retrieval-augmented generation
Also called: RAG
Retrieval-augmented generation (RAG) is an AI technique where a language model fetches relevant documents from an external source at query time and grounds its answer in that retrieved text, instead of relying only on what it memorized during training.
How RAG works
Two parts do the job: a retriever and a generator. When a question arrives, the retriever searches an external index (a vector database, a search index, or the live web) and pulls back the passages that best match the query. The generator, usually a large language model, reads those passages alongside the question and writes an answer conditioned on them. The 2020 paper that named the technique described this as pairing “parametric memory” (what the model learned in training) with “non-parametric memory” (the documents it looks up on demand).
Why bother? A base model only knows what was in its training data, frozen at a cutoff date. RAG lets it answer with current or private information without retraining, and it can point back to the exact source it used. That source attribution is the practical payoff: the model has fewer chances to invent facts, and a reader can check the claim.
Example flow:
user question -> retriever finds top passages -> LLM reads passages + question -> grounded answer + citationsMost consumer AI answer tools run on this pattern. Google’s AI Overviews, ChatGPT search, and Perplexity fetch pages, summarize them, then cite. Grounding is not the same as being correct (retrieved text can still be misread or misquoted), but it moves an answer closer to the source material and creates the citation slots your content can win.
How it affects your traffic
RAG is the engine behind AI search, and it changes where your traffic comes from. If an answer engine can retrieve, quote, and cite your page, you earn a citation and a referral even when the user never scrolls a traditional blue-link SERP. Pages that are hard to retrieve (thin, unstructured, blocked to AI crawlers, or buried) get skipped, and a competitor's clearer passage gets quoted instead. Our AI search and GEO work makes your content easy for these systems to pull and attribute, so you show up inside the answers, not just under them.
Get AI search / GEO hub that moves the needle
We turn terms like this into ranked pages and qualified pipeline. Start with a free Initial SEO Strategy.