What Is Retrieval-Augmented Generation (RAG)? Definition for Marketers
Definition: Retrieval-augmented generation (RAG) is a technique in which an AI system first retrieves relevant text from a search index or knowledge base, then uses it as context to write its answer, instead of relying only on what the model learned in training.
How it works
When a question arrives, the system searches for relevant passages and places them in the model's context alongside the question. The model then writes an answer grounded in that retrieved text. It is the architecture behind most AI search products that use live search, such as Perplexity, Google AI Overviews and ChatGPT's search feature, although each vendor's exact implementation is not public.
Why it matters for brands
RAG is why a page published after a model's training cutoff can still be cited. The relevant question is not whether the model knows about your site but whether the retrieval step can find, read and select it, which depends on crawlability and indexing rather than on training data.
What to do next
Make key pages crawlable in plain HTML, keep them indexable, and structure content in clear, quotable chunks. See llms.txt and chunk-formatted content.
Frequently asked questions
What does RAG mean in AI search?
Retrieval-augmented generation: the system retrieves relevant passages first and then generates an answer grounded in them, rather than answering only from its training.
Can my new website appear in AI answers if the model was trained before it existed?
Yes, for systems that use live retrieval. They can fetch and cite pages published after training, provided those pages can be crawled and indexed.
Related terms
Want this done for your brand?
Start with a free audit: a written scope and a baseline of how AI assistants describe your brand today.
Get my free audit →