Short answer: ARR growth tells you how much revenue a SaaS company added. Net revenue retention and capital efficiency tell you whether that revenue was worth adding. NRR shows whether the installed base grows on its own. Capital efficiency shows what each new dollar of ARR actually cost. Growth without either one is rented, not owned.
ARR growth was always a proxy, and everyone forgot that
Sit through enough board meetings and you notice the same slide opens every deck: ARR, up and to the right, with a growth percentage in bold at the top. Then the conversation moves on.
The reason that slide got so much airtime is historical, not analytical. Between 2020 and 2022, capital was cheap and growth was the cleanest available signal that a product had found demand. If you were adding revenue fast, the assumption was that the rest would sort itself out. Burn multiples north of 3x got waved through. Payback periods stretched past two years and nobody blinked.
That world is gone and it has not come back. Median growth across private B2B SaaS has been sliding for years. SaaS Capital’s annual survey of more than a thousand private companies put the median growth rate at 25%, down from 30% two years earlier, and for bootstrapped companies in the $3M to $20M ARR range the median has fallen to around 15%. Meanwhile the median Rule of 40 score across publicly traded SaaS sits closer to 28% than 40%, with only about one in five actively traded names clearing the threshold. The interesting part is why: margins have mostly improved since 2022. It is growth that fell off.
So growth stopped being a differentiator. What replaced it is a pair of questions that ARR growth cannot answer on its own. Does the revenue you already have grow by itself? And how much cash did you set on fire to get the rest?
What net revenue retention actually measures
Net revenue retention is the percentage of recurring revenue you keep from an existing cohort over a period, after churn and downgrades, including expansion from upsells, cross-sells, seat additions and usage growth.
The formula:
NRR = (Starting ARR + Expansion – Contraction – Churn) / Starting ARR
Notice what is missing. New logos do not appear anywhere. That is the whole point. NRR isolates the behaviour of the base you have already paid to acquire, which makes it the only metric on the dashboard that tells you whether your product compounds or leaks.
Its quieter sibling, gross revenue retention, strips expansion out entirely and is capped at 100%. You need both. GRR is the leak detector. NRR is the compounding engine. A company running 118% NRR on 78% GRR is not a healthy business, it is a company outrunning severe churn with aggressive upselling, and that only works while the upsell motion holds.
Benchmarks vary by source, which tells you something about how much segmentation matters here. Broadly:
| Segment | Median NRR | Top quartile |
| SMB (under $25K ACV) | around 97% | 110%+ |
| Mid-market ($25K to $100K ACV) | around 108% | 120% to 125% |
| Enterprise (above $100K ACV) | around 118% | 130%+ |
| Bootstrapped, $3M to $20M ARR | around 103% | 118% (90th percentile) |
Figures drawn from ChartMogul subscription data, the Optifai 2026 pipeline study and SaaS Capital’s private company survey. The blended median across everything lands somewhere near 101% to 108% depending on whose dataset you trust, which is exactly why the blended median is useless to you. Pick the row that matches your ACV.
One number is worth pausing on. SaaS Capital found that growth rate correlates with NRR exponentially rather than linearly. Companies in the highest NRR band reported median growth 83% above the population median. Moving from the 90% to 100% band into the 100% to 110% band added roughly five points of growth on its own. That is not a marketing insight, it is arithmetic showing up in survey data.
The compounding math nobody puts on a slide
Here is the version of this argument that actually changes minds. Two companies, both at $10M ARR today.
Company A runs 95% NRR and sells hard. It closes $5M of new ARR in year one and grows new-ARR production 20% a year by continually adding sales capacity.
Company B runs 120% NRR and sells modestly. It closes $2M of new ARR in year one, growing that by 10% a year.
Year one, Company A looks like the better business. It grows 45% against Company B’s 40%. A board would reward A and question B’s ambition.
Run it five years:
| Year | Company A ARR | Company B ARR |
| 0 | $10.0M | $10.0M |
| 1 | $14.5M | $14.0M |
| 2 | $19.8M | $19.0M |
| 3 | $26.0M | $25.2M |
| 4 | $33.3M | $32.9M |
| 5 | $42.0M | $42.4M |
They finish in the same place. Now look at what it cost.
