Usage-Based vs Seat-Based Pricing: Which Is Better for NRR?
Seat-based and usage-based pricing compared on net revenue retention, forecasting, sales and billing, with how hybrid flat-plus-usage pricing works and how to switch models without hurting revenue.Load more

Usage-based vs seat-based pricing comes down to what grows with customer value. Seat-based pricing charges per user and suits tools whose value rises with headcount. Usage-based pricing charges per unit consumed and suits APIs, data and AI products. Many B2B SaaS companies now run a hybrid: a flat platform or seat fee plus usage.
Key takeaways
- Choose the metric that grows when the customer gets more value, not the one competitors use.
- Usage-based pricing expands faster in healthy accounts and contracts faster in struggling ones.
- Seat-based revenue is easier to forecast but falls when customers cut headcount or automate work.
- Hybrid flat-plus-usage pricing gives a predictable floor with expansion on top, which is why it is now common.
- Move models with new customers first, and grandfather existing customers before changing their bill.
We test your value metric against your own usage and deal data.
What is the difference between usage-based and seat-based pricing?
Seat-based pricing charges for who uses the product; usage-based pricing charges for how much they use it. Both are value metrics, the unit that makes a bill grow. Tiered pricing is a third term people mix in, but it is packaging that can sit on either metric, as our overview of SaaS pricing models explains.
| Subscription (seat-based) | Consumption (usage-based) | Tiered | |
|---|---|---|---|
| What is charged | A recurring fee per user | A rate per unit used | A package at a set price |
| When it is billed | In advance, monthly or annually | In arrears, or from prepaid credits | In advance |
| What grows revenue | More users | More consumption | Upgrades to a higher tier |
| Typical products | Collaboration, CRM, workflow | APIs, data, infrastructure, AI | Most B2B SaaS, on top of either |
| Main risk | Login sharing, headcount cuts | Bill shock, volatile revenue | Unclear fences between tiers |
Which should you choose? A quick verdict by product type
Pick seats when value follows people, usage when value follows volume, and a hybrid when it follows both. The verdicts below are where we would start a pricing review, before looking at your data.
- Collaboration and workflow tools: seat-based, often per active user, because each extra person gets their own value.
- APIs and developer infrastructure: usage-based, because engineers expect to pay per call or per resource and want to start small. Our devtools go-to-market page covers how this shapes the wider engine.
- Data and analytics platforms: usage or credits for compute and storage, with seats for named users if admin matters.
- Products with AI agents or heavy automation: hybrid, with seats for people and a per-use or per-outcome charge for the work the AI does.
- Systems of record sold to enterprises: a platform fee plus seats, with usage allowances that rarely bind.
Is seat-based or usage-based pricing better for NRR?
Usage-based pricing raises the ceiling on net revenue retention, and seat-based pricing raises the floor. Usage revenue grows as soon as a customer consumes more, with no purchase decision. It also falls as soon as they consume less. Seat revenue moves only at renewal or when someone adds users, so it is steadier in both directions.
That is why the answer depends on your customers, not on the model. If usage in your best accounts grows reliably, usage-based pricing will show it in NRR. If usage is seasonal or concentrated in a few customers, a seat or platform floor protects you. The net revenue retention definition in our glossary sets out how to calculate it by cohort.
| What happens in the account | Seat-based | Usage-based | Hybrid (platform fee + usage) |
|---|---|---|---|
| Usage doubles, team size flat | No change until more seats | Revenue roughly doubles | Revenue rises on the usage portion |
| Team cuts 20% of users, usage flat | Revenue falls at renewal | No change | Small fall if seats are part of the fee |
| AI automates work previously done by people | Seats fall over time | Revenue rises with automated volume | Usage portion rises |
| Seasonal quiet quarter | No change | Revenue dips that quarter | Floor holds, usage dips |
● Illustrative.
How do the two models compare on forecasting, sales and billing?
