Skip to content
Omnikalc

Erlang C calculator: how many agents does your call center need?

Enter calls per hour, handle time, agents and a target answer time to get the service level, the chance a caller waits, the average wait and the occupancy.

Updated on
Your inputs

Value out of range

seconds

Value out of range

Value out of range

seconds

Value out of range

%

Value out of range

Results

Service level (answered within the target)

65.4%

Below the usual 80/20 target

Average speed of answer

0:49

Traffic intensity
8.00 Erlangs
Agent occupancy
80.0%
Probability a call has to wait
40.9%
Agents to schedule, including shrinkage
14.3

Call centers staff to a service target, and the Erlang C formula links the number of agents to the wait. This Erlang C calculator converts your call volume and handle time into traffic, then shows the probability that a caller waits, the service level for your target answer time, the average speed of answer and the agent occupancy for the number of agents you enter, including shrinkage.

How the calculation works

  • Traffic intensity (in Erlangs) = calls per hour × average handle time ÷ 3,600.
  • Erlang B gives the chance that all agents are busy, and Erlang C converts it into the probability that a call waits.
  • Service level = 1 − Erlang C × e^(−(agents − traffic) × target time ÷ handle time).
  • Average speed of answer = Erlang C × handle time ÷ (agents − traffic).
  • Agents to schedule = agents on the phones ÷ (1 − shrinkage).

Example

A queue receives 120 calls per hour with a 240 second handle time, which is 8 Erlangs of traffic. With 10 agents, occupancy is 80 %, 41 % of calls wait, and only 65 % are answered within 20 seconds, with an average speed of answer of about 49 seconds. With 30 % shrinkage you must schedule about 14.3 agents to have 10 on the phones.

Why it matters

Understaffing makes callers wait and leave, while overstaffing wastes salary. The relationship is not linear: one extra agent at a busy time can lift the service level by many points, while the same agent in a quiet hour adds almost nothing. Testing staffing levels before you write the schedule shows where the money does the most good.

Good to know

Erlang C assumes calls arrive randomly, handle times are similar and no one hangs up while waiting, so use it as a planning guide. Work interval by interval, because a day-long average hides the peaks. Occupancy above about 85 percent leaves agents with little rest and tends to raise handle time. Add agents until the service level meets your goal, then add shrinkage for the schedule. Chat and email follow different rules, since agents can handle several contacts at once.

The queue wait time calculator covers simple queues, the server capacity calculator sizes web servers, and the team capacity calculator plans sprint hours.

Frequently asked questions

What is Erlang C?

A queueing formula that gives the chance a call must wait and the average wait in a call center where callers queue until an agent is free.

What service level should I aim for?

The classic target is 80 percent of calls answered within 20 seconds, but the right target depends on your customers and your budget.

What is shrinkage?

The share of paid time agents are unavailable because of breaks, meetings, training, vacation and absence. Staff above the number you need on the phones.

Does Erlang C account for abandoned calls?

No. It assumes callers wait forever, so real queues with abandonment perform slightly better than the formula predicts.

Was this calculator helpful?

Thanks for your feedback!

Results are estimates provided for information purposes. They do not replace professional advice.