GEO glossary

The practice of making a brand, its products, and its evidence visible, citable, and recommendable inside generative AI answers. It is not SEO renamed; it turns content into citable answer assets.

Related: GEO-001 · GEO-049

Structured content a model can quote, compare, and transact on: answer units, product entities, evidence, policies, and localized language.

Related: GEO-037 · GEO-097

An 80–120 word self-contained answer on a page: claim, evidence, difference, next step. Its job is to let AI extract an accurate sentence without guessing.

Related: GEO-037 · GEO-038

How much a brand appears in AI answers, usually split into mention, citation, recommendation, and conversion rather than a vague “are we in AI?”

Related: GEO-011 · GEO-109

The share of citations a brand earns within a fixed prompt set, relative to all citations in the category. It is closer to trust than raw mentions.

Related: GEO-059 · GEO-095

A fixed, reproducible set of buyer questions used to measure AI visibility over time. A good set covers markets, languages, purchase stages, and categories, not just brand names.

Related: GEO-011 · GEO-114

A graph of products as entities, variants, attributes, alternatives, compatibility, and accessories. AI sees these relationships, not a single SKU row.

Related: GEO-025 · GEO-067

An independently identifiable object: a brand, product, category, person, or organization. If the entity is missing or inconsistent, AI struggles to cite you reliably.

Related: GEO-009 · GEO-033

A Schema.org property that tells machines a set of URLs describe the same entity. It is one of the cheapest tools for entity disambiguation.

Related: GEO-009 · GEO-092

A Markdown index at the site root that tells generative engines what exists, using titles, summaries, and links. It is documentation, not a ranking switch.

Related: GEO-008

Tells search engines and AI how language or regional versions of the same page relate. It solves “which page matches which”, not “every page is crawlable”.

Related: GEO-020 · GEO-078

Machine-readable fields for products, prices, stock, ratings, and organizations using vocabularies such as Schema.org. The goal is confirmed facts, not a passing validator.

Related: GEO-026 · GEO-056

A structured product list for marketplaces, comparison engines, ads, and agents. In agentic commerce it increasingly acts as the new landing page.

Related: GEO-080 · GEO-098

Shopping where an AI agent searches, compares, orders, and tracks fulfillment on the buyer’s behalf. It requires real-time price, stock, shipping, returns, and payment interfaces.

Related: GEO-097 · GEO-099

Visits from AI interfaces such as ChatGPT, Perplexity, and Gemini. Referrers are often lost and fall into Direct, so logs, UTMs, and surveys are needed together.

Related: GEO-012 · GEO-107

The outcome that happened because of an action, not what would have happened anyway. GEO incrementality needs controls, holdouts, or time-series estimates.

Related: GEO-072 · GEO-110

Creating citable third-party facts through original data, expert commentary, and media relationships. AI cites facts and viewpoints, not press releases.

Related: GEO-086 · GEO-090

A set of quality signals: experience, expertise, authoritativeness, and trust. In GEO it shows up as authors, evidence, third-party mentions, and corrections.

Related: GEO-045 · GEO-049

The share of crawlable URLs actually fetched by AI crawlers. It is the first leak in the crawl → cite → convert funnel.

Related: GEO-023 · GEO-024

The share of AI answers where you appear among the 3–5 shortlisted brands. It is closer to pre-purchase decisions than rank position.

Related: GEO-061 · GEO-071