Response to "Generative Engine Optimization (GEO)" (arXiv:2311.09735): Toward Auditable Metrics for AI Search Visibility
Year: 2025
Type: article
Abstract: This paper responds to Generative Engine Optimization (GEO) (arXiv:2311.09735), which proposes methods to increase citation prominence of web content inside large language model (LLM) outputs. While GEO demonstrates that content adaptations can influence generative answers, its evaluation relies on opaque, LLM-scored metrics and equates citation prominence with visibility. The paper identifies three limitations: lack of auditability, narrow scope of “visibility,” and absence of temporal or cross-model analysis. As an alternative, we propose the Prompt-Space Occupancy Score (PSOS™) under the AIVO Standard™, a governance-ready KPI that measures brand visibility as a function of breadth, persistence, decay adjustment, and cross-model consistency. PSOS is designed for transparency, replication, and adoption in enterprise governance contexts. This response situates GEO as an important early research step, while outlining the requirements for auditable, compliance-ready metrics to guide brand strategy and board-level oversight in the era of AI search.
Author: de Rosen, Timothy
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7078894325
FWCI:
Cites:
Cited by:
Related to: