The AI search glossary: GEO, AEO and AI Overviews defined

Generative Engine Optimization is the practice of getting a brand named, quoted and linked inside the answers that AI engines write, rather than inside the ten blue links below them.

Answer Engine Optimization is the practice of structuring content so that a search or AI system can lift a direct answer from it and attribute that answer to you.

AI Overviews is the Google feature that places a generated summary with linked sources at the top of the results page, above the traditional links.

Entity SEO is the work of making a brand exist as a distinct, well described thing in the machine readable graphs that search and AI systems rely on, rather than as a string of characters.

An LLM citation is a moment when a large language model names or links your brand inside a generated answer, either as a recommendation or as the source it drew a fact from.

AI share of voice is the percentage of a fixed prompt set where a brand appears in the generated answer, measured across engines and compared against named competitors.

A prompt set is the fixed list of buyer questions used to measure AI visibility, run on the same schedule across the same engines so the results can be compared over time.

Retrieval augmented generation is the pattern where a model searches a live index first, then writes its answer using the retrieved passages, instead of relying only on what it memorised during training.

Schema markup is structured data added to a page in JSON-LD so that machines can read what the page is about without guessing from the prose.

A knowledge graph is a database of entities and the relationships between them, used by search and AI systems to decide what a name refers to and what is true about it.

A featured snippet is the block Google places at the top of results containing an answer extracted verbatim from a page, with a link to that page.

People Also Ask is the expandable list of related questions on a Google results page, each opening to an extracted answer from a ranking page.

The llms.txt file is a plain text file at the root of a site that points AI systems at the pages and data a publisher considers most useful.

A hallucination is a confident statement from an AI model that is not supported by any real source, such as an invented product feature, a wrong price, or a claim about a brand that no one ever published.

Grounding is the practice of tying a generated answer to specific retrieved sources so the claims in it can be traced back to something real.

A zero-click search is a query answered on the results page itself, where the user gets what they needed without visiting any website.

A citation gap is the set of prompts where competitors are named in AI answers and you are not, mapped against the sources the engines used to name them.

Source authority is how much weight a retrieval system gives a particular website when deciding what to quote, based on signals like independence, topical consistency, freshness and how often others rely on it.

E-E-A-T stands for experience, expertise, authoritativeness and trust, the quality framework Google's raters use and the closest published description of what both search and AI systems reward.

A brand entity is the machine readable version of a company: a single identity with attributes, relationships and identifiers that search and AI systems can resolve consistently.

Chunk retrieval is the step where a system splits pages into passages, indexes those passages, and pulls the individual chunks that best match a query rather than whole documents.