What are healthy SaaS unit economics in 2026?
Median B2B SaaS net revenue retention is 82%; companies holding NRR at or above 100% grow ~48% YoY — roughly 3.5× faster than the sub-60% NRR cohort.
ChartMogul panel of ~2,700 B2B companies ≥ $250k ARR (2025) and cohort studies 2023–2026 · verified as of
Net revenue retention by segment
NRR is the single strongest growth predictor in every panel below. The AI-native figure is the surprise: AI products monetize attention fast but retain revenue poorly.
| Metric | Value | Segment | Period | Source |
|---|---|---|---|---|
| B2B SaaS — median NRR | 82% | ~2,700 companies ≥ $250k ARR | 2025 | chartmogul.com |
| B2C SaaS — median NRR | 49% | ~600 companies ≥ $250k ARR | 2025 | chartmogul.com |
| AI-native SaaS — median NRR | 48% | ~200 companies ≥ $250k ARR | 2025 | chartmogul.com |
| Top-quartile NRR, $1–3M ARR | 94% | growth-stage cohort | 2023 | chartmogul.com |
| Top-quartile NRR, ARPA > $1k | 110% | B2B, higher price points | 12 mo ending Mar 2023 | chartmogul.com |
| Top-quartile gross revenue retention | 86% | any stage | 2023 | chartmogul.com |
Growth rate by retention tier
Retention compounds into growth. The spread between retention tiers is wider than the spread between funding stages.
| Metric | Value | Segment | Period | Source |
|---|---|---|---|---|
| Median YoY growth, NRR ≥ 100% | 48% | high-retention cohort | 2024 | chartmogul.com |
| Median YoY growth, NRR < 60% | 13.1% | low-retention cohort | 2023 | chartmogul.com |
| Growth of efficient cohort (high NRR + CAC payback low) | 71% | median; Rule of 40 score 47 | 2025 | highalpha.com |
| Growth of inefficient cohort (NRR < 98% + CAC payback > 15 mo) | 10% | median YoY | 2025 | highalpha.com |
| Growth uplift from deep AI embedding vs AI-as-feature | +70% | $1–5M ARR companies | 2025 | highalpha.com |
Trial conversion and churn
The single most actionable lever in the data: requiring a card at trial start nearly quadruples free-to-paid conversion.
| Metric | Value | Segment | Period | Source |
|---|---|---|---|---|
| Median free-to-paid conversion, self-serve trials | 8% | opt-in (no card) | study Jan 2026 | chartmogul.com |
| Free-to-paid conversion, card-gated trials | 30% | opt-out (card required) | study Jan 2026 | chartmogul.com |
| Healthy band, card-gated trials | 25–35% | GOOD range | study Jan 2026 | chartmogul.com |
| Median monthly customer churn, early stage | 3–4% | pre/at product-market fit | 12 mo ending Mar 2023 | chartmogul.com |
| Median monthly churn, low-NRR cohort | 7% | ≈ 2× the churn of NRR ≥ 100% cohort | H1 2024 | chartmogul.com |
Customer acquisition cost
CAC keeps climbing across the industry. The widely-quoted 26-month payback figure is a 2021 survey — treat it as historical context, not a current benchmark.
| Metric | Value | Segment | Period | Source |
|---|---|---|---|---|
| CAC increase vs 5 years prior | +60% | industry-wide | 2026 | paddle.com |
| Average CAC payback periodhistorical | 26 mo | historical reference | 2021 survey | paddle.com |
Why do the growth-uplift numbers differ?
You will see the NRR growth advantage quoted three ways: "2.5× faster" (High Alpha, 2024), "43.6% vs 13.1%" (ChartMogul cohorts, 2022) and "48% vs 13.1%" (ChartMogul, 2023–24). These are different panels, years and cut-offs — not contradictions. Every version agrees on the direction and rough magnitude: NRR ≥ 100% companies grow 3–4× faster than the lowest-retention cohort. When a single number matters for your model, use the panel closest to your segment — and check its sample size first.
Methodology
Every number on this page is a fact node in the God of Startups Atlas — a knowledge graph where each metric carries its source URL, reporting period and evidence tier, and is re-verified against the source before publication. Numbers are never generated or extrapolated by AI. Tier 2 = established industry panels (ChartMogul, Paddle, High Alpha) with disclosed sample sizes; where a figure is dated, we say so instead of relabeling it with the current year.
Related benchmarks
Where does your model sit on these curves?
Map your startup’s numbers against this data inside God of Startups. Your idea is checked against the market, never shared with it.