Source: Stripe Blog
AI vendors are discovering that unit economics built on compute costs, token consumption, and unpredictable usage patterns don't fit traditional per-seat or tiered SaaS models—forcing experimentation with usage-based pricing, reserve capacity, and hybrid structures that few payment systems were designed to handle. The shift upends what "predictable recurring revenue" means, forcing CFOs and finance teams at enterprise customers to model AI spend like cloud infrastructure rather than software, with all the budget volatility that entails. Stripe's emphasis on this at Monetize 2026 signals that payment and billing infrastructure itself has become a competitive advantage for whoever can accurately measure, meter, and monetize AI at scale.