What average handle time really measures, why chasing it directly backfires, and how to reduce AHT by removing friction rather than rushing agents off calls.
What average handle time is
Average handle time is the mean duration of a contact from the moment an agent picks it up to the moment they finish the after-call work behind it — talk time, hold time and wrap-up combined. It is one of the most watched numbers in a contact center because it drives staffing directly: with the same volume, a longer handle time means more agents, which is why the metric is really a cost lever wearing an efficiency badge.
That is also why it is the most abused number in the building. Handle time is easy to measure and easy to pressure, and pressuring it directly produces exactly the behaviour you do not want.
Why chasing AHT directly backfires
Tell agents to lower their handle time and they will — by rushing customers, skipping steps, transferring calls they could have resolved, and cutting the after-call notes that the next agent needs. The number improves and the operation gets worse, because the calls come back. A contact resolved slowly is cheaper than the same contact handled twice, and handle time counts only the first. This is the trap: AHT is a real cost driver, but treating it as a target rather than an outcome optimises for speed at the expense of resolution.
Where handle time actually comes from
The honest way to reduce handle time is to remove the friction that inflates it rather than to hurry the agent. Most long calls are long for structural reasons: an agent hunting across systems for information that should be on one screen, hold time waiting for another department, a knowledge base that does not answer the question, a process that requires re-verifying a customer three times, or a product problem that generates a hard conversation. Fix those and handle time falls without anyone being rushed, because the wasted time was never the agent's to give back.
AHT and staffing
Handle time is one of the two inputs, with volume, that decide how many agents a queue needs, which is why a small change in AHT moves headcount noticeably. If you want to see the effect concretely, our staffing calculator takes volume and handle time and returns the agents required, and stepping the handle time up or down shows how sensitive the staffing is to it. That sensitivity is the real reason to care about AHT — and the reason to reduce it structurally rather than by pressure, because a rushed reduction that raises repeat contacts costs more staff, not less.
The right way to use it
Treat average handle time as a diagnostic, not a target. Watch it by contact type and by agent to find outliers worth investigating, ask why the long ones are long, and fix the structural causes. Pair it with first-contact resolution so a falling handle time that is really rising repeat contacts shows up immediately. A program that reports AHT alongside resolution and quality is using it correctly; one that ranks agents on handle time alone is training them to lose customers efficiently.
When you outsource
Ask a provider how it treats handle time. A provider that leads with low AHT as a selling point may be optimising for the wrong thing; one that talks about resolution, friction removal and AHT as an outcome understands the work. Our first-contact resolution guide covers the metric that keeps AHT honest, and the customer service metrics guide puts it in the wider scorecard.
Frequently asked questions
What is a good average handle time?
There is no universal target, because the right handle time depends entirely on the contact type. A password reset and a complex technical diagnosis are not the same unit of work, so comparing their handle times or holding them to one number is meaningless. The useful benchmark is your own: watch AHT by contact type over time, investigate the outliers, and judge it against first-contact resolution. A handle time is good if the contact was resolved and the customer was satisfied, not because it hit a round figure.
Why is it bad to set an AHT target for agents?
Because agents will hit it by rushing customers, skipping steps, transferring resolvable calls and cutting after-call notes — improving the number while making the operation worse, since the calls come back. Handle time counts only the first contact, so a fast handle time that raises repeat contacts costs more, not less. AHT is a real cost driver but it is an outcome, not a target; pressuring it directly optimises for speed at the expense of resolution, which is what loses customers.
How do you actually reduce average handle time?
By removing the friction that inflates it rather than hurrying the agent. Most long calls are long for structural reasons: hunting across systems for information, hold time waiting for another department, a knowledge base that does not answer the question, repeated verification, or a product problem behind the call. Fix those and handle time falls without anyone being rushed. The wasted time was never the agent's to give back, so structural fixes reduce AHT durably where pressure only hides it.
How does handle time affect staffing?
It is one of the two inputs, with volume, that set how many agents a queue needs, so a small change in AHT moves headcount noticeably. Longer handle time means more agents for the same volume, which is why AHT is really a cost lever. A staffing calculator that takes volume and handle time shows the sensitivity directly. It is also why a rushed AHT reduction that raises repeat contacts backfires on staffing — the repeat volume needs more agents than the shorter calls saved.
