Abstract:

Demand forecasting is an essential process in the routine operations of retail organizations. Ineffective demand forecasting leads directly to overstocking, customer dissatisfaction, and poor performance within the retail sector. Methods to accurately forecast are a daily challenge for retail organizations and manufacturers. Recently, Prophet has become an area of interest and is used to provide more accurate forecasts and identify hidden patterns in data. The primary subject matter of this study is to develop and propose a Prophet model for more accessible item demand forecasting based on sales, time, products, trends, seasonality, holidays, and other essential data. Two performance metrics, Mean Absolute Percentage Error (MAPE) and Mean Squared Error (MSE), have also been implemented to test the model's performance. The results obtained from our experiment show that Prophet is an excellent tool for predicting the demand quantity of items in retail chain organizations, with 3.454% in MAPE. Furthermore, based on the literature review, several conclusions and recommendations are suggested for retail chain organizations encountering these challenges during their operational processes.