While AI expands software's capabilities, vendors may not capture the value. Companies could use AI to build solutions in-house more cheaply. Furthermore, traditional "per-seat" pricing models are undermined when AI reduces the number of employees required, potentially shrinking revenue even as the software delivers more value.

Traditional SaaS companies are trapped by their per-seat pricing model. Their own AI agents, if successful, would reduce the number of human seats needed, cannibalizing their core revenue. AI-native startups exploit this by using value-based pricing (e.g., tasks completed), aligning their success with customer automation goals.

The value in software is shifting from SaaS platforms (like CRMs) to the AI agent layer that automates work on top of them. This will turn established SaaS companies into simple data repositories, or "hooks," diminishing their stickiness and pricing power as agents can easily migrate data.

In categories like customer support, where AI can handle the vast majority of queries, charging per human agent ('per seat') no longer makes sense. The business model is shifting to be outcome-based, where customers pay for the value delivered, such as per ticket resolved or per successful interaction.

The primary moat for many SaaS companies was the complexity and high cost of migrating away from their product. AI agents can now automate this process, eroding that advantage, increasing competition, and giving buyers significant leverage to renegotiate contracts.

The fundamental business model of many SaaS companies is based on per-user pricing. AI agents pose an existential threat to this model by enabling smaller teams to achieve the same output as larger ones. As companies wonder why they should pay for 100 seats when 10 people can do the work, the entire economic foundation of the SaaS industry faces a crisis.

AI agents can easily siphon off value from SaaS products priced on per-seat utility by automating tasks previously done by humans (e.g., support tickets). In contrast, deeply embedded systems of record (ERP, CRM) are insulated by career-limiting switching costs and the immense challenge of migrating timeless, critical data.

If AI agents are delegated to choose the optimal software for a task, they will constantly evaluate and switch between vendors based on performance and cost. This dynamic breaks the long-term customer relationships and enterprise lock-in that SaaS companies rely on, effectively commoditizing the software market and destroying brand loyalty.

As AI agents perform more work and human headcount decreases, the traditional seat-based pricing model becomes obsolete. The value is no longer tied to human users. SaaS companies must transition to consumption-based models that charge for the automated work performed and value generated by AI.