Shift from seat to usage pricing

This pricing shift is an opening for AI-native support startups because they can charge for work completed, while Intercom has to protect a legacy model that charges for the humans AI is supposed to replace. Intercom still sells seats for the help desk, then layers on Fin at 99 cents per resolved ticket. That creates product and packaging complexity that a company like Decagon avoids by starting with per conversation or per resolution pricing from day one.1 2 3 4

• Intercom has openly described the tension. As AI handles more tickets, support teams can shrink, which pressures seat expansion. The company expects to cannibalize some seat revenue and is redesigning pricing for a post AI support market with simpler, more predictable billing.2 3
• Decagon does not carry that baggage. Its pricing is built around either each inbound conversation or only the cases fully resolved by the AI. That is easier for a buyer to map to ROI because the bill rises when automation volume rises, not when more agents are added.4
• The broader market is moving the same way. AI support agents are increasingly sold on an outcome basis, roughly $0.99 to $1.50 per resolution, because customers compare them to the $10 to $15 cost of a human handled ticket. That makes usage pricing feel native for challengers and transitional for incumbents.3 4

The likely end state is that support software stops being priced mainly by how many reps log in, and starts being priced by how many customer problems get solved. Intercom can still win if Fin pulls enough volume onto its platform, but AI-first challengers have a cleaner path because their pricing already matches the behavior customers want more of, which is automated resolution.1 2 3 4