Demand forecast adjustments in the Sri Lankan FMCG industry : a field experiment

Abstract

In FMCG industries, demand forecasting is a vital element of the planning process in supply chain. The most widely used approach of forecasting demand is to generate preliminary forecasts using historical data, trends, growth and seasonal patterns in these businesses. Afterwards, company's demand planners make subjective adjustments for these forecasts, which account for unpredictable events anticipated throughout the planning horizon. Applying these adjustments require substantial human hours and work. However, do they enhance accuracy, what types of adjustments can be found? and what forms of adjustments more productive than others? To study this, we gathered data from FMCG company using field experiment and outcomes were analyzed by statistical software. The findings showed that judgmental adjustments improved accuracy in generally. Nevertheless, a thorough examination showed that although the larger adjustments often resulted in lower average accuracy and the smaller adjustments frequently increase accuracy. Furthermore, there is no correlation of accuracy among negative adjustments and positive adjustments, which involved for forecast upward direction. However, forecasters tend to adjust forecasts in an upward direction (positive adjustments) rather than downward direction (negative adjustments) as per the data found from our examination. This research study examines strategies designed to increase the effectiveness of judgmental adjustments, involving statistical analysis, contextual factor observation and forecasting method evaluation.