Call center utilization forecasting

Posted on Posted in For business

How to calculate how many agents do we need on the shifts?

If we place too many agents, then they will be “idle” and contact center will incur extra costs. Or vice versa, if we place not enough agents on the shifts, agents will be overloaded and will not be able to provide high-quality customer service and, as a result, missed calls will occur.


It is difficult to predict the actions of the customer, his mood and the desire to call the contact center on the hotline.

Many questions arise:

  • When will the customer call?
  • What is the purpose of the call?
  • What preceded the customer’s call?

Many factors affect the number of incoming calls. For example: day of the week (the number of incoming calls between Monday and Friday, weekday and weekend significantly differ), holidays, seasonality, promotions, marketing campaign activities, promotional events, etc.


Due to the above factors, one of the most important roles in the effective operation of the contact center is the prediction of the workload of agents. Forecasting is the tool that allows us to solve this problem.

Forecast is based on historical data.

Forecasts have different horizons: hours, days, weeks and months.

For more detailed forecasting, you can take even 15-minute intervals.


We use the following data to make a forecast:

  • The number of calls for the same hour / day from the last week, month, year.
  • Average duration of 1 call.
  • Average post-processing time of 1 call.
  • Average waiting time of agent’s response.

Next, using Erlang’s formula, we calculate how many calls can be received by the contact center during the forecast billing period. After that, we distribute the received number of calls by days of the week and hours per day.

As a result, we get how many agents are needed on the shifts for every hour, so that customers do not wait for a long connection with the agent, and the agents have the opportunity to rest, take a break and at the same time do not stay idle.


A deviation from the forecast of no more than 5% is considered acceptable. If the percentage of deviation in fact turns out to be greater, it is necessary to analyze in details which factors influenced the increase or decrease in the traffic, and to be sure to adjust the forecast taking into account identified factors.

Accurate forecast is one of the main drivers of the effective operation of the contact center and great resource optimization tool.

Other articles