Forecasting is used to predict or describe what will happen, and the use of such forecasts in planning
would help in making a good decision about the most attractive alternatives for the company. The
forecasting tasks is even more important in developing a new product since the actual sales performance
of this new product has not been known, while many decisions should be made to manage the product
into a desired growth. This situation is faced by Unilever as it has just started developing new products
which have a short life span characteristic. This characteristic is a new experience for the company to
cope with. The excess stock tolerance of this type of product is smaller than of the non-perishable
products, while out of stock condition has never been an option in the company’s policy for any type of
its product. This leads the company to consider having a ‘good’ forecasting. Therefore, this research is
conducted as an attempt to seek what forecasting techniques are appropriate to support decisions
making in the operational level of supply chain when dealing with short life span products to avoid
undesired stock conditions.
To answer this question, literature study of forecasting and information gathering about the current
forecasting practices are done during the research. The literature study provides knowledge about the
forecasting method ranges and their applications. Whilst, studying the current forecasting practices gives
knowledge about what have been done in the company and the company’s expectations with respect to
this matter. The company’s expectations are related to the general objectives of the forecasting process
which is described through an objective tree approach and the solutions for the company’s specific
forecasting problems which are identified through a system diagram and causal diagram approaches.
The main general objective of the forecasting process is determined in the objective tree, which is to
have a good forecasting. The next is to perform further decomposition of the main general objective into
the sub objectives that contributing to it in order to determine the forecasting specification. This
functional specification results in a fact that forecast accuracy is not the only important parameter.
Practical factors are also expected to be considered in selecting the forecasting techniques, for examples
the easiness of using the method and the easiness of understanding the model. It creates a possible
trade off between forecast accuracy and ‘simplicity’ of the method/model.