Why Net Revenue Retention Predicts Everything
If a venture investor could see only one metric before deciding whether to back a SaaS company, it would not be growth rate, gross margin, or even revenue. It would be net revenue retention (NRR). NRR measures what happens to revenue from your existing customer base over time, after accounting for churn, downgrades, and expansion. It is the cleanest possible test of whether your product creates compounding value.
An NRR above 100 percent means your existing customers generate more revenue over time than you lose to churn and downgrades — your business grows even if you never sign another new customer. An NRR below 100 percent means you are leaking value and must acquire new customers just to stand still. The difference between 95 percent and 110 percent NRR, sustained over five years, is the difference between a company that triples and one that decays.
Acquisition is the engine. Retention is the fuel efficiency. A fast engine with a leaking tank still runs out of gas. NRR tells you whether your tank holds.
The Precise Formula
**Net Revenue Retention (%) = [(Starting MRR − Churned MRR − Contraction MRR + Expansion MRR) ÷ Starting MRR] × 100**
The denominator is the MRR from your existing customer base at the start of the period — specifically, customers who were already paying at the beginning. New customers acquired during the period are excluded from both numerator and denominator, because NRR measures the behavior of the base you already had, not the growth from new acquisition.
The four components each tell a distinct story. Churned MRR is revenue lost to full cancellations. Contraction MRR is revenue lost to downgrades — customers who stay but move to a cheaper plan or reduce seats. Expansion MRR is revenue gained from existing customers upgrading, adding seats, or purchasing add-ons. The net of these against the starting base is your retention rate.
Benchmark NRR by Segment
NRR benchmarks vary sharply by customer segment because expansion mechanics differ. Enterprise customers expand through seat growth and tier upgrades; SMB customers are more volatile and expand less predictably.
- **Enterprise SaaS:** 110–130% NRR. Large accounts with seat-based pricing expand naturally as teams adopt the product. High switching costs limit churn.
- **Mid-market SaaS:** 100–115% NRR. Moderate expansion, moderate churn. The segment where NRR most reliably distinguishes good businesses from average ones.
- **SMB SaaS:** 90–105% NRR. Higher churn is partially offset by expansion, but many SMB products lack natural expansion vectors.
- **Consumer/prosumer SaaS:** 70–95% NRR. Individual users downgrade and churn freely. Expansion is limited without a team or usage-based component.
A blended NRR across segments can hide structural problems. A company with 120% NRR in enterprise and 80% NRR in self-serve SMB might report a blended 105% that looks healthy while the SMB segment is quietly bleeding. Always segment NRR by customer cohort and plan tier, not just in aggregate.
The Expansion Motions That Drive NRR Above 100
Achieving NRR above 100 percent requires expansion revenue to exceed the combined loss from churn and contraction. The most reliable expansion motions are structural — built into how the product is priced and adopted rather than dependent on a sales team pushing upgrades.
- **Seat-based expansion.** Products priced per user expand automatically as a customer's team grows. Slack, GitHub, and Figma all rely on this. The key is low enough friction that new users can be added without a procurement cycle.
- **Usage-based expansion.** Products priced by consumption (API calls, compute, storage, messages) expand as the customer's usage grows. AWS, Stripe, and Snowflake exemplify this. The risk is that usage can contract just as easily as it expands.
- **Tier-based upsell.** Customers move from a basic to a premium tier as their needs grow. This requires clear value differentiation between tiers and a natural trigger that makes the upgrade obvious.
- **Add-on cross-sell.** Selling adjacent products to the existing base. Higher friction than the other motions but effective when the add-on solves a problem the customer already has.
The strongest NRR engines combine two or more of these motions. A seat-based product with a usage-based component expands on both axes simultaneously — more users and more usage per user. This is why investors pay a premium for companies with hybrid pricing models.
Common NRR Calculation Errors
NRR is frequently miscalculated in ways that flatter the business. The most common error is including new customers in the starting base, which inflates the denominator and masks poor retention among existing customers. The second is ignoring contraction MRR, treating downgrades as if they did not happen. The third is measuring over inconsistent time windows — mixing a 30-day cohort with a 90-day one.
The fix is discipline: fix the cohort (customers paying at the start of the period), fix the window (a consistent monthly or annual period), and include all four components (churn, contraction, expansion, starting base). Compute NRR monthly for operational visibility and annually for investor reporting. Track it by segment, not just in aggregate. Use the SaaS Churn calculator on this site to decompose your MRR movement and isolate the expansion component that drives your NRR.