Global searching is not enabled.
Skip to main content

Topic outline

  • Course Overview

    In this course we learn how customer requests arrive at a call center, as well as to understand fluctuations in arrivals and how to simulate them.

    This leads us to the so-called Poisson distribution. This is crucial for the understanding of many parts of WFM, like forecasting or for safety staffing calculations.
    The Poisson distribution is an important assumption underlying the famous Erlang C formula.

    • Introduction

      In this topic we see where fluctuations in arrivals come from. Please watch the clip and repeat the demo with 100 customers, each having a 0.01 probability of calling. Then submit the resulting Excel sheet.

    • The Poisson distribution

      Restricted Not available unless: The activity Introduction Clip is marked complete

      In this topic we continue the Excel demo and analyse the resulting distribution. Please repeat again the demo and submit the Excel sheet.

    • Working with the Poisson distribution

      Restricted Not available unless: The activity Poisson Demo is marked complete
      Please watch the clip on the Poisson distribution and do the quiz afterwards.
    • Practical implications

      Restricted Not available unless: You achieve a required score in Quiz (Poisson distribution)

      In this topic we discuss the practical implications of the theory presented. Please watch the video and then make the final assignment.

    • Final assignment

      Restricted Not available unless: The activity Implications Clip is marked complete
      Download the Excel sheet. It contains forecasts and actuals for a whole day at quarter level for 20 lines. Consider the following definitions: 
      •  forecast error = FCE = forecast - actual 
      •  percentage error = PE = FCE / actual 
      •  absolute percentage error = APE = absolute value of PE (i.e., without the minus sign)
      •  mean average percentage error = MAPE = average APE over all forecasted periods

      Calculate the APE for the overall call volume and the MAPE for every line separate and all lines aggregated.
      Additionally, which lines have a forecasting error based on the following rule of thumb: 

      • A line has a forecasting error if there are two consecutive periods with an error larger than twice the square root of the forecast
      Using only the data until 09:00, which lines have a forecasting error based on the following rule of thumb 

      • A line has a forecasting error if there are three consecutive periods with an error larger than the square root of the forecast in the same direction.

    • Feedback and Certificate

      Before you move on to the next course, we would like to hear your feedback on this one.

      To obtain the Certificate of Completion your average grade of all topics should be higher than 80%. You can retry every single topic as much as you would like. There is no limit in time or number of tries.