Net Revenue Retention (NRR)

Gross Revenue Retention (GRR) measures only the revenue you retain from existing customers — it excludes expansion revenue and can never exceed 100%. It tells you the worst-case outcome: 'If none of our customers upgraded or bought more, what percentage of our revenue would we keep?' Net Revenue Retention (NRR) includes expansion revenue on top of that retained base, so it can exceed 100%. NRR tells you the net outcome: 'After accounting for all changes — losses and gains — in our existing customer base, are we growing or shrinking?' Both metrics matter. GRR reveals how well you are retaining customers regardless of upsell success; NRR reveals the total commercial relationship health. A company can have impressive NRR but poor GRR if its strong upsell motion is masking high churn — a fragile position that warrants scrutiny.

While there is no hard regulatory requirement, public market investors have developed strong preferences based on the NRR profiles of successful SaaS IPOs. The most admired and highly valued SaaS businesses at IPO — Snowflake, Datadog, Crowdstrike, Okta — all showed NRR above 120% at or near the time of their offerings. NRR between 110% and 120% is generally viewed as solid and sufficient for a successful IPO, especially when combined with strong growth rates and large addressable markets. NRR between 100% and 110% will attract some investor skepticism and typically results in lower revenue multiple valuations. NRR below 100% at IPO would require extraordinary justification (very high growth rate, massive market, credible improvement trajectory) to succeed. Investment bankers running SaaS IPO processes typically target 110%+ NRR as a minimum threshold.

Improving NRR requires working both sides of the equation: maximizing retention (reducing churn and contraction) and maximizing expansion (growing revenue within existing accounts). On the retention side, investing in customer onboarding to ensure customers achieve value quickly is highest leverage — most churn decisions are made in the first 90 days. Health scoring systems that proactively identify at-risk accounts allow customer success teams to intervene before cancellations occur. On the expansion side, designing pricing that scales naturally with customer success (per-seat, per-usage, or per-outcome models) creates organic upsell without requiring hard selling. Building a product roadmap that unlocks new use cases within the same account turns one department's adoption into company-wide adoption. Dedicated expansion sales roles, separate from new business acquisition, can focus exclusively on growing existing account revenue.

Technically, a company with NRR above 100% will grow revenue indefinitely even with zero new customer acquisition — as long as the NRR holds. In practice, however, this is not a strategy any company should pursue. First, NRR is calculated on a base that itself depends on past new customer acquisition — you need a large enough customer base for expansion revenue to matter. Second, NRR is not static; it can deteriorate if you stop investing in new products, customer success, and market expansion. Third, new customer acquisition brings in customers whose expansion revenue contributes to future NRR, compounding the growth engine. Think of NRR as a multiplier on top of new customer growth, not a replacement for it. The magic happens when you combine strong new customer growth with NRR above 100% — the two compound together into explosive revenue growth.

NRR should be presented clearly with a precise definition of what is included and what period it covers. Best practice is to present trailing 12-month NRR using the standard formula: (cohort starting MRR + expansion − churn − contraction) ÷ cohort starting MRR × 100. Always disclose whether you are measuring revenue retention on an MRR basis or ARR basis, and whether you are using a rolling 12-month cohort or a point-in-time cohort. Present both NRR and GRR side by side so investors can separately assess churn quality and upsell quality. If your NRR is improving, showing a trend chart over 8–12 quarters is compelling evidence of business improvement. Investors are sophisticated enough to notice if companies change their NRR definition between quarters — consistency is non-negotiable.

By definition, new customers acquired during the measurement year are excluded from NRR calculation. NRR measures only what happens to the cohort of customers who existed at the start of the period. This is an important distinction because it isolates the quality of your existing customer relationships from the quantity of new customer acquisition. New customers will, however, become part of the cohort in future NRR calculations — they become next year's 'existing customers.' This is why rapidly growing companies with many new customers can sometimes show elevated NRR temporarily: the cohort used in the numerator (12 months later) may include revenue from newer customers who have expanded, while the cohort used in the denominator was smaller. Being precise about cohort definitions avoids these distortions.

Early-stage startups (under $2M ARR) often lack sufficient data to calculate statistically meaningful NRR — the customer base is too small and the time period too short. The most common proxy at early stages is cohort retention curves: how much of the revenue from the first month's customer cohort is still active at month 3, 6, 12? For an early-stage startup, targeting NRR above 100% is aspirational but may be premature before achieving product-market fit and building a customer success function. A more practical early target is minimizing gross churn — keep GRR above 85% as a starting point, then build expansion programs once you have a stable retained base to expand from. NRR becomes increasingly meaningful and measurable as you cross $1M–$2M ARR with at least 12 months of customer history to analyze.