DESIGN OF A WEB-BASED SALES DECISION SUPPORT SYSTEM USINGTHE DOUBLE EXPONENTIAL SMOOTHING METHOD (CASE STUDY:UMKM TAPE MAYANG MADU JEMBER)

Abstract

Small and Medium Enterprises (UMKM) producing Tape Mayang Madu in Jember face the challenge of fluctuating demand, which often causes imbalances between production and the market. Manual planning based on experience risks causing excess or insufficient stock, thereby increasing costs or losing sales opportunities. To address this issue, this study designed a web-based decision support system using the Double Exponential Smoothing (DES) method to improve sales forecasting accuracy. The system was developed using the waterfall model of the Software Development Life Cycle (SDLC) through the planning, analysis, design, implementation, testing, and deployment stages. The website, named PrakiraJual, was built using HTML, CSS, and JavaScript for the interface, and PHP and MySQL for the backend and database. Key features include data input, updates, deletion, and filtering of sales data, sales and raw material price forecasting, conversion of forecast results into raw material requirements, and production cost calculations. Blackbox testing results confirmed that all functions operate properly, while UAT demonstrated that the system is user-friendly and meets expectations. A MAPE value of 0.35% confirms the system's high accuracy. In conclusion, PrakiraJual can assist UMKM in accurately planning production, reducing the risk of losses, and has the potential for further development with production time prediction features and financial data integration.