Enhancing Sales Performance through ARIMA-Based Predictive Modeling: Insights and Applications Model

Abstract

Gociko Snack, a Micro, Small, and Medium Enterprise (MSME), often faces significant challenges in managing its inventory due to the unpredictable nature of market demand. Accurate sales forecasting is crucial for Gociko Snack to optimize stock levels, reduce storage costs, and avoid out-of-stock or overstock situations. Traditional methods of sales prediction are often unable to cope with the dynamic and complex market environment in which Gociko Snack operates. in solving the case This research uses the ARIMA (AutoRegressive Integrated Moving Average) model for forecasting and application modeling using the CodeIgniter framework in a structured Waterfall system development methodology. Through rigorous testing and evaluation, the Mean Absolute Percentage Error (MAPE) was set at 9.18, which shows the effectiveness of the application in predicting sales trends with a high success rate. This research contributes valuable knowledge and practical solutions to empower businesses to navigate and utilize data-driven decision making for long-term success and resilience.

Keywords

• autoregressive integrated moving average, arima, forecasting, waterfall.