Intercom's Fin AI agent resolves about two-thirds of support conversations on average

Intercom's Fin answers customer support conversations for thousands of companies. Intercom reports an average resolution rate of about 66–67% across its customer base, rising over time.

Organization: Intercom (and its customers)
Industry: Customer Service Software
When: 2025
Evidence: Vendor case study
Agents and tools: Fin Claude

Reported outcome

Average resolution rate of 66–67% across about 6,000 customers, up from about 51% at launch.

66–67%
Avg. resolution rate
~6,000
Customers
~51%
Out-of-box (earlier)

Caveats

Published by the company that sells the agent. Treat numbers as marketing until confirmed elsewhere.

Intercom sells Fin and defines 'resolution' itself.
Results vary a lot by industry, question type, and how good the help content is.

What the agent does

Fin reads a company's help center and support content, answers customer questions in chat and email, and can take some actions through connected systems. It hands off to human agents when it cannot resolve the issue.

Why this is a useful benchmark

Fin runs across thousands of companies, so its average is a rough industry baseline for AI support. If your bot resolves far less than about two-thirds, check your help content first.

What to copy

Fix your help center before tuning the bot. Most failures are missing or outdated content.
Review unresolved conversations weekly and write the missing articles.
Report resolution rate and CSAT together.

Starter prompt

Written by us. Use it to audit your help content.

Here are 50 support conversations the AI could not resolve:
{conversations}

Group them by root cause:
- missing article
- outdated article
- needs an account action the AI cannot take
- needs a human (judgment, exception)

For each "missing" or "outdated" group, draft the article title and
a 5-bullet outline.