How it works
For each new conversation, first response time is:
time of first human reply − time of the customer’s first message
Teams then summarize across conversations. A median is more honest than an average, because one thread that waited all weekend can distort the mean. Many teams also track a high percentile, such as the 90th, to see how bad the slow cases get.
Decide up front what counts as “new”. In messaging, the same customer may write again weeks later in the same thread. Many teams treat a message after a set period of silence, or after the conversation was closed, as a new conversation for measurement purposes.
Example
A pet-grooming salon receives five new SMS conversations on a Thursday. The first human replies come after 3, 6, 12, 25 and 190 minutes. The median first response time is 12 minutes. The average is about 47 minutes, pulled up by one message that arrived during a lunch rush and sat for over three hours.
Both numbers are useful. The median shows the typical experience. The outlier shows a gap: nobody covers the inbox from noon to 1 PM.
First response time vs. related metrics
- Speed to lead is first response time for new prospects only.
- A response time SLA is a target (“reply within 15 minutes during business hours”). First response time is the measurement you compare against it.
- Average response time includes every reply in a conversation, not just the first.
Why it matters for messaging teams
In chat, the first reply sets expectations for the whole exchange. It is also the easiest response-time metric to act on: if FRT is slow, the cause is usually coverage (nobody watching at certain hours) or ownership (nobody sure whose message it is).
Common mistakes
- Counting automated acknowledgments as the first response.
- Mixing business hours and after-hours conversations in one number without saying so.
- Looking only at the average.
- Measuring it per team but never per person or per channel, so you can’t see where the delays are.
Response-time analytics in inChat are in development.