What is average handle time?

Average handle time (AHT) is the mean time a support team spends on each customer contact, counting talk or chat time, hold time and the after-contact work of notes and follow-ups. Contact centres use it to plan staffing and spot inefficiency. Once AI agents take the simple contacts, the people's AHT rises, and that is usually good news.

Why does AHT matter once agents act?

AHT has always been a capacity measure: how many staff are needed for how many contacts. It has also been misused as a performance target, encouraging staff to hurry customers off the phone.

AI agents change what the figure means. An agent that handles password resets and balance queries removes the shortest contacts from the human queue. The contacts that reach people are now the complaints, the bereavements, the disputes and the cases the agent was not allowed to handle. Human AHT goes up, not because staff got slower, but because the work got harder. A team that still targets AHT will push staff to rush exactly the conversations that most need time.

How is AHT calculated and used in practice?

The formula is simple: total talk or chat time, plus hold time, plus after-contact work, divided by the number of contacts handled. It is reported by channel, team, contact type and individual.

Used well, it answers planning questions: how many people are needed on Monday morning, how long a new process adds per call. Used badly, it becomes a target that trades quality for speed. With agents in the mix, measure it separately for the agent and for people, and by contact type, or an average score will hide the failure that matters.

An example from a utility's support line

Take an energy supplier that deploys a support agent for meter readings and payment date changes. Before the agent, human AHT is around six minutes. Afterwards, those simple contacts mostly go to the agent, and human AHT rises to around nine minutes.

The supplier's first reaction is concern. A closer look shows the human queue is now mostly billing disputes and customers in payment difficulty, each needing more care. Complaints about rushed calls fall. The right response is to re-baseline AHT by contact type, not to push it back down.

How is AHT different from time to resolution?

AHT measures how long staff spend on a single contact. Time to resolution measures how long the customer waits from first raising an issue until it is actually solved, across however many contacts that takes.

They can move in opposite directions. Cutting AHT by ending calls quickly can lengthen time to resolution, because the customer has to call back. An AI agent with a very short handle time that cannot resolve the issue only moves the wait elsewhere. For a customer, time to resolution is what counts.

Which terms sit next to AHT?

• After-contact work: the notes and tasks after a contact ends, often where agents can help staff most by drafting summaries.
• Occupancy: the share of paid time staff spend handling contacts.
• Commitments made during a contact, such as when an AI agent commits to a delivery time, create follow-up work that AHT does not capture.

Where Pryme fits

Pryme does not treat speed as evidence of quality. A support agent that knows when to say no is certified on whether it answers correctly, refuses what is outside its role and hands over when it should. Handle time is an operational figure; it never earns an agent depth.