If you've ever opened your Intercom invoice and thought "where did this number come from?" — you're not imagining it.

Intercom's AI agent, Fin, is billed on resolutions. The price looks simple on the pricing page: $0.99 per resolution. The problem isn't the number. The problem is the word. "Resolution" turns out to be a surprisingly loose concept, and that looseness is exactly what makes your monthly spend hard to predict.

This post does two things: it explains, in plain English, what Intercom actually counts as a resolution, and it shows why that definition leaves so many teams frustrated — and what a more predictable model looks like.

Quick answer

• Fin costs $0.99 per resolution, with a minimum of 50 resolutions per month on the standalone plan. See Fin AI Agent pricing.
• A "resolution" is counted when a customer confirms Fin's answer helped or simply leaves the conversation without asking for more help. The second case is the one that surprises people. See Fin AI Agent outcomes.
• Because resolutions scale with AI adoption rather than headcount, the more your customers use Fin, the more you pay — which makes the bill genuinely hard to forecast. See ClearFeed's pricing analysis.

If you're already evaluating a change, jump to the alternative. Otherwise, let's unpack the definition first.

What Intercom actually means by a "resolution"

Intercom's own documentation defines a resolution as a type of outcome. In their words, a resolution "is counted when, following Fin's last answer in a conversation, the customer either confirms the answer was satisfactory (confirmed resolution), or exits the conversation without requesting further assistance (assumed resolution)." See Fin AI Agent outcomes.

So there are two flavors:

• Confirmed resolution — the customer says something affirmative, like "Ok thanks" or "That helped." Clear enough.
• Assumed resolution — the customer just... stops replying. They don't ask for more help, so Intercom assumes the issue was resolved and bills you.

A few other rules are worth knowing:

• One outcome per conversation. No matter how many questions Fin answers in a single thread, you're charged once. That part is fair.
• Greetings don't count. If Fin only responds to "hi," that isn't a resolution.
• Human handoffs don't count. If the customer asks for a person, or Fin's procedure fails, you aren't billed for that conversation.
• Resolutions aren't the only billable outcome. Intercom also charges $0.99 for a "procedure handoff" and a "disqualification," and $9.99 for a "qualification." Most teams don't realize a single AI conversation can carry a price tag ten times higher than the headline rate. See Fin AI Agent outcomes.

On paper, this is "outcome-based pricing," and Intercom frames it as paying only for value delivered. See Intercom pricing. The reality is more complicated.

The part that frustrates people: "assumed resolution"

Here's the crux. An assumed resolution bills you when a customer leaves without asking for more help — but a customer leaving is not the same as a customer being helped.

People abandon chats for all kinds of reasons: they got distracted, they gave up, they found the answer somewhere else, or Fin's reply was close enough that they stopped bothering. Intercom's documentation doesn't attach a specific time window to "exits the conversation"; third-party analyses note it's commonly operationalized as roughly 24 hours of no response counting as an assumed resolution. See ClearFeed's analysis, which puts it bluntly: "a quiet customer is not always a satisfied customer."

That single design choice means your bill is partly driven by customer silence — one of the noisiest, least reliable signals there is. You can be charged $0.99 for a conversation where nothing actually got resolved, simply because the person didn't come back to complain.

Why the bill is so hard to predict

Per-seat pricing is annoying but predictable: you know how many agents you have. Resolution pricing breaks that predictability because it scales with how much your customers use the AI, not with anything you directly control.

The real-world numbers tell the story. According to ClearFeed's review of Intercom pricing:

• One Reddit user reported their bill jumping from $4,000 to $9,000 per month.
• A legacy customer saw costs rise from $119 to $854 per month after migrating to the new model.
• Another user projected an increase from $1,200 to $10,000 per month.
• A modeled five-person team came out to ~$2,745/month — far above the $29–$85 per-seat headline rates.

The pattern is consistent: the more successful your AI deployment, the bigger the bill. As ClearFeed summarizes, "the pay-per-resolution model makes expenses harder to predict at higher volumes because more AI adoption directly increases costs."

