This diploma thesis addresses the problem of regular sales forecasting adjusted by the effect of cannibalization.
Sales forecasting is essential for companies because they need to always have products in stock, and have limited storage space in the stores.
Having one product on sale affects the sales of other products. Because of this we need to adjust sales forecasts which don't take into account the effect of cannibalization on sales.
In this diploma thesis the moving average method adjusted by the effects of cannibalization will be used for forecasting. This method will be used because the moving average method is already used by Mercator, it's simple and it gives reasonably accurate forecasts.
The sales data of Mercator and Walmart will be used for forecasting one year sales based on the sales data of previous years.
Finally we will evaluate the accuracy of the forecast together, first for all data and then by categories. In the end we will analyze if the forecast with cannibalization is more accurate than without it.
|
