Average Handle Time (AHT)
Definition
Average Handle Time (AHT)
Average handle time (AHT) is the mean duration of one customer contact. It pools talk time, hold time, and after-call work, then divides by the total contacts handled in a period. Contact center leaders read the trend weekly, not the single number.
The metric predates omnichannel support. When the phone was the only inbound door, AHT was the shift manager’s compass. Today chat, email, and social sit beside it — reported next to first contact resolution, CSAT, and average speed of answer.
A healthy AHT balances speed against resolution quality. Cut it recklessly and customer experience pays. Tracked honestly, it surfaces training gaps, staffing pinch points, and the process friction padding every interaction.
Key takeaways
- AHT = (total talk + total hold + after-call work) ÷ total contacts handled.
- Global voice benchmarks land near six minutes; banking and technical support run 20–40% longer.
- Cutting AHT without matching investment in training and knowledge bases erodes first contact resolution.
- Modern AHT reporting spans voice, chat, and email; phone-only dashboards understate the real handle load.
- After-call work alone can eat 15–20% of total AHT when CRM forms lag behind the conversation.
How it works
Average handle time measures the full length of a customer contact, then averages it across a period. The formula pools three inputs: talk, hold, and after-call work, divided by contacts handled. Read the trend, not the raw number.
AHT = (total talk time + total hold time + after-call work) ÷ total contacts
Each input carries a different pressure point:
- Talk time: the live conversation between agent and caller.
- Hold time: any stretch the caller waits while the agent researches or transfers.
- After-call work (ACW): logging notes, updating the CRM, tagging the disposition code.
- Total contacts: voice calls, chats, or emails counted by your workforce management (WFM) tool.
Global benchmarks put voice AHT near six minutes — though the range is wide. Talkdesk, a cloud contact center software vendor, clocked technical support at 8–12 minutes in its 2024 CX report, while retail order queries cleared in under three.
Research from Cornell University ILR, the university’s industrial and labor relations school, finds after-call work alone can consume 15–20% of total handle time when knowledge bases lag.
Read AHT honestly by splitting it by channel and complaint type. A five-minute average across voice and email means little if voice sits at eight minutes and email at two.
Skills-based automatic call distribution (ACD) routes each contact to the agent best equipped to close it, which trims wasted transfers and hold time.
Shallow interactive voice response (IVR) menus help too. Deep menus push callers into the wrong queue, and every misroute lands as extra handle time on someone’s dashboard.
| Channel | Typical AHT (2024) | Main driver |
|---|---|---|
| Voice, retail | 2:30–3:30 | CRM screen-pop, tight script |
| Voice, technical | 8:00–12:00 | Diagnosis and escalation |
| Live chat | 8:00–15:00 | Concurrent conversations |
| 3:00–6:00 | Drafting and review only |
Examples
AHT plays out differently by industry. A bank fielding fraud disputes needs longer conversations to verify identity and settle chargebacks. An online retailer taking order-status calls can hold handle time under three minutes with a decent CRM screen-pop.
Retail account, Manila (2024). A business process outsourcing (BPO) partner serving a US online retailer cut AHT from 5:40 to 4:10 after deploying skills-based routing and rewriting the agent script.
First contact resolution held at 82% through that quarter, so the faster calls weren’t simply bouncing back into the queue a day later.
Healthcare payer, Cebu (2023). An insurance BPO deliberately raised AHT from 7:20 to 8:50 after adding a mandatory eligibility recheck. Complaint volume fell 30% because fewer members had to call back.
Telecom, US. Zendesk’s 2024 CX Trends report, published by the customer service software firm, puts tier-2 technical support inside US telecoms near 10 minutes per contact — against 4 minutes on tier-1 activation calls.
That split lets workforce management teams staff each tier on a separate curve, rather than averaging two very different jobs into one forecast.
SaaS live chat (2024). The same benchmark put chat handle time between 8 and 15 minutes per resolved case, because agents typically juggle two or three concurrent conversations. Voice AHT on identical tickets averaged 4:30.
Call parking is another lever some centers use to keep queues moving while an agent researches. A held call resumes with any teammate on the floor instead of waiting for the original agent.
Related terms
AHT never gets read alone. These neighboring metrics and systems either feed it, constrain it, or catch the damage when it’s chased too hard, so a contact center scorecard usually carries all of them side by side.
- First Contact Resolution: the share of contacts closed without a callback, transfer, or reopened ticket.
- Service Level Agreement (SLA): the contractual response and resolution targets agreed between a contact center and its client.
- Customer Satisfaction (CSAT): the post-contact score paired with AHT to guard against speed-over-quality reporting.
- Interactive Voice Response (IVR): the pre-agent routing menu that lifts or drops AHT depending on menu depth.
- Business Process Outsourcing (BPO): the third-party delivery model most global AHT reporting standards were built around.
- Workforce Management (WFM): the scheduling and forecasting discipline that consumes AHT as a core input.
- Automatic Call Distribution (ACD): the routing engine that assigns each inbound call to the right agent queue.
FAQ
What is a good average handle time?
Six minutes is the loose global benchmark for voice contacts, per Talkdesk’s 2024 industry report. Good depends on channel and complexity: technical support, banking, and insurance run 8 to 12 minutes, while retail order queries clear in under three.
Does lower AHT always mean better service?
No. Rushing calls to shave AHT typically drops first contact resolution and CSAT within a quarter. A flat AHT paired with rising FCR is the healthier pattern, since it means agents resolve more rather than talk less.
How is AHT different from average speed of answer (ASA)?
ASA measures the caller’s wait before an agent picks up, while AHT measures everything after. Both feed service level agreements, but they point at different bottlenecks. ASA is a staffing problem — AHT is a training-plus-tooling problem.
What counts as after-call work?
Everything an agent does between hanging up and taking the next contact: case notes, CRM updates, disposition tagging, and follow-up emails. Poorly designed forms are the single biggest inflator.
How do you calculate AHT for chat and email?
Same formula, different clock. For chat, handle time starts at the first agent message and ends when the case closes, with post-chat notes counted as after-call work. Email AHT usually excludes queue wait and counts only active drafting.
Which tools help contact centers manage AHT?
Modern WFM suites, ACD platforms, CRM screen-pops, speech analytics, and post-call automation all shape AHT, and most centers layer three or four of them rather than betting on one vendor.
The OA hubs directory maps thousands of BPO providers by location, service line, and headcount if you want to size an offshore contact center team against your AHT target.
Independent
