Online Store Web Software Engineering with Sales Forecasting Implementation

Abstract

This research focuses on the development of a web-based online store system with the implementation of a sales forecasting feature to support the business operations of UMKM Fay Brownies. The system was designed using the Extreme Programming (XP) software development method, which emphasizes an iterative, collaborative, and adaptive approach to changing user requirements. Several technology stacks were utilized in the development,
including Next.js, Laravel, Django, and MySQL. The main features of the application include user authentication, shopping cart, ordering, payment, shipping cost calculation, automatic notifications via WhatsApp, and an admin dashboard for managing business aspects such as products and sales. The system is equipped with a sales forecasting module that employs seven methods: ARIMA, Single Exponential Smoothing, Simple Moving Average, Double
Moving Average, Weighted Moving Average, Long Short-Term Memory (LSTM), and Auto Regressive models to predict product sales based on historical data. System evaluation showed that the application successfully meets user needs in conducting transactions easily and securely, while also providing accurate sales forecasts to support decision-making regarding raw material stock and business planning. The results of this study are also expected to serve as a reference for the development of e-commerce applications with forecasting capabilities that can be adopted by other small and medium-sized businesses.