What is usage-based pricing — and when should you switch from per-seat?

TL;DR

• UBP charges for consumption, not headcount; ~46% of cloud-native SaaS now has some usage component in 2027, up from 26% in 2020 (OpenView 2024 PLG Index).
• Pure-usage suits ~15% of B2B SaaS (infra, API, data); pure-seat suits ~35% (workflow); the other 50% should run hybrid — seat base plus usage on one metered dimension.
• Switch when value scales with consumption, when seat-count is a proxy mismatch, or when license friction is blocking champion-led expansion.
• Do not switch for workflow products like Slack or Salesforce CRM; do not switch without spend caps and forecast tools — surprise bills churn customers.
• Three killers of UBP rollouts: no spend caps, sales comp not rewired, and broken metering you cannot reconcile against invoices.

The 3 Pricing Models and When Each Wins

The 2027 landscape has settled into three archetypes. The mistake most RevOps teams make is forcing one model onto a product that wants another.

The honest read: pure usage is right for ~15% of B2B SaaS (infrastructure, API, data). Pure seat is right for ~35% (workflow tools where a person does the work). The remaining 50% should run hybrid — the Datadog template: a per-host base finance can forecast, plus a per-event meter that captures upside when telemetry volume explodes.

What Changes When You Switch

Switching pricing is an operating-model project, not a billing project. Four functions move in lockstep.

Sales comp gets rewired first. If AEs sell a pre-paid commit but get paid only on first-year ACV, they sandbag commits and the flywheel never spins. 2027 best practice: pay AEs on the commit at booking, with a true-up accelerator on overage realized in the same fiscal year.

Finance forecasting changes shape. ARR becomes "committed ARR plus run-rate consumption," and guidance has to acknowledge consumption can swing 15-25% quarter to quarter on macro alone. Snowflake guides product revenue, not ARR, because consumption-led businesses cannot pretend monthly run-rate is locked.

Customer success flips from renewal-led to consumption-led. CSMs own the burn-down curve: if a customer is 60% through the quarter and only 30% through their commit, that is a save motion. The playbook becomes weekly consumption reviews and expansion triggered at 70-80% burn — not annual QBRs.

Billing infrastructure is the unglamorous foundation: reliable event ingestion, idempotent metering, and a usage ledger finance can reconcile against invoice lines. This is why Metronome, Lago, Stripe Billing with Usage Records, Maxio (Chargify+SaaSOptics merger), and Subskribe's native usage CPQ exist — building it yourself is a multi-engineer-year distraction.

The 3 Implementation Failures

The first killer is launching usage without spend caps and alerts. A customer signs a pre-paid commit, a misconfigured pipeline emits ten times the events, and four weeks later they get a surprise $400K overage invoice. They do not pay, they churn, and NRR takes a hit. Every UBP rollout needs hard caps, soft caps with alerts at 50/80/100% of commit, and a self-serve dashboard.

The second killer is sales comp that has not caught up. AEs who sold seats for a decade will keep selling seats — quoting "estimated usage" as a seat count, the customer under-commits, expansion never materializes. Fix this before launch, not after the first miss.

The third killer is metering you cannot trust. If engineering cannot reconcile billed events against emitted events, you cannot defend an invoice when a customer disputes it. Treat metering as a tier-zero system: SLAs, dual-write reconciliation, quarterly audit against product telemetry.

Related on PULSE

• [When should I move from per-seat to usage-based pricing?](/knowledge/q76)
• [How is AI agent pricing shifting from per-seat to consumption and outcome-based models in 2027?](/knowledge/q13077)
• [Why is SaaS pricing shifting from per-seat to usage- and outcome-based in 2026?](/knowledge/q12132)
• [Should Salesforce kill the per-seat pricing model?](/knowledge/q1527)
• [Fractional CRO vs Full-Time CRO: When Do I Make the Switch?](/knowledge/q15647)
• [CPI Security equipment lock-in in 2027 — can you switch providers?](/knowledge/q11044)

Common Hybrid Models That Bridge Per-Seat and Usage-Based Pricing

Many teams assume switching to usage-based pricing means going all-in on pure consumption billing, but the most successful transitions use hybrid models that blend both approaches. The most common hybrid structures include:

Tiered usage bundles: Customers buy a base package of seats (say, 10 users) that includes a fixed amount of usage (e.g., 1,000 API calls per user per month). Additional usage beyond the bundle is metered and billed separately. This works well for products like collaboration tools where some teams need heavy API access while others just need basic functionality.

