Ask a contact center leader what the biggest lever on cost is, and many will say headcount. But headcount is mostly determined by something else: average handle time (AHT). Small changes in AHT ripple through every interval of every day.
What's in AHT?
AHT is the average total time to complete a contact:
AHT = Talk time + Hold time + After-call work (wrap-up)
All three count toward the agent being unavailable for the next call, which is why all three matter for staffing.
The Math: AHT Is Half of Your Workload
The workload an interval generates is calls multiplied by handle time. For 100 calls in a half-hour interval at 300 seconds, that's 100 × 300 = 30,000 seconds of work, or about 16.7 agents' worth. Change either number and the workload moves proportionally. Then Erlang C adds the extra agents needed to absorb randomness on top.
Here's what required staffing looks like at an 80% in 20 seconds target for that same 100-call interval:
| AHT | Agents needed | Change vs. 5:00 |
|---|---|---|
| 4:00 (240 sec) | 17 | −4 |
| 4:30 (270 sec) | 19 | −2 |
| 5:00 (300 sec) | 21 | — |
| 5:30 (330 sec) | 23 | +2 |
| 6:00 (360 sec) | 25 | +4 |
Thirty seconds of AHT — in either direction — moves the requirement by two agents. That's roughly 10% of the interval's headcount for a 10% change in AHT. Multiply that across a full day and you can see why a handle-time improvement program pays for itself so quickly.
Why AHT Drift Is Dangerous
The flip side is that AHT quietly creeping up is one of the most common causes of "we hit our forecast but missed service level." If your staffing was built on 5:00 AHT and it's actually running 5:30, you are two agents short in this interval, and nothing about call volume looks wrong.
Check actual AHT against your planning AHT weekly. If they diverge, update the plan, not just the report.
Want to see this with real numbers? Open the calculator with a sample call center already loaded, then swap in your own.
Try it with sample data →AHT Varies More Than You Think
Don't use a single AHT for the whole operation if the real numbers differ:
- By time of day. Morning calls may be simple status checks while afternoon calls are more complex.
- By day of week. Mondays often carry longer, more complicated calls from weekend issues.
- By tenure. New hires typically run higher handle times. A big class of new agents will push the average up.
- By contact reason. A billing question and a technical troubleshooting call are not the same workload.
Where AHT Reductions Actually Come From
Pushing agents to talk faster is the least effective approach and often hurts quality and repeat contacts. Better sources:
- Reduce after-call work. Auto-populated notes, shorter wrap-up forms, and disposition codes that take two clicks instead of six.
- Fix the tools. Slow screens and clunky CRM navigation add seconds to every call. Ask agents what slows them down.
- Improve knowledge access. Hold time often comes from agents searching for answers. A good knowledge base cuts it.
- Coach the outliers. Look at the top and bottom of the AHT distribution — the best performers may have habits worth teaching.
- Deflect avoidable contacts. Removing simple repeat calls through self-service can reduce volume but may raise AHT on what's left. Plan for that shift.
Watch the tradeoff: Cutting AHT at the expense of first-contact resolution just converts one long call into two shorter ones. Always track repeat contacts alongside AHT.
The Bottom Line
AHT is the multiplier on every staffing calculation you do. Keep your planning AHT honest, break it out by interval where you can, and test "what if AHT changes by 30 seconds" before it does. For the full calculation walkthrough, see how to calculate call center staffing.