AI Enables Outcome-Based SaaS Pricing

This marks the real pricing break between old SaaS and AI native software. In the old model, vendors sold access to a worker's software seat. Here, Intercom is selling the completed support job itself. That matters because a buyer can now compare Fin directly to the fully loaded cost of a human agent, not just to another help desk license, which makes value easier to justify and seat cannibalization easier to absorb.

• Fin turns support into a metered output. Intercom charges about $0.99 when the bot resolves a conversation, while its seat plans still range from roughly $29 to $132 per user per month. That lets Intercom keep revenue even if some human seats disappear, as long as automated resolutions scale faster than seat loss.
  2 sacra 3 sacra
• This is different from earlier bots that mainly routed tickets or matched keywords. Fin can read a company's help docs, answer in natural language, and resolve a large share of inbound questions without the heavy rule writing that older systems needed. That makes outcome pricing possible, because the product is doing a complete unit of work, not just assisting a person.
  1 sacra 4 sacra
• The broader shift is from charging for software availability to charging for labor replacement. Zapier added pay as you go task pricing for flexibility, and finance buyers increasingly prefer usage models because cost maps more clearly to work done. In support, that means budgets can move from salaries and underused seats toward per resolution spend.
  1 sacra 5 sacra

Going forward, the winners in B2B SaaS will be the companies that can price against a business result and deliver it reliably at scale. In customer support, that pushes the market toward AI agents with measurable resolution rates, tighter integrations, and lower cost per solved issue, while seat only vendors get pulled toward a smaller and weaker role.