Seat-based pricing is simpler to sell, bill and forecast; usage-based pricing is simpler to start and to expand. The table compares them on the criteria a CFO and a sales leader will both ask about.
| Criterion | Seat-based | Usage-based |
|---|---|---|
| Revenue predictability | High: contracted seats | Lower: follows activity unless committed |
| Expansion | Needs a new seat purchase | Automatic as consumption grows |
| Entry price | Fixed per user | Can start very small |
| Buyer objection | Paying for inactive users | Fear of an unpredictable bill |
| Sales comp | Simple: contract value | Harder: committed vs actual usage |
| Billing systems | Standard subscription billing | Metering, rating and usage alerts |
| Fit with AI features | Weak: AI cost grows per use | Strong: price tracks cost and value |
Surge45 assessment; qualitative comparison, not a benchmark.
How does hybrid flat-plus-usage pricing work?
Hybrid flat-plus-usage pricing charges a fixed fee that includes an allowance, then bills usage above the allowance. The fixed part is the platform or seat fee; the variable part is overage, credits or a per-outcome charge.
- Platform fee plus overage. A fixed monthly fee with included usage, and a published rate for anything above it.
- Seats plus credits. Users pay per seat, and AI or automation features draw down a credit pool included with each tier.
- Seats plus per-outcome. People pay per seat, and the AI agent is charged per completed task or resolution.
- Commit plus on-demand. A lower rate for committed annual volume, with a higher on-demand rate above it.
Design the allowance so a typical customer in each tier uses most of it but rarely exceeds it in the first months. That keeps the first invoices predictable and makes overage a sign of success rather than a surprise. Where the allowance sits inside each plan is a packaging decision covered in our guide to SaaS pricing tiers.
How do you choose a good usage metric?
A good usage metric grows with the customer's success, is easy for them to predict and is cheap for you to measure accurately. If a metric fails any of the three tests, it will create disputes, procurement friction or churn, however well it tracks your costs.
| Test | Question to ask | Good sign | Warning sign |
|---|---|---|---|
| Value | Does it rise when the customer gets more value? | More usage means more work done for them | Usage rises when the product works badly, such as retries |
| Predictability | Can a buyer estimate next quarter's bill? | Tracks a business volume they already forecast | Driven by technical behaviour they cannot see |
| Measurability | Can you meter it per account, reliably? | Already logged and visible in the product | Needs new tracking or manual counting |
| Fairness | Would a customer accept it as reasonable? | Matches how they think about the job | Feels like a tax on adoption |
Messages sent, records processed, hosts monitored and tasks run all pass these tests for their products. Raw compute time often fails the predictability test for business buyers, which is why many vendors translate it into credits or bundles that map to work the buyer recognises.
What do AI features change about seat-based pricing?
AI features break the logic of pure seat pricing, because each use carries a real cost and the AI can do work that used to need more seats. A seat-only price either loses money on heavy users or charges light users for capacity they never touch.
- Cost per use is real. Every AI request has an inference cost, so unlimited use inside a fixed seat price erodes gross margin as adoption grows.
- Seats can shrink as value grows. If an AI agent resolves work that people used to do, a customer may need fewer seats while getting more value from the product.
- Outcome pricing is now practical. When the product can tell whether a task was completed, charging per outcome ties price to the result the buyer wants, as Intercom does for its AI agent.
- Credits keep it simple. A pool of credits included in each tier, with top-ups, lets one price book cover several AI features without a separate meter for each.
What do seat, usage and hybrid pricing look like in practice?
Public pricing pages show all three models, and most large vendors have moved towards hybrid. Each example was checked on the company's live pricing page or documentation as of November 2026.
| Company | Model | How it charges |
|---|---|---|
| Slack | Seat-based | Per active user per month; members inactive for 28 days earn a prorated credit |
| Linear | Seat-based | Basic $10 and Business $16 per user per month, billed yearly |
| Twilio | Usage-based | US SMS $0.0083 per segment plus carrier fees; committed-use discounts by arrangement |
| Snowflake | Usage-based (credits) | Compute in credits billed per second, 60-second minimum; storage at a flat rate per TB |
| Datadog | Usage (per host) with commit | Infrastructure Pro $15 per host per month billed annually, $18 on demand |
| Intercom | Hybrid: seats + outcomes | Per-seat helpdesk plans, plus the Fin AI agent at $0.99 per outcome |
| HubSpot Marketing Hub | Hybrid: seats + contacts + credits | Tiers include core seats, marketing contacts and HubSpot Credits; extra seats priced separately |
Each company's public pricing page or documentation, checked November 2026.
