Every published SaaS churn benchmark blends per-seat, usage-based, flat-rate, and tiered pricing into a single number. That number is useless for portfolio operators comparing companies with fundamentally different billing models. A 4% monthly churn rate is alarming for a usage-based product and perfectly normal for a flat-rate one — but the blended benchmark treats them identically. Here's the segmented data that most reports omit.
SaaS churn benchmarks by pricing model — what the data shows
Churn rates vary 3–5× by pricing model. A flat-rate SaaS product with a single tier churns at 4–6% monthly — roughly triple the rate of a usage-based product at 1.5–2.5%. The gap isn't noise. It reflects structural differences in how customers engage with the product, how deeply they integrate it into workflows, and how much switching cost the pricing model creates.
Yet the benchmarks most investors cite — from OpenView, Bessemer, SaaS Capital — report a single median churn rate across all pricing models. When a portfolio company reports 3.5% monthly churn, the first question shouldn't be "is that good?" It should be "what pricing model are they on?"
Customer Churn Rate
Percentage of customers lost in a given period — the clearest signal of product-market fit.
Why blended churn benchmarks mislead
Blended benchmarks average across pricing models that produce structurally different churn profiles. The result is a number that describes no actual company. It's the equivalent of averaging the fuel efficiency of sedans and pickup trucks and calling it the benchmark for "vehicles."
The problem is worse for portfolio operators. When you're comparing a per-seat collaboration tool against a usage-based data platform, a blended benchmark tells you nothing about which company has a retention problem. One might be outperforming its model's norm while looking bad against the blend. The other might be underperforming while looking fine.
Per-seat vs usage-based vs flat-rate — fundamentally different churn dynamics
Per-seat pricing ties cost to headcount. When a customer downsizes, they reduce seats before they cancel — creating a contraction signal that precedes churn by weeks or months. The churn event itself is a lagging indicator. Median monthly churn lands at 3–4%, but the contraction-to-churn pipeline makes it partially predictable.
Usage-based pricing creates natural stickiness. Customers who consume more pay more, which means the product is embedded in a workflow that generates usage. Cancellation requires replacing that workflow, not just finding a cheaper tool. Monthly churn runs 1.5–2.5% at the median — roughly half the per-seat rate.
Flat-rate pricing has the weakest retention mechanics. A single tier at a fixed price offers no expansion path, no usage lock-in, and no contraction buffer. Customers either stay or leave — there's no middle ground. That binary dynamic pushes monthly churn to 4–6%, the highest of any model.
Churn benchmarks by pricing model
| Pricing Model | Median Monthly Churn | Top Quartile | Key Driver |
|---|---|---|---|
| Per-seat | 3–4% | < 2% | Seat contraction precedes cancellation |
| Usage-based | 1.5–2.5% | < 1% | Workflow lock-in via consumption |
| Flat-rate / single-tier | 4–6% | < 3% | No expansion path, binary stay/leave |
| Tiered | 2.5–3.5% | < 1.5% | Upgrade/downgrade flexibility absorbs churn pressure |
Per-seat SaaS — median monthly churn 3–4%
Per-seat products see the clearest early-warning signals. A customer dropping from 50 seats to 30 over two months is broadcasting intent. The churn event — full cancellation — typically follows 2–4 months after sustained contraction begins. Top-quartile per-seat companies hold monthly churn below 2% by intercepting the contraction signal with re-engagement campaigns before it converts to cancellation.
The failure mode is ignoring contraction data. Companies that track only logo churn miss the revenue erosion happening inside retained accounts. A customer who went from $5,000/mo to $1,500/mo is still "retained" in logo terms but has effectively churned 70% of their revenue.
Usage-based SaaS — median monthly churn 1.5–2.5% (higher stickiness)
Usage-based products benefit from what investors call "passive retention" — the customer doesn't have to make an active decision to stay each month. As long as the workflow that generates usage continues, revenue continues. This structural advantage shows up clearly in the data: median monthly churn of 1.5–2.5%, with top-quartile companies running below 1%.
The risk is different. Usage-based churn often presents as a gradual decline in consumption rather than a cancellation event. A customer whose monthly usage drops from $8,000 to $800 over six months hasn't churned in any dashboard — but the revenue impact is equivalent. Tracking usage velocity alongside churn rate catches this pattern early.
Flat-rate / single-tier — median monthly churn 4–6%
Flat-rate products have the highest churn and the least forewarning. There's no contraction signal, no usage decline to track — the customer is paying a fixed price until the day they cancel. Monthly churn of 4–6% at the median means annual churn of 39–53%, which is not sustainable for most business models without exceptionally low acquisition costs.
Top-quartile flat-rate companies (below 3% monthly) achieve that by adding switching costs outside the pricing model itself: data lock-in, team workflows, integrations. The pricing may be flat, but the product creates stickiness through other mechanisms.
