From Persuasion to Computation: Reconceptualizing Marketing in Agent-Mediated Markets
Abstract
For over 125 years, the marketing funnel—rooted in the AIDA model (Attention, Interest, Desire, Action)—has served as the dominant paradigm for understanding consumer purchase behavior. This conceptual framework, developed during the era of mass advertising, assumes a human consumer progressing through sequential psychological stages from product awareness to final purchase. We argue that the emergence of agentic commerce—economic transactions mediated by autonomous AI agents acting on behalf of consumers—fundamentally invalidates this paradigm. AI agents possess no psychological journey; they cannot be made “aware” through advertising, develop “interest” via storytelling, or feel “desire” through emotional branding. Instead, these algorithmic intermediaries execute logical, deterministic processes that bear no resemblance to human decision-making.
This paper introduces an alternative theoretical framework: the API Call Model. Rather than psychological persuasion, agent-mediated commerce operates through three computational phases: Query (structured data requests to multiple sources), Analysis (systematic attribute comparison against objective functions), and Execution (automated transaction completion). We develop seven propositions examining how this architectural shift transforms marketing from a persuasion discipline centered on human psychology to an engineering discipline centered on machine-readable data infrastructure. Our framework challenges fundamental assumptions in marketing theory regarding the nature of consumer decision-making, the role of brands, and the mechanisms of competitive advantage in increasingly algorithm-mediated markets.
Keywords: agentic commerce, AI agents, marketing funnel, AIDA model, consumer behavior, algorithmic intermediation, agent-mediated markets, marketing theory