Intercom AI Pricing vs Alternatives: Outcomes Explained

For years, customer service pricing has been measured in simple terms — users, seats or licenses. Occasionally, a challenger like Gorgias would price per ticket to disrupt the model. With the advent of AI however, and the slow erosion of large contact centres with lots of staff, platforms like Intercom are shifting toward outcome-based pricing, where you pay based on what the AI successfully completes.

Initially, it makes sense: pay for value delivered and reduce your cost to serve. But in practice, it’s more complicated.

Yes, AI can significantly reduce the cost of human labour. But that doesn’t mean your overall cost disappears; it shifts. Vendors are evolving their pricing models to ensure they continue to capture value as automation increases.

This guide breaks down:

• How Intercom’s Outcome pricing works
• How interaction-based pricing compares
• The 40% tipping point where costs diverge

What is outcome-based pricing?

Outcome-based pricing shifts AI billing away from usage and toward perceived value delivered. As of March 2026, Intercom prices its AI (Fin) based on Outcomes rather than resolutions.

It’s not the only provider taking this approach. Platforms like Zendesk and Gorgias use similar models, charging per Resolution — typically defined as a customer issue handled successfully by AI without human intervention.

While terminology differs (Outcomes vs Resolutions), the principle is the same: you are charged based on a defined measure of successful automation.

What counts as an Outcome?

An outcome typically includes:

• A fully AI-resolved conversation
• A successfully completed AI process
• Certain structured interactions where the AI delivers a defined result

However, not every AI interaction becomes an outcome:

• Some customer-led escalations are not billed
• The exact classification depends on configuration

On first analysis, a pricing model like this seems fair and balanced. If an AI interaction doesn’t meet the criteria for an outcome — for example, it escalates to a human outside of a specific process — you’re not charged. Those interactions are effectively “free” from an AI pricing perspective.

But this is where the model becomes more nuanced, because what qualifies as an outcome is tied directly to how “successful” your AI is. That means the proportion of billable interactions increases as your automation improves.

• Early stage AI → fewer outcomes → lower cost
• Mature AI → more outcomes → higher cost

What is interaction-based AI pricing?

Interaction-based pricing (used by platforms like Gnatta) is simpler –  you’re paying for all of your AI usage, but at a much lower unit cost. That means:

• Every AI conversation = one predictable unit of cost
• No dependency on how “successful” the AI was
• No ambiguity in billing definitions

This model aligns closely with contact volumes – a metric every customer service team already understands and tracks. That means it’s easier to forecast, simpler to explain and justify internally, and more stable as AI usage scales. Usage based models like this also typically start with a lower unit cost, creating a more accessible point of entry for teams beginning their AI rollout.

Learn more about AI Agent Interaction Pricing here.

The 40% tipping point: when outcome pricing becomes more expensive

This is where the models clearly diverge. At approximately ~40% outcomes vs total AI-handled interactions, outcome-based pricing becomes less cost-effective than interaction-based pricing. Intercom currently reports an Outcome rate of 67%. 

Example scenario

• 1,000 AI-handled interactions
• 670 of those become billable Outcomes

At this point:

• Intercom bills $663 (based on $0.99 per Outcome)
• Gnatta bills $377 (based on £0.28 per Interaction)

For every additional Outcome after this point, the price gap becomes ever wider – even though only a portion of interactions are billed under the Outcomes model. Most teams building an AI solution are actively trying to achieve high automation rates, and continuously improving AI effectiveness. That means Outcome-based pricing will inevitably become a more expensive model – and that assumes no further ‘Outcome’ types will be added to the current definition.

The trade-offs

The bottom line on your bill isn’t the only factor in this kind of decision. Choosing an Outcome-based model carries two major implications:

The Success Penalty

In an environment where improving your AI solution increases costs, your team will feel tension as AI costs erode the savings they’re trying to make elsewhere. By comparison, an upfront usage-based model makes costs clear from the outset, with AI performance creating genuine savings that belong to your team. They can actively pursue greater automation performance without penalty.

Cost Volatility

By it’s nature, the definition of an Outcome is pretty muddy. How it’s defined is going to depend highly on your specific business structure, and how well (or poorly) your team have implemented Procedures inside Intercom’s AI builder. That’s going to make estimating those costs incredibly tricky, and the truth is you’ll never be really sure how much it’s going to cost until you get the bill. After years of creating accurate volume forecasts, its a big step back in cost-transparency for most contact centres.

Want to compare costs for your own setup?

If you’re currently evaluating AI pricing models, the easiest way to understand the difference is to model it against your own data.

We can help you:

• Estimate AI coverage
• Project outcome rates
• Compare interaction vs outcome costs
• Identify where the tipping point sits for your business

👉 Get in touch to run a simple cost comparison based on your contact volumes