Impact of AI-Driven Personalization on Retail Banking Customer Experience

Thesis Abstract

The rapid digital transformation of retail banking has heightened customer expectations for personalized experiences, yet banks' deployment of AI-driven personalization remains uneven, raising concerns about effectiveness, privacy, and equitable service delivery. This study addresses the problem of how AI-driven personalization influences customer experience in retail banking and whether personalization translates into improved satisfaction, trust, perceived value, and loyalty while maintaining ethical considerations around data use. The aim is to examine the impact of AI-based personalization on customer experience in retail banking, with specific objectives to (1) assess the relationship between AI-driven personalization and customer satisfaction, (2) evaluate how personalization affects trust and perceived privacy risk, (3) analyze the impact on perceived value and loyalty intentions, (4) compare differences across customer segments (demographics and digital maturity), and (5) explore moderating roles of perceived data governance and transparency. The study adopts a mixed-methods, sequential explanatory design grounded in Technology Acceptance Theory (TAM) and the Expectation-Disconfirmation Theory to elucidate how AI-driven personalization shapes perceived usefulness, ease of use, and satisfaction, and how these in turn influence loyalty behaviors. A quantitative phase collects data from 600 customers of three major retail banks in a metropolitan region, selected via stratified random sampling to ensure representation by age, income, and digital banking usage. An instrument comprising validated scales for AI personalization perception, user satisfaction, trust, perceived privacy risk, perceived value, and loyalty intentions, will be administered through an online survey with a response rate target of 28–32%. Reliability will be assessed using Cronbach’s alpha and composite reliability, with construct validity examined via confirmatory factor analysis. Structural equation modeling (SEM) will test the hypothesized relationships, including mediation effects of satisfaction and trust, and moderation effects of data governance transparency and privacy controls. The qualitative phase will conduct 20 in-depth interviews with a purposive sub-sample of survey respondents to explore nuanced experiences, including perceived fairness, algorithmic explainability, and control over personalization settings. Thematic analysis will identify patterns related to ethical considerations, trust development, and perceived value. Expected findings include positive associations between AI-driven personalization and customer satisfaction, trust, and perceived value, leading to stronger loyalty intentions. It is anticipated that privacy concerns and perceived data governance practices will moderate the strength of these relationships, with higher transparency and opt-out capabilities mitigating perceived privacy risk and enhancing acceptance of personalization. The study also foresees differential effects across demographic and digital maturity segments; younger, digitally native customers may exhibit higher responsiveness to personalization, while older or privacy-conscious segments may require stronger governance and clear explanations of data use. The conceptual model will likely reveal partial mediation by satisfaction and trust, with full mediation contingent on perceived transparency and control. The study contributes to knowledge by integrating TAM and Expectation-Disconfirmation Theory within a micro-foundations perspective on AI personalization, offering a nuanced understanding of how algorithm-driven experiences shape consumer behavior in retail banking. It advances measurement by validating constructs specific to AI personalization in financial services and contributes to policy discourse on data governance, transparency, and customer control. Practically, the research provides bank managers with actionable insights on optimizing personalization strategies, aligning AI capabilities with customer expectations, and implementing governance frameworks that enhance trust and loyalty without compromising privacy. The main conclusion is that AI-driven personalization can enhance customer experience and loyalty in retail banking when combined with transparent data practices, robust governance mechanisms, and customer empowerment features. Recommendations include designing explicit opt-in/opt-out controls, implementing explainable-AI interfaces to clarify why certain recommendations are offered, embedding privacy-by-design principles in personalization pipelines, and communicating governance policies clearly to customers. Future research could extend the model to cross-country contexts, examine long-term behavioral outcomes, and explore normative ethical frameworks governing AI personalization in financial services.