What is first response time?

Why it matters

On a website, a visitor who asks a question is usually deciding something: whether the product fits, whether the delivery date works, whether to pay. Every minute of silence makes leaving more likely. In live chat, expectations are measured in seconds to a couple of minutes, not hours as with email.

FRT does not measure whether the problem was solved — that is resolution rate. A team can reply fast with nothing useful. Read the two together.

Average or median

FRT has a long tail: a few conversations that arrived at night or during a rush can wait hours. The average is pulled up by them; the median shows the typical wait. Report both, and look at FRT by hour of day to see whether the tail comes from staffing gaps.

First response time in TapCub

TapCub measures FRT from the start of a conversation to the first reply from an agent, in the site's time zone. The support report shows the average and the median, the trend over time, and the average per agent in the workload table. Conversations that the AI assistant answered and closed without a person are counted separately as AI-handled and AI-resolved, so AI replies do not hide slow human ones.

A weekday × hour heatmap of when conversations start helps match office hours and shifts to demand; outside office hours the widget can collect offline messages instead of leaving visitors waiting.

How to bring it down

• Staff the hours when conversations actually start, not the hours that are convenient.
• Let an AI assistant answer common questions from your knowledge base, and hand over to a person when it cannot.
• Set a concurrency limit so no agent is buried while others are free.
• Outside office hours, say so and collect the question rather than leaving a chat unanswered.

Related guides: Does Live Chat Increase Conversions? How to Measure It Honestly.

See also: Customer experience metrics.

See also: Customer service training: what to cover, a 30-day plan and how to measure it.