When
Where
Daniel Zantedeschi, Assistant Professor, School of Information Systems, Muma College of Business, University of South Florida
Strategic Visibility in AI-Mediated Search: From Retrieval Incentives to Influence Governance
Abstract: AI-mediated search changes how organizations compete for visibility. In retrieval-augmented generation, a small set of selected documents supplies the context for an AI answer. Producers therefore have an incentive to adapt their content to gain inclusion, while platforms have to weigh both the visibility they allocate and the evidence used in downstream decisions.
This talk presents two complementary studies. The first develops an economic model of retrieval governance in which producers invest in strategic positioning, and the platform chooses how strongly scores determine inclusion. We bound the returns to positioning, characterize equilibrium investment and platform policy, and illustrate the trade-off between maintaining a retriever’s ranking and limiting the survival of an optimized passage. The second, From SEO to REO, introduces Retrieval Engine Optimization and the RAG Visibility Governance framework. Across five retrieval architectures and two business domains, it analyzes how concrete content adaptations change inclusion and how those gains transfer across models. A generation-stage intervention reduces observed mock-brand influence while keeping answer relevance comparable, at the cost of lower faithfulness to the retrieved evidence. Together, the studies connect platform design to organizational visibility and to the governance of AI-generated answers.
Bio: Daniel Zantedeschi is an Assistant Professor in the School of Information Systems at the University of South Florida, Muma College of Business. He studies how AI systems shape information flows and the decisions organizations make with them. As models increasingly decide what gets surfaced, his work asks what incentives that creates for the people who produce the content.
His current research centers on strategic visibility in AI-mediated search: how organizations optimize for retrieval rather than ranking, and how platforms can govern that influence without distorting it. He pairs methodological developments with controlled and field experiments to propose evaluation criteria and design principles that editors, platform operators, and regulators can apply. A related line, on how AI should enter peer review, recently appeared in Management Science as a Discussion Paper.
Trained first as an engineer in Milan and then as a statistician in the United States, he earned his PhD at the University of Texas at Austin’s McCombs School of Business and was a Mars Postdoctoral Fellow at Wharton, where he worked with corporate partners on advertising measurement. That early work on attribution, identifying which channel actually caused a sale when the data arrive incomplete and out of order, drew him to how system design shapes information quality. He has also taught at Ohio State’s Fisher College of Business.
Outside the academic commitments, he serves as treasurer of the USTA Florida Foundation, which funds youth and community tennis courts and programs statewide.