AHT Calculator & Guide

What Is Average Handle Time?

Average handle time (AHT) is the average total time an agent spends handling a customer interaction, measured from the moment the call is answered through the completion of any after-call work.

It has three components:

• Talk time: the live conversation between agent and customer.
• Hold time: any time the customer spends on hold during the call, including warm-transfer waits.
• After-call work (ACW): the wrap-up that follows, such as writing notes, setting dispositions, updating the CRM, or creating follow-up tickets.

AHT matters because it sits underneath nearly every operational number in a call center. It drives how many agents you need on the schedule (it is a direct input to Erlang staffing models), what each call costs, and how long customers wait in queue. Small changes in AHT compound: shaving 30 seconds off a queue that handles 50,000 calls a month returns roughly 25,000 agent minutes.

The AHT Formula

The standard formula is:

AHT = (Total Talk Time + Total Hold Time + Total After-Call Work) / Number of Calls Handled

All three time totals must cover the same set of calls as the denominator. Most contact center platforms report each component, so the calculation is usually a matter of pulling the right report rather than tracking anything new.

Worked example. Suppose your team handled 1,000 calls last week with these totals:

• Talk time: 4,000 minutes
• Hold time: 500 minutes
• After-call work: 900 minutes

Total handling time is 4,000 + 500 + 900 = 5,400 minutes. Divide by 1,000 calls and AHT is 5.4 minutes per call, or 5 minutes 24 seconds. The breakdown also tells you where the time goes: about 74% talking, 9% on hold, and 17% in wrap-up. That breakdown is often more actionable than the headline number.

What Is a Good AHT? Why Benchmarks Mislead

There is no universal benchmark for average handle time, and most published numbers are not comparable to your operation. Commonly cited industry averages cluster around a few minutes per call, but the spread behind those averages is enormous, for reasons that have nothing to do with agent performance:

• Call complexity. A password reset and an insurance claim are different jobs. Technical support and healthcare calls legitimately run far longer than retail order-status calls.
• What gets counted. Some teams include after-call work, some do not. Some exclude hold. Two centers with identical operations can report very different AHT purely on definition.
• Channel deflection. When self-service absorbs the easy contacts, the calls that remain are the hard ones, and AHT rises even as the operation improves.
• Process burden. Compliance scripts, identity verification, and mandatory disclosures add time that agents cannot coach away.

The honest benchmark is your own history: track AHT by call type and by agent, watch the trend, and investigate changes. A rising AHT on the same call mix is a signal worth chasing. Being a minute above a number from a vendor infographic is not.

What Drives AHT Up

When handle time creeps upward, the causes usually fall into a few buckets:

• Knowledge gaps. Agents put customers on hold to find answers, ask colleagues, or search a disorganized knowledge base. This shows up as hold time and repeated dead air.
• Weak discovery. Agents who do not diagnose the real issue early spend the back half of the call redoing the front half.
• Manual after-call work. Hand-written summaries, duplicate data entry across systems, and unclear disposition rules can quietly add minutes of wrap-up to every call.
• Tooling friction. Slow systems, too many screens, and missing customer context force agents to make the customer wait while software catches up.
• Transfers and escalations. Every misroute adds a second greeting, a second verification, and a second explanation of the problem.
• New-hire ramp. A wave of new agents raises team AHT for months. That is a training curve, not a performance problem, and it needs coaching rather than pressure.

How to Reduce AHT Without Wrecking Quality

The classic mistake is managing AHT as an agent speed problem. Tell agents their handle time is too high and they will find the fastest fix available: end calls sooner. Issues go unresolved, repeat calls and transfers climb, and total workload gets worse while the AHT dashboard turns green. Speed you get by cutting resolution is borrowed, not earned.

Durable AHT reduction attacks causes:

• Find out where the time actually goes. Split AHT into talk, hold, and wrap by call type. A wrap-time problem and a hold-time problem have completely different fixes.
• Coach from real calls, not averages. Compare how your shortest effective calls differ from your longest ones on the same call type. The difference is usually specific, coachable behavior: a tighter opening, better discovery questions, confident answers instead of holds. This is where call scoring earns its keep: scoring calls against defined criteria turns "be faster" into "here is the behavior to change, and here is a call where you did it well."
• Kill after-call work with automation. Auto-generated call summaries and dispositions can remove most manual wrap-up without touching the conversation at all. This is often the largest AHT win available, and it costs quality nothing.
• Fix knowledge access. Every hold to go find an answer is a knowledge base or training gap wearing an AHT costume.
• Guard the counterweight. Any AHT initiative should be paired with quality scores and first-call resolution on the same dashboard. If AHT falls while quality falls with it, you did not improve anything; you moved cost from the phone queue to the callback queue.

This is the approach behind Voxjar's call center quality assurance software: because AI evaluates 100% of calls instead of a tiny manual sample, you can see exactly which behaviors correlate with long handle times, and coach those behaviors with the agent's own calls as evidence. Handle time comes down because calls get better, not shorter.

AHT vs Other Call Center Metrics

AHT is one efficiency metric inside a larger measurement system, and it only makes sense read alongside its neighbors.

• AHT vs average talk time. Talk time is one component of AHT: just the conversation, with no hold or wrap-up. Teams sometimes report talk time as AHT, which understates true handling cost.
• AHT vs service level. Service level measures how fast calls get answered; AHT measures how long they take once answered. They are linked through staffing: higher AHT means more agents needed to hit the same service level.
• AHT vs occupancy. Occupancy is the share of logged-in time agents spend handling calls. AHT is per call; occupancy is per agent shift.
• AHT vs first-call resolution. FCR is the natural counterweight. A low AHT with falling FCR means agents are ending calls, not resolving them, and total contact volume will rise to prove it.

If you use AHT for staffing math, our Erlang C staffing calculator takes AHT as a direct input to compute how many agents you need.