How many agents does 100 calls an hour need at a 90/30 service level?

Eleven agents on the phones hit a 90/30 target at 100 calls an hour with a 240 second handle time, actually delivering 94.6 percent within 30 seconds at a 5.1 second average answer. Keeping 11 available takes 16 on the schedule at 30 percent shrinkage.

How much more staff does 90/30 need than 80/20 at this volume?

Exactly one agent on the phones: 80/20 needs 10 and 90/30 needs 11 at 100 calls an hour with a 240 second AHT. After 30 percent shrinkage, that is 15 scheduled versus 16. The upgrade is a single head, not a rebuild of the roster.

Why is the jump from 80/20 to 90/30 so cheap?

Because the last few points of service level are bought from a queue that is already short. At 100 calls an hour the step from 10 to 11 agents carries you past 90/30 to 94.6 percent answered in 30 seconds. At smaller volumes the integer math can make it free: 25 calls an hour needs the same 4 agents for 80/20 and 90/30.

Does a 90/30 target change how hard agents work at 100 calls an hour?

It eases the pace. Eleven agents for 100 calls an hour run at 60.6 percent occupancy, because the same headroom that shortens the queue also adds recovery time between calls. A stricter service level quietly buys a calmer floor, which belongs in the accounting of what the eleventh agent purchases.

When is a 90/30 target worth paying for?

It pays where seconds of waiting convert to outcomes: sales and revenue lines, queues under contractual SLA penalties, and urgent or emergency response lines. At 100 calls an hour the premium is one extra scheduled agent, 16 instead of 15, so the bar for justifying it is low. A general support queue with no penalty for a slightly slower answer often does better keeping 80/20 and improving IVR routing or self-service instead.

The call center staffing ladder

Call center staffing

How many agents do you need for 100 calls an hour at a 90/30 service level?

11 agents on the phones, 16 scheduled at 30 percent shrinkage, and only one more head than the same line at 80/20.

The short answer

Handling 100 calls an hour at a 4-minute average handle time with a 90/30 service level takes 11 agents on the phones; at 30 percent shrinkage you schedule 16.

Calls per hour: 100
Offered load: 6.7 erlangs
Agents on phones: 11
Scheduled agents: 16
Service level hit: 94.6%
Occupancy: 60.6%

Why the math lands there

At 100 calls an hour with a 240 second AHT, Erlang C puts 11 agents on the phones to answer 90 percent of calls within 30 seconds. Eleven actually overshoots, delivering 94.6 percent in 30 seconds with an average speed of answer of 5.1 seconds, because near the target each added agent absorbs a large share of whatever queueing remains. Ten agents clear 80/20 at this volume. The eleventh is the entire price of the stricter target.

Tightening from 80/20 to 90/30 sounds like a budget conversation, and at 100 calls an hour it costs exactly one agent on the phones, 10 to 11, and one scheduled body, 15 to 16 at 30 percent shrinkage. Occupancy lands at 60.6 percent, so the extra agent is not standing around either. The team runs at a sustainable pace while answering nearly everything inside half a minute.

The harder question is whether the line earns it. A revenue line, a contractual SLA, or an urgent-response queue can convert faster answers into money or compliance, and one extra head is cheap insurance. A general support queue often gets more from staying at 80/20 and spending the difference on call deflection or IVR routing. Integer effects cut both ways, too: at 25 calls an hour, 90/30 needs the same 4 agents as 80/20, so at small scale the stricter target costs nothing at all.

1100 calls an hour times 240 seconds of handle time is 6.7 erlangs of offered load: 6.7 hours of phone work arriving every hour.
2Counting up from 7 agents, 11 is the first count where Erlang C clears the 90/30 target: 94.6 percent of calls answered within 30 seconds, with an average wait of 5.1 seconds.
3At 11 agents, occupancy is 60.6 percent: the share of each logged-in hour spent handling calls rather than waiting for the next one.
4Breaks, meetings, training, and absence take agents off the phones for 30 percent of paid time, so keeping 11 on the phones means scheduling 16.

One agent either way

Erlang math is a cliff, not a slope. This is the same hour with one or two agents more or fewer on the phones.