Tiered pricing — median monthly churn 2.5–3.5%
Tiered pricing splits the difference. Customers who hit the ceiling of their current tier can upgrade instead of leaving for a competitor. Customers under cost pressure can downgrade instead of cancelling. This flexibility absorbs churn pressure in both directions, landing median monthly churn at 2.5–3.5%.
The downgrade path is the key mechanism. Companies that make downgrading easy see higher retention than those that force customers into a binary keep-or-cancel decision. A customer who downgrades from $500/mo to $200/mo is still generating revenue and still reachable for future expansion — a cancelled customer is neither.
The lifecycle timing dimension — when churn happens
Pricing model explains the "how much" of churn. Lifecycle timing explains the "when" — and the two dimensions interact in ways that change the intervention strategy entirely.
Month 1–3 churn vs month 12+ churn (completely different causes)
40–60% of total churn happens in the first 90 days. This isn't a retention problem — it's an activation problem. Customers who churn in months 1–3 never reached the value moment. They signed up, encountered friction, and left before the product proved itself. No amount of dunning or re-engagement fixes this; the intervention point is onboarding, not retention.
Churn after month 12 is a fundamentally different signal. These customers found value, integrated the product, and stayed for a year — then left anyway. The causes are competitive displacement, budget cuts, or a change in the underlying workflow the product supports. The response is different: win-back campaigns, competitive analysis, and product investment in the features that long-tenure customers depend on.
For portfolio operators, the early/late split reveals whether a company's churn problem is fixable with operational improvements (onboarding, activation) or requires strategic intervention (product, positioning). A company with 5% monthly churn concentrated in months 1–3 has a better prognosis than one with 3% monthly churn spread evenly — the former can be fixed with better onboarding, the latter has a deeper product-market fit issue.
Net MRR Retention
Revenue retained from existing customers including expansion, contraction, and churn.
Involuntary churn rates by payment method
Involuntary churn — customers lost to failed payments rather than deliberate cancellation — accounts for 20–40% of total churn at the median SaaS company. But the rate varies sharply by payment method, and companies that don't segment by method are applying the wrong recovery strategy to half their failures.
Credit card vs ACH vs invoice
Credit card payments fail at 4–8% per charge attempt. The primary causes — expired cards, insufficient funds, bank declines — are well-understood and partially automatable. Card updater services (Visa Account Updater, Mastercard ABU) resolve 20–30% of expired-card failures automatically. Smart retry logic recovers another 15–25%. Combined with a dunning sequence, total recovery rates of 50–70% are achievable.
ACH and bank transfer payments fail at 1–3% — lower frequency but harder to recover. There's no card updater equivalent for bank accounts. When an ACH payment fails, recovery depends entirely on customer action: providing updated banking details. Recovery rates sit at 20–40%, roughly half the credit card rate.
Invoice-based billing eliminates mechanical payment failure entirely but introduces a different retention risk: non-payment. Net-30 invoices that stretch to net-60 or net-90 create cash flow drag that doesn't appear in churn metrics but erodes working capital. For enterprise customers on invoice billing, the "churn" equivalent is the invoice that never gets paid — and the 6–12 month lag before it's written off as bad debt.
The portfolio implication: companies with high credit card billing mix have higher failure frequency but better recovery potential. Companies relying on ACH or invoice billing have fewer failures but less mechanical recovery upside. The optimal strategy differs — and a blended involuntary churn rate obscures which lever to pull.
Implications for portfolio operators
Segmented churn data changes three decisions that portfolio operators make regularly: which companies to flag for intervention, what kind of intervention to deploy, and how to set churn targets in operating plans.
First, flagging. A company reporting 4% monthly churn on a flat-rate model is performing at the median for its pricing category. The same rate on a usage-based model would put it in the bottom quartile. Without the pricing-model context, portfolio reviews generate false positives (flagging normal flat-rate churn) and false negatives (missing alarming usage-based churn).
Second, intervention type. Early-lifecycle churn concentrated in months 1–3 calls for onboarding and activation work — operational improvements that a portfolio ops team can drive. Late churn spread across months 12–24+ points to product or competitive issues that require strategic investment. Involuntary churn from failed payments is the most mechanically recoverable, but only if the company has the right payment recovery infrastructure in place.
Third, target-setting. Operating plans that set a single churn target without adjusting for pricing model are setting the wrong bar. A 2% monthly churn target is ambitious for a per-seat product (top quartile), reasonable for a tiered one (slightly above median), and unambitious for a usage-based product (bottom half). Segmented benchmarks let you set targets that are appropriately challenging for each company's model.
The companies that get this right — segmenting churn by pricing model, lifecycle stage, and payment method — make better allocation decisions, set more accurate targets, and catch problems earlier. The ones that rely on blended benchmarks are managing to a number that describes none of their companies accurately.