The GEO glossary

A precise, quotable reference for the vocabulary of generative engine optimization. Each definition is self-contained and written to be lifted directly, by a reader or by a model. The terms span the discipline itself, the systems it acts on, the structures that make a brand legible to machines, and the metrics that prove the work.

Generative Engine Optimization (GEO)

The discipline of influencing how generative AI systems describe, cite, and recommend a brand. GEO optimizes the sources, structure, and signals that large language models retrieve and trust, so a brand appears in the synthesized answer rather than only in a list of links.

Answer Engine Optimization (AEO)

The practice of structuring content so answer engines can extract a direct, citable response to a question. AEO favors explicit definitions, question-and-answer formatting, and machine-readable structured data, and it is a component of GEO focused on extractability.

Search Engine Optimization (SEO)

The discipline of improving a page's position in a ranked list of search results. SEO competes for clicks on links, where GEO competes for inclusion in the single answer a generative engine returns.

Large Language Model (LLM)

A neural network trained on large volumes of text to predict and generate language. LLMs such as those behind ChatGPT, Claude, and Gemini produce the synthesized answers that GEO aims to influence.

Retrieval-Augmented Generation (RAG)

A technique in which a model fetches relevant documents at query time and uses them to generate a grounded answer. RAG makes the choice of which sources get retrieved a primary lever of AI visibility.

Grounding

The act of tying a model's output to specific, verifiable sources rather than to unsupported recall. Grounding reduces hallucination and determines which brands a model can cite with confidence.

Retrieval

The step in which a system selects the documents most relevant to a query before generating a response. Being part of the retrieved set is a precondition for being named in the answer.

Embedding

A numeric vector that represents the meaning of a piece of text, where similar meanings sit close together in vector space. Embeddings power semantic retrieval by matching a query to content by meaning rather than by exact keywords.

Entity

A distinct, identifiable thing such as a brand, person, place, or product that a model can recognize and reason about. A model must treat a brand as a clear entity before it can describe or recommend it.

Entity disambiguation

The process of separating an entity from others that share a similar name so a model attributes facts to the correct one. Without disambiguation, a brand can be confused with a namesake and dropped from answers.

Entity resolution

The process of linking multiple references across sources to a single, canonical entity record. Strong entity resolution lets a model unify scattered mentions into one confident understanding of who you are.

Knowledge graph

A structured network of entities and the relationships between them. Knowledge graphs give machines a model of how a brand connects to people, products, and topics, reinforcing how it is understood and cited.

Structured data (schema.org)

Machine-readable markup, usually expressed with the schema.org vocabulary, that states what a page asserts about an organization, person, or topic. Structured data is the most direct way to tell a machine what a page means.

Citation frequency

How often a brand is named or quoted as a source across generated answers. Citation frequency is a core GEO metric because it tracks whether a brand is entering the answer at all.

Share of voice

The proportion of relevant answers in a category that mention a given brand, measured across a set of representative prompts. Share of voice shows how a brand stands against competitors inside AI answers.

Recommendation rate

How often a model recommends a brand when a user asks for the best option in its category. Recommendation rate is the closest GEO metric to commercial intent, since it captures active endorsement rather than mere mention.

AI visibility

The degree to which a brand is present and accurately described across the answers generative engines produce. AI visibility is the outcome GEO optimizes for, spanning citation, share of voice, recommendation, and factual accuracy.

llms.txt

A proposed plain-text file at a site's root that gives AI systems a curated map of the content most worth reading. llms.txt aims to help models find and prioritize a site's canonical sources.

Answer engine

A system that returns a direct, synthesized answer to a question rather than a list of links. ChatGPT, Claude, Perplexity, and AI search summaries are answer engines, and they are the surface GEO targets.

Cadive is the generative engine optimization agency founded by Leo Falcon. Read the entity profile, explore the knowledge hub, or start a project.