Net Revenue Retention Above 120%: Why It Matters More Than New Logo Growth at Series B

Net Revenue Retention (NRR) above 120% is the single strongest signal at Series B that your product compounds without permission — existing customers expand faster than they churn, which means every new logo becomes a flywheel rather than a leaky bucket. According to Bessemer Venture Partners' Cloud 100 Benchmarks Report 2025 and SaaS Capital's 2025 Private SaaS Company Valuations data, companies clearing 120% NRR consistently command 2–2.5x higher ARR multiples than peers stuck at the 101% median. This guide breaks down why the metric matters more than logo velocity at Series B, how the top quartile actually hits it, and the step by step framework to engineer it into your business.

The Math: Why Series B Investors Now Lead With NRR, Not Logos

At seed and Series A, new-logo growth is the dominant signal because the absolute ARR base is too small for retention math to matter. At Series B — typically $3–10M ARR, with a 2025 median round size of $28M at a $120–160M post-money according to Value Add VC's 2025 Series B funding analysis — the math inverts. The installed base is now large enough that retention compounds harder than acquisition.

Consider two companies, both at $8M ARR raising Series B:

Company A: 100% new logo growth, 95% NRR. Year 1: $8M × 0.95 + $8M new = $15.6M. Year 2: $15.6M × 0.95 + $16M new = $30.8M.
Company B: 50% new logo growth, 130% NRR. Year 1: $8M × 1.30 + $4M new = $14.4M. Year 2: $14.4M × 1.30 + $7.2M new = $25.9M.

Year two looks similar. But by year four, Company B compounds past Company A because acquisition costs decline (less sales spend per dollar) and the gross margin on expansion revenue is dramatically higher — typically 85–90% versus 60–70% on new-logo revenue once you back out CAC. Tomasz Tunguz's Q3 2025 Datadog analysis shows this dynamic playing out at scale: Datadog hit 120% net dollar retention in Q3 2025 while growing 28% year-over-year, with operating margins expanding because expansion ARR carries near-zero incremental CAC.

Takeaway: At Series B, run the compounding model both ways. If your NRR is below 110%, every new-logo dollar is fighting churn instead of stacking on top of it — and your Series C investors will run the same spreadsheet before they price the round.

The 120% Benchmark: Who Actually Hits It

The 120% number isn't aspirational marketing — it's the public-comp threshold separating top-quartile from median. Recent SEC filings make the gap concrete:

Snowflake: 127% NRR as of October 31, 2024, settling to 124–126% through fiscal year 2025 per their Form 10-Q filed for the period ending October 31, 2025. Consumption pricing structurally lifts NRR because customers naturally use more as their data volumes grow.
Datadog: 120% net dollar retention in Q3 2025, per their Q3 2025 earnings release. The recovery from sub-120% in 2023 came from adding new product lines (Cloud SIEM, LLM Observability) that existing customers attached to their existing contracts.
Cloudflare: 118% dollar-based net retention in their most recent 2025 quarter, driven by seat expansion and Zero Trust attach.
monday.com: 111% NRR in Q3 2024 with enterprise (>$50K ARR) customer count up 39% year-over-year and enterprise ARR up 47%, per their Q3 2024 Form 6-K. The enterprise-segment NRR is much higher; SMB drags the blended number.
HubSpot: Approximately 102% in Q1 2025, with management guiding for "a couple of points" of improvement, per SaaStr's complete history of HubSpot's NRR. HubSpot peaked at 115% during the 2021 cycle and has spent four years rebuilding the expansion motion.

The pattern is unambiguous: infrastructure and consumption-priced products clear 120%+ structurally. Seat-based SMB products fight gravity. According to Benchmarkit's 2025 B2B SaaS Performance Benchmark Report covering 2,000 private SaaS companies, the median NRR sits at 101% — meaning half the market is barely treading water once you net out churn.

Takeaway: Before you set an NRR target, name your three closest public comps and pull their last four quarters of NRR from 10-Qs. If your model is consumption-based and your comps are at 125%, your Series B pitch should not feature a 110% projection.

The Five Components of NRR — And Where the 20 Points of Upside Live

NRR is mechanical. It is: (Starting ARR + Expansion + Upsell – Downgrades – Churn) ÷ Starting ARR, measured on the same cohort 12 months apart. The five levers, and the realistic point contribution each can deliver between Series A and Series B:

Gross logo churn (worth 3–8 points): Top quartile sits at 5% annual logo churn or below, per SaaS Capital's 2025 benchmark data. Every 1 point of logo churn you cut adds roughly 1 point of NRR.
Gross revenue churn (worth 2–5 points): Downsells on renewal — same logo, smaller contract. Usually a packaging problem (you sold too many seats) or a pricing-model problem (flat-rate when usage dropped).
Seat/usage expansion (worth 8–15 points): The biggest lever. Snowflake's 127% NRR is essentially all usage expansion. For seat-based products, this means embedding inside the customer's hiring plan.
Cross-sell into new product lines (worth 4–10 points): Datadog added 27 modules across Infrastructure, APM, Logs, Security, and LLM Observability — each is a cross-sell vector into the same buyer.
Annual price increases (worth 2–5 points): The most underused lever pre-Series B. A 5% annual escalator on a multi-year contract is now standard at the enterprise tier.

