Average Handle Time

In depth

Average handle time (AHT) measures how long it takes to handle a single customer interaction from start to finish. As the formula shows, it adds together the time spent in the conversation itself, any time the customer is placed on hold, and the after-call work that follows, then divides the total by the number of interactions handled over the period.

The metric began in voice call centres, where "talk time" is literal, but it applies just as well to any channel. For messaging and email support, the talk component becomes the time spent reading, composing, and exchanging replies, while hold time maps to any spell where the conversation is paused waiting on the customer or a colleague.

It is easy to confuse with resolution time, but the two measure different things. Average handle time counts only the minutes an agent is actively working on an interaction. Average resolution time measures the whole journey until the issue is solved, which can stretch across several interactions and several days, including all the time a ticket sits waiting in between. A short handle time can sit alongside a long resolution time, which is why teams watch both.

What counts in average handle time?

Three components make up the figure:

• Talk or handling time. The active time spent in the conversation, whether that is a phone call or the back-and-forth of a chat or email thread.
• Hold or wait time. Any period the customer is held while the agent checks something or consults a colleague.
• After-call work. The admin that follows each interaction, such as writing notes, tagging the ticket, and setting up any follow-ups.

Because after-call work is bundled in, average handle time captures the true cost of an interaction, not just the visible conversation. If wrap-up is creeping up, the handle time will reflect it even when the conversation itself is quick.

What's a good average handle time?

There is no single benchmark that fits every team. What counts as good depends on how complex your queries are, which channels they arrive on, and the kind of support you offer. A technical B2B team will handle far longer conversations than one answering simple retail questions, and neither number is "better" in isolation.

It is also worth resisting the urge to drive the figure down for its own sake. A very low handle time can mean agents are rushing, closing conversations before the issue is genuinely resolved and creating repeat contacts. The most useful reading pairs handle time with first contact resolution and a satisfaction score, so you can tell efficient handling apart from hurried handling.

The reliable way to bring it down is to make handling easier: give agents the right answer quickly, cut down on retyping, and take routine questions off the queue entirely so the interactions that remain get proper attention.