Reflexive digital twins and behavioral governance in data-driven systems

Digital twins are increasingly used to model and optimise complex systems, yet their behavioural and institutional implications remain fragmented across disciplines. Here, we develop Reflexive Digital Twin Marketing Theory (RDTMT), conceptualising marketing as a socio-technical learning system that co-evolves with consumers, algorithms, and governance arrangements through continuous feedback. Using an Empirics-First approach, we conduct a PRISMA-ScR–guided scoping review across Web of Science, Scopus, PubMed, and Google Scholar (2010–2025), identifying 742 records and synthesising 55 studies at the intersection of healthcare digital twins, marketing analytics, and ethical AI governance. Co-occurrence mapping (VOS) reveals four convergent clusters—continuous synchronisation, adaptive intelligence, reflexive governance, and societal co-evolution—indicating that “twin-like” systems reshape behaviour by structuring how people are represented, predicted, and acted upon. Building on these regularities, RDTMT introduces three constructs—Digital Reflexivity, Ethical Feedback Elasticity and the Value Coherence Index—to explain when predictive personalisation aligns with stakeholder trust and social legitimacy. This framework reframes marketing not as unidirectional influence but as reflexive governance in data-rich environments, offering testable propositions for responsible behavioural prediction and feedback-driven institutional design.