What is Net Revenue Retention (NRR)?

The NRR formula starts with Monthly Recurring Revenue (MRR) from a fixed cohort of existing customers at the beginning of a period. You add any expansion MRR those customers generated — from upsells, seat additions, or cross-sells. Then you subtract contraction MRR (customers who downgraded) and churned MRR (customers who cancelled entirely). Finally, divide by the starting MRR and multiply by 100.

For example: if you start the month with $500,000 MRR, upsells add $75,000, downgrades cost $20,000, and cancellations remove $30,000, your NRR is ($500K + $75K − $20K − $30K) ÷ $500K × 100 = 105%.

Critically, NRR is always calculated on a cohort of existing customers — new logos are excluded entirely. This makes it a pure measure of account health and expansion capacity, not an acquisition metric. Most companies calculate NRR monthly and then annualize, though some (especially those with annual contracts) measure it over a 12-month cohort window.

Gross Revenue Retention (GRR) measures revenue kept from existing customers after losses from churn and downgrades only — expansion is excluded. Because it strips out upsells, GRR can never exceed 100%. It is a floor-and-ceiling metric: the best GRR can do is hold the line. Median GRR across SaaS companies was approximately 91% in 2025, per Stripe's industry data.

NRR adds back expansion revenue, so it can exceed 100% and acts as a net growth signal. NRR is always at least as high as GRR. The gap between the two — the NRR-GRR spread — is a diagnostic that analysts and investors scrutinize closely. A wide spread (30+ points) may indicate that a handful of large accounts drive most expansion while the broader base erodes — a concentration risk where the headline NRR looks strong but collapses if those anchor accounts churn, per Gainsight.

You need both metrics together. A company can show 120% NRR while GRR sits at 75%, meaning it's losing a quarter of its base every year but masking the damage with heavy expansion from the accounts that stay. NRR alone flatters; NRR and GRR together reveal the underlying health of the cohort.

NRR is the primary compounding lever in B2B SaaS. A company with 120% annualized NRR grows its revenue base approximately 2.5x over five years from existing customers alone, with no new acquisition spend required. A company at 95% NRR sees its existing base shrink to 77% of its original size over that same period — forcing it onto a treadmill where new logo acquisition must continuously outpace base erosion just to stay flat.

Investors price this compounding directly into valuations. Software companies with NRR above 120% command 30–50% higher valuation multiples compared to peers at 100% NRR, even with identical ARR and growth rates, per m3ter's 2026 analysis. A 10-point improvement in NRR can translate to a 20–30% valuation uplift. In private M&A transactions, crossing from 105% to 110% NRR frequently adds meaningful incremental value to buyer offers.

Customer expansion now accounts for approximately 52% of new revenue in B2B companies in 2025, up significantly from prior years (McKinsey 2025). This shift means NRR has become a forward indicator of total revenue growth, not just a retention efficiency metric — making it a critical input for board reporting, fundraising, and M&A positioning.

Three forces push NRR above 100%: upsells to a higher tier or more seats, cross-sells of adjacent products, and usage-based growth where spend rises naturally with adoption. The operational challenge is detecting expansion moments before they pass and reaching the right person with relevant context within days.

The most reliable external signals are firmographic changes: a customer raising a new funding round brings new budgets and strategic mandates; a hiring surge in a specific department signals a need for more licenses; an executive change, product launch, or acquisition can open conversations that were previously closed. Existing customers buy at a 60–70% success rate versus 5–20% for new prospects (research cited by OpenView), partly because the relationship, integration, and trust already exist — which is why expansion deals also close 2–3x faster than equivalent new-logo deals.

Signal timing is decisive. Expansion teams that act within five business days of a trigger event convert at substantially higher rates than those who follow up two or more weeks later. When a business signal (such as new headcount) converges with a product signal (such as high utilization nearing capacity), conversion rates are meaningfully higher than single-trigger outreach.

Benchmarks depend heavily on customer segment and must be read in context. Enterprise SaaS (ACV above $100K) carries a median NRR of approximately 118%, mid-market ($25K–$100K ACV) lands at approximately 108%, and SMB (below $25K ACV) sits at approximately 97%, per m3ter's 2026 benchmarking analysis. Best-in-class is generally defined as ≥130% across segments.

For private SaaS companies, KeyBanc Capital Markets' 16th Annual Private SaaS Survey (2025, 100+ private firms) found median net retention at approximately 101%, reflecting a compression trend from approximately 105% in 2021. ChartMogul's dataset of 2,500+ companies placed the top-quartile threshold above $500 ARPA at approximately 100% NRR in H1 2024, with expansion revenue accounting for over 50% of total revenue at top-quartile companies. A company approaching an IPO should expect investors to benchmark against ≥107–110% as a baseline, with anything above 120% commanding a meaningful multiple premium.

Below 90% is a warning signal requiring urgent intervention — at that level, the business is losing revenue from its existing base faster than it can grow, and new-logo acquisition must run faster just to avoid top-line contraction.

NRR is ultimately a signals-and-timing problem. The accounts most likely to expand are already in your CRM — the challenge is knowing exactly when they're ready, and reaching them before the moment passes. Expansion signals like funding rounds, new executive hires, headcount surges, and product launches at customer companies are public and predictable, but monitoring hundreds of accounts manually across a full book of business is not scalable for human reps.

Komo's AI Revenue Engine automates the signal-monitoring layer for existing accounts: it watches your customer base for firmographic changes, surfaces the right expansion triggers at the right moment, drafts personalized outreach anchored to the specific signal, and queues those messages for a human to review and send. Account managers and customer success teams can act on expansion moments across their full book of business — not just the loudest or largest accounts.

The human-in-the-loop design matters specifically for expansion. Expansion conversations with existing customers are relationship-sensitive; a poorly timed or generic nudge can damage a relationship that took months to build. Komo keeps a person on every send that matters, so the speed of AI signal detection is paired with the judgment of a human rep who knows the account context.