Data-Driven Retail Strategy: Insights from Market Basket Analysis, Customer Segmentation, and Demand Forecasting

Abstract

The retail industry, particularly small and medium- sized businesses (SMBs), faces significant challenges in leveraging transactional data to optimize inventory management, customer engagement, and sales forecasting. Existing tools often lack integration between analytical techniques, resulting in fragmented insights. This paper introduces a comprehensive retail analytics system that unifies Market Basket Analysis (MBA), customer segmentation, and machine learning– driven sales forecasting into a single platform. The system employs the Apriori and FP-Growth algorithms to identify product associations, RFM (Recency, Frequency, Monetary) analysis with K- Means clustering for customer segmentation, and XGBoost for demand prediction. A Streamlitbased dashboard translates these insights into actionable visualizations for non-technical users. Experimental results demonstrate robust performance: XGBoost achieves an R² score of 0.89 in sales forecasting, K- Means clustering yields distinct customer segments (silhouette score > 0.6), and FP-Growth generates high-lift product association rules (e.g., {Bread → Butter} with lift=1.45). By bridging the gap between advanced analytics and practical usability, Shoplytics empowers retailers to make data-driven decisions efficiently