Unlocking the promise of AI in personalized marketing: mapping the hidden barriers to adoption

Abstract

Purpose This study aims to examine the interdependent barriers impeding the effective adoption of artificial intelligence (AI) in personalized marketing. By integrating the technology–organization–environment (TOE) framework, the theory of planned behavior (TPB) and institutional theory, it develops a systems-oriented understanding of how technological, organizational, behavioral and ethical factors interact to constrain AI’s strategic potential in marketing contexts. Design/methodology/approach A mixed, two-stage design was used. Ten barriers were identified through a structured literature review and validated via a Delphi process involving 33 domain experts. Subsequently, data from 187 professionals engaged in AI-driven marketing were analyzed using interpretive structural modeling (ISM) and MICMAC analysis to determine hierarchical re lationships, driving–dependence power and systemic dynamics. Findings The ISM hierarchy revealed eight structural levels, positioning infrastructure and cost constraints as root drivers and ROI uncertainty and ethical concerns as dependent outcomes. MICMAC results confirmed data quality, system integration, awareness and infrastructure as high-driving factors, while resistance to change and privacy concerns functioned as volatile linkages. The results demonstrate that technological readiness and AI literacy form the foundation for overcoming downstream governance and performance uncertainties. Research limitations/implications Given its cross-sectional and expert-dependent design, the study captures a temporal snapshot of AI adoption. Longitudinal and cross-sector research could further test the stability and contextual variation of the identified hierarchies, especially under emerging generative AI conditions. Practical implications Managers and policymakers should prioritize interventions targeting high-driving enablers – particularly data quality, infrastructure and workforce capability – while evolving adaptive ethical and regulatory frameworks. Emphasizing these foundational levers can enhance scalability, mitigate ROI risk and build stakeholder trust in AI-enabled personalization. Originality/value This research advances AI adoption and service innovation theory by reconceptualizing adoption barriers as a hierarchically interdependent system rather than discrete determinants. The integrated TOE–TPB–institutional model and ISM–MICMAC framework together illuminate how structural, behavioral and legitimacy factors co-evolve to shape responsible and scalable AI implementation in marketing ecosystems.