SaaS Churn Rate Benchmarks 2026: By Industry, ARR & Stage
- Median NRR for B2B SaaS ≈ 105%; top quartile clears 115%, best-in-class DevTools 130%+.
- Healthy monthly logo churn: <1% mid-market, <0.5% enterprise. SMB sits at 3–5%/month.
- GRR median ≈ 88–92% for venture-backed B2B SaaS; below 80% signals a structural retention problem.
- Industry spread is huge: DevTools NRR 110–130% vs. SMB-Tools 90–100%.
- Most "bad" benchmarks are methodology mismatches — monthly vs. annual, customer- vs. revenue-weighted.
Every B2B SaaS operator googles SaaS churn rate benchmarks at least once per quarter — usually the night before a board meeting. The problem: most benchmark posts cherry-pick a single number, ignore segment differences, and conflate monthly with annual figures. This guide aggregates publicly available 2026 benchmarks across four dimensions — ARR stage, industry, customer segment, and metric type — so you can find the row that actually matches your business.
Numbers below are aggregated ranges synthesized from public benchmark reports including the KeyBanc Capital Markets / Capstone SaaS Survey, ChartMogul SaaS Benchmarks, and SaaS Capital Survey. They reflect 2025/2026 data; vendor and cohort definitions vary, so treat ranges as directional, not exact.
What counts as a "good" SaaS churn rate in 2026
Three metrics, three different conversations. Mixing them up is the most common cause of confused board discussions.
- Logo churn rate. Number of customers lost over a period, divided by customers at the start of the period. Counts a $50/mo and a $50K/yr customer equally.
- Gross Revenue Retention (GRR). Revenue retained from the existing book of business, excluding any expansion. Caps at 100%. Exposes the underlying retention engine.
- Net Revenue Retention (NRR). Revenue retained including expansion (upsells, seat growth, usage growth). Can exceed 100%. Shows the combined effect of retention and account growth.
For 2026, the median venture-backed B2B SaaS company runs roughly 105% NRR, 88–92% GRR, and 10–14% annual logo churn. Top-quartile companies materially outperform on all three. Below median is not catastrophic — most companies sit there — but it usually means at least one of three things: weak onboarding, weak expansion motion, or a structural fit problem in a customer segment you continue to sell to.
Logo churn benchmarks by ARR stage
Logo churn drops sharply as companies move upmarket — partly because larger contracts come with annual commitments and partly because the customers themselves churn out of business less often. The cliff between sub-$1M ARR and $10M+ ARR is the single largest gap in any SaaS benchmark.
| ARR stage | Monthly logo churn | Annualized logo churn | Typical contract |
|---|---|---|---|
| < $1M ARR | 3.0% – 5.0% | 30% – 45% | Monthly self-serve |
| $1M – $10M ARR | 1.5% – 2.5% | 18% – 26% | Mix of monthly + annual |
| $10M – $50M ARR | 0.8% – 1.5% | 9% – 17% | Annual, sales-led |
| $50M+ ARR | 0.4% – 0.9% | 5% – 10% | Annual / multi-year |
Ranges aggregated from public 2025/2026 benchmark reports; post-onboarding cohort, customer-weighted.
The mistake at every stage is benchmarking against the next stage up. Sub-$1M ARR teams compare themselves to mid-market numbers, conclude they're broken, and start over-engineering retention before product-market fit. The correct comparison is your stage row — and the trajectory across stages, not the level within one.
Gross & net revenue retention by ARR stage
Revenue retention tells the more honest story. Two companies with identical 15% logo churn can have wildly different revenue impact depending on which logos churn — small accounts churning is normal, top-decile accounts churning is structural.
| ARR stage | GRR (median range) | NRR (median range) | Top-quartile NRR |
|---|---|---|---|
| < $1M ARR | 75% – 85% | 85% – 95% | 105%+ |
| $1M – $10M ARR | 82% – 90% | 95% – 110% | 118%+ |
| $10M – $50M ARR | 88% – 93% | 105% – 118% | 125%+ |
| $50M+ ARR | 90% – 95% | 110% – 125% | 135%+ |
Ranges aggregated from public 2025/2026 benchmark reports; revenue-weighted, post-onboarding cohort.
The clearest signal across stages is GRR. Healthy GRR > 90% means your retention engine works without expansion having to compensate. NRR over 110% with a GRR below 85% usually means a few large expansion deals are masking real retention problems — the kind that surface the moment expansion slows.
Churn benchmarks by industry
Industry effects are enormous and often dwarf stage effects. A $5M ARR DevTools company will materially outperform a $20M ARR SMB-Tools company on NRR, and that gap is structural, not operational.
| Industry / Vertical | NRR (median range) | Annual logo churn | Notes |
|---|---|---|---|
| DevTools / Infrastructure | 110% – 130% | 8% – 15% | Consumption pricing → strong NRR |
| B2B FinTech | 105% – 120% | 10% – 18% | High switching cost, regulated |
| Vertical SaaS | 102% – 115% | 8% – 14% | Industry lock-in, low expansion ceiling |
| HR-Tech | 100% – 110% | 14% – 22% | Headcount-tied, expansion follows growth |
| MarTech | 95% – 108% | 16% – 26% | Crowded category, frequent re-evaluations |
| SMB-Tools (horizontal) | 90% – 100% | 22% – 35% | SMB customers churn out of business |
Ranges aggregated from public 2025/2026 benchmark reports; revenue-weighted (NRR), customer-weighted (logo).
