B is acquiring customers just to lose them, then bleeding expansion from survivors to mask it. Investors who do diligence look at GRR — and Company B gets a brutal valuation discount.

Try it

Run the numbers.

Pressure-test the concept against your own knowledge — answer the challenge or try the live scenario.

🧪

Knowledge Check

Challenge coming soon for this concept.

Industry benchmarks

Is your number good?

Calibrate against real-world tiers. Use these ranges as targets — not absolutes.

Net Revenue Retention (Public SaaS)

Public B2B SaaS, $50M+ ARR

Best-in-Class

> 130%

Strong

115-130%

Acceptable

100-115%

Net Contraction

85-100%

Distressed

< 85%

Source: Bessemer Cloud Index / Meritech Capital

Gross Revenue Retention

B2B SaaS by segment

Enterprise Best

> 95%

Strong

90-95%

Average SMB

80-90%

Weak

70-80%

Critical

< 70%

Source: OpenView Partners / KeyBanc SaaS Survey

Real-world cases

Companies that lived this.

Verified narratives with the numbers that prove (or break) the concept.

❄️

Snowflake

2020-2024

mixed

Snowflake's IPO in September 2020 disclosed Net Retention of 158% — at the time, the highest ever for a public software company. Peak NRR hit 178% in early 2022. The driver: consumption-based pricing meant that as customer data grew, Snowflake revenue grew automatically. Customers signed $1M deals and grew to $5M+ within 24 months without sales lifting a finger. By 2024, NRR moderated to ~127% as customers learned cost optimization (FinOps tools, reserved capacity). Stock peaked at $400 (90x ARR) and fell to ~$130 as the NRR story normalized — a 60% drop driven primarily by expansion deceleration.

NRR at IPO (2020)

158%

Peak NRR (2022)

178%

Current NRR (2024)

~127%

GRR (consistently)

~95%

NRR > 150% is rarely durable for more than 2-3 years. Consumption pricing creates incredible expansion but invites optimization. Plan for NRR to mean-revert toward 110-130%; don't model permanent 150%+.

Source ↗

🧲

HubSpot

2018-2024

success

HubSpot's NRR has been a textbook 'durable expansion' story — consistently 105-110% for years. Not flashy, but extremely sustainable. Their secret: Multi-Hub strategy. Customers start with Marketing Hub, then add Sales Hub, then Service Hub, then CMS Hub, then Operations Hub. Each hub is a $10K-$100K+ ACV addition. GRR is ~92% (SMB-skewed). The model works because expansion comes from PRODUCT additions, not seat growth or consumption — making it less volatile than consumption-based NRR.

Avg NRR (2018-2024)

~107%

GRR

~92%

Multi-Hub Customers

~36% (2024)

Stock CAGR (5yr)

~25%

Durable mid-NRR (105-115%) from product expansion beats volatile high-NRR (150%+) from consumption. The market rewards predictability — HubSpot trades at premium multiples despite never hitting Snowflake-level NRR.

Source ↗

Related concepts