AI-Powered Marketing Strategies and Customer Satisfaction among Lazada Users in Malaysia

Abstract

Artificial intelligence (AI) is increasingly embedded in e-commerce marketing through recommendation systems and predictive analytics. This study examines the relationships between personalized recommendations, predictive analytics, and customer satisfaction among Lazada users in Malaysia. Because the original manuscript’s use of the Technology Acceptance Model (TAM) was not directly aligned with the measured constructs, the revised study adopts an adapted Information Systems Success perspective, in which AI-enabled information features are examined as antecedents of user satisfaction. Data were obtained from 381 respondents recruited through the Lazada MY Community using convenience sampling. Multiple regression analysis indicated that predictive analytics was positively and significantly associated with customer satisfaction (β = .985, p < .001), whereas personalized recommendations were not statistically significant (β = −.006, p = .385). The reported model yielded R² = .993; however, this exceptionally high value requires additional diagnostic verification, including multicollinearity, common-method bias, item overlap, residual assumptions and model specification. The findings suggest that predictive analytics may be particularly important for customer satisfaction in the sampled context, while the value of personalized recommendations may depend on the relevance and perceived usefulness of the recommendations. Given the non-probability sample and single-community recruitment, the findings should be interpreted as context-specific rather than representative of Malaysian e-commerce users generally.