Average handle time (AHT): how to calculate it

Average handle time (AHT) is the average time an agent spends on one customer contact, from the moment the call starts to the moment the agent has finished all the work it created.

In a call center, AHT is one of the two numbers a staffing formula needs, the other being how many calls arrive. Get it wrong and your staffing plan is wrong with it. This guide covers the AHT formula, how it is measured, why a single average can mislead, and how to improve it without making service worse.

The AHT formula

AHT = (talk time + hold time + after-call work) ÷ number of calls handled

• Talk time is the time the agent is speaking with the customer.
• Hold time is the time the customer is placed on hold while the agent looks something up.
• After-call work (ACW), also called wrap-up, is the time spent on notes, CRM updates and follow-up tasks after the call ends.

Here is a small example. An agent handles 50 calls:

• Talk time: 3 minutes on average, so 150 minutes in total.
• Hold time: 20 minutes in total.
• After-call work: 1.5 minutes on average, so 75 minutes in total.

AHT = (150 + 20 + 75) ÷ 50 = 4.9 minutes, or 294 seconds.

AHT vs average talk time and call duration

These get confused. Average talk time counts only the conversation. Average call duration usually runs from pick-up to hang-up, so it includes hold but not wrap-up. AHT is the widest measure because it also includes after-call work. When you are staffing, use AHT: the agent is not free to take the next call during wrap-up, so it counts as handling time.

Why AHT matters for staffing

AHT converts call volume into workload. Calls per hour multiplied by AHT in seconds, divided by 3,600, gives the traffic in erlangs: the number of agents that would be busy all the time if calls arrived perfectly evenly. That is the starting point of the Erlang C calculation and the Erlang C calculator.

Here is what a change in AHT does to a team handling 120 calls an hour, aiming to answer 80% within 20 seconds, with 30% shrinkage:

AHT	Traffic (erlangs)	Agents on the phones	Agents to schedule
200 seconds	6.7	10	15
240 seconds	8.0	11	16
280 seconds	9.3	13	19
320 seconds	10.7	15	22
360 seconds	12.0	16	23

A 10% rise in AHT, from 240 to 264 seconds, takes the same team from 11 agents to 12 on the phones, and from 16 to 18 on the schedule. A few seconds per call is real money once you multiply it by every call, every day.

Why a single average misleads

An average can hide as much as it shows.

It hides the spread. Two agents can both average five minutes, one with every call at five minutes and one with half at one minute and half at nine. The second has a very different workload and training need.

It mixes call types. Billing questions and fault reports sit in one number. Measure AHT by queue or call type, then combine them correctly.

Combining needs weighting. Suppose three call types:

Call type	Share of calls	AHT
Billing	40%	4 minutes
Technical	35%	8 minutes
Sales	25%	5 minutes

The overall AHT is 0.40 × 4 + 0.35 × 8 + 0.25 × 5 = 5.65 minutes. Averaging the three figures directly gives 5.67, which looks close here, but it drifts further from the truth as the mix gets lopsided. Always weight by the number of calls.

It changes with the day. Mondays and the day after an outage often run long. Use AHT by interval for the staffing plan, not one figure for the week.

The trap of chasing a low number

If agents are judged on AHT alone, they have a reason to rush, transfer calls away or end them early. The cost shows up as repeat calls and unhappy customers, and a repeat call adds a whole extra handle time. Treat AHT as a planning input and a diagnostic, and track first-contact resolution and satisfaction next to it. The target is the right handle time for each call type, not the shortest one.

How to improve AHT without hurting quality

• Give agents the information on screen. A CRM screen pop with the customer's history saves the opening questions.
• Shorten the path in. Long IVR menus delay the moment the real conversation starts.
• Fix the cause of repeat questions. If one issue fills a third of calls, fix it at the source.
• Improve the knowledge base, so agents find answers faster and hold times drop.
• Trim after-call work. Templates, call dispositions and automatic call logging cut the wrap-up.
• Coach with call recordings from your best agents, so good habits spread.
• Look at the long tail. The slowest 10% of calls often point to a process or training gap.

What a 30-second saving is worth

Small changes add up. A team that handles 1,000 calls a day and saves 30 seconds on each frees 1,000 × 30 ÷ 3,600 = 8.3 hours of agent time a day, about one full-time agent. That is the real prize for trimming after-call work or a clumsy IVR step, as long as the saving comes from removing waste rather than rushing customers. Check the effect on repeat calls before you count the win.

Use AHT for coaching, not punishment

AHT is most useful as a conversation starter. An agent whose AHT is far above their peers might need product training, a better knowledge base or help with wrap-up. An agent far below might be ending calls too soon. Compare agents handling the same call types, look at a few recordings, and ask what is behind the number. Leaderboards that rank agents by AHT alone push behaviour in the wrong direction.

How to measure it

Pull AHT from your ACD or dialer reports, by queue and by interval, over several weeks. Check that the report includes hold and after-call work, because some dashboards show talk time only and label it as handle time. Compare the figure with a sample of recordings now and then, to make sure what the system calls wrap-up really is wrap-up.

What to do next

Measure AHT by queue for a few weeks, then use the figures in the Erlang C calculator. Remember to add shrinkage afterwards, and to check the result against your service level target.