Glossary

Glossary: the terms of AI search, defined

The terms of AI search and answer engines, each defined plainly and explained in its own note.

Each term is defined in a line here and explained in its own note.

A

AI crawler: An AI crawler is a bot run by an AI company that fetches web pages to train models, to build an AI search index or to answer a user live. The main crawlers by job, how robots.txt applies to each, an example and common misconceptions.
AI share of voice: AI share of voice is a brand's portion of the mentions or citations in AI answers to a set of questions, compared with everyone else in those answers. The three ways to count it, why raw counts mislead, an example and common misconceptions.
AI visibility: AI visibility is how often and how prominently a brand appears in AI answers, as a mention or as a cited source. How it is measured, what Google Search Console now reports, a per-engine example from our scans and common misconceptions.
Answer engine optimization (AEO): Answer engine optimization (AEO) is the work of getting a brand named or cited when AI engines answer questions. How it works, what Google says about it, an example from our scans and the common misconceptions.

Canonical URL: A canonical URL is the one address a search engine treats as the main version of a page that exists at several URLs. How canonicalization works, how to signal it, and why duplicate URLs can split or misdirect AI citations.
Citation (AI answers): A citation in an AI answer is a link to a source page that the engine attaches to its response. How each engine shows citations, how they differ from mentions, how many each engine gives per answer, and an example of a misleading one.

E-E-A-T: E-E-A-T stands for experience, expertise, authoritativeness and trustworthiness, the qualities Google's quality raters look for. What each part means, how Google uses it, and why it carries into AI answers.
Entity (search): An entity is a distinct thing, such as a person, place, company or product, that a search engine recognizes whatever words are used to name it. How entities work in Google and why they matter for AI answers.

Featured snippet: A featured snippet is a Google result that shows an extract from a page first and the link second. How Google picks one, how to opt out, and why the same snippet controls decide what AI Overviews and AI Mode can show.

Generative engine optimization (GEO): Generative engine optimization (GEO) means shaping content so AI engines that write answers from web sources use and cite it more. Where the term came from, what the original study tested and found, and the common misconceptions.
Google AI Mode: Google AI Mode is a conversational search mode in Google Search that answers with an AI response, supporting links and room for follow-up questions. How it works, how it differs from AI Overviews and what it cited in our scans.
Google AI Overviews: Google AI Overviews are AI-generated snapshots shown in some Google Search results, with links to supporting pages. How they are built, which pages can be linked, the new Search Console control and report, and what they cited in our scans.

IndexNow: IndexNow is an open protocol that lets a site tell participating search engines, including Bing, that a URL was added, updated or deleted. How it works, who supports it, and why Microsoft ties it to fresher AI answers.

Knowledge Graph: Google's Knowledge Graph is its database of entities and the facts that link them, launched in 2012 and behind knowledge panels. How it is built, