"Is our churn rate good?"
It's the question every SaaS founder eventually asks — and the answer is more nuanced than most benchmarks suggest. The right churn rate for a $1M ARR startup selling to SMBs is very different from the right churn rate for a $10M ARR company selling to enterprises.
This guide compiles the 2025 churn benchmarks for B2B SaaS across ARR tiers, market segments, and product categories. More importantly, it explains why the top performers hit their numbers — and what you can do if yours don't measure up.
The B2B SaaS Churn Benchmark Table (2025)
By ARR Tier
| ARR Range | Median Annual Churn | Top Quartile | Bottom Quartile | |-----------|-------------------|--------------|-----------------| | < $1M | 12–20% | < 10% | > 25% | | $1M–$5M | 8–15% | < 6% | > 18% | | $5M–$20M | 5–10% | < 4% | > 12% | | $20M–$50M | 4–7% | < 3% | > 9% | | > $50M | 2–5% | < 2% | > 7% |
Key insight: Churn is inversely correlated with ARR. This isn't just because bigger companies have more resources to fight churn — it's because larger ARR usually correlates with higher ACVs (average contract values), longer contracts, and more enterprise-grade customers who churn less by nature.
By Market Segment
| Segment | Annual Churn Benchmark | Notes | |---------|----------------------|-------| | Enterprise (ACV > $50K) | 3–7% | Long contracts, high switching costs | | Mid-Market (ACV $10–50K) | 5–10% | Most variable; depends heavily on product fit | | SMB (ACV $1–10K) | 10–20% | High churn, but high volume; more recoverable | | PLG / Self-Serve (ACV < $1K) | 20–40% | Expected; offset by CAC efficiency | | Vertical SaaS | 4–8% | Specialist tools with high switching costs |
By Product Category
| Category | Annual Churn | Why | |----------|-------------|-----| | Infrastructure / DevOps | 3–6% | Deep integrations, high switching cost | | CRM / Sales | 7–12% | High competition, frequent evaluation | | Analytics / BI | 5–9% | Data lock-in helps; tool sprawl hurts | | Marketing / Growth | 10–18% | Attribution fragmentation, frequent tool swaps | | Finance / Accounting | 4–8% | Regulatory needs create stickiness | | HR / People Ops | 6–11% | Seasonal churning (layoffs, restructuring) | | Customer Success | 8–14% | High competition from established players | | Collaboration / Productivity | 12–20% | Extremely competitive; winner-take-most |
Monthly vs Annual Churn: How to Convert
Most benchmarks quote annual churn, but if you're tracking monthly, here's the conversion:
Monthly → Annual: Annual churn ≈ 1 - (1 - monthly churn)^12
At 1.5% monthly churn → ~16.8% annual At 2% monthly churn → ~21.5% annual At 0.5% monthly churn → ~5.8% annual
Annual → Monthly: Monthly churn ≈ 1 - (1 - annual churn)^(1/12)
At 10% annual → ~0.87% monthly At 5% annual → ~0.43% monthly
Why this matters: Many founders track monthly churn and compare it to annual benchmarks incorrectly. A 2% monthly churn rate looks modest — but it's equivalent to 21% annual churn, which puts you in the bottom quartile for most segments.
The Hidden Churn: Contraction MRR
Annual and monthly churn measure customers who cancel completely. But revenue churn — the metric that actually matters — also includes contraction: customers who downgrade, reduce seats, or remove add-ons.
Net Revenue Retention (NRR) is the most important churn metric for growth-stage SaaS:
NRR = (Beginning MRR + Expansion - Contraction - Churn) ÷ Beginning MRR × 100
NRR Benchmarks (B2B SaaS, 2025)
| Benchmark | NRR | |-----------|-----| | World-class (Snowflake, Datadog) | 130%+ | | Excellent | 115–130% | | Good | 105–115% | | Average | 95–105% | | Problem zone | < 95% |
An NRR above 100% means you're growing revenue from existing customers even before counting new ones. This is the holy grail of SaaS — and companies with NRR > 120% can often grow without adding a single new customer.