Company A sold roughly $37M of new ARR across those five years. Company B sold about $12M. Same destination, three times the new business closed. At the blended cost of roughly $1.30 of sales and marketing spend per dollar of new ARR that recent GTM benchmark data reports, that gap is around $48M of S&M for A against $16M for B. Thirty-two million dollars of difference, for identical ARR.
And that comparison flatters Company A, because it charges both companies the same rate per dollar of new ARR. In reality expansion revenue is far cheaper to win than a new logo. New-name CAC ratios run meaningfully above blended.
Every one of those extra dollars A spent came from somewhere: a round, a credit facility, deferred profitability. Company A is a fundraising story. Company B is a business. Both show the same ARR chart.
Capital efficiency: the four numbers that matter
Capital efficiency is not one metric. It is a small set of ratios that answer the same question from different angles. Four are worth tracking properly.
Burn multiple is net cash burned divided by net new ARR in the same period. Popularised by David Sacks at Craft Ventures, it is the bluntest instrument here and the most useful. Below 1.0x is strong. Around 1.2x is a normal Series A median. Past $10M ARR with a burn multiple above 3x, you have a problem you cannot explain away with a market narrative.
CAC payback period is CAC divided by monthly gross profit per customer. This is where 2026 gets uncomfortable. Blended payback across $5M to $50M ARR companies has stretched from roughly 12 to 15 months in 2023 out to around 18 months, according to High Alpha and OpenView benchmark data. Paid auctions inflated, content cycles lengthened, and payback quietly slipped. Bessemer’s efficiency bar sits at LTV:CAC above 3:1 with payback inside 18 months, and the order matters: a 5:1 ratio with 24-month payback is a worse business than 3:1 with 8-month payback, because payback is real cash and LTV is a projection.
Magic number is quarterly net new ARR times four, divided by prior-quarter S&M spend. Above 1.0 means your go-to-market pays for itself inside a year. The 2025 median crossed 1.0 for the first time in several years, which is genuinely good news, though most sub-$25M ARR companies sit nearer 0.7.
ARR per employee is the one founders resist and acquirers love, because headcount is where SaaS cash actually goes.
None of these are new. What changed is their weight in diligence. Burn multiple went from a nice-to-have to a metric most late-stage investors now call critical, and efficient companies have been trading at a clear premium: Bessemer’s Cloud 100 work found companies with sub-1x burn multiples and Rule of 40 scores at or above 40 commanding roughly 2.3x the revenue multiples of inefficient peers.
Why valuation follows retention, not growth
The valuation case is where this stops being philosophy.
McKinsey’s analysis of more than 100 B2B SaaS companies found top-quartile NRR performers trading at a median 24x EV/Revenue while bottom-quartile peers sat at 5x. Nearly a fivefold spread, tracking a single metric. KeyBanc’s survey work points the same direction from the operating side: companies above 110% NRR grew roughly 2.3x faster than peers stuck at 95% to 100%.
The logic is not complicated. A business with 120% NRR has a growing asset even if sales stops entirely for a quarter. A business at 95% has a melting one, and every dollar of growth has to be bought again next year. Investors are not paying for last year’s growth rate. They are paying for how much of next year’s revenue arrives without new spend.
Where organic search fits into the efficiency story
This is the part most metrics posts skip, and it is the part marketing leaders can actually act on.
Capital efficiency is a channel-mix problem before it is a finance problem. CAC varies by more than 100x across channels. Brand search and direct sit at the bottom, roughly $200 to $800 per customer with payback in one to three months. Organic search runs around $500 to $3,000 with payback in two to six months. Businesses can further improve acquisition efficiency by ensuring customer address data is accurate before it enters their CRM or marketing systems. Using the PostGrid Bulk Address Verification API helps validate large address datasets, reducing failed deliveries, improving customer data quality, and eliminating operational costs caused by inaccurate addresses. LinkedIn and ABM run into the thousands or tens of thousands, which is defensible when ACV supports it and ruinous when it does not. Blended CAC is a weighted average of those choices, so it is a decision, not a fact.
