Meta-analytic insights into AI chatbots adoption in the service industry: uncovering dynamics and contextual insights using a meta-SEM approach

Purpose This study comprehensively analyses factors influencing the adoption of artificial intelligence (AI) chatbots in the services industry by employing meta-analytic structural equation modeling (meta-SEM). It provides a comprehensive framework for understanding the complex dynamics of chatbot adoption and offers valuable insights for organizations in the services sector. Design/methodology/approach The meta-analysis draws on 561 correlation coefficients from 56 studies, encompassing 60 samples with an aggregated sample size of 25,511 participants. It employs a multivariate framework to examine the key enablers and inhibitors of AI chatbot effectiveness, leveraging synthesized data to reveal patterns and insights across diverse contexts. Findings Trust and performance expectancy emerge as the most significant factors driving users’ behavioral intention to adopt AI chatbots. Hedonic value, social influence, and effort expectancy also contribute positively, though the impact of effort expectancy is comparatively weaker, whereas anthropomorphism and perceived risk have insignificant effects. Behavioral intention emerges as a strong predictor of AI chatbot usage behavior. Further, moderator analysis reveals that economic status strengthens the influence of performance expectancy and trust in developed markets. In contrast, hedonic value and social influence play a greater role in developing economies, reflecting differences in user priorities. Additionally, low digital adoption levels amplify the effects of effort expectancy and social influence. In contrast, high Internet penetration further boosts trust and performance expectancy, showing a preference for reliable, utility-driven interactions in highly connected regions. Originality/value This study provides a concise and clear overview of the factors influencing AI chatbot adoption and usage in the services industry. Notably, it contributes to the literature by being one of the first meta-analyses to explore the moderating impact of economic status, Internet penetration level, and digital adaptation status on AI chatbot usage, thus adding valuable knowledge to the field of chatbots in the service industry.