Per-seat with usage add-ons: The core subscription remains per-seat pricing, but premium features or high-volume actions are metered. For example, a project management tool might charge $15/user/month for the base product, then $0.01 per automation run or $5 per 1,000 external integrations. This lets you keep the predictable revenue of per-seat while capturing value from power users without raising base prices.

Usage floors with overages: Customers commit to a minimum monthly usage amount (e.g., $500 worth of compute credits) but pay per unit for anything above that floor. This protects your revenue predictability while giving customers flexibility to scale up. It’s common in infrastructure tools where usage fluctuates seasonally.

Seat-based pricing with usage caps: You charge per user but include a generous usage allowance (e.g., 10 GB storage per user). If a customer exceeds that cap, they either upgrade to a higher tier or pay overage fees. This prevents the “one user consuming 100x more resources” problem that makes pure per-seat pricing unsustainable.

The right hybrid model depends on your customer profile. Enterprise buyers often prefer tiered bundles because they can forecast costs, while startups may favor usage floors with overages to avoid paying for unused capacity. A good rule of thumb: if your top 20% of customers use 80% of your infrastructure, a hybrid model with overage pricing is usually safer than pure UBP.

How Usage-Based Pricing Changes Your Go-to-Market Motion

Switching to usage-based pricing isn’t just a billing change — it fundamentally alters how you sell, market, and support customers. Here are the key shifts you need to prepare for:

Sales compensation becomes more complex: Under per-seat, reps know their commission when they close a 100-seat deal at $50/seat. With UBP, initial deals are smaller (often $500–$2,000/month) because customers start with low usage. Reps need compensation plans that reward landing new accounts AND growing usage over time — typically a mix of upfront commission on committed minimums and residual commissions on usage growth. Without this, reps will resist UBP because it reduces their immediate payout.

Customer success becomes revenue-critical: In per-seat models, churn happens when users leave. In UBP, churn can happen silently — a customer who stops using your product doesn’t cancel, they just stop paying. You need real-time usage monitoring and proactive outreach when usage drops 20% or more. Many UBP companies assign CSMs based on usage tiers (e.g., accounts spending >$5K/month get dedicated CSMs) rather than company size.

Pricing transparency becomes a competitive weapon: Per-seat pricing hides the true cost of heavy usage. UBP exposes it — which can be scary for customers but also builds trust. The best UBP companies publish pricing calculators and provide usage dashboards so customers can forecast their bills. This reduces support tickets about unexpected charges and increases customer confidence in scaling their usage.

Marketing shifts from “users” to “use cases”: Per-seat marketing focuses on headcount growth (“Add more users!”). UBP marketing focuses on consumption patterns (“Run 10x more API calls without worrying about user limits”). Your content, case studies, and sales demos should highlight how customers save money by paying only for what they use, not for seats they never fill.

Self-serve onboarding becomes essential: With per-seat, you can afford white-glove onboarding because you’re charging per user. With UBP, customers often start small and scale up — so they need to be able to sign up, integrate, and start using your product without a sales call. Companies like Twilio and Stripe built their entire growth engine on self-serve onboarding, then add sales support for accounts hitting $10K+/month in usage.

When to Stick with Per-Seat Pricing (And When to Avoid UBP)

Usage-based pricing isn’t always the right answer. Here are scenarios where per-seat pricing still wins — and warning signs that UBP might hurt your business:

Stick with per-seat when your product’s value is tied to user count. Collaboration tools (Slack, Notion, Figma) where each additional user creates network effects are classic per-seat products. Trying to bill Figma by “design operations” would confuse customers and reduce adoption. Similarly, HR software, CRM systems, and project management tools where value scales with team size should stay per-seat.

