21-Point NRR Gap: What Good Actually Looks Like by Stage and Segment in 2026

According to McKinsey analysis of more than 100 B2B SaaS companies, top-quartile NRR performers trade at 24x EV/Revenue. Bottom-quartile companies sit at 5x. The difference between those two cohorts is roughly 15 percentage points of NRR.

That gap does not happen by accident. It is the compounded result of CS organizations that know which NRR benchmark applies to their stage and segment, and build toward the right one.

Most CS Directors walk into board meetings with a single NRR number. Boards compare it to something they saw in an investor memo, which almost certainly used the private SaaS industry average of roughly 101-106% without stage or segment context. The result is a conversation about whether the number is good or bad, with neither side having the information needed to answer that question.

The data tells a more useful story. Here is what it says.

The Segment Gap Is the Biggest Variable

Before looking at stage benchmarks, the most important variable to control for is customer segment. Annual contract value (ACV) predicts NRR better than almost any other structural factor, because it determines how much expansion surface each account carries.

According to Optifai's analysis of 939 B2B SaaS companies, the median NRR by customer segment in 2025 breaks down as follows:

That 21-point spread between SMB and enterprise medians is structural, not a performance gap. Enterprise accounts have more expansion surface: additional seats, more business units to land in, usage-based overages, new product modules. When expansion is structurally available and a CSM is actively working it, NRR climbs.

SMB accounts run on tighter budgets, carry higher logo churn volatility, and have fewer natural expansion triggers. Holding 97% NRR in an SMB-focused product often represents excellent CS execution given those constraints.

The 106% industry average is a blend of all three segments. It describes no real company in particular.

A 97% NRR at an SMB company and a 97% NRR at an enterprise company look identical on a slide. The interpretation is opposite. The first is approximately median for its cohort. The second is a serious warning sign, one that typically signals churn or contraction well above segment peers.

This is the most important adjustment CS Directors need to make before any board conversation. Segment comes first.

NRR by Stage: What Investors Actually Expect

Segment explains most of the benchmark variance. Stage explains the rest.

ChartMogul's "The New Normal for SaaS" Retention Report, which covers retention data from 2,100 VC-backed SaaS companies, tracks NRR expectations across ARR stages. Synthesized alongside data from the KeyBanc Capital Markets 2025 Private SaaS Company Survey and SaaS Capital's retention research, the stage expectations map looks like this:

A few patterns in this table are worth pulling out.

Crossing 100% NRR is the defining milestone at Series A. Below that threshold, the existing customer base is shrinking. Above it, expansion has started to compound. The Bessemer Venture Partners framework from their Atlas documentation (100% = good, 110% = better, 120%+ = best) applies most precisely to Series B and beyond, where the expansion motion is expected to be mature and intentional.

NRR often compresses at scale. Companies at $100M+ ARR can see NRR drop slightly from the growth-stage peak, not because CS teams lose effectiveness, but because core accounts are more fully penetrated and the customer mix has broadened to include segments with less natural expansion potential. A moderate compression at scale is normal, not alarming.

A 95% NRR at seed stage is fine if new customer acquisition is strong and there is a credible path to improvement. The same number at Series B signals a product-market fit or retention problem that a CS process change alone will not fix.

The Quartile Spread: Where the Board Conversation Gets Useful

Knowing your segment median and stage target sets the baseline. The quartile spread within your cohort is where the conversation becomes genuinely actionable.

McKinsey's analysis of more than 100 B2B SaaS companies across Q1 2019 to Q4 2024 found:

That 24x vs. 5x multiple difference, holding growth rate constant, is the most concrete argument for investing in CS operations that a board will encounter. The companies at the top quartile are not running fundamentally different businesses. They are running better retention and expansion programs on top of comparable product bases.

For a company at $30M ARR, the spread between top-quartile NRR (113%) and bottom-quartile NRR (98%) translates to roughly $4.5M in annual organic revenue difference. The top-quartile company generates approximately $3.9M from net expansion. The bottom-quartile company loses $600K from its existing base. That swing appears in valuation, in capital efficiency, and in how much the sales team has to run just to hold the company flat.

Here's the thing. The board question should not be "is our NRR good?" It should be "are we in the top quartile for our stage and segment, and if not, what specifically are we doing to close the gap?"