Two details are worth copying. Slack's fair billing credit removes the main objection to per-seat pricing, paying for people who do not use the product. Intercom charges for its AI agent only when it resolves something, which ties the usage price directly to the outcome the buyer wants.
What does the market data say about usage-based pricing?
Usage-based pricing is widespread but no longer accelerating, and hybrids are where most companies landed. The longest-running survey data comes from OpenView's annual SaaS benchmarks.
Source: Kyle Poyar, Growth Unhinged, from OpenView SaaS benchmark surveys (700+ respondents in 2023), February 2024.
Poyar also noted that usage-based businesses saw revenue growth slow faster than subscription businesses during the downturn. That is the NRR trade-off in practice: the model that expands fastest in good times contracts fastest in bad ones. It is the strongest argument for a fixed floor under any usage component.
How do you move from seats to usage without hurting revenue?
Measure usage first, shadow-bill it, then launch to new customers before touching existing ones. This is the sequence we use, because the biggest risk in any pricing change is a big-bang switch.
Instrument the usage metric
Make sure the product records the unit you plan to charge on, per account, and that customers can see it.Shadow-bill for a quarter
Calculate what each customer would pay under the new model without charging it. Find the winners, the losers and the outliers.Design the floor and the allowance
Set a platform or seat fee that protects revenue, and an allowance that covers typical usage in each tier.Launch to new customers first
Watch win rate, average discount and first-year expansion before anything changes for the installed base.Migrate existing customers with notice
Give a grandfathering period, show each customer their projected bill, and add usage alerts and caps.
If you are also reconsidering your free plan or trial in the same change, read freemium vs free trial first, because the free motion changes which usage you should meter.
If expansion is flat while customers use the product more, or AI features are costing more than they earn, the value metric is the place to look. Surge45 tests seat, usage and hybrid options against your own usage and deal data in the Engine Diagnostic, then builds the change and measures NRR, win rate and discount.
Frequently asked questions
Is seat-based or usage-based pricing better for NRR?
Usage-based pricing usually expands faster in healthy accounts, because revenue grows as soon as consumption grows, with no new purchase decision. It also contracts faster when usage falls. Seat-based NRR is steadier but tied to headcount. Most B2B SaaS companies get the best of both with a hybrid.
What is hybrid flat-plus-usage pricing?
Hybrid flat-plus-usage pricing charges a fixed platform or seat fee that includes a usage allowance, then bills usage above it at a published rate. The fixed part gives the vendor predictable revenue and the buyer a predictable floor; the usage part lets revenue grow with value without renegotiating the contract.
What is the difference between consumption, subscription and tiered pricing?
Subscription pricing charges a recurring fee for access, often per seat. Consumption pricing charges for metered use, usually in arrears. Tiered pricing is packaging: plans at rising price points that can sit on either metric. Most real price books combine all three, such as tiers with seats and usage allowances.
Does usage-based pricing make revenue harder to forecast?
Yes, unless you design for it. Pure pay-as-you-go revenue moves with customer activity, so finance forecasts from usage trends rather than contract values. Annual commitments, prepaid credits and a platform fee restore most of the predictability while keeping the expansion that usage brings.
When should a SaaS company move away from per-seat pricing?
Consider it when value no longer tracks the number of people using the product, for example when automation or AI does work that used to need seats, or when customers share logins to avoid paying. Test a usage component with new customers first, and keep seats where they still reflect value.
Sources
- Kyle Poyar, The state of usage-based pricing, Growth Unhinged, February 2024
- Slack pricing, November 2026
- Linear pricing, November 2026
- Twilio US SMS pricing, November 2026
- Snowflake, Understanding overall cost, November 2026
- Datadog pricing, November 2026
- Intercom pricing, November 2026
- HubSpot Marketing Hub pricing, November 2026
About Nathan Mzumara
B2B SaaS growth specialist · SEO, digital marketing and SaaS
Nathan Mzumara is a B2B SaaS growth specialist with a 12+ year career in SEO, digital marketing and SaaS. He writes for Surge45, offering groundbreaking B2B insights for B2B SaaS businesses on GTM engineering, AI search, attribution and pipeline. He is also the founder of Search Marketing Club, where he runs a community of search marketing professionals.
Related Articles
Get the next playbook first
New teardowns, playbooks and research for B2B SaaS revenue leaders, when they publish.



