Look up the terms used in AI search visibility work.
Search a glossary of GEO, AEO, and AI-search terms — from citation and grounding to share of voice and structured data — each with a precise definition and links to related tools.
How to use this tool
Type a term or partial word into the filter box, such as "citation" or "RAG".
Read the definition, written to be technically precise rather than marketing shorthand.
Check the "why it matters" note where present to see how the term connects to AI visibility work.
Follow the related tool link on relevant entries to put the concept into practice.
Clear the filter to browse the full alphabetized list.
Bookmark the page as a shared reference for your team when discussing AI-search strategy.
A glossary of terms used across GEO (Generative Engine Optimization) and AEO (Answer Engine Optimization) work. Definitions aim to be technically precise rather than marketing shorthand — terminology in this space is still settling, so exact usage can vary slightly between vendors and publications.
AEO (Answer Engine Optimization)
The practice of structuring content so an AI system — a chatbot, voice assistant, or answer engine — can extract a direct, correct answer to a specific question, rather than optimizing primarily for a ranked list of links.
Why it matters: AEO overlaps heavily with GEO but is narrower in focus: clear question-and-answer structure and a direct first-sentence answer are AEO tactics that feed into broader AI visibility.
Google's AI-generated summary shown above traditional search results for some queries, synthesizing information from multiple sources and typically linking out to a handful of them as citations.
Answer Presence
Whether a brand, product, or fact appears at all in an AI system's response to a given prompt, regardless of sentiment or ranking — the most basic unit of AI-visibility measurement.
Why it matters: Answer presence is binary (mentioned or not) and is usually measured before more nuanced signals like sentiment or share of voice.
A link or explicit source attribution an AI system provides alongside a generated answer, pointing to the page or document it drew a claim from.
Why it matters: Citations are the closest AI-search equivalent to a search ranking; being the cited source can drive both traffic and perceived credibility.
How recently a page was published or meaningfully updated, used by crawlers and retrieval systems as one signal of how current — and sometimes how reliable — its information is.
Corroboration
The degree to which a claim about a brand or product is confirmed by multiple independent sources rather than appearing on only one page.
Why it matters: Some AI systems weight corroborated claims as more trustworthy, so a fact repeated consistently across your site, review platforms, and press coverage is more likely to be treated as reliable than one found in a single place.
An automated program that fetches web pages, identified by a user-agent string (for example GPTBot or ClaudeBot) sent with each request, which site owners can allow or block via robots.txt.
Why it matters: Robots.txt rules for a crawler's user-agent directly determine whether that AI company's training or retrieval systems can read your content at all.
The removal of a page or an entire site from a search engine's or AI system's index, meaning it can no longer be retrieved or cited, often caused by a noindex directive, a manual penalty, or persistent crawl errors.
Entity
A distinct, uniquely identifiable thing — a person, organization, product, or place — that a knowledge graph or language model can recognize and reason about consistently, independent of the exact words used to describe it.
Why it matters: Strong entity grounding, meaning a clear name, category, and consistent facts, helps a model connect mentions of your brand across many sources into one coherent identity.
A highlighted answer box shown above standard results on a traditional search engine results page, extracted directly from the text of a ranking page.
Why it matters: Featured snippets are a precursor to AI Overviews; content structured to win one (clear, concise, direct answers) tends to also work well for AI answer extraction.
GEO (Generative Engine Optimization)
The practice of structuring and distributing content so it is more likely to be retrieved, cited, and represented accurately by generative AI systems — chatbots, AI search, AI Overviews — as distinct from optimizing for a traditional ranked list of links.
The technique of anchoring an AI model's response in specific, retrieved source documents rather than relying solely on patterns learned during training, typically implemented via retrieval-augmented generation.
Why it matters: Well-grounded answers are more likely to include accurate citations; poor grounding is a common cause of hallucination.
Hallucination
A factually incorrect or fabricated statement generated by an AI model and presented with the same confidence as a true one, arising because the model is predicting plausible text rather than verifying facts.
Why it matters: Brand-related hallucinations, such as wrong pricing or invented features, can spread through AI answers with no single wrong page to correct, which is why monitoring what models actually say about your brand matters.
A structured database of entities and the relationships between them (for example, "Company X" → "founded by" → "Person Y"), used by search engines and some AI systems to answer factual questions without reading a full page.