The implication for churn: A 10% gross churn rate can be acceptable if expansion revenue (upgrades, seat additions, add-ons) exceeds it. A 5% gross churn rate is a problem if there's no expansion offsetting it.
Voluntary vs Involuntary Churn: Where Is Your Revenue Going?
Understanding your churn type changes your retention strategy completely.
Voluntary churn: Customer actively cancels. Reason: product isn't delivering value, found a better alternative, budget cut, company shutdown.
Involuntary churn (failed payments): Credit card declined. Customer didn't intentionally cancel — they were churned by their own bank.
The Split
In 2025 data from subscription analytics platforms:
- 20–40% of all SaaS churn is involuntary (failed payments)
- For PLG/self-serve products, this can be as high as 50%
- For enterprise products, it's typically 10–20% (annual invoicing reduces card failures)
This is critical: If you're losing $50K/year to churn, up to $20K of that might be recoverable with basic dunning — automated retry logic, payment method update emails, and SMS reminders.
ChurnRecovery handles this automatically. Unlike tools that charge per recovery, we're free — meaning recovering $20K costs you nothing.
Why Some Companies Have 3x Better Churn Than Their Peers
Looking at the top-quartile performers in each segment reveals consistent patterns:
1. They Do Proper Onboarding
The companies with the lowest churn rates are almost always the companies with the most deliberate, structured onboarding. Not a generic email sequence — guided activation with defined milestones.
The data: Customers who reach a predefined "activation event" within the first 14 days have 40–60% lower churn than those who don't. The activation event varies: it might be "ran first report," "connected first integration," or "invited a teammate."
Companies with low churn have identified their activation event and built their entire onboarding flow to get customers there as fast as possible.
2. They Monitor Usage and Intervene Early
Top performers don't wait for customers to cancel. They monitor usage patterns and reach out proactively when engagement drops — before the customer has decided to leave.
A customer who has logged in 10 times per week and drops to twice per week is exhibiting pre-churn behavior. Catching that moment and sending a check-in (from a CSM or even an automated email) recovers a meaningful percentage of at-risk accounts.
The tools: ChurnScore, product analytics, health scoring in customer success platforms. But the principle works even with a simple spreadsheet if you're under $1M ARR.
3. They Have Cancel Flows That Actually Work
The worst cancel flows are "Are you sure? ← Yes / No →". These don't surface the cancellation reason, don't present retention offers, and don't create any friction for impulsive cancellations.
The best cancel flows:
- Ask for the cancellation reason (8–12 options)
- Route to relevant offers based on reason: discount for price, tutorial for confusion, pause for temporary situations
- Offer a plan downgrade as a middle path
- Present loss framing (what you'll lose access to)
The benchmark: Well-designed cancel flows recover 15–30% of attempted cancellations. Poorly designed or absent cancel flows recover 2–5%.
4. They Build Products With High Switching Costs
Not "artificial" switching costs (export fees, data lock-in) — those just generate resentment. Real switching costs:
- Network effects (teammates using the tool)
- Deep integrations (connected to 5 other tools in the stack)
- Data accumulation (3 years of history in the product)
- Workflow automation (custom rules, triggers, templates)
The more customers invest in your product, the higher the psychological cost of leaving. Design for depth of usage, not just breadth.
5. They Segment and Price Correctly
Churn often has a pricing root cause that looks like a product problem. Customers on plans that don't match their use case are chronically at risk — either they're overpaying for features they don't use, or they're hitting plan limits that create friction.
Top performers regularly audit which plan segments have highest churn and investigate whether a pricing or packaging change would solve it.
