Authors: Devansh Indrodiya, Shivangi Patel
Abstract: The integration of Large Language Models (LLMs) into modern search engines has significantly transformed digital discoverability, shifting search behavior from deterministic webpage ranking to probabilistic entity citation within AI-generated responses. Unlike traditional search engines that present ordered lists of hyperlinks, generative search systems synthesize contextual answers and selectively cite businesses based on semantic relevance, trust signals, review sentiment, and inferred user intent. This transformation challenges conventional Search Engine Optimization (SEO) strategies that were originally designed to optimize positional ranking rather than inclusion within generative responses. This paper introduces Generative Engine Optimization (GEO), a geospatial artificial intelligence framework designed to model, measure, and improve business visibility in generative search environments. The proposed framework integrates geospatial analysis, semantic entity recognition, and machine learning–based prediction models to evaluate discoverability within AI-generated responses. A monitoring system called GeoRank360 is developed to track business citations across multiple generative platforms and compute a unified metric termed the Generative Visibility Score (GVS), which incorporates citation frequency, semantic prominence, sentiment strength, entity consistency, and temporal stability. An empirical evaluation conducted across 100 local businesses, five generative search platforms, 500 query variations, and over 4,000 geo-grid coordinates reveals spatial visibility volatility ranging from 35% to 60%, substantially higher than fluctuations observed in traditional search rankings. Predictive modeling achieves up to 87.1% accuracy in forecasting generative citation outcomes. The results indicate that semantic relevance exerts greater influence than geographic proximity in determining visibility within generative search responses. The proposed GEO framework establishes a foundation for future research in generative search visibility modeling, semantic ranking analysis, and AI-driven local discovery systems.
