Skip to main content
First response time (FRT) is the time from the first inbound customer message to the first human reply on a thread. It is one of the metrics your SLA policies track.

​
How it’s calculated

FRT = time from the first inbound customer message to the first outbound human reply.
  • The clock starts on the first inbound customer message, not on thread creation. These differ when a thread is created programmatically before the customer sends anything.
  • Threads whose first reply came from a machine user, such as an AI agent or an API integration, are left out entirely, even if a human replies later.
  • Threads with no inbound customer message are left out.
  • For charting purposes, threads are grouped by when the first reply was sent, not when the thread was created.
Median and 90th percentile are quantile(0.5) and quantile(0.9) across all FRT values in the window.

​
Statistical views

Median: the midpoint FRT across all threads in the selected period. Half of threads got a first reply faster than this, half slower. Less skewed by outliers than the average. 90th percentile: 90% of threads received a first reply within this time. Use it to see how the slowest 10% of threads are performing.

​
Team member view

In Team reporting, FRT covers only threads where that team member sent the first reply.

​
SLA tracking

A thread breaches its FRT SLA if no human reply arrives before the configured deadline. The SLA clock starts when the thread is created, not at the first inbound message, so a thread’s SLA outcome and its FRT can disagree. See SLA compliance for how the breach rate is calculated.