Retrieval-Augmented Generation (RAG) is the pattern behind most AI answers that cite live sources. When a user asks a question, the system first retrieves relevant passages — from the web or a knowledge base — then generates an answer grounded in those passages, often with citations. Perplexity and Google AI Overviews are consumer-facing examples.
RAG is why GEO works: if your content is retrievable, well-structured, and unambiguous, it can be pulled into the grounding set and quoted, even for topics the base model knows little about. If it is thin, contradictory, or hard to parse, it is passed over in favor of sources the system trusts more.
The practical implication for a provider is to write extractable, well-sourced answers to the exact questions your audience asks — content a retrieval step can find and a generation step can safely quote without hallucinating.