Seats, credits, or outcomes? The question to ask before you change your pricing model

Open LinkedIn on any morning and someone is announcing the end of per-seat pricingSeat-Based PricingSeat-based pricing charges a per-user fee for each person who accesses the product. It's one of the most common B2B SaaS pricing models, offering simplicity and predictability for both vendor and buyer.Read more →. The argument is a fair one. AI agents do work without occupying a seat, so headcount stops tracking the value a customer gets. As one Andreessen Horowitz partner wrote: "Per-seat is no longer the atomic unit of software."¹

In a 2026 survey of 300 B2B SaaS CEOs, 97% said they were likely to retire seat-based pricing within two years. In the same survey, 94% said seat-based pricing still reflects their product's value today.²

So most of these CEOs expect to drop a model they say still works today. They are planning for where they think the market is going, before their buyers have shown them what should replace it.

The fair case against seats

Per seat is the most common value metricValue MetricA value metric is the unit of measurement that determines how a customer is charged. It's the axis along which price scales — such as users, contacts, API calls, revenue managed, or storage.Read more →, and often the laziest choice. It scales with headcount, and a bigger team doesn't always get more value. As Netlify's CEO put it: "When the price is per person, every new user is a cost decision."³ That slows adoption of the product you most want them to depend on.

Seat pricing also works against a product that makes teams more efficient: the better it works, the fewer seats the customer needs.

So moving away from seats often makes sense. The harder question is what to move to.

What teams are moving to

Kyle Poyar's 2026 State of B2B Monetization survey covered more than 230 software companies.⁴ Three in four had changed their pricing or packaging in the previous year. Hybrid pricingHybrid PricingHybrid pricing combines two or more pricing models — typically a base subscription fee plus usage-based or seat-based components — to balance revenue predictability with value alignment.Read more → was the most common model, and AI credits were already in use at 29% of companies, with another 33% planning to introduce them.

The report also describes how that is going. Companies with hybrid pricing "are generally the happiest, yet struggle to explain their pricing to customers." And credits, in Poyar's words, "can become a nightmare for customers, especially once teams have to manage different credit models across dozens of their vendors."

Credits only work when four things are true. They are tied to something the customer values, that thing can be measured, the buyer knows how many credits equal that value before they commit, and that rate doesn't change against them during the contract. The last two are the hardest to meet. Each SaaS company defines a credit differently, so a buyer can't compare one against another or say in advance how many they will use. Seat pricing is at least easy to budget for: seats times price gives you next year's bill.

A value metric has two jobs

The first job is to scale with the progress your customer is making. The second is to give the buyer a number they can forecast and defend inside their own company.

Seats do the second job and often fail the first. Credits and raw usage can do the first and fail the second. Outcome-based pricingOutcome-Based PricingOutcome-based pricing is a model where the price a customer pays is tied directly to measurable business results the product delivers, such as revenue generated, costs saved, leads qualified, or other quantifiable outcomes. It represents the purest form of value alignment between vendor and buyer.Read more → does the first job best, but the SaaS company and the customer have to agree on when an outcome has happened.

This is why the HOPE framework scores a value metric on two dimensions. Customer Clarity asks whether the metric is connected to value, feels fair and familiar, and is predictable. Operational Ease asks whether you can track it, bill for it and scale on it.

Most of the current debate is being argued from the operational side: which unit follows our costs, and which protects our margin as usage grows. Those are real questions, but buyers decide on the clarity side, and they still buy the way they always have. As one founder wrote of AI buyers, "They still operate on annual budgets, need accurate forecasting, and plan around stable, recurring costs."⁵

Three questions before you change the metric

Can a buyer estimate next year's bill in under a minute, without your help? If the answer needs a calculator you built, a call with your team, or a month of usage data, the metric is failing on predictability.

Does the bill only increase when the customer gets more of what they hired you for? Start with the job. If spend can increase while progress stands still, the metric will feel like a tax, and renewal is when you will hear about it.

Could your champion explain the unit to their CFO in one sentence? "We pay per resolved conversation" passes. "We pay for credits, which convert at different rates depending on the action" doesn't.

A metric that fails one of these can still be the right one. It needs predictability designed around it: a platform fee that sets the floor, an included allowance sized for normal use, caps and alerts, and a place where the customer can see their usage before the invoice arrives. With credits, it also means pooling them across the customer's organization, forecasting their usage for them, and right-sizing the commitment when the forecast and the purchase don't match. If a customer pays for credits they never use, the SaaS company either oversold or didn't get them to value.

Test it with buyers before you launch

Run a purchase simulationPurchase SimulationA purchase simulation is a research method where potential buyers walk through a realistic buying scenario with proposed pricing, tiers, and packaging — allowing you to observe their decision-making process and validate your pricing before launch.Read more → with existing customers: show them the structure, watch where they hesitate, then ask them to estimate what they would pay next year.

If they can't estimate it, the metric needs more work before it ships.

Key takeaways

• The debate over seats, credits and outcomes asks which metric to charge on. Buyers ask a different question: what will this cost me, and can I explain it?

• So a value metric has two jobs: scale with customer progress, and stay predictable enough to budget for.

• No model does both by default. Seats are predictable but often disconnected from value. Credits and usage can follow value but are hard to forecast.

• That is why you score candidate metrics on customer clarity before operational ease.

• Then test the metric with real buyers before launch. If they can't estimate their own bill, it isn't ready.

To close

Pricing models change every few years. Perpetual licences gave way to subscriptions, subscriptions made room for usage, and usage now sits alongside credits and outcomes.

Buyers have asked the same two questions throughout: what am I paying for, and what will it cost me next year?

If your customers can't answer those today, switching metrics won't help unless the new one is easier to predict. Working that out with buyers before anything ships is the core of the PACE System™.

Sources

1. James da Costa, Andreessen Horowitz. LinkedIn post, January 2025.

2. Cruxy survey of 300 B2B SaaS CEOs in the UK and US, reported by IT Brief, 16 April 2026.

3. Mathias Biilmann Christensen, CEO, Netlify. LinkedIn post, May 2026.

4. Kyle Poyar, "The state of B2B monetization in 2026." Growth Unhinged, 13 May 2026.

5. Alvaro Morales, "Usage-Based Pricing vs. Predictability: The False Tradeoff in AI Monetization." LinkedIn, 15 January 2026.