The SaaS Metrics Glossary: MRR, ARR, NRR, CAC, LTV, Payback (2026)
Every core SaaS metric defined with its formula, healthy benchmarks by stage, and what each number actually predicts. The reference sheet for founders, operators, and investors.
TL;DR. SaaS businesses run on a small set of interlocking metrics: MRR/ARR (revenue scale), NRR/GRR (revenue durability), churn (leakage), CAC and payback (cost of growth), LTV:CAC (unit economics), burn multiple (capital efficiency), and Rule of 40 (growth-profit balance). This glossary defines each one, gives the formula, and states benchmarks by stage — because a "good" number at $1M ARR is a warning sign at $20M.
Revenue metrics
MRR (Monthly Recurring Revenue)
MRR is the normalized monthly value of all active subscriptions. Annual contracts divide by 12; one-time fees, services, and overage charges are excluded.
MRR moves through four channels, and reporting them separately is what makes MRR useful:
- New MRR — from new customers
- Expansion MRR — upgrades, seats, cross-sell to existing customers
- Contraction MRR — downgrades
- Churned MRR — cancellations
Net New MRR = New + Expansion − Contraction − Churned
ARR (Annual Recurring Revenue)
ARR = MRR × 12. Used for scale conversations ($1M ARR, $10M ARR milestones). The trap: counting non-recurring revenue (services, one-time setup) in ARR. Investors will strip it out in diligence; better to report it clean yourself.
Retention metrics
GRR (Gross Revenue Retention)
GRR measures how much recurring revenue you keep from an existing cohort, ignoring expansion. Formula: (starting MRR − churn − contraction) / starting MRR, over 12 months. Ceiling is 100%.
Benchmarks: 90%+ is healthy for SMB products, 95%+ for enterprise. GRR below 80% means the product is leaking faster than most GTM motions can refill.
NRR (Net Revenue Retention)
NRR is GRR plus expansion — how much a cohort's revenue grows even with zero new logos. Formula: (starting MRR − churn − contraction + expansion) / starting MRR.
| Stage / segment | Healthy NRR |
|---|---|
| SMB-focused SaaS | 95–105% |
| Mid-market | 105–115% |
| Enterprise | 110–130% |
NRR above 100% means the business grows without new sales — the single strongest predictor of durable SaaS value. Public-market leaders (Snowflake, Datadog in their growth years) ran 130%+.
Churn
Logo churn = % of customers lost per period. Revenue churn = % of MRR lost. They diverge whenever customers differ in size — losing 10 tiny accounts can matter less than one enterprise contraction. Always state which churn you mean and over what period; a "2% churn" that's monthly is ~22% annually.
Acquisition metrics
CAC (Customer Acquisition Cost)
CAC = total sales + marketing spend / new customers acquired in the period. Include salaries, tools, and program spend — not just ad budget. A "blended CAC" mixes paid and organic; report paid CAC separately or the blend hides deteriorating paid efficiency behind organic growth.
CAC Payback Period
Payback = CAC / (monthly recurring gross profit per customer). Note gross profit, not revenue — divide by gross margin.
| Motion | Healthy payback |
|---|---|
| Self-serve / PLG | under 12 months |
| Mid-market sales-led | 12–18 months |
| Enterprise | 18–24 months |
Payback is the most operationally honest acquisition metric: unlike LTV, it uses no forward-looking assumptions. See Product-Led Growth for how PLG motions compress it.
LTV (Customer Lifetime Value) and LTV:CAC
LTV = average monthly gross profit per customer × average customer lifetime in months, where lifetime ≈ 1 / monthly churn rate. The classic health bar is LTV:CAC ≥ 3.
Handle with care: at low churn rates the formula extrapolates decades of retention from months of data. Early-stage companies should lean on payback and NRR; LTV becomes trustworthy once cohorts are 2+ years old.
Efficiency metrics
Burn Multiple
Burn multiple = net burn / net new ARR. How many dollars you burn to add a dollar of ARR. Under 1x is excellent, 1–1.5x good, 2x+ concerning past Series A. It's the metric that catches "growth at any cost" — a company can show strong ARR growth while the burn multiple quietly reveals the growth is bought, not earned.
Rule of 40
Revenue growth rate % + profit margin % ≥ 40. A 60%-growth company can burn 20% margins; a 10%-growth company must run 30% profitable. Below ~$10M ARR the rule is noisy; it becomes a real gate in growth and late stage.
Magic Number
Net new ARR in a quarter × 4 / sales & marketing spend of the prior quarter. Above 0.75, sales efficiency supports increased investment; below 0.5, fix the funnel — see ICP definition — before adding spend.
Which metrics matter at which stage?
- Pre-PMF (under $1M ARR): activation, retention cohorts, logo churn. Ignore LTV. Watch product-market fit signals.
- $1–10M ARR: NRR, CAC payback, burn multiple. This is where unit economics get set.
- $10M+ ARR: NRR, Rule of 40, magic number, paid vs. blended CAC. Efficiency now prices the company.
Related reading
- All SaaS metrics articles — the growing cluster.
- The Complete Guide to OKRs for SaaS Teams — turning these metrics into quarterly goals.
- Need a metrics deep-dive for your board deck? Talk to us.
Related reading
How to Build an Ideal Customer Profile (ICP): Step-by-Step for B2B
A practical, data-first method for defining your Ideal Customer Profile — the 6-step process, the attributes that actually predict fit, ICP vs buyer persona, and a worked SaaS example.
OKR Examples for B2B Marketing Teams (with Grading)
12 real-world OKR examples for B2B marketing teams — demand gen, ABM, content, and brand — with graded end-of-quarter scores and the reasoning behind each one.
OKRs vs KPIs: When to Use Which (and How They Work Together)
OKRs are time-boxed change goals; KPIs are standing health metrics. The difference explained with a decision table, SaaS examples, and the three-step way to run both without confusing your team.
What Is Product-Led Growth? Definition, Examples, and Playbook (2026)
Product-led growth explained — what PLG actually is, how it differs from sales-led, the metrics that define it (activation, TTV, PQLs), real company examples, and when PLG is the wrong choice.