Agents on phones	Answered in 30s	Average wait	Occupancy
9	76.6%	32.2s	74.1%
10	88.5%	12.6s	66.7%
11the answer	94.6%	5.1s	60.6%
12	97.6%	2.1s	55.6%
13	99%	0.8s	51.3%

If your handle time or target differs

Agents on the phones for 100 calls an hour, across handle times and the two most common service level targets. Adjust every input live in our Erlang C calculator.

Average handle time	80/20 target	90/30 target
3 minutes	8 agents	9 agents
4 minutesthis page	10 agents	11 agents
5 minutes	12 agents	13 agents
6 minutes	14 agents	15 agents

Scheduled headcount sits above every number here: this page assumes 30 percent shrinkage, and your own rate comes out of our shrinkage tool.

What Erlang C does not tell you

The one-agent price holds at this exact volume and handle time. It is not a general rule you can carry to another queue, and it describes a single steady hour at exactly 100 calls. Your SLA is probably reported daily or monthly across every interval, including the ones that run over forecast, so hitting 94.6 percent in the modeled hour does not guarantee the reported month lands above 90. The gap between 94.6 and 90 buys some forecast tolerance, but not a bad Monday.

Shrinkage is the bigger threat to this plan, because the entire upgrade is one agent. Any hour that loses one person to a meeting or an overlapping lunch leaves 10 on the phones and silently downgrades the line to roughly the 80/20 service the cheaper roster would have bought. Sixteen scheduled covers the average day, not the badly placed hour, so protect the eleventh seat in the schedule rather than treating it as slack.

Erlang C also assumes the clock is honest. At an average speed of answer of 5.1 seconds, abandonment is a rounding error while you are staffed, so the model's no-abandonment assumption is safer here than on a strained queue. What moves the reported number more at this configuration is measurement definition: whether the 30 seconds starts when the caller enters the queue or after the IVR finishes, and whether short abandons count against you. Settle those definitions before you conclude that one extra agent is all the upgrade requires.

The agent count is the input; the schedule is the work. Turning 16 scheduled agents into shifts that follow your call curve, respect breaks, and survive swaps is what Soon builds automatically, with intraday coverage tracking when the forecast misses.

Frequently asked questions

How many agents does 100 calls an hour need at a 90/30 service level?: Eleven agents on the phones hit a 90/30 target at 100 calls an hour with a 240 second handle time, actually delivering 94.6 percent within 30 seconds at a 5.1 second average answer. Keeping 11 available takes 16 on the schedule at 30 percent shrinkage.
How much more staff does 90/30 need than 80/20 at this volume?: Exactly one agent on the phones: 80/20 needs 10 and 90/30 needs 11 at 100 calls an hour with a 240 second AHT. After 30 percent shrinkage, that is 15 scheduled versus 16. The upgrade is a single head, not a rebuild of the roster.
Why is the jump from 80/20 to 90/30 so cheap?: Because the last few points of service level are bought from a queue that is already short. At 100 calls an hour the step from 10 to 11 agents carries you past 90/30 to 94.6 percent answered in 30 seconds. At smaller volumes the integer math can make it free: 25 calls an hour needs the same 4 agents for 80/20 and 90/30.
Does a 90/30 target change how hard agents work at 100 calls an hour?: It eases the pace. Eleven agents for 100 calls an hour run at 60.6 percent occupancy, because the same headroom that shortens the queue also adds recovery time between calls. A stricter service level quietly buys a calmer floor, which belongs in the accounting of what the eleventh agent purchases.
When is a 90/30 target worth paying for?: It pays where seconds of waiting convert to outcomes: sales and revenue lines, queues under contractual SLA penalties, and urgent or emergency response lines. At 100 calls an hour the premium is one extra scheduled agent, 16 instead of 15, so the bar for justifying it is low. A general support queue with no penalty for a slightly slower answer often does better keeping 80/20 and improving IVR routing or self-service instead.

How many agents do you need for 100 calls an hour?

10 agents on the phones, 15 on the schedule, and a service level cliff hiding one absence away.

10 on phones · 15 scheduled · 80/20

See the math →

How many agents do you need for 100 calls an hour with a 6-minute handle time?

At this volume, every 30 seconds of average handle time is roughly one more agent on the phones.

14 on phones · 20 scheduled · 80/20

See the math →

How many agents do you need for 25 calls an hour?

4 agents on the phones, 6 on the schedule, and an occupancy number you will need to defend.

4 on phones · 6 scheduled · 80/20

See the math →