What is average handle time?

Average handle time (AHT), explained

AHT originated in phone contact centres, where it is typically calculated as talk time plus hold time plus after-call work, divided by the number of calls handled. For chat and email, the equivalent is active handling time plus wrap-up work per conversation.

It is primarily a planning number. If you know how many conversations to expect and how long each takes, you can estimate how many people you need and what support costs per contact.

As a performance target it is risky. Agents rushed to shorten conversations transfer more, skip checks and close issues that are not fixed, which lowers first contact resolution and creates repeat contacts that cost more in total.

AI changes AHT in two opposite ways. Handing routine questions to a bot leaves humans with harder conversations, so human AHT can rise even as the team does better overall. Good handoffs, where the agent receives the full history and context, can bring it back down by removing the discovery phase of each conversation.

How to measure it

Compare AHT only within the same channel, and read it next to first contact resolution and CSAT.

Common mistake

How SuperBot relates

SuperBot's human handoff passes the full conversation history to your team in one shared inbox, so agents can skip re-asking what the customer already told the bot.

• Support cost calculator
• $25 AI agent vs $200 help desk: the honest comparison

Related terms

• First response time (FRT)First response time (FRT) is the time between a customer contacting support and receiving the first reply. It is usually reported as a median or average over a period. Automated acknowledgements are often excluded, because FRT is meant to show how long customers wait before someone, or something, actually starts helping.
• First contact resolution (FCR)First contact resolution (FCR) is the percentage of customer issues fully resolved in the first interaction, with no follow-up, transfer or repeat contact needed. It is one of the most customer-centred support metrics because having to contact a business again about the same problem is a common source of frustration.
• Human handoff (escalation)Human handoff is the point where an automated system such as a chatbot passes a conversation to a person, ideally with the full conversation history and context attached. Good handoff rules trigger on low confidence, sensitive topics, explicit requests for a human or signs of frustration, and tell the customer what happens next and how long it will take.
• Shared inboxA shared inbox is a single queue where a whole team can see, assign and reply to customer conversations, instead of messages sitting in one person's email. It shows who is handling what, keeps the conversation history in one place, and prevents both duplicate replies and conversations that nobody answers.
• Service level agreement (SLA)A service level agreement (SLA) in customer support is a defined commitment for how quickly a team will respond to or resolve customer requests, such as a first reply within four business hours. SLAs can be contractual promises to customers or internal targets, and are usually set per channel and priority level.