Use the standard formula: total talk time plus hold time plus after-call work, divided by contacts handled. Then split where time is spent: talking, waiting, writing, documentation, or handoff. That keeps AHT from becoming a speed target without quality control.
Start with the full AHT formula
Use active handle time plus after-call work divided by resolved tickets. This keeps documentation, wrap-up, and follow-up work inside the calculation.
- Split conversation time, hold time, and ACW when possible.
- Compare the same channels and case types.
- Use the same time window as your staffing data.
Use the example calculation as a sanity check
With 4,200 talk minutes, 600 hold minutes, 900 ACW minutes, and 700 contacts handled, the output is 8.1 minutes AHT. If your result looks extreme, check that every input covers the same period and unit before changing targets.
- Make sure ACW is not counted twice.
- Keep escalated tickets separate.
- Label peak weeks before comparing them with normal weeks.
Read benchmarks by channel, not as one hard rule
A lower AHT is only healthy when first contact resolution, reopen rate, and CSAT stay stable. Use benchmarks as diagnostic signals, not agent pressure.
- Order-status questions should not be compared with complaints.
- Chat, email, and voice each need their own baseline.
- Tie AHT decisions to quality metrics.
Turn AHT into capacity impact
The calculator shows how many hours a 10% lower AHT would free in the ticket set you entered. Use that to choose between AI draft replies, better routing, or staffing changes.
- High ACW: improve draft replies and documentation.
- High active time: improve knowledge, context, and macros.
- Low buffer: check staffing and occupancy.
Calculating AHT for non-voice channels
AHT comes from the call centre, where a conversation has a clean start and end. Email and chat do not, so the metric has to be redefined before it means anything. For email, count the time an agent actually spends composing and researching a reply, summed across every reply in the thread, plus wrap-up. For chat, count active time on the conversation and exclude the gaps where the agent handled another chat concurrently. Waiting time is the trap: the hours a ticket sits in the queue belong to resolution time, never to handle time.
- Email: sum agent-active minutes across all replies in one thread, then add wrap-up.
- Chat: use active time only, and record concurrency separately as an occupancy metric.
- Never include queue waiting time; that is resolution time, not handle time.
- Report non-voice AHT next to a thread-count metric, or long threads stay invisible.
Where ecommerce handle time actually goes
In an ecommerce queue, most of the handle time is not the typing. It is the lookup: which order, which shipment, what was promised in the previous reply. An ecommerce help desk that shows order context beside the conversation removes the largest part of that lookup, which is why it moves AHT more than any scripting technique.
ecommerce help desk software