Intercom Adds Usage-Based AI Pricing

Intercom is turning AI from a feature into a second pricing engine. Instead of letting automation shrink seat revenue, it keeps the help desk sold per agent while charging separately when Fin actually resolves a ticket. That lets Intercom monetize both the human workflow and the automated outcome, which helps explain the jump from $382M ARR at the end of 2025 to $419M by April 2026 and the sharp lift in retention.1 2 3 4

• Fin changes the unit of value from agent access to problem solved. Intercom priced Fin at $0.99 per AI resolved ticket, after initially launching at $1.90, which gives customers a direct comparison against human support that can cost roughly $10 to $15 per resolution.2 3 4
• The bundle matters. Intercom sells Fin alongside inboxes, ticketing, knowledge base, workflows, reporting, and agent assist. When Fin cannot answer, a human can step in, train the bot, and update docs in the same system, which is harder to replicate with a standalone AI layer sitting on top of another help desk.2 3 4
• This is also a defensive move against AI native rivals like Sierra and Decagon. Those companies sell autonomous resolution first, while Intercom uses its installed base and existing seat contracts to cross sell AI into thousands of support teams, especially B2B software companies where each conversation is lower volume but higher value.2 3 4 5

The next step is a fuller mix of seat, usage, and outcome pricing across chat, email, voice, and proactive support. If Intercom keeps attaching AI revenue to its installed base faster than AI natives can displace the help desk, growth will come less from adding agents and more from owning every resolved conversation.2 3 4 5