What is first-contact resolution?

First-contact resolution (FCR) is the share of customer issues that are fully resolved in the customer's first interaction, with no need to call back, chase or be contacted again. It is one of the measures most closely linked to customer experience. With AI agents answering first, the hard part is defining "resolved" truthfully.

Why does first-contact resolution matter once agents act?

Customers care about having to come back more than almost anything else. Every repeat contact is extra effort for them and extra cost for the firm. FCR captures that directly, which makes it a better guide than speed or containment.

AI agents can raise FCR for real, by having the right sources and enough authority to finish common tasks in one go. They can also inflate it. An agent that gives a confident but wrong answer ends the first contact; the customer only discovers the problem later. An agent that refuses to hand over may mark a conversation as finished when the customer has simply given up. Both look like first-contact resolution until the repeat contacts are counted.

How is FCR measured in practice?

There are three common methods, and serious teams combine them:

• Repeat contact window: an issue counts as resolved at first contact if the same customer does not contact again about the same issue within a set period, such as seven days, in any channel.
• Customer confirmation: a survey asking whether the issue was resolved.
• Case review: quality reviewers check a sample of conversations for whether the resolution was correct and complete.

The repeat window must cover every channel. A customer who chats with an agent and then phones the next day was not resolved at first contact, even if the chat system says so. Our guide on reporting on contact drivers with an AI agent shows how to link contacts about the same issue.

An example from a water company

Take a water company whose AI agent handles billing queries. Its chat-only FCR looks excellent. When the team links chat to phone and email by customer and issue, the figure drops sharply for one intent: customers asking to set up a payment plan. The agent explains the options well but cannot set the plan up, so customers phone to complete it.

The firm certifies the agent to set up standard payment plans within limits, with anything outside those limits passed to a person. FCR for that intent rises, measured across all channels. Our guide on resolving complaints at first contact with an AI agent applies the same idea to complaints.

How is FCR different from resolution rate?

Resolution rate is the share of issues eventually resolved, however many contacts it takes. FCR is the stricter subset resolved in the first contact.

A team can have a high resolution rate and poor FCR: everything gets sorted in the end, after several calls. FCR is the measure that rewards getting it right first time, which is also why it is the one most tempting to inflate by declaring resolution early.

Which terms sit next to first-contact resolution?

• Time to resolution: how long the customer waits until the issue is solved.
• Repeat contact rate: the inverse view, how often customers come back about the same issue.
• Deflection is not resolution: avoiding a contact is not the same as solving the problem.

Where Pryme fits

FCR improves most when an agent has authority for the whole task. In Pryme that authority is earned per agent: the support agent that knows when to say no gets depth 3 for setting up a standard payment plan only after Attest shows it does so correctly on the firm's own past cases.