What is average handle time (AHT)?
Average handle time (AHT) is the average duration of a customer call — including talk time, hold time, and after-call work. It's a core contact center efficiency metric that AI voice agents reshape significantly.
Written By Catherine Weir
Last updated About 3 hours ago
Average handle time, usually shortened to AHT, is the average total duration of a customer call — including the time the caller is talking with the agent, any time spent on hold, and the after-call work the agent does to wrap things up. AHT is one of the most widely used efficiency metrics in contact centers.
Lower AHT means more calls handled per agent-hour, which means lower cost per call. But AHT can be pushed too low — agents rushing calls damages first call resolution and customer satisfaction. The goal isn't minimum AHT; it's the AHT that's consistent with resolving calls properly.
How AHT is calculated
AHT = (total talk time + total hold time + total after-call work) / number of calls
Talk time — actual caller/agent conversation
Hold time — caller on hold while the agent looks something up or consults someone
After-call work (ACW) — time the agent spends after the call ends updating records, sending follow-ups, etc.
AHT is measured per agent or per queue or per call type — aggregate AHT is less useful than segmented AHT.
Typical AHT by industry
Retail and e-commerce: 4–8 minutes
Technology support: 8–15 minutes
Healthcare and insurance: 10–20 minutes
Financial services: 8–15 minutes
Telecom: 10–20 minutes
Simple appointment booking: 1–3 minutes
These are rough benchmarks — any specific business's AHT depends heavily on call complexity and agent training.
The AHT vs. FCR trade-off
Optimizing for AHT alone tends to hurt first call resolution. Agents rushing off calls take shortcuts, skip verifications, and promise callbacks they shouldn't — all of which create repeat calls and drop FCR. The best contact centers optimize AHT at a given FCR level, not AHT by itself.
How AI voice agents affect AHT
AI voice agents reshape AHT in three ways:
Simple calls resolved instantly — the AI handles appointment booking, FAQ responses, and message-taking in 1–3 minutes end-to-end, regardless of what a human's AHT would have been. This pulls overall AHT down.
Complex calls escalated with context — when the AI escalates, the human picks up the call with a summary of what's been discussed, reducing the human's effective AHT compared to a cold call
Zero after-call work for AI-handled calls — the AI generates the call summary, updates the CRM, and sends follow-up messages automatically. ACW is near zero for fully-handled calls.
The result for a typical SMB deploying AI voice: overall AHT drops by 40–60%, and human agents can focus on the complex calls where their time is best spent.
Related concepts
First call resolution (FCR) — the companion quality metric
Call deflection — the biggest lever on aggregate AHT
Call containment rate — the AI-specific metric
See it in action
The analytics dashboard for the Receptionist Agent at 365agents reports AHT by call type, along with containment, deflection, and escalation. You can see exactly where the AI is saving time and where calls are taking longer than they should.