Skip to main content
Retrieval and citation are two different events in AI search, defined in Peec’s Feb 2026 study of over 1 million citations (full study). A retrieval is when a webpage is fetched by the LLM and makes it into the candidate set the model might use. A citation is when that page is actually used and clearly referenced in the AI answer. The relationship between the two is the citation rate — on average, how often a URL is cited per answer. Why the distinction matters:
  • Being retrieved is necessary but not sufficient. On Perplexity, 64% of retrieved URLs are never cited in the answer at all (Feb 2026 data).
  • Retrieval and brand visibility diverge too: in a Jul 2026 analysis, Kimi ranked top-5 as a retrieved source (11.2% retrieval share) but only sixth in brand mentions — see Kimi’s AI visibility content strategy.
  • ChatGPT’s Deep Research agent reads far more pages than it cites — see Being read is not being cited.
For benchmarks on what a good citation rate looks like per engine — ChatGPT 2.0+, Google AI Mode 1.1–1.5, Perplexity 1.5–2.0 — see Citation rate benchmarks by engine. Source: What does a good citation rate look like? Benchmarks from over 1 million AI citations, Peec (Tom Wells), Feb 2026.