Two practical implications. First: if you're an SMB-Tools company benchmarking against DevTools, you'll always look bad — the underlying customer base churns differently. Compare to your industry, not to the leaderboard. Second: industry-relative position matters more than absolute number. A MarTech company at 105% NRR is in the top quartile of MarTech; a DevTools company at 105% NRR is below median.
Benchmarks by customer segment
Customer segment is the dimension most often missed in board reporting. A blended company-wide NRR of 102% can hide a 130% NRR in the enterprise book and an 80% NRR in the SMB book — two completely different businesses inside one P&L.
| Segment | Annual logo churn | GRR (median) | NRR (median) |
|---|---|---|---|
| SMB (<$10K ACV) | 30% – 50% | 70% – 82% | 80% – 95% |
| Mid-Market ($10K–$100K ACV) | 12% – 20% | 85% – 92% | 100% – 115% |
| Enterprise ($100K+ ACV) | 5% – 10% | 90% – 96% | 110% – 130% |
Ranges aggregated from public 2025/2026 benchmark reports; ACV bands reflect typical industry usage.
If you sell across all three segments, report all three rows separately. A single blended number is a vanity metric — it can't be acted on, because the levers in each segment are different. SMB churn is fixed by onboarding and pricing; mid-market churn by CSM motion and product depth; enterprise churn by executive alignment and renewal pipeline discipline. Same number, three completely different playbooks.
Why your churn rate looks worse than the benchmark
Most teams whose churn looks 2× worse than benchmark don't have a 2× retention problem — they have a methodology mismatch. The four most common:
- Monthly vs. annualized. A 2% monthly churn is not "twice the 12% annual benchmark" — it compounds to 22% annually. Always compare apples to apples.
- Customer-weighted vs. revenue-weighted. If your top decile of accounts retains at 98% and the bottom decile at 70%, customer-weighted churn looks bad while revenue-weighted churn looks fine. Benchmarks usually report revenue-weighted.
- Trial / onboarding contamination. Including trial dropouts and customers in the first 90 days inflates churn 2–3× compared to public reports, which use post-onboarding cohorts.
- Cohort definition. Some benchmarks track only paying customers continuously active 12+ months. If you include every customer ever signed, your number looks worse for structural reasons that have nothing to do with your retention engine.
Before concluding you have a retention problem, normalize your measurement to match the benchmark's methodology. Half the time the gap closes. The other half, you still have a problem — and it's worth reading our complete guide to reducing SaaS churn for the seven tactics with the clearest track record. If the bottleneck is simply seeing at-risk accounts before they cancel, our comparison of churn prediction tools covers what's available in the category.
For broader business-model context, see the Wikipedia overview of churn rate and software as a service.
Frequently asked questions
What is a good SaaS churn rate in 2026?
For B2B SaaS in 2026, healthy monthly logo churn sits below 1% for mid-market and below 0.5% for enterprise. Annualized that translates to roughly 5–12% logo loss for healthy companies. Anything above 2% monthly compounds to 22%+ annually, which most SaaS businesses cannot out-grow.
What is the difference between gross and net revenue retention?
GRR measures only revenue lost from churn and downgrades; expansion is excluded and the metric caps at 100%. NRR includes expansion and can exceed 100%. GRR exposes the underlying retention engine; NRR shows retention plus upsell. A healthy NRR can hide a leaky GRR — always look at both.
What is a good NRR for a B2B SaaS company?
Median NRR for venture-backed B2B SaaS sits around 105% in 2026. Top-quartile companies are above 115%, best-in-class DevTools and infrastructure SaaS often clear 130%. SMB-focused tools usually run lower (90–100%) because expansion is structurally harder.
How does churn rate vary by industry?
DevTools and B2B FinTech tend to show the highest NRR (110–130%) thanks to consumption-based pricing and high switching cost. HR-Tech and MarTech sit closer to median (95–110%). SMB-focused horizontal tools have the highest annual logo churn (22–35%) because SMB buyers themselves churn out of business at higher rates.
Why is my churn rate higher than the benchmark?
Most often it's a methodology mismatch, not a real performance gap. Public benchmarks usually report annualized, revenue-weighted, post-onboarding cohorts. If you measure monthly, customer-weighted, or include trial dropouts, your number can look 2–3× worse than the benchmark even if your retention engine is healthy.
How often should I benchmark my churn rate?
Quarterly is enough for board reporting. The benchmarks themselves shift slowly — annual updates from KeyBanc, ChartMogul, and SaaS Capital are sufficient to recalibrate. Re-running benchmarks more often than that adds noise without signal.