AEO Glossary

This glossary defines the working vocabulary of answer engine optimization as the terms are actually used, including where usage is inconsistent. Each entry links to a full page with the definition, the mechanism behind it, and the disputes around it.

Why the vocabulary is unstable

AEO is a young field named from several directions at once, and its terminology shows it. Three forces keep the vocabulary unsettled.

First, the field is newer than its subject. Practitioners were optimizing for featured snippets and voice assistants years before large language models made generated answers the default interface, and the old labels were stretched to cover the new behavior. A term like answer engine predates the systems it now mostly describes, so its meaning shifted underneath the people using it.

Second, independent groups coined overlapping names for the same activity. Academic work introduced generative engine optimization in a research context, while marketing practice converged on answer engine optimization, and both communities kept their own term. The result is two labels whose definitions overlap by perhaps ninety percent, separated mostly by who is speaking. AI SEO and LLM SEO circulate as informal third and fourth options.

Third, vendors coin terms as positioning. A tracking tool benefits from naming the metric it sells, and an agency benefits from naming the discipline it claims to lead, so terms arrive with commercial gravity attached. Some of these names describe something real and stick; others describe nothing that existing words did not already cover. A glossary written now has to distinguish terms that carry a mechanism from terms that carry only a trademark, and this one tries to do exactly that.

A reader can protect themselves with three questions when a new term arrives. Does it name a mechanism that existing terms do not cover? If the definition could be swapped with an existing term's definition without loss, it is a label, not a concept. Does it have a traceable origin? A term with a citable first use, the way GEO traces to a 2023 research paper, can at least be pinned to an initial definition when usage drifts. And who profits from its adoption? A term coined by a measurement vendor will quietly define itself in ways their tool can measure, which is not dishonest but is worth noticing before the definition settles into your own vocabulary.

None of this is unusual. Early SEO went through the same phase before settling on shared vocabulary, and the same settling should be expected here. In the meantime, precise definitions matter more, not less: when two people say citation and mean different things, the conversation silently fails.

The terms

Answer engine: a system that responds to a query with a composed answer, built by retrieving indexed content and synthesizing it, rather than with a ranked list of links.

Generative engine optimization (GEO): the practice of improving content visibility inside responses produced by generative engines; in practice a near-synonym of AEO with an academic origin.

Entity authority: the degree to which an entity is resolvable and its claims are corroborated across independent sources, such that an engine can identify it and cite material about it with confidence.

Citation rate: the share of sampled generated responses, across a defined query set, in which a given source is cited.

How entries are written

Every entry follows the same discipline. The definition comes first, in one or two sentences that could be lifted whole. The mechanism comes second: what actually happens in the retrieval and synthesis pipeline that makes the term matter. Disagreements in usage are stated rather than smoothed over, and no entry cites a statistic that cannot be traced to a verifiable source. Terms are added when they carry a distinct mechanism, not when a vendor coins a new label for an existing one.

The set is deliberately small. Four terms cover the core of the field as it stands: the system being optimized for, the competing name for the discipline, the trust structure that decides whether a source gets used, and the metric that tells you whether any of it worked. New entries will be added as the vocabulary stabilizes, at the pace the field earns them.