Research Output

Forecasting error measures in multiple product situations.

  As a result of qualitative research work carried out in a major brewing company, three main issues
arose. First, there is a difference in the value of a forecast and the way it is used or abused. Second,
a clear indication of the social and political influences on planning and the forecasting process also
emerged. Third, although the importance of monitoring forecast accuracy is recognised, no one seems
to care about it.
A statistical assessment of forecasting performance has been carried out using historical sales data in
order to assess which error measure is most useful to the company. In doing this, period and product
variation in forecasting accuracy was assessed and explained. A variety of monitoring techniques are
then reviewed and suggestions made for the choice of error measure and monitoring procedure. Recommendations
are made as to how to better manage the forecasts, and also, from general observation, how
better housekeeping can improve accuracy

  • Publication Status:


  • Library of Congress:

    HD61 Risk Management

  • Dewey Decimal Classification:

    338 Production


Canduela, J. (2003). Forecasting error measures in multiple product situations



Forecasting; product variation; error measure; monitoring;

Monthly Views:

Available Documents