Customer and Product Profitability Analytics for Data‑Driven Financial Planning in U.S. Retail: Integrating Basket-Level Transactions and Accounting Fundamentals

Abstract

Retailers operate on razor-thin margins and must leverage data analytics to improve profitability. This study proposes an integrated framework linking customer and product-level profitability analysis with firm-level financial performance for strategic planning. We utilize an open retail transaction dataset of 2,500 U.S. grocery households over two years (the “Complete Journey” data) and public financial statement data (SEC XBRL filings) of a major retail company. We apply descriptive analytics to identify profitable customer segments and products, and we develop machine-learning models (logistic regression, random forests, XGBoost) to predict customer churn and lifetime value. These insights are then connected to corporate financial metrics (e.g. profit margins, return on assets) to evaluate their impact on firm performance. Results show a strong disparity in customer profitability – a small fraction of customers drive a large share of sales (consistent with the Pareto 80/20 rule) – and key products (e.g. staple categories like pasta sauces) contribute disproportionately to revenue. The XGBoost model yields an AUC of ~0.85 in predicting customer churn, enabling targeting of at-risk customers. Scenario analysis indicates that a 5% improvement in customer retention could boost net profit by ~25% (aligning with prior findings that small retention gains can amplify profits significantly). We discuss how these data-driven insights inform financial planning decisions such as budgeting for loyalty programs, product mix optimization, and capital allocation. The study demonstrates the value of integrating granular customer analytics with accounting fundamentals to support strategic financial planning in the retail sector.