Article

GEO: Generative Engine Optimization (the original paper)

Aggarwal, Murahari et al. (Princeton & IIT Delhi) · arxiv.org

The peer-reviewed experiment that coined GEO and proved which content tweaks raise citation visibility 22-41%.

Open on arxiv.org ↗

It answers 2 questions

  1. How do AI assistants actually choose which sources to cite in an answer?
  2. Should I publish comparison pages and 'best X' listicles to win AI recommendations?

Read with it

  1. Large-scale evidence that mentions and citations are different things, and how each is earned. semrush.com Article How AI Search Really Works: Findings from Our AI Visibility Study Semrush
  2. Shows even one assistant cites differently by mode, so you must test prompts, not assume. semrush.com Article Only 25% of cited sources overlap between ChatGPT's reasoning modes Semrush
  3. 3.7M citations analyzed: 91% of cited URLs appear in only one engine, which changes your whole strategy. growth-memo.com Article The Consensus Gap Kevin Indig (Growth Memo)
  4. Video How AI Search Engines Work (Ahrefs AEO Course 1.1) Ahrefs (Sam Oh)
  5. Video How Ranking in Google AI Overviews, ChatGPT, and Perplexity Differ (Ahrefs AEO Course 1.2) Ahrefs (Sam Oh)
  6. A developer's technical teardown of Perplexity's multi-stage pipeline, from candidate pages to inline citations. ziptie.dev Article How Perplexity AI Answers Work: Retrieval, Ranking, and Citation Pipeline ZipTie.dev
eChai Partner Brands The partnership →