G2 Report: Only 9% of Buyers Call Seat Pricing Fairest for AI Agents as Pay-As-You-Go Rises
Key takeaways
- G2's 2026 AI Agents report found only 9% of B2B buyers consider per-seat pricing the fairest model for AI agents.
- Support for pay-as-you-go pricing rose from 38% in 2025 to 57% in 2026.
- Outcome-based pricing gained ground too, while fixed pricing fell.
- 46% of agent buyers report higher-than-expected operating costs, so spend controls matter.
Companies in this story: G2
Seat-based pricing has been the default for SaaS for two decades. For AI agents, buyers are rejecting it. According to G2's 2026 AI Agents report, based on a July 2026 survey of 1,019 B2B decision-makers, only 9% of buyers consider per-seat pricing the fairest model for AI agents.
The key numbers
| Pricing preference | 2025 | 2026 |
|---|---|---|
| Pay-as-you-go (consumption or per task) | 38% | 57% |
| Pay per outcome | 11% | 19% |
| Fixed pricing | 35% | 21% |
Figures from G2's press release and SaaStr's coverage of the report.
Why seats don't fit agents
A seat measures people who log in. An AI agent does not log in; it works continuously in the background, sometimes doing the work of several people and sometimes very little. Charging per seat for that work feels arbitrary to buyers, which is why they increasingly prefer paying for usage or results.
The catch: unpredictable bills
Usage-based pricing is not a free win. G2 found that 46% of agent buyers report higher-than-expected operating costs. Buyers want to pay for what they use, but they also want to know what they will pay. That tension is shaping how vendors package agents:
- Spending caps and alerts
- Prepaid credit bundles
- Hybrid plans with a platform fee plus usage
- Outcome pricing tied to a clear, measurable result
G2's own recommendation to vendors is to test subscription, pay-as-you-go and outcome-based options rather than forcing one model on every customer.
What it means for SaaS companies
- Audit your AI packaging. If agent features sit inside per-seat plans, buyers may see poor value.
- Make costs predictable. Clear limits, calculators and alerts reduce the fear of surprise bills.
- Define the outcome. If you price on results, make the definition of a result unambiguous.
- Test before you switch. Pricing changes affect conversion and plan mix. Run controlled tests on your pricing page first.
Rethinking how you package and sell AI features? Our CRO for SaaS team helps test pricing and packaging changes safely.
Survey figures as reported by G2 and SaaStr.
Ready to find your North Star?
Book a growth call and we will map the fastest path to predictable revenue.