Sales Forecasting Using Statistical Methods: A Case Study of Papee Pork Sausage, Mueang District, Lampang Province

Abstract

This research aims to examine the information, production format and product sales of the Papee Pork Sausage factory in order to apply statistical methods to forecast sales and to reduce the problem of mismatched production and sales volumes that do not align with customer demand in each retail store. The study utilized sales data from selected retail stores of the Papee Pork Sausage factory, obtained through purposive sampling. Time-series forecasting methods including (1) moving average method, (2) weighted moving average method and (3) exponential smoothing method were applied. The error value from the forecast was assessed using the 3 methods: (1) Mean Absolute Deviation (MAD), (2) Mean Square Error (MSE) and (3) Mean Absolute Percent Error (MAPE). Microsoft Excel was used to develop the forecasting models. The results indicated that the four-week moving average forecasting method (n=4) estimated the required production quantity at 324 units, with the lowest average forecasting error of 23%. Consequently, the production quantity for the sample product was determined to be 324 units, which yields a production level closely aligned with actual sales. The results of the comparison between the statistical forecast models and the actual sales data from retail stores showed that Mean Absolute Percentage Error (MAPE) of the forecasted sales or required production quantity was 1.09%. This allows the business to plan production according to customer demand, reduce the volume of expired products returned to the factory, and decrease product costs from 33,175 to 29,456 baht per week. The resulting reduction in production costs contributes to increased profitability for the factory.