Intercom Fin Pricing: How Per-Resolution Billing Works

Per outcome pricing is an honest idea. You pay when the thing you wanted actually happened, which is a fairer deal than paying for messages sent or seats idle. Plenty of buyers prefer it, and for a support team with a real deflection problem it can work out well.

It is also harder to forecast than a flat plan, because your bill now moves with your ticket volume. This guide covers what triggers a charge, what does not, how the two meters stack, and which parts to model carefully before you commit.

Every figure below is what Intercom's own pages stated on 17 August 2026. Pricing moves, so treat this as a method for modelling rather than a quote, and confirm the current numbers on their pricing page before you sign anything.

What does Intercom charge per outcome?

Intercom's pricing page, read on 17 August 2026, listed $0.99 per outcome. The same page stated you are charged at most once per conversation, even if Fin answers several questions inside it.

That per conversation cap is worth understanding properly. A long support thread where Fin handles five separate questions costs the same as a one line answer. It is a meaningful protection, and it means your bill tracks conversation count rather than message volume.

What counts as a billable outcome?

Intercom's pricing page described an outcome as being counted when the customer confirms their issue is resolved, when they do not ask for more help after Fin responds, or when Fin completes a procedure including handoffs.

Their separate outcomes documentation gives the fuller picture, including the outcomes that cost nothing. This is the table to keep in front of you when you design workflows, because the way you build a path changes which column it lands in.

What do the seats cost on top?

Fin sits on top of an Intercom subscription, so the seat meter runs regardless. Intercom's pricing page, read on 17 August 2026, listed three plans billed annually.

The page also noted a minimum of one Full seat on Intercom plans, and, for teams bringing Fin alongside an existing helpdesk, a minimum monthly commitment with 50 outcomes given as the example figure.

What does a real month actually cost?

Multiply the outcomes by the rate, add the seats, and the shape becomes obvious quickly. The table below is straightforward arithmetic on the figures above, using the Advanced plan at $85 per seat per month and $0.99 per outcome, both as listed on 17 August 2026.

The pattern is the point. In all three rows the outcome meter is the larger number, and it grows with your success rather than your headcount. That is the deal per outcome pricing offers: cost tracks value delivered, and it also tracks ticket volume you may not control.

That last part deserves a fair hearing rather than a complaint. If a resolution genuinely replaces work a person would have done, a dollar is cheap for it, and the maths works in your favour almost immediately. The risk is volume you did not plan for, not the rate itself.

Which parts should you model carefully?

Three variables move the number most, and only one of them is the price. Get these right in your forecast and the rest is arithmetic.

• Assumed resolutions. Intercom's own wording counts a customer who does not ask for more help as an outcome. Most people close the tab whether or not they were helped, so this category is usually a large share of your billed volume.
• Conversation volume, not ticket volume. Because the charge caps at once per conversation, what matters is how many distinct conversations start, not how chatty they are.
• Where you route Fin. Escalations are listed as non-billable, so the audience rules and workflows you build directly shape the bill. Sending Fin at a segment it cannot help costs you escalations, which is cheap, and sending it at one it can help costs you outcomes, which is the point.

One more thing to ask Intercom directly. Independent analyses of Fin pricing describe reopened conversations being credited back, and describe a silence window before an assumed resolution is counted. Intercom's own pages I read did not state either mechanic in those terms, so get the specifics from them in writing rather than from a blog, including this one.

Is there a trial before any of this starts?

Intercom's Fin FAQ page, read on 17 August 2026, referred to a 14 day free trial before charges begin. Use it properly rather than as a look around, because a trial is the only cheap way to learn your own outcome volume.

The number you want out of the trial is not whether the answers were good. It is how many conversations reached an outcome per week, because that figure multiplied by the rate is your recurring cost, and nothing else in the model matters as much.

How does a flat plan compare?

This is where our own product sits differently, and it is a genuine trade rather than a better deal for everyone. At Sistava you hire an AI Employee on a flat monthly plan, from $50/month, and tell it in plain English what to handle. The bill does not move when a good month brings more customers through the door.

The honest comparison depends on your volume and your variance. A flat plan is predictable and rewards heavy use, which suits a small team with spiky traffic. A per outcome rate rewards low volume and aligns cost with delivered value, which suits a team that wants spend to scale down as well as up.

Work out your own break-even before you decide anything. Take your monthly conversation count, multiply by a realistic outcome rate, multiply by the per outcome price, add the seats, and compare that to a flat plan. If the numbers are close, pick on predictability rather than on price.

Sources

• Intercom, Pricing, read 17 August 2026
• Intercom Help, Fin AI Agent outcomes
• Intercom Help, Fin AI Agent FAQs
• Macha, Intercom Fin pricing, the per-resolution cost breakdown, dated 3 July 2026