For a finance team trying to forecast, that's the worst combination — a per-unit cost tied to an external, fuzzy trigger you can't fully see or steer.

The deeper problem: the model can fight your goals

Step back and the structural issue becomes clear. With resolution pricing:

• Success costs more. Every additional question Fin handles well is another charge. You're effectively taxed on the thing you bought the tool to do.
• The unit is ambiguous. Because "resolution" includes silent abandonment, you can't cleanly tie spend to value.
• Forecasting requires constant babysitting. Teams end up reviewing what Fin "resolved" weekly just to understand the invoice, on top of maintaining docs, workflow logic, and escalation rules.

None of this means Fin is a bad support tool — it's a capable one. It means the pricing model introduces friction and uncertainty that a lot of teams, especially lean B2B SaaS teams, don't want to manage.

What predictable pricing should look like

If you're shopping for an alternative because of the pricing model (not just the features), here's what to look for:

• A flat, knowable monthly cost you can put in a budget and forget.
• Pricing tied to a unit you control (like a plan tier or identified visitors), not a fuzzy "outcome."
• No penalty for success — using the AI more shouldn't quietly inflate the invoice.
• Alignment with your actual goal, whether that's deflecting support, or — for many SaaS teams — converting and activating buyers.

That last point matters. Many teams using Intercom for the website are really trying to engage and convert visitors, not just deflect tickets. If that's you, it's worth reading Intercom vs. Aimdoc and Intercom Fin vs Qualified Piper to see how support-first and revenue-first tools differ.

How Aimdoc approaches this differently

Aimdoc is built for the connected B2B SaaS buyer journey — engaging website visitors, qualifying intent, routing to sales, and helping users activate inside the product. And it's priced to be predictable.

Instead of charging per fuzzy "resolution," Aimdoc uses a flat monthly plan. The Convert plan starts at $999/month and includes a defined allotment (like visitor identifications) — so you know your number before the month starts, and using the AI more doesn't quietly raise the bill.

That model fits the two jobs most SaaS teams actually care about:

• On the website, Aimdoc Engage answers product and pricing questions, qualifies intent, and routes high-value visitors to the right next step.
• Inside the product, Aimdoc Activate helps new users complete onboarding and reach value.

The point isn't that you should never pay for outcomes. It's that your AI bill should be something you can predict, tied to value you can actually see — not to whether a customer happened to stay quiet for 24 hours.

If resolution pricing is what pushed you to look, that frustration is a useful signal. It usually means the tool is priced around its model of value, not yours.

Tired of guessing what next month's AI bill will be? See how Aimdoc Engage and Aimdoc Activate work on a flat, predictable plan — explore Aimdoc as an Intercom alternative, check pricing, or book a demo.

FAQ

What counts as a resolution in Intercom?

A resolution is counted when, after Fin's last answer, the customer either confirms the answer helped (a "confirmed resolution") or leaves the conversation without asking for more help (an "assumed resolution"). You're charged once per conversation. See Fin AI Agent outcomes.

How much does Intercom Fin cost per resolution?

Fin is priced at $0.99 per resolution, with a minimum of 50 resolutions per month on the standalone plan. Other outcomes are also billable — including $9.99 for a "qualification." See Fin AI Agent pricing.

Why is my Intercom bill so unpredictable?

Because resolution pricing scales with how much your customers use the AI rather than with your team size or a fixed fee. Reported bill increases include jumps from $4,000 to $9,000 per month and from $1,200 to a projected $10,000 per month. See ClearFeed's analysis.

Does Intercom charge for resolutions that didn't actually resolve anything?

It can. An "assumed resolution" is billed when a customer simply stops replying — even though, as analysts point out, "a quiet customer is not always a satisfied customer." See ClearFeed's analysis.

What's a more predictable alternative to Intercom's resolution pricing?

Look for a flat monthly plan tied to a unit you control. Aimdoc, for example, uses flat-rate plans (the Convert plan starts at $999/month) rather than per-resolution charges. See Aimdoc pricing and the Intercom alternative page.