Intercom pricing explained: seats, resolutions, and what you actually pay

Intercom makes an excellent product and a genuinely confusing bill. The list price looks small, the invoice does not, and the gap between the two is the part most teams only understand after a few months of usage. This post lays out the published pricing model, works one honest example, and shows what the same workload costs on a platform that includes AI resolutions in the seat price. All Intercom numbers here are from Intercom's own published pricing; where details vary by plan or change over time, we say so instead of guessing.

How does Intercom pricing work?

Intercom charges along two axes at once: a per-seat subscription that starts at $29 per seat per month, and usage pricing for its AI agent, Fin, at $0.99 per resolution. Seats scale with the size of your team, resolutions scale with the volume of conversations the AI handles, and the two are added together on the same invoice. There is a 14-day free trial.

This structure is worth sitting with for a second, because it means your bill has no ceiling by default. Hiring a support agent raises the seat line. A good month for the AI, where it resolves more conversations on its own, raises the usage line. Both lines grow when things go well, which is exactly when finance starts asking questions.

How much does Intercom actually cost per month?

For a five-person team whose AI resolves 1,000 conversations a month, the published math is $145 in seats plus $990 in Fin resolutions, so from $1,135 per month before any add-ons. That is the entry-level seat price; higher plans cost more per seat.

The worked example

• Seats: 5 seats at $29 per seat per month, from $145 per month.
• AI resolutions: 1,000 resolutions at $0.99 each, $990 per month.
• Total: from $1,135 per month, roughly $13,620 per year.

Notice which line dominates. At this volume the AI usage fee is nearly seven times the seat cost, and it is the line you control least. You can cap seats by not hiring. You cannot easily cap resolutions without making the AI worse on purpose, which defeats the point of buying it.

There is also a budgeting problem hiding in the variance, separate from the total. A flat bill is the same in January and December. A metered bill moves with seasonality, launches, incidents, and press, which means the months when support is under the most pressure are also the months when the invoice spikes. If you have ever had to explain a surprise usage bill to a founder or a finance lead, you already know that predictability has a value of its own, independent of the average monthly cost.

What counts as a resolution?

Intercom bills Fin per resolution, which in its published terms is a conversation outcome where the AI handled the customer's issue, either confirmed by the customer or inferred when the customer leaves without asking for more help. The precise definition and its edge cases are Intercom's to set, and they matter, because the definition of a billable outcome is effectively the definition of your bill.

The practical questions to ask any per-resolution vendor before signing: does an answer the customer never confirms count, does a conversation that later reopens still count, and can you audit the list of billed resolutions line by line? None of these questions have obvious answers from a pricing page, and the difference between generous and strict interpretations is real money at four figures of monthly volume.

What hidden costs should you watch for?

Beyond seats and resolutions, Intercom prices several capabilities separately, and the entry price does not include everything on the feature tour. Budget for the plan tier you actually need, plus any add-ons your workflow depends on, before comparing vendors on the headline number.

We will not quote numbers for individual add-ons here, because they change and vary by plan; the honest move is to price your specific cart on Intercom's published pricing page. The pattern to watch is structural: entry pricing gets you in the door, and the capabilities that make the product sticky, more channels, more automation, more volume, tend to live on higher tiers or metered lines. Ask for the total at your team size and volume, in writing, before you commit.

When does per-seat pricing make sense?

Per-seat pricing is rational when human labor is the product and each additional agent creates proportional value, which described the helpdesk era well. If you run a large team where most conversations are handled by people, seats track value reasonably, and per-seat pricing is a fair deal.

The model strains once an AI agent does the volume. The AI is not a seat, so vendors bolt on per-resolution pricing to capture its value, and you end up paying twice: for the people and for the automation that was supposed to reduce the people cost. Small teams feel this hardest, since five seats of tooling overhead lands on five people. We wrote more about that tradeoff in AI customer support for small teams.

What does an AI-included alternative look like?

The alternative is a seat price that already contains the AI. On HelpYap, the Pro plan is $49 per seat per month and every seat carries 1,000 AI conversations, so the workload from the worked example above, 5 people and 1,000 AI conversations, costs $245 instead of from $1,135. The AI resolving more conversations does not change the bill, because the resolutions are not metered.

Seat pricing without a meter is not magic; it is a different bet. HelpYap bets that a predictable bill at a lower price wins more small and mid-sized teams than metered upside wins enterprises. Each seat carries its own AI conversation pool, Starter at $19 per seat includes 300 conversations per seat, Pro at $49 includes 1,000 per seat, Business at $89 includes 2,500 per seat, so the pool grows when the team does, and hitting a limit moves you to the next tier or the next seat, not onto a meter. The full matrix is on the pricing page, and the feature-by-feature comparison is at HelpYap vs Intercom.

The capability set is the part people assume must be missing at that price, so to be concrete: the AI agent answers from your docs with source citations, escalates to a shared team inbox with Slack threads, and the whole thing installs with one script tag:

<script src="https://www.helpyap.com/widget.js" data-project="your-project"></script>

Both products offer a free trial, 14 days on Intercom and 7 days on HelpYap, so the cheapest way to settle the comparison is to run your real questions through both and read the answers next to the quotes.

The bottom line

Intercom's published pricing is from $29 per seat per month plus $0.99 per Fin resolution, which for a five-person team at 1,000 AI resolutions works out to from $1,135 per month, with the usage line growing every time the AI succeeds. Per-seat plus per-outcome pricing can be fine for large human-heavy teams, but it punishes small teams for automating well. HelpYap Pro covers the same team, with five times the AI volume included, for $245 per month, and nothing in that number is metered. Run the math on your own numbers, in writing, before you sign either contract.