SaaS Capital's ACV-tier breakdown of the $25K-$50K mid-market cohort confirms how wide within-cohort variation gets: bottom quartile at 97%, median at 102%, top quartile at 111%. A 14-point spread within a single ACV tier means operational choices, not structural factors, explain most of the variation at this level.

Bottom line: operational discipline, not company size, separates top-quartile NRR performers from median ones within the same segment.

The 2021-2024 Compression: Context for Your Current Number

One more piece of context matters before any board conversation about NRR trajectory.

Private SaaS median NRR declined for three consecutive years. Benchmarkit's 2025 B2B SaaS Performance Metrics report (covering 1,600+ private SaaS companies) and KeyBanc Capital Markets' 16th Annual Private SaaS Survey both document this trend:

The decline reflects two overlapping forces: post-pandemic SaaS rationalization (buyers cutting seats and consolidating tools after 2020-2021 over-purchasing) and softer expansion motions as budget scrutiny increased. The 2025 data suggests stabilization, with gross retention (GRR) recovering from a 2023 low near 86% back to roughly 90%.

For CS Directors, this context matters because many investor expectations were calibrated to 2021-era benchmarks (105-110% as a standard for growth-stage SaaS). Those benchmarks no longer reflect current conditions. A company holding 102% NRR in 2026 is performing better relative to peers than the same number would have suggested in 2021.

What This Means for Your Board Conversations

CS Directors who frame NRR against the industry average are giving boards a number that is simultaneously too high to be a useful baseline for enterprise companies and too low to be meaningful for SMB-focused ones.

The more useful framing is cohort position: where are you relative to companies at your ACV level and ARR stage? Are you at the median, below it, or approaching the top quartile? Is the gap a GRR problem (churn) or an NRR minus GRR problem (expansion)? And what would it take to move one quartile up?

That conversation is one a board can act on. "We're above the industry average" is not.

More on the early engagement signals that predict NRR movement before the quarter closes is in research on why health scores miss the signals that matter most. For the foundational NRR formula, the mechanics of GRR vs. NRR, and how valuation connects to retention, the NRR explainer covers those in detail. For the CS execution patterns that separate top-quartile from median teams, the analysis of the playbook execution gap is directly relevant.

What to Do

Find your actual cohort. Segment (ACV range) and stage (ARR) are the two variables. Use those to identify your peer median and top-quartile target. Stop using the industry-wide average.

Separate churn from expansion. GRR measures churn control. The difference between your NRR and GRR measures expansion capture. If NRR is below your cohort median, diagnose which component is driving the gap. Churn and expansion require different interventions and different KPIs to track.

Build the board narrative around quartile position. Present three numbers: current NRR, cohort median, and your top-quartile target. Then explain what you are doing to close the gap. That is a board-level CS narrative. "We're above the industry average" is not.

Detect expansion signals earlier. The companies at the top quartile within each cohort are not running harder. They are detecting expansion opportunity earlier, from engagement data in communications, open commitments, and account threading depth, before a formal renewal cycle forces the issue. Cuelock's AI agents connect to Slack, Gmail, and calls to surface those signals without manual data entry. The AI Readiness Gap Finder shows where manual monitoring is currently concentrated in your workflow.

Use Cuelock's free NRR Benchmark Calculator to compare your current NRR against segment and stage-matched benchmarks in under two minutes.

---

Methodology / Sources: Segment benchmarks from Optifai's analysis of 939 B2B SaaS companies, corroborated by SaaS Capital's B2B SaaS Retention Benchmarks. Stage benchmarks synthesized from ChartMogul "The New Normal for SaaS" Retention Report (N=2,100 VC-backed SaaS companies), KeyBanc Capital Markets 2025 Private SaaS Company Survey (16th Annual), and SaaS Capital retention research. The Concerning/Acceptable/Good/Exceptional framework is a synthesis of benchmarks from these three sources and is not a direct quote from any single report. Quartile and valuation data from McKinsey "The Net Revenue Retention Advantage: Driving Success in B2B Tech" (analysis of 100+ B2B SaaS companies, Q1 2019 to Q4 2024). NRR trend data from Benchmarkit 2025 B2B SaaS Performance Metrics (1,600+ companies). Stage-specific Bessemer framework references BVP Atlas documentation. All figures are approximate benchmarks for orientation purposes.