There’s an industry shift in sentiment around seat-based SaaS pricing and usage-based models, which reflects changing IT expectations and monetisation. As usage and integrations into operational workflows continue to rise, AI’s flexibility, value and accountability are all falling under deeper scrutiny.
Why now? The unit of value in software is changing. SaaS is typically designed around organisations being allotted one seat per user, which means the organisation would need to buy more seats for each additional employee they want to provide access.
Go deeper with GlobalData
Access deeper industry intelligence
Experience unmatched clarity with a single platform that combines unique data, AI, and human expertise.
However, the growing integration of AI into business processes has made the seat-based pricing model less profitable for vendors. AI enables an individual user to trigger a significantly higher number of actions, which makes the potential production from one seat immensely greater than before.
Foundational AI providers naturally want to ensure their company maintains a path towards long-term profitability. Large language models are computationally intensive, and token-based or API call pricing reflects the underlying infrastructure costs much more accurately than per-seat licences.
Numerous market drivers such as AI-driven automation, potentially tighter budgets and CFO scrutiny are also pushing buyers to demand pricing aligned to measurable outcomes rather than headcount. Some buyers may welcome paying for software services priced based on usage, but they may also worry about cost volatility and forecasting challenges.
While usage-based models are not the best option for every use case, they will better support the organisations where their whole set of operational processes, such as engineering workflows and AI usage guidelines, are tightly defined and managed.
Benefits go both ways
The benefits of usage-based models are tangible: closer alignment between vendor value and customer impact, lower barrier to entry and easier scaling during growth or contraction. This is particularly useful during economically or technologically dynamic times.
Organisations can gain increased flexibility since usage-based pricing lowers upfront commitment, enabling experimentation before broader rollout. For example, many teams may be piloting AI in an area like customer support before scaling it further across the business.
Those with more mature AI workflows likely have been, or will be, very carefully scrutinising each potential use case to ensure they are efficient, economical and fall within acceptable budgetary constraints.
As AI agents continue to streamline workflows and automate tasks for humans, seat-based pricing becomes even less logical. For vendors, charging per task completed or per API call better reflects outcome delivery, but it’s on the customer to ensure they are managing these tasks and API calls appropriately. Failure to do so may result in expensive bills abruptly curtailing badly designed workflows.
Usage-based pricing must be predictable and transparent to avoid honest or unwitting customer mistakes, as well as immediate distrust resulting in churn from the less mature section of the customer base. SaaS providers must offer clear guardrails and reporting so that users don’t easily over-consume.
To ensure trust and transparency, usage-based pricing will only work if customers can easily understand what is being measured and why costs may rise. It is essential to provide customers with dashboards, alerts and very clear billing, along with guidance on how to keep usage and spending under control. This is particularly important where users are making use of internal AI support and don’t have the team skills or savvy to really understand workflows that have been vibe coded, or are new and complex. Vendors may be inadvertently blamed for poor system design.
A value-based future
In digital operations, there’s also a complex set of customer environments, use cases and appetites in which hybrid models are likely to persist. The future may lie in flexible frameworks that combine platform access with consumption-based elements. Vendors can combine platform access fees with usage tiers, particularly for enterprise contracts that require predictability alongside scalability.
The industry will see a profound shift as AI continues to embed within operational workflows. Revenue models will increasingly follow compute consumption and business outcomes rather than headcount alone. A likely model for many providers will be to charge a base platform fee plus usage for AI-heavy or automated features. Vendors who fail to make both their value and pricing transparent will struggle.