Cohort Analysis: The Only Churn Analysis That Matters
Aggregate churn rates are misleading. A 10% annual churn rate looks consistent — until you look at cohorts:
Scenario: A company's overall churn is 10% annually.
| Cohort | Annual Churn | |--------|-------------| | Customers acquired in Q1 2024 | 22% | | Customers acquired in Q2 2024 | 18% | | Customers acquired in Q3 2024 | 12% | | Customers acquired in Q4 2024 | 6% |
The company improved dramatically. But the aggregate number hides it.
The inverse also happens: a worsening business can look fine in aggregate because high-quality early cohorts are masking new cohort deterioration.
How to read cohort data:
- Plot each cohort's churn curve (% retained by month)
- Compare curves across cohorts: are they improving?
- Look for the "churn cliff" — the month where each cohort shows its steepest drop
- Identify if churn is front-loaded (bad onboarding) or late-loaded (product loses value over time)
Benchmarking Yourself Correctly
Most founders compare themselves to the wrong peer group. "SaaS churn benchmarks" lumped together include consumer apps, horizontal tools, enterprise software, and bootstrapped indie products — none of which should be measured the same way.
The right comparison peer group:
- Same ARR tier (±1 tier)
- Same market segment (SMB / MM / Enterprise)
- Similar ACV
- Similar product category
A $3M ARR marketing analytics tool for agencies should not be comparing itself to Salesforce's churn rate. It should be comparing itself to other $3–5M ARR marketing analytics tools with similar ACVs.
Where to find peer data:
- OpenView Partners SaaS Benchmarks (annual)
- Bessemer Venture Partners State of the Cloud (annual)
- SaaS Capital surveys
- ChartMogul SaaS Benchmarks
What to Do If Your Churn Rate Is Too High
Step 1: Diagnose the type. Split voluntary vs involuntary. Fix involuntary churn first — it's the fastest win. Dunning and cancel flow implementation typically takes 1–2 weeks and recovers 5–15% of churning revenue with minimal effort.
Step 2: Identify the cohort pattern. Is churn front-loaded (first 60 days) or back-loaded (months 6–12+)? Front-loaded churn is an onboarding problem. Back-loaded churn is a product depth problem.
Step 3: Survey churned customers. Exit surveys with 3–5 questions, sent within 48 hours of cancellation. The insights from churned customers are worth more than any analytics tool.
Step 4: Instrument your activation event. Define your activation event and measure what % of new customers reach it within 14 days. If it's below 60%, that's your priority.
Step 5: Build or improve your cancel flow. If you don't have a cancel flow with reason tracking and retention offers, implement one. The ROI is consistently 10–30x the implementation cost.
Step 6: Set up health scoring. Even a simple model (logins/week + feature breadth + integrations connected) lets you identify at-risk customers before they cancel.
The Bottom Line on Benchmarks
Churn benchmarks are useful for calibration, not judgment. A 15% annual churn rate is catastrophic for a $30M ARR enterprise company — and completely normal for a $500K ARR PLG product.
The questions that matter:
- Is your churn improving quarter over quarter?
- Is your NRR above 100%?
- Are you in the top quartile for your segment and ARR tier?
- Are you capturing the low-hanging fruit of involuntary churn recovery?
If you're not recovering failed payments and you don't have a cancel flow, start there. These are the highest-ROI churn interventions available — and with tools like ChurnRecovery, they're free to implement.
How ChurnRecovery Helps You Hit Better Benchmarks
ChurnRecovery addresses both types of churn:
Voluntary churn: Cancel flow with reason detection and retention offers. When a customer clicks "cancel," we intercept with a branded flow that asks why and presents relevant save offers. No code required to get started.
Involuntary churn: Dunning sequences that automatically retry failed payments, send card update emails, and escalate through SMS before cancelling a subscription.
Analytics: See your save rate, offer acceptance by reason, revenue recovered, and cohort-level churn data — all in one dashboard.
No monthly fees. No setup fees. No per-recovery charges.
Get early access to ChurnRecovery →
How does your churn rate compare? Try our free churn calculator to see how much revenue you're losing — and how much you could recover.