The per-seat pricing model that defined the last generation of enterprise software is breaking down. As AI capabilities become central to how software delivers value, the relationship between headcount and value extraction no longer holds. Companies are experimenting with new pricing models that better align vendor incentives with customer outcomes, creating both opportunities and challenges for SaaS businesses navigating the transition.
Per-seat pricing made sense when software was primarily a tool for human productivity. More users meant more value extracted, and the correlation was strong enough that counting seats served as a reasonable proxy for value delivered. But AI-powered software often delivers value by reducing the number of humans needed for a task, creating a fundamental tension: the more valuable the product becomes, the fewer seats customers need.
Usage-based pricing has emerged as one alternative. Under this model, customers pay for what they consume—API calls, compute cycles, transactions processed, or other measurable units. OpenAI popularized this approach with their token-based pricing, and many AI-native companies have followed suit. The appeal is clear: pricing scales with actual usage, customers only pay for value received, and there's no awkward negotiation about seat counts that don't reflect how the product is actually being used.
But usage-based pricing brings its own challenges. Forecasting revenue becomes more difficult when consumption varies. Sales teams struggle with incentives when deal sizes fluctuate post-signature. And some customers resist unpredictable costs, particularly in enterprise environments where budgets are set annually. The most sophisticated companies are developing hybrid models that combine predictable base fees with usage-based components—providing stability while still aligning incentives.
Outcome-based pricing represents an even more radical shift. Under this model, customers pay based on results achieved—revenue generated, costs saved, deals closed, or other measurable outcomes. The vendor shares directly in the upside they create, and customers only pay when value is demonstrably delivered. This approach is most common in sales and marketing automation, where outcomes are relatively easy to measure and attribute.
Implementing outcome-based pricing requires solving difficult measurement problems. What counts as an outcome? How do you attribute results to the software versus other factors? What happens when outcomes are delayed or hard to measure? Companies pursuing this model invest heavily in measurement infrastructure and often start with simpler outcome definitions before evolving toward more sophisticated approaches.
The pricing evolution has significant implications for company building. Unit economics become more complex to calculate and communicate to investors. Go-to-market motions must adapt—product-led growth works differently when there's no natural seat expansion dynamic. Customer success becomes even more critical, since revenue depends on customers actually using and benefiting from the product rather than simply maintaining their subscription.
For customers, the shift is largely positive. Pricing models that align vendor success with customer outcomes should produce better products, more responsive support, and healthier long-term relationships. But procurement teams must develop new capabilities for evaluating and managing these arrangements. The predictability of traditional SaaS pricing had real benefits, even if the underlying model was imperfect. Finding the right balance between alignment and predictability will define successful vendor-customer relationships in the AI era.