The purpose of this thesis is to analyze the effectiveness of a consumer packaged goods (CPG) company’s current demand forecasting practices and provide recommendations for future improvement. Using three years of bias and mean absolute percent error (MAPE) data for two prominent compromised skin brands, overarching trends between lag and forecast error were be identified. A quick cost analysis revealed the cost of over and under forecasting in terms of holding cost and the cost of lost revenue and how this relates to lag. This thesis concluded that although there is a correlation between MAPE and the number of months prior to the shipment the forecast was made (lag), there is no correlation between lowest MAPE and lowest total cost.
