Retrieval-Augmented Generation (RAG)
RAG is the pattern of retrieving relevant documents first and then having a language model answer using them.
In detail
It exists because models are static and often wrong about specifics. Retrieval supplies current, verifiable text; generation turns it into an answer.
For publishers this is the whole game: if your page is not retrieved, it cannot be used, no matter how good it is.
Example
An assistant fetching three pages about conversion tracking and composing an answer that cites them.
Why it matters
It makes crawlability, clean HTML and self-contained sections directly commercially relevant.
Common mistakes
- Splitting a single answer across pages so no chunk is useful alone.
- Relying on client-side rendering for the substance of the page.
Related questions
Related articles
- How to Get Your Website Found in AI Search Results
A practical sequence for becoming a source AI answers can retrieve, quote and attribute — starting with the things that block you entirely.
Related guides
- The Complete AI SEO GuideGuide
How to stay visible when answers are generated: retrievability, answer-shaped content, entities, structured data, and honest measurement.
Related terms
- Large Language Model (LLM)
A large language model is a generative model trained on very large amounts of text to predict and produce language.
- AI Search
AI search is any search experience where a generative model answers the query directly instead of only listing links.
- Generative Engine Optimization (GEO)
GEO is optimising content so generative answer engines can retrieve, understand, quote and attribute it correctly.
- Schema Markup
Schema markup is structured data added to a page in a machine-readable format that describes what the page contains.
Related questions
- What is GEO (Generative Engine Optimization)?
GEO is the practice of making content easy for generative answer engines to retrieve, quote and attribute. It overlaps heavily with SEO: crawlable pages, clear structure, unambiguous entities and factual precision.