Takeaway: Decompose your current NRR into all five components for the last four quarters in a spreadsheet model. If you can't, you don't have an NRR strategy — you have an NRR number. Build the decomposition first; the answer of where to invest will become obvious.

The Series B Pitch: How to Frame NRR vs. New Logo Growth

Series B investors in 2025 expect ARR between $3M and $10M with NRR above 110% as table stakes, per the T2D3 2025 B2B SaaS Performance benchmark. The deals getting priced at 18x forward ARR rather than 8x share three structural features:

NRR above 120% for at least two consecutive quarters (proves it isn't a single-cohort fluke).
A "net new ARR mix" disclosure showing >50% of net new ARR comes from existing customers. This is the slide that closes Series B rounds.
A Rule of 40 score above 50, since the Bessemer Cloud 100 Benchmarks Report 2025 notes the bar has moved from Rule of 40 to Rule of 50–60 for AI-native companies.

The pitch reframe sounds like this: "We grew new logos 60% year-over-year, but the more important number is that 58% of our net new ARR came from existing customers expanding their contracts. Our 124% NRR means we can hit the plan even if new-logo growth slows — and we'd rather spend the next dollar deepening Tier 1 accounts than chasing logo count."

That reframe maps to how investors actually underwrite: a 120%+ NRR business with slower logo growth gets priced higher than a 95% NRR business with faster logo growth, because the former has structurally lower forward CAC requirements. According to FE International's 2026 SaaS valuation guide, moving NRR from 100% to 120% combined with sub-5% logo churn can move a multiple from roughly 4x ARR to 6x or more.

Takeaway: Build a single slide that shows, by quarter, the split of net new ARR between (a) new logos, (b) seat/usage expansion, and (c) cross-sell. If expansion is below 40% of net new ARR, focus the next two quarters on that — it will move the round more than 10% extra logo growth.

The 90-Day Plan to Move NRR by 10 Points Before You Raise

You cannot fix NRR retroactively — it's measured on a cohort 12 months out. But you can engineer the leading indicators in a single quarter:

Days 1–15: Build the decomposition. Pull every renewal event from the past 12 months. Tag each as: renewed flat, expanded (with $ delta), contracted (with $ delta), or churned. Calculate the five-component breakdown above. This is the artifact your Series B deck will be built on.
Days 16–30: Segment your book. Identify the top 20% of customers by ARR. These accounts typically drive 60–70% of expansion revenue. Assign every Tier 1 account a named owner (CSM or AE) with a 90-day expansion plan documented in a one-page account brief.
Days 31–60: Ship one expansion-trigger product feature. Usage-based metering, a new module, seat-tier upgrades, a premium SLA. Datadog's playbook of adding modules customers can attach mid-contract is the canonical example.
Days 61–75: Add a contractual price escalator. 5% annual increase clause in every new contract and every renewal. This compounds for years; the math is staggering by year three.
Days 76–90: Build the NRR forecast model. A spreadsheet model with cohort-level forward NRR projections by segment, used to (a) brief your board, (b) anchor your Series B deck, and (c) drive weekly forecast reviews with your CS and Sales leads.

Takeaway: The 90-day plan does not move your reported NRR — it moves the trajectory. A board deck that shows NRR at 108% with a credible decomposition, an expansion product shipped, escalators in place, and a forward model projecting 118% in four quarters is a fundable story. A 115% number with no infrastructure behind it is not.

Conclusion: Build the Model Once, Use It Forever

Net Revenue Retention is not a vanity metric — it is the cash-on-cash return of your installed base, and at Series B it determines whether your next round is priced at 8x or 18x forward ARR. The companies clearing 120% (Snowflake, Datadog, Cloudflare) didn't get there by accident. They built the decomposition, segmented the book, shipped expansion-trigger features, and ran disciplined forecast cycles for years.

The hardest part for most operators is not the strategy — it is the spreadsheet model that ties cohort-level retention, segment-level expansion, churn waterfall, and forward NRR projections into a single board-ready artifact. Building that model from scratch takes 40+ hours of finance work. ModelStack's SaaS Series B Readiness pack includes a pre-built NRR decomposition model, a cohort-level retention waterfall, an expansion forecast template, and the exact slide formats top-quartile founders use in their Series B decks. Skip the spreadsheet engineering and spend that time actually moving the metric.