Intercom · Strategy · Mid · 45 to 60 minutes

Should Fin still be priced per resolution?

Intercom is a customer service software company that helps businesses support customers, including with an AI agent called Fin. Answer this product management case and get it graded line by line against a weighted rubric.

pricing strategy
product judgment
analytical reasoning
systems thinking
customer empathy

The situation

You are the PM responsible for Fin monetization. Intercom currently charges for Fin when it fully resolves a customer support conversation without human involvement. That model has been easy for support leaders to understand, easy to compare against human support costs, and it aligns Intercom's revenue with Fin's performance. Adoption has grown to more than 12,000 teams, and Fin's average resolution rate is now 76 percent. Now customers are using Fin in more complex workflows, such as subscription changes, billing issues, and disputes. In these cases, Fin may gather context, read or write to external systems, and complete several steps before handing off to a human agent for confirmation because of policy, risk, or compliance needs. Some customers argue that Fin is delivering meaningful value in these conversations, even though they do not count under the current pricing metric. Others still want pricing to stay simple, fair, and predictable. Your GM asks for a recommendation on whether Intercom should keep the current value metric or redefine what it charges for as Fin's job expands.

What to produce

Write a 1 to 2 page strategy memo.

Your answer must

State a clear recommendation on whether to keep the current pricing metric or change it, and explain why now.
Define the value metric you would use, including what should count, what should not count, and at least two example workflow types.
Analyze the main trade offs across customer value, trust, pricing simplicity, ROI clarity, and Intercom's incentives.
Identify the biggest risks, including billing disputes or gaming, and propose mitigations.
Outline a rollout plan with success measures for the first 6 to 12 months.

Constraints

Your approach must keep pricing tied to customer value delivered, not raw activity such as messages or tokens.
It must work for both fully automated support and workflows that intentionally include a human agent.
It cannot depend on manual review or custom pricing logic for every customer workflow.
It must remain simple enough that a support leader can reasonably predict bills and estimate ROI.

Format

Use short sections with headings such as problem, options, recommendation, metric definition, risks, and rollout.

How Intercom actually played it

Intercom prices its Fin AI agent per resolution, and explains why that unit needs to evolve as Fin takes on more complex work and success is no longer binary. The post notes the average resolution rate now stands at 76 percent. It is a thoughtful look at revisiting the value metric behind outcome-based pricing.

A value metric that fits version one can limit version two.
Outcome-based pricing needs a clear definition of the outcome.
Revisit pricing when the product's job expands.

Built from From resolutions to outcomes: Evolving how Fin delivers value by Darragh Curran.