Abstract:
This thesis aims to study the relationship of factors which are expected to impact promotional sales and this will lead to develop a suitable forecasting model for each product category. The result of this research will improve the accuracy of forecasting sales so that the business can manage inventory level effectively. The case study of a retailer who sales variety of products through different type of branches such are Department Store, Extra Store and Hypermarket. This study focuses on 4 product categories of General Merchandise consist of Pillow and Bolster, Bed Set, Container and Storage and Hanger, covering March 2016 to February 2018. Multiple Linear Regression Analysis was applied in this research. The results show that there were different factors which impact promotional sales in each product category thus the business cannot consider the same factors to forecast all categories. The result of the relationship analysis by Multiple Linear Regression Analysis found that there were different forecasting model in each product group when it was sold in different store format which can be developed 12 models. After that, the models were applied in order to verify forecasting accuracy by comparing with traditional forecast. It was found that applied the models can reduce forecasting error which Hanger (Plastic) has the highest error reduction which accounted for 36.44% in average because this category got the highest Adjusted R Square (91%) from Multiple Linear Regression Analysis which means that the selected factors into these models can explain the change in promotional sales by 91% in average.