Answers are grounded in this paper and cite their sources.
The accelerating deployment of artificial intelligence in marketing has deepened a structural paradox: organizations that have gained unprecedented technical capacity to reach and convert audiences remain largely incapable of enabling genuine communicative participation. Bibliometric mapping of 467 peer-reviewed publications (2015–2026) confirms that communication-specific outcome constructs: dialogue quality (5.9%), message co-authorship (3.9%), and audience agency (3.2%) are structurally absent from a field dominated by efficiency and reach metrics. This paper proposes the Communication-Centered Dynamic Participatory Co-creation (CC-DPC) framework as an open innovation architecture that addresses this paradox by theorizing how AI-driven marketing systems can generate communicative participation, not merely transmission throughput. The framework integrates service-dominant logic, dynamic capabilities theory, stakeholder theory, and dialogic communication science into four sequential layers: strategic communication alignment, AI-mediated infrastructure, participation message co-creation, and Ethical Communication Governance—supported by the three interlocking mechanisms of participation, execution, and adaptation of the Dynamic Participatory Co-creation model (DPC). The CC-DPC framework introduces communicative capital as a theoretically bounded construct distinct from social capital and brand equity and formalizes five propositions guiding an empirical research agenda. Based on the core argument of open innovation theory that sustainable competitive advantage is derived from structured access to external knowledge, the framework demonstrates why transmission-optimizing AI systems cannot replicate the inimitable communicative capital generated through participatory co-creation architectures.
No takes yet. Share an insight, caveat, or question.
Pum et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: