How many agents do you need to answer 500 calls an hour?

The Erlang C floor for 500 calls an hour at a 240-second handle time is 39 agents on the phones, good for 84.2 percent answered within 20 seconds at a 10.8-second average wait. Treat 39 as a floor rather than a target, because it runs agents at 85.5 percent occupancy, and schedule 56 at 30 percent shrinkage to keep it staffed. That works out to 7.8 agents on the phones per 100 calls.

How many erlangs is 500 calls an hour?

At a 240-second average handle time, 500 calls an hour is 33.3 erlangs of offered load (500 times 240 seconds, divided by 3,600), meaning 33.3 hours of talk, hold, and wrap-up work arrive every hour. The 80/20 answer of 39 agents sits above that load because queues need spare capacity to absorb randomly clustered arrivals. Staffing at or barely above the raw erlang load leaves the queue unstable, so the gap between 33.3 and 39 is the price of answering quickly.

Is 85.5 percent occupancy sustainable for agents handling 500 calls an hour?

Not for long. It sits above the 85 percent line most planners treat as the ceiling, and it is the strongest argument for staffing above the Erlang minimum of 39: the 40th agent buys occupancy relief and service level headroom in the same move, and the queue stops depending on nobody ever stepping away.

How fragile is the 39-agent floor at 500 calls an hour?

One agent off the phones takes the service level from 84.2 to 77 percent, and two take it to 68, both well under an 80/20 target at a 240-second handle time. With occupancy already at 85.5 percent there is no elastic left in the group, so a meeting that runs long produces the same hour as deliberate understaffing.

Should I split 500 calls an hour into separate skill-based queues?

Splitting one pooled queue into two separate queues of 250 calls an hour raises the requirement from 39 agents to 21 per queue, 42 in total, so the split costs 3 extra agents before any real-world imbalance between the queues. That 3-agent figure assumes both streams stay perfectly balanced, which rarely holds. If you need skill-based routing, keep cross-skill overflow in place to recover part of the pooling benefit.

The call center staffing ladder

Call center staffing

How many agents do you need for 500 calls an hour?

The Erlang answer is 39 agents on the phones and 56 scheduled, but 85.5 percent occupancy makes 39 a floor to staff above, not a target.

The short answer

Handling 500 calls an hour at a 4-minute average handle time with an 80/20 service level takes 39 agents on the phones; at 30 percent shrinkage you schedule 56.

Calls per hour: 500
Offered load: 33.3 erlangs
Agents on phones: 39
Scheduled agents: 56
Service level hit: 84.2%
Occupancy: 85.5%

Why the math lands there

Five hundred calls an hour at a 240-second average handle time puts 33.3 erlangs of offered load on the queue. Erlang C says 39 agents on the phones clears an 80/20 target, delivering 84.2 percent of calls answered within 20 seconds at an average speed of answer of 10.8 seconds. That is 7.8 agents per 100 calls, a leaner ratio than any smaller queue can manage, because at this volume the random peaks and troughs of arrivals largely cancel out across the pool.

The number to watch is not 39, it is 85.5 percent. That is the occupancy those 39 agents run at, past the point where recovery time between calls disappears, after-call work bleeds into the next conversation, and sick rates start climbing. Teams hold 85.5 percent occupancy for weeks, not quarters. So treat 39 as a floor to deliberately staff above, and treat the 40th and 41st agents as burnout insurance rather than slack.

Getting 39 people on the phones takes 56 on the schedule at 30 percent shrinkage, since breaks, coaching, meetings, and absence claim almost a third of paid time. And 56 is the headcount where this volume starts inviting structural mistakes: someone proposes splitting the pool by skill or product line, and the queueing math quietly charges for it. Two separate queues of 250 calls an hour need 21 agents each, 42 total, three more than the single pooled queue.

1500 calls an hour times 240 seconds of handle time is 33.3 erlangs of offered load: 33.3 hours of phone work arriving every hour.
2Counting up from 34 agents, 39 is the first count where Erlang C clears the 80/20 target: 84.2 percent of calls answered within 20 seconds, with an average wait of 10.8 seconds.
3At 39 agents, occupancy is 85.5 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 39 on the phones means scheduling 56.

