From Citation Selection to Citation Absorption: A Measurement Framework for Generative Engine Optimization Across AI Search Platforms

A new arXiv preprint from researchers Zhang Kai and Yao Jingang introduces a measurement framework for Generative Engine Optimization (GEO) that goes beyond simple citation counts. The study analyzes the geo-citation-lab dataset covering 602 controlled prompts across ChatGPT, Google AI Overview/Gemini, and Perplexity, with 21,143 search-layer citations and 23,745 citation-level feature records. The key finding: citation breadth and depth diverge significantly across platforms.

Perplexity and Google cite more sources on average, but ChatGPT shows substantially higher average citation influence among fetched pages. High-influence pages tend to be longer, more structured, semantically aligned, and rich in extractable evidence such as definitions, numerical facts, comparisons, and procedural steps. The authors argue that GEO should measure answer-level absorption as a separate outcome from citation counts, providing a more nuanced understanding of how AI search engines actually use content.