What is the AHT full form?
The AHT full form is Average Handle Time. It is the average time an agent spends on one call from start to finish, including the work after it. The formula is: AHT = (total talk time + total hold time + total after-call work) / number of calls handled. Lower is not always better.
AHT is one of the first terms a new agent hears on a BPO floor, usually from a team leader pointing at a screen. It is also one of the most misused. Used well, it tells you how much work each call creates, which is the foundation of staffing and pricing. Used badly, it becomes a stopwatch that pushes agents to rush customers off the phone.
You can work out AHT for any team in a few seconds with our free AHT calculator. The rest of this post explains what goes into the number and what to do with it.
What goes into the AHT formula?
The formula has three parts on top and one part underneath. Each needs a clear definition, or two reports for the same team will show two different AHTs.
- Talk time: the time the agent is connected and speaking with the customer, from the moment the call is connected to the moment it ends or is transferred.
- Hold time: the time the customer spends on hold during the call while the agent checks something or asks a supervisor.
- After-call work (ACW): the time after the call ends that the agent spends on notes, dispositions, forms or CRM updates before becoming available again. Also called wrap-up time.
- Calls handled: the number of calls the agent actually answered or connected with. Missed, abandoned and unanswered calls are not included.
How do you calculate AHT? A worked example
Take one agent's shift on a verification process. During the shift the agent handled 40 calls. The dialer report shows 180 minutes of total talk time, 20 minutes of total hold time and 40 minutes of after-call work.
Add the three: 180 + 20 + 40 = 240 minutes. Divide by calls handled: 240 / 40 = 6 minutes. The agent's AHT is 6 minutes, or 360 seconds.
Now look at the team. If ten agents handled 380 calls between them with 2,280 minutes of combined talk, hold and wrap time, the team AHT is 2,280 / 380 = 6 minutes. Always calculate team AHT from the totals, not by averaging each agent's AHT, because agents with very few calls would otherwise count as much as agents with many.
Most dialers, including ViciDial, report talk, pause and wrap time per agent, so you can pull these figures straight from the agent time report. Check whether your dialer counts dispo time as part of the call or as pause, because that changes the ACW figure.
What is a good AHT for fronting, verification and support?
There is no single good AHT. It depends on what the call has to achieve. The ranges below are general guidance from typical processes, not targets to copy without checking your own data and your client's expectations.
- Fronting (live transfer): often 2 to 4 minutes per connected call that reaches qualification, since the job is a short opener, a few questions and a transfer. Calls that end early pull the average down.
- Verification: often 5 to 10 minutes, because the agent confirms details, reads disclosures and captures information accurately.
- Inbound customer support: often 4 to 8 minutes for general queries, and longer for technical support or billing disputes.
- Closers: usually the longest, often 15 minutes or more for licensed insurance sales, because the sale happens on that call.
Why does AHT matter for staffing and cost?
AHT turns call volume into workload. If a queue receives 600 calls in an hour and AHT is 5 minutes, that is 3,000 minutes of work, or 50 hours of agent time, in that hour. Before you add any buffer for answering speed, you need 50 agents fully busy. That is why AHT is one of the main inputs to every staffing model, including Erlang C.
Small changes add up. On the same queue, a 30-second rise in AHT adds 300 minutes of work per hour, which is five more agents. Over a month of shifts, that is a real line on the payroll. Our Erlang calculator shows how AHT, volume and service level combine into an agent count.
For outbound fronting floors, AHT also decides how many transfers a seat can produce in a shift. A fronter with a shorter AHT and the same transfer rate makes more transfers, as long as the shorter calls are not just rushed ones. See our guide to call center metrics for how AHT sits alongside transfer rate and conversion.
How is AHT measured differently on inbound and outbound?
On an inbound queue, a call is handled when an agent answers it, so AHT runs from answer to the end of wrap-up. Calls that abandon in the queue never reach an agent and are left out. That makes inbound AHT fairly clean to measure.
On an outbound floor, the picture is messier. The dialer places many calls that never reach a person: no answers, busy lines, voicemail and disconnected numbers. Those should not count as handled calls, or AHT will look far lower than the real work per conversation. Count only calls connected to an agent, and decide whether very short calls, such as people who hang up within a few seconds of hearing the opener, belong in the figure. Many floors report two numbers: AHT for all connected calls, and AHT for calls that reached qualification or a transfer.
Transfers raise another question. When a fronter warm-transfers to a closer, the fronter's handle time should normally stop when they leave the call, and the closer's starts when they join. If the fronter stays on to introduce the customer, that overlap belongs to the fronter. Agree the rule once and write it down, so team leaders do not argue about it every month.
Finally, pause codes matter. If agents use a pause code for wrap-up instead of the dialer's dispo screen, that time can disappear from ACW and show up as break time instead. Check how your floor actually works before trusting the report.
When is a low AHT a bad sign?
A very low AHT can mean an agent is excellent, or it can mean they are cutting corners. On support processes, it often shows up as customers calling back because the issue was not solved. On fronting processes, it can mean skipped qualifying questions, which turns into returned transfers and unhappy buyers.
That is why AHT should always be read next to a quality measure. For support, look at first call resolution and QA scores. For fronting, look at transfer acceptance and returns. For verification, look at error rates in the captured details. An agent with a slightly higher AHT and clean work is almost always worth more than one with a low AHT and rework. Set AHT as a guide for coaching rather than a hard limit that triggers warnings, and agents will stop gaming it.
Watch for unusual patterns too. An agent whose AHT suddenly drops by a third, or whose calls cluster just under a target, deserves a listen before a reward.
How can you reduce AHT without hurting quality?
The best AHT reductions come from removing waste around the call, not from asking agents to talk faster. Start by listening to ten of the longest calls on a process and writing down where the time actually went: long holds, repeated questions, slow screens, or a customer who did not understand the opener. The fix usually becomes obvious. These are the changes that help most often.
- Tighten the script so each question is short and asked once.
- Give agents quick reference sheets for the five most common questions or objections, so they do not put people on hold.
- Cut after-call work by pre-filling forms, using standard disposition codes and removing duplicate data entry.
- Fix slow systems. A CRM that takes ten seconds to load a screen adds minutes to every hour.
- Coach individuals using recorded calls, focusing on the longest parts of their calls.
- Route complex calls to experienced agents rather than letting new hires struggle through them.
How do bots affect AHT on outbound floors?
On a bot-run seat, talk time is set mostly by the script and the person on the other end, and after-call work is close to zero because the bot records answers as it goes. That makes AHT on bot seats very consistent from hour to hour, which makes planning easier.
The B3 Voice Verifier bot pitches and verifies, captures the details and writes them to your CRM and the ViciDial comments, so the human closer does not have to type them again. The Fronter bot qualifies on your script and warm-transfers to a live agent. Both are priced per bot per month, with no per-minute cost, so a longer verification call does not change the bill.
If you run human and bot seats side by side, keep their AHT in separate reports. Mixing them hides what is happening on each, and the comparison between the two is one of the most useful things a pilot can tell you.
