AI challengers avoid seat based transition

AI first challengers can price directly around work done, which lets them attack incumbent revenue models instead of protecting them. A company like Decagon can start with per conversation or per resolution billing from day one, while Intercom has to layer that onto an installed base built around seats, plans, and add ons. That creates real product and sales complexity, even when the end market is clearly moving toward paying for resolved outcomes.

• Intercom has openly described the tension. As support teams shrink, seat revenue comes under pressure, so Fin has been added as a per resolution charge on top of tiered seat pricing. That is a migration problem, not just a pricing page problem, because it changes how revenue is forecast, packaged, and defended in renewals.
  1 sacra 2 sacra 3 sacra
• Decagon does not carry that baggage. Its model is natively usage based, charging either per conversation or only when the AI fully resolves the issue. That is easier for buyers to map to ROI, because the bill tracks inbound support volume and deflection rather than how many human agents still need seats.
  4 sacra 5 sacra
• The offset for Intercom is proof and bundle. Fin has scale data across thousands of customers, and Intercom can sell AI together with inbox, knowledge base, workflows, reporting, and human handoff. That makes it harder to rip out, even if a standalone AI rival looks cleaner on pricing alone.
  1 sacra 2 sacra 5 sacra

The market is heading toward outcome based support pricing, with seats becoming the wrapper around human oversight rather than the core bill. The winners will be the vendors that can make AI spend feel as concrete and predictable as payroll savings, while still owning the broader support workflow after the first wave of ticket deflection.