Retrieval-Augmented Generation (RAG)
RAG, Retrieval-Augmented Generation (RAG) is a technique where an AI model retrieves relevant documents from the live web or a database and uses them to ground its answer, instead of relying only on what it memorized during training.
RAG is why an assistant can name a business it was never explicitly trained on and cite a source for it. When you ask 'best dentist in Austin', a RAG-enabled assistant fetches current pages, directories, reviews, local lists, and composes its answer from them.
For businesses this is the mechanism you can influence. Training data is fixed and opaque, but the documents a model retrieves at answer time are the live web. Being present and well-described on the pages it retrieves is how you get pulled into a RAG answer.
Key points
- →The model fetches live sources, then writes the answer from them.
- →Explains citations and up-to-date, location-specific recommendations.
- →The retrieved pages are the surface you can actually optimize.
Related terms
See your own AI visibility
Run a free scan and see how ChatGPT, Claude, Gemini & Perplexity describe your business, with a plan to get named more often.
Run my free scan →