AI search, defined.

Answer Engine Optimization (AEO) is the practice of structuring content so that a search or AI system selects it as the direct answer to a question, rather than as one link among many. It focuses on concise, self-contained answers, explicit question-and-answer formatting, and structured data.

Generative Engine Optimization (GEO) is the practice of increasing the likelihood that a large language model cites a given source when generating an answer. It depends heavily on third-party mentions the model retrieves, not only on the brand's own website.

llms.txt is a proposed plain-text file placed at the root of a domain that summarises the site and lists its most important pages for large language models. It is a voluntary convention, not a standard, and no engine is obliged to read it.

AI Overviews are AI-generated summaries that appear above traditional results on Google, synthesising information from multiple web sources with links to those sources. They were previously tested under the name Search Generative Experience (SGE).

AI crawlers are automated agents operated by AI companies that fetch web content either to build training corpora or to retrieve sources at the moment a user asks a question. The two purposes use different user agents and are controlled separately in robots.txt.

A zero-click search is a search that ends without the user visiting any external website, because the result page itself satisfied the query through a featured snippet, knowledge panel, or AI-generated summary.

Citation share is the proportion of tracked questions in which a given brand or domain is cited as a source by an AI engine, relative to the total number of questions tracked or to competitors cited on the same questions.

Structured data is machine-readable markup, usually JSON-LD following the schema.org vocabulary, that states explicitly what a page is about — its type, author, price, questions and answers — instead of leaving those facts to be inferred from prose.

Retrieval-augmented generation (RAG) is a technique where a language model retrieves documents from an external source at the time a question is asked, then writes its answer from those documents rather than from its training data alone. The retrieved documents are what the model can cite.

Semantic search retrieves documents by meaning rather than by exact keyword match, comparing numerical representations called embeddings so that a query and a document can match even when they share no words in common.

Entity SEO is the practice of making a brand, person or product resolvable as a single distinct thing across the web, so that search engines and language models can connect every mention of it to one identity rather than treating each mention as unrelated text.

A knowledge graph is a database of entities and the relationships between them — people, companies, products and places stored as nodes with typed connections — used by search engines and language models to reason about things rather than matching text.

A featured snippet is a block at the top of Google's results that extracts a passage from a single ranking page to answer a query directly, showing the extracted text with a link to its source. It predates AI Overviews and is a different feature.

E-E-A-T stands for Experience, Expertise, Authoritativeness and Trust — the framework in Google's Search Quality Rater Guidelines used to assess content quality. It is a set of criteria human raters apply, not a score any algorithm computes or any tool can report.

A hallucination is output from a language model that is fluent, confident and false — an invented statistic, a misattributed quote, a citation to a source that does not exist, or a claim about a company that is simply untrue.

Share of voice in AI search is the proportion of tracked questions in a category where a brand is named or cited, measured against the total for all brands appearing. It is a relative measure: it moves when competitors gain or lose visibility, not only when the brand does.