Why it matters: Wikidata and Google's Knowledge Graph are prominent examples; structured, consistent facts about your organization make it easier for a knowledge graph to represent you correctly.
A machine learning model trained on large volumes of text to predict and generate language; the underlying technology behind chatbots such as ChatGPT, Claude, and Gemini.
llms.txt
A proposed plain-text Markdown file, placed at a site's root, that gives language models a concise, curated overview of a site and links to its most important pages, intended as an alternative to relying on a full crawl.
Why it matters: It is an emerging, not yet universally adopted convention, and support for reading it varies by AI system.
How easily a specific paragraph or passage, rather than a whole page, can be extracted and quoted as a self-contained answer, based on factors like whether it states a claim clearly, includes a number or date, and doesn't depend heavily on surrounding context.
Why it matters: Retrieval systems often pull individual passages, not full pages, so a passage that only makes sense with prior context is less likely to be cited cleanly.
The input text a user, or a system testing AI visibility, sends to an AI model to elicit a response; in AI-visibility work, often a realistic buyer or research question a target customer might ask.
Why it matters: Which prompts a brand shows up for, and how, is the basic unit most AI-visibility measurement is built on.
An architecture where a model retrieves relevant documents or passages from an external source at query time and includes them in its context before generating an answer, rather than relying purely on what it learned during training.
Why it matters: Most AI search and AI Overview products use some form of RAG, which is why current, retrievable web content can influence answers even for models with an older training cutoff.
Schema.org
A shared vocabulary of structured-data types — Organization, Product, Article, and others — that site owners can embed as JSON-LD to describe page content in a machine-readable format that search engines and some AI systems can parse directly.
Why it matters: Structured data doesn't guarantee a citation, but it does reduce ambiguity about what a page is actually describing.
Automated classification of whether a piece of text — a review, a social post, or an AI-generated answer — expresses a positive, negative, or neutral opinion about a brand.
Why it matters: In AI-visibility monitoring, sentiment analysis is often applied to the AI-generated answers themselves, since being mentioned frequently but unfavorably is a different problem than not being mentioned at all.
The page a search engine returns in response to a query — traditionally a ranked list of links, now often including AI Overviews, featured snippets, and other non-link elements alongside or above the organic results.
Share of Voice
The proportion of AI-generated answers, across a defined set of prompts, in which a given brand is mentioned, usually measured relative to named competitors and expressed as a percentage.
Why it matters: It quantifies AI visibility over time and against competitors, analogous to share-of-voice metrics used in traditional media and paid search.
Machine-readable markup, most commonly JSON-LD using Schema.org vocabulary, embedded in a web page to explicitly describe its content, as opposed to leaving a model to infer meaning from unstructured prose alone.
The distinction between facts a model "knows" because they appeared in the dataset it was trained on (static, with a cutoff date) versus facts it accesses at query time through retrieval or browsing (current, but dependent on what is retrievable).
Why it matters: This distinction explains why a model can state outdated information from training even when current information is publicly available through retrieval, and why some products only browse for certain query types.
Zero-Click Search
A search, or an AI-assisted query, that fully satisfies the user's need on the results page itself, so the user never clicks through to any source website.
Why it matters: As AI Overviews and answer engines grow, more queries may end in zero clicks, making earning a citation (for visibility, even without a click) more important relative to earning a ranked link alone.
Useful context for applying this tool to your site.
What's the difference between GEO and AEO?
GEO (Generative Engine Optimization) is the broader term for structuring and distributing content so generative AI systems retrieve, cite, and represent it accurately. AEO (Answer Engine Optimization) is closely related but emphasizes structuring content so it can be extracted as a direct answer to a specific question; the two terms overlap significantly and are sometimes used interchangeably.
Are these definitions the industry-standard, agreed-upon versions?
Terminology in AI search and GEO is still settling, and different vendors or publications sometimes use these terms slightly differently. These definitions aim to be technically accurate and consistent with common usage as of today, rather than reflecting one vendor's marketing framing.
Why does the glossary link some terms to other tools?
Terms like citation, structured data, or share of voice connect directly to a specific, practical task, such as generating schema markup or calculating your AI visibility. The related link takes you from the definition straight to a tool that puts the concept into practice.
Does this glossary cover traditional SEO terms too?
Only where a term is genuinely relevant to AI search, such as SERP and featured snippet, which are traditional search concepts that also underpin how AI Overviews and answer engines work. It does not attempt to be a complete general SEO glossary.