Pressure Mounts on Per-Seat SaaS Pricing in Enterprise AI
Commentary from the B2B software ecosystem indicates that traditional per-seat SaaS revenue models are under increasing pressure. As AI creates more input-constrained roles, the value proposition is shifting. For enterprise AI startups, this trend highlights the growing importance of usage-based or hybrid pricing models that better align with customer ROI.
- Recent industry data highlights a rapid decline in per-seat pricing, with adoption dropping from 21% to 15% among B2B companies in one year, while hybrid models surged from 27% to 41%. - Enterprise search competitors are still navigating this shift; Glean maintains a per-user model starting around $50/user/month plus a $15 add-on for AI features, while Hebbia targets high-value finance and legal use cases with "power seats" costing $10,000/year. - Unlike per-seat models, foundation model providers like Cohere and OpenAI primarily charge for their APIs based on consumption, typically measured in input and output tokens, which directly aligns with the computational cost of inference. - The gross margins for AI companies average 50-60%, significantly lower than the 80-90% common for traditional SaaS, which pressures vendors to adopt pricing that tracks variable compute costs. - Per-seat pricing can penalize vendors for product improvement; as AI becomes more autonomous and reduces the need for human operators, it directly erodes the vendor's licensable user base. - Foundation model providers are experimenting with hybrid enterprise offerings; Anthropic recently lowered its per-seat fee for its Claude AI but removed API discounts and now requires mandatory consumption commitments. - Some vendors are exploring outcome-based pricing, where fees are tied to specific business results like cost savings or revenue generated, though this model is still nascent and represents a small fraction of the market. - The move away from per-seat pricing is also a response to AI's impact on user roles, where a single "power user" can trigger automated workflows that consume exponentially more resources than dozens of casual users, making a flat per-user fee unprofitable.