Avoid UBP when your infrastructure costs don’t scale with usage. If your product is mostly fixed-cost (e.g., a SaaS app where adding 1,000 users costs the same as adding 10), pure UBP will compress your margins on heavy users. You’ll either lose money or need to raise prices, which defeats the purpose. In this case, tiered per-seat pricing (e.g., Basic/Pro/Enterprise) is simpler and more profitable.

Don’t switch if your customers can’t predict their usage. Enterprise buyers need to budget annually. If your product’s usage is highly variable and unpredictable (e.g., a security tool that only gets used during breaches), UBP creates budget uncertainty that procurement teams will reject. Hybrid models with committed minimums can work here, but pure UBP is a tough sell.

Avoid UBP if your sales cycle is long and high-touch. Per-seat pricing is simple to quote and negotiate. UBP requires detailed discussions about expected usage, overage rates, and pricing guarantees. If your average deal takes 6+ months and involves legal review, adding usage complexity will slow things down further. Stick with per-seat until you have a self-serve motion or shorter sales cycles.

Watch out for “usage cliffs” where customers suddenly stop using your product. Seasonal businesses, project-based work, or products used only during specific events can create massive revenue swings. If your churn rate would spike 40%+ during low-usage months, UBP will destabilize your cash flow. In these cases, annual contracts with usage floors or per-seat pricing with usage caps are safer.

The best approach is often to start with per-seat pricing, then introduce usage-based elements gradually — first as add-ons, then as optional tiers, and finally as the primary model once you’ve validated that customers understand and prefer it.

FAQ

Is usage-based pricing always better than per-seat? No. Per-seat works well when value correlates with the number of users, like collaboration tools or HR platforms. Usage-based pricing shines when customer value scales with consumption—API calls, data volume, or compute—and seat counts don't reflect actual product usage.

When is the right time to switch from per-seat to usage-based? Typically when you notice that seat counts are a poor proxy for value, when customers want to start small and expand based on usage, or when seat-license friction slows down adoption inside accounts. Many companies start with a hybrid model rather than a full switch.

Will usage-based pricing hurt my revenue predictability? It can, if you rely solely on pure usage billing. That’s why most B2B companies adopt a hybrid model—a base subscription plus usage overage—to maintain predictable recurring revenue while capturing expansion from heavy users.

How do I set usage-based prices without undercharging? Start with customer interviews to understand willingness to pay per unit, then benchmark against competitors in your category. A common approach is to price the first tier low to drive adoption, then increase per-unit rates for higher usage bands to capture more value from power users.

Do customers prefer usage-based pricing over flat fees? Many customers appreciate the flexibility to pay only for what they use, especially in early stages. However, some prefer predictable flat fees to avoid budget surprises. The best approach is often offering both options—a flat per-seat plan and a usage-based plan—so customers can choose.

What metrics should I track to make usage-based pricing work? Track unit economics like cost per API call or GB stored, customer usage patterns (median, P90, P99), and expansion revenue from overages. Also monitor churn tied to usage spikes—if customers leave after a high bill, your pricing thresholds may need adjustment.

Sources

• OpenView Partners, 2024 PLG Index — UBP adoption rose from 26% (2020) to ~46% (2024) of cloud-native SaaS.
• Bessemer Venture Partners, State of the Cloud 2024 — hybrid dominant; pure-usage ~12%, pure-seat ~35%.
• Tomasz Tunguz, "The Hybrid Pricing Era" (2024) — seat+usage convergence across public SaaS.
• Metronome, 2024 UBP Benchmark Report — commit-and-true-up patterns, AE comp structures.
• Snowflake FY2024 Q4 earnings call — consumption guidance methodology; product revenue vs ARR.
• Datadog 2024 Investor Day — host-plus-event hybrid model and net retention dynamics.
• Pavilion 2024 Pricing and Packaging Survey — 800+ RevOps leaders on post-UBP comp rewiring.
• a16z, "The New Business of Consumption" (2024) — finance/procurement implications.