Retrieval
Retrieval is the step where an AI system searches an index to find the most relevant passages for a query before generating an answer, and it decides which content is even eligible to be cited.
Retrieval is the "find" half of how answer engines work. Before a model writes anything, the system searches its index for passages relevant to the question and pulls back a candidate set. Only content that survives retrieval can be quoted or cited — everything else is invisible to the answer, no matter how good it is.
That makes retrieval the gate every AEO tactic aims at. Being crawlable gets you into the index; being a clear, self-contained, on-topic passage gets you retrieved for the right questions. Modern retrieval blends meaning-based and keyword-based methods and is usually followed by reranking, which trims the candidates down to the few the model actually reads.
Example. For "how long does it take to charge an EV," the retrieval step might gather 40 candidate passages from across the web. If yours isn't among them, you can't be cited — winning retrieval is step one of being the answer.
Relevant pillar
Related terms
- RAG (Retrieval-Augmented Generation)RAG is the technique behind most AI answer engines, where the model first retrieves relevant documents from the live web or an index and then generates an answer grounded in what it found.
- Passage RetrievalPassage retrieval is the practice of finding and returning specific relevant passages from within documents, rather than whole pages, which is why AI engines cite paragraphs instead of articles.
- RerankingReranking is a second pass in retrieval where an initial set of candidate passages is reordered by a more precise relevance model, deciding which few actually make it into the AI's answer.