FCR counts an interaction as resolved only if the customer doesn't come back with the same issue. A ticket that gets closed but reopens two days later because the fix didn't actually work doesn't count, which makes FCR a stricter and more honest measure than raw resolution rate.

For an AI agent, a high FCR means it's genuinely solving the underlying problem, not just providing an answer that sounds complete. Vendors sometimes report resolution rate, how many tickets the agent closed, without disclosing how many of those reopened, so FCR is worth asking about specifically.

Frequently Asked Questions

How is FCR different from resolution rate?

Resolution rate often just means the ticket got closed. FCR specifically checks whether the customer had to come back about the same issue afterward, a much stricter and more meaningful bar.

How long after the first contact do you wait to count it as resolved?

Common windows are 7 or 30 days: if the same customer doesn't contact support again about the same issue within that window, it counts as first-contact-resolved. Shorter windows inflate the score; a 30-day window is a more honest measure.

Why might an AI agent have a lower FCR than a human agent?

If it's confidently answering questions it doesn't have complete context for, rather than escalating uncertain cases, it will close tickets quickly but generate repeat contacts. This is why escalation discipline matters as much as raw capability.

Should a vendor's advertised resolution rate be trusted at face value?

Ask specifically whether the number is FCR (checked against reopens) or simple ticket closure. The two can differ substantially, and a vendor citing an impressive closure rate without a reopen window attached is a reasonable thing to press on.