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 20s	Average wait	Occupancy
37	68.2%	28.3s	90.1%
38	77.4%	17.1s	87.7%
39the answer	84.2%	10.8s	85.5%
40	89.1%	6.9s	83.3%
41	92.5%	4.4s	81.3%

Occupancy check: at 39agents, 85.5 percent of every logged-in hour is call handling. Above 85 percent, a queue that looks calm on paper is brutal in the chair; plan headcount above the Erlang minimum or the roster will shed people faster than you can hire them.

If your handle time or target differs

Agents on the phones for 500 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	30 agents	31 agents
4 minutesthis page	39 agents	40 agents
5 minutes	48 agents	49 agents
6 minutes	57 agents	58 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

Erlang C models agents as servers that work at a constant rate forever. It has no concept of fatigue, so it will happily return a staffing level that runs people at 85.5 percent occupancy hour after hour. In practice agents defend themselves: wrap-up stretches, unavailable states get manufactured, and average handle time drifts upward. That drift raises the offered load, which pushes occupancy even higher on the same 39 heads. The spiral is invisible in the formula and very visible by week six of the quarter, which is why occupancy management at this volume is a staffing decision, not a coaching topic.

The split-queue penalty of 3 agents is a floor, not the full bill. It assumes the 500 calls divide into two perfectly balanced streams of 250 an hour, which they will not. The moment one queue runs hot while the other sits light, you pay idle time on one side and a blown service level on the other, with no way to move the surplus across the wall. Overflow rules and cross-training claw back part of the pooling benefit, but every routing boundary you draw at this volume is a staffing cost, whether or not anyone budgets it as one.

The marginal numbers explain why staffing above the floor is affordable insurance. Dropping from 39 to 38 agents on the phones cuts the service level from 84.2 to 77 percent, and 37 gives you 68, so a badly stacked break wave or an overrunning huddle across 56 scheduled people shows up on the wallboard within the hour. The 40th agent works the other way: a modest addition against 56 salaries already committed, buying both service level headroom and relief on that 85.5 percent occupancy.

The agent count is the input; the schedule is the work. Turning 56 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 do you need to answer 500 calls an hour?: The Erlang C floor for 500 calls an hour at a 240-second handle time is 39 agents on the phones, good for 84.2 percent answered within 20 seconds at a 10.8-second average wait. Treat 39 as a floor rather than a target, because it runs agents at 85.5 percent occupancy, and schedule 56 at 30 percent shrinkage to keep it staffed. That works out to 7.8 agents on the phones per 100 calls.
How many erlangs is 500 calls an hour?: At a 240-second average handle time, 500 calls an hour is 33.3 erlangs of offered load (500 times 240 seconds, divided by 3,600), meaning 33.3 hours of talk, hold, and wrap-up work arrive every hour. The 80/20 answer of 39 agents sits above that load because queues need spare capacity to absorb randomly clustered arrivals. Staffing at or barely above the raw erlang load leaves the queue unstable, so the gap between 33.3 and 39 is the price of answering quickly.
Is 85.5 percent occupancy sustainable for agents handling 500 calls an hour?: Not for long. It sits above the 85 percent line most planners treat as the ceiling, and it is the strongest argument for staffing above the Erlang minimum of 39: the 40th agent buys occupancy relief and service level headroom in the same move, and the queue stops depending on nobody ever stepping away.
How fragile is the 39-agent floor at 500 calls an hour?: One agent off the phones takes the service level from 84.2 to 77 percent, and two take it to 68, both well under an 80/20 target at a 240-second handle time. With occupancy already at 85.5 percent there is no elastic left in the group, so a meeting that runs long produces the same hour as deliberate understaffing.
Should I split 500 calls an hour into separate skill-based queues?: Splitting one pooled queue into two separate queues of 250 calls an hour raises the requirement from 39 agents to 21 per queue, 42 in total, so the split costs 3 extra agents before any real-world imbalance between the queues. That 3-agent figure assumes both streams stay perfectly balanced, which rarely holds. If you need skill-based routing, keep cross-skill overflow in place to recover part of the pooling benefit.

How many agents do you need for 200 calls an hour?

17 agents on the phones, 25 on the schedule, and exactly one point of margin over the 80/20 target.

17 on phones · 25 scheduled · 80/20

See the math →

How many agents do you need for 1000 calls an hour?

Erlang C says 74 agents on the phones and 106 scheduled, and at this volume that answer is where the planning starts, not where it ends.

74 on phones · 106 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 →