Augmented cross-selling through explainable AI -- a case from energy retailing

Explainability can improve real-world business outcomes: The study demonstrates that SHAP, an explainable AI method, provided explanations that actually matched the behavior of real customers who made cross-purchases, suggesting that understanding why an AI makes recommendations—not just accepting predictions—matters for practical business applications.

High accuracy alone isn't enough for adoption: While the ML model achieved 86% correctness (AUC) in predicting cross-purchases, the researchers emphasize that complex models "hide the reasoning behind the prediction from their users, which can lead to lower adoption and lack of insight," indicating that businesses need transparency alongside accuracy to trust and implement AI systems.

XAI bridges the gap between technical capability and business reality: The paper addresses a critical gap in AI research by providing "little empirical evidence of the benefits to traditional businesses" before this study, showing that explainable AI isn't just a technical nice-to-have but has measurable value in real commercial settings like energy retailing.

Transparency supports relationship marketing: By making AI recommendations explainable, companies can better understand and communicate with customers about cross-selling opportunities, potentially building trust and improving customer relationships rather than appearing to use opaque algorithmic manipulation.