Attention-guided deep learning for effective customer loyalty management and multi-criteria decision analysis

In an era where customer satisfaction and loyalty are pivotal for business success, understanding the myriad factors influencing consumer behavior presents a significant challenge. This paper introduces an innovative methodology leveraging deep learning (DL) with an attention mechanism to analyze and evaluate the determinants of product loyalty. The proposed approach focuses on customer ratings and preferences in the mobile data domain, employing long short-term memory (LSTM) layers combined with an advanced attention mechanism. The attention mechanism enhances the model’s ability to focus on critical features, while the LSTM layers effectively capture temporal relationships in sequential data. A key contribution of this research is the incorporation of the halo effect—a cognitive bias that shapes consumer perceptions and decision-making—into the analysis. This inclusion provides a more comprehensive understanding of the intricate and impactful factors driving customer loyalty. To assess the relative importance of these factors, we integrate traditional and novel multi-criteria decision-making (MCDM) techniques, such as the analytic hierarchy process (AHP) and step-wise weight assessment ratio analysis (SWARA). The study validates the efficacy and accuracy of the proposed approach by comparing these conventional methods with our DL-based model. The results highlight the model’s capability to identify and evaluate the key drivers of consumer loyalty, offering businesses actionable insights to refine their marketing strategies and enhance customer engagement. This research provides a robust framework for understanding the dynamics of customer loyalty, equipping organizations with the tools to achieve sustained success in competitive markets.