Two things follow.
First, cutting content and SEO to fund paid almost always raises blended Customer acquisition cost (CAC) within two quarters, even when the paid dashboard looks unchanged. The organic flywheel decelerates and the cheap volume it was contributing quietly disappears from the mix. Paid acquisition’s share of B2B SaaS pipeline has been falling while organic, content and answer-engine visibility rose, partly because AI Overviews absorbed top-of-funnel queries that paid search used to catch, leaving paid to compete for a narrower and more expensive set of high-intent terms.
Second, organic is one of very few acquisition channels whose cost does not scale linearly with volume. That property is what a burn multiple rewards. A page that ranks and gets cited by answer engines keeps producing pipeline after the spend stops. A paid campaign stops the day the card declines.
And on the retention side: the same content operation that wins search visibility also drives product adoption. Documentation, comparison pages, workflow guides and integration content reduce time-to-value for customers you already have. Adoption drives expansion, expansion drives NRR. That link rarely gets attributed properly in any dashboard, which is why content budgets get cut first and blended CAC deteriorates second.
Rebuilding the reporting
Practical version. Reorder the board deck so the story runs: growth, then efficiency, then quality, then forward view.
- Growth: net new ARR, split explicitly into new logo, expansion and churn. If those three are not broken out, you cannot see NRR.
- Efficiency: burn multiple, CAC payback, magic number.
- Quality: GRR, NRR by ACV tier and cohort, gross margin, ARR per employee.
- Forward view: Rule of 40 trajectory and the specific lever being pulled next quarter.
Report NRR by segment and cohort, never blended alone. A blended NRR of 105% can hide enterprise at 120% and SMB at 88%, which are two completely different companies wearing one number.
How these metrics get quietly gamed
Worth naming, because it happens constantly.
Excluding churned customers from the NRR denominator inflates the number and is the most common offence. Mismatching time horizons between CAC and LTV so that CAC reflects this quarter’s spend while LTV assumes an older, more loyal cohort’s churn rate. Reporting burn multiple on a quarter with unusual working capital timing. Counting a one-off services deal inside net new ARR. Using a blended CAC that averages $200 brand-search signups with $30,000 ABM accounts and calling the result an insight.
If a metric only looks good under one specific definition, that is the finding.
FAQ
What is a good net revenue retention rate for B2B SaaS? It depends entirely on ACV. SMB SaaS under $25K ACV typically sits near 97%, mid-market around 108%, and enterprise above $100K ACV around 118%. Above 120% is best-in-class in any segment. Compare yourself to your ACV tier, not the blended industry median.
Is NRR more important than ARR growth? NRR is more predictive. ARR growth tells you what happened last year; NRR tells you how much of next year’s revenue arrives without new acquisition spend. Companies above 110% NRR have been shown to grow roughly 2.3x faster than peers in the 95% to 100% band, so NRR largely drives growth rather than competing with it.
What is a good burn multiple in 2026? Below 1.0x is strong. Roughly 1.2x is a normal Series A median and 1.4x is a common growth-stage target. Above 3x past $10M ARR signals a real capital efficiency problem.
How is capital efficiency different from profitability? Profitability asks whether you make money today. Capital efficiency asks how much cash it takes to produce a dollar of new recurring revenue. An unprofitable company can be highly capital efficient if each dollar of burn buys more than a dollar of durable ARR.
Does the Rule of 40 still apply? As shorthand, yes, but it is a poor fit below roughly $50M revenue, where healthy companies routinely score under 40 while growing fast on negative margins. Median public SaaS sits near 28% and median private SaaS far lower. Use it for trajectory, not as a pass/fail gate.
Which metric should an early-stage SaaS company prioritise first? Gross revenue retention. Expansion cannot durably offset a leaking base, so if GRR is below your segment floor, any NRR target you set is aspirational rather than achievable.
