How do you calculate average handle time?
For phone calls, add total talk time, total hold time and total after-call work, then divide by the number of calls handled. Microsoft’s Dynamics 365 documentation uses this formula for voice, and for chat it adds active chat time to wrap-up time and divides by the number of chats, as checked on 8 October 2026.
For example, if an agent handles 20 calls with 100 minutes of talk time, 10 minutes of hold time and 30 minutes of after-call work, AHT is 140 minutes divided by 20, or 7 minutes.
After-call work, also called wrap-up time, covers notes, tagging and follow-up messages. Time a customer spends waiting in a queue before anyone answers is usually measured separately, as wait time or speed of answer.
Why is handle time hard to read for an AI agent?
Pushing AHT down can backfire even with human agents. Rushed interactions can leave problems half solved, which come back later as repeat contacts and lower first contact resolution.
An AI voice agent has no hold time or after-call work, so its handle time is simply the length of the conversation. A long call isn’t necessarily a problem if the caller was working through a setup step, and a short one isn’t automatically good: it can mean a quick answer or a caller who gave up. Read call length beside what happened in the conversation.
How does Kawl Now handle call length?
Kawl Now saves how long each AI conversation lasted and shows it with the transcript in your dashboard, and the REST API and MCP server return it too. There are no handle time targets or staff metrics. If you turn on handover to a person, an early release, the human part of the call happens on your team’s phone after the AI conversation ends, and it isn’t included in that length.
Every browser call stops at 10 minutes, and each workspace takes up to 3 calls at once and 12 new calls a minute. Capacity for AI conversations comes from those limits rather than from how many people you have free.
Use the lengths to choose which conversations to read first. Very short calls can show a caller who gave up, and calls that reach the limit can show a topic your answers don’t cover well.
Common questions
Is a lower average handle time always better?
No. A lower AHT can mean faster answers or rushed ones. Track it beside first contact resolution and customer satisfaction, and find out why long interactions run long before you try to shorten them.
Does average handle time include hold time?
Yes. The standard formula is talk time plus hold time plus after-call work, divided by the number of interactions handled. Time spent waiting in a queue before anyone answers is usually measured separately, as wait time.