Ai-Driven Personalization And Its Influence On Customer Loyalty In Online Retail

Abstract

The proliferation of artificial intelligence (AI) in online retail has fundamentally transformed the manner in which 
businesses interact with customers, primarily through AI-driven personalisation engines that dynamically tailor 
product recommendations, pricing, content, and communications to individual consumer preferences. This study 
empirically examines the influence of AI-driven personalisation on customer loyalty in the context of Coimbatore City 
one of Tamil Nadu's fastest-growing urban retail hubs. Grounded in the Technology Acceptance Model (TAM), 
Expectation Confirmation Theory (ECT), and the S-O-R (Stimulus–Organism–Response) framework, this research 
adopts a quantitative positivist approach. Primary data were collected from 1,270 active online shoppers in 
Coimbatore using a structured questionnaire administered through stratified random sampling. The measurement 
model was validated using Confirmatory Factor Analysis (CFA) in IBM AMOS 26, while the hypothesised structural 
relationships were tested through Structural Equation Modelling (SEM). Supplementary analyses including 
Exploratory Factor Analysis (EFA), multiple regression, Pearson correlation, and reliability diagnostics were 
conducted using IBM SPSS 27 and Python 3.11 (scikit-learn, pandas, seaborn, stats models). The results confirm that 
AI-driven personalisation significantly and positively influences customer trust (β = 0.673, p < 0.001), perceived 
value (β = 0.581, p < 0.001), and customer satisfaction (β = 0.612, p < 0.001), which in turn collectively explain 
71.4% of the variance in customer loyalty (R² = 0.714). The SEM model demonstrates an acceptable goodness-of-fit 
(CFI = 0.952, TLI = 0.947, RMSEA = 0.048, SRMR = 0.052). Privacy concerns emerged as a significant moderating 
variable, partially attenuating the personalisation–loyalty relationship. Practical implications for e-retailers 
operating in Tier-II Indian cities are discussed alongside recommendations for AI ethics governance.

DOI: https://doi-ds.org/doilink/07.2026-74599521