What is average handle time (AHT)?

AHT is one of the oldest numbers in customer service. It started in the call center, where every minute on the phone was a minute of paid agent time, and it moved to chat and email with the same meaning: how long one contact takes to handle, start to finish.

How to calculate AHT

Add up the time spent handling contacts in a period, then divide by the number of contacts handled.

AHT = (total talk or chat time + total hold time + total wrap-up time) / contacts handled

In one week a support team handled 1,200 chats. Agents spent 9,600 minutes in active chats and 2,400 minutes on notes and follow-up afterwards: 12,000 minutes divided by 1,200 chats is an AHT of 10 minutes.

Three choices decide what the number means:

• Start and stop. Handle time starts when an agent picks up the contact, not when the customer joins the queue. Time spent waiting to be answered is a separate metric.
• Wrap-up. Include the notes, tags and follow-up tasks that come after the conversation, or AHT will look shorter than the work.
• Exclusions. Remove spam, test contacts and transfers counted twice.

In Aide, part of that wrap-up happens without the agent: every customer message is tagged with its topic automatically, and Copilot can summarise the conversation into an internal note from the reply box.

AHT in a call center

In a call center, AHT covers three parts of every call: talk time, hold time while the agent checks something, and after-call work such as notes and dispositions. It drives how many agents a shift needs, so it sits at the center of workforce planning.

Outsourcing (BPO) contracts and staffing plans are often built around a target AHT, which is why the term turns up so often in BPO job descriptions and reports. The calculation is the same; the stakes are higher because hours are billed or scheduled against it.

AHT in chat and email

AHT means the same thing in chat and email, average handle time, but two differences change how it reads.

• Concurrency. A chat agent often handles two or three conversations at once, so each chat's handle time runs longer while the agent still clears more work per hour.
• Elapsed time vs work time. An email ticket can stay open for two days while the agent works on it for six minutes. Handle time should count the six minutes, not the two days.

For written channels, many teams track AHT next to resolution time, which measures how long the customer waited for an answer.

What is a good AHT?

There is no universal good AHT, because it depends on the channel and the requests in the queue. Published benchmarks vary widely by industry, so the useful comparison is your own AHT by request type, over time.

In a Shopify store's queue, a where is my order chat answered from tracking takes a couple of minutes. A damaged item claim, with photos, a replacement and a carrier report, takes far longer. A team whose mix shifts toward damage claims will see AHT rise even if every agent works faster.

In Aide, every customer message carries a topic, and analytics can be filtered by topic with each number opening the exact conversations it counts, so a team can see which request types are growing in its queue before reading too much into one average.

Why AHT alone misleads

AHT measures speed, not whether the customer's problem was solved, and pushing it down on its own makes service worse.

• Rushed contacts come back. A customer who gets a quick, incomplete answer writes again, and the second contact is often longer than the first would have been.
• Transfers hide time. Passing a hard contact to another team lowers one agent's AHT and raises the customer's total time.
• The mix changes. When AI or self-service takes the simple requests, the contacts left for people are harder, so human AHT rises while the team is doing better work.

The practice this page rejects is setting AHT as an agent's target. Use it to plan staffing, and judge service by first contact resolution, reopen rate and CSAT.

Where Aide fits

Aide works on chat, email and helpdesk tickets, not phone calls. In Zendesk and Front, the Aide panel shows the customer's Shopify orders beside the ticket with a draft reply and fitting macros, and agents can refund, cancel or change the shipping address from it instead of switching to Shopify. Aide's analytics separate the conversations Aide resolved on its own from the ones a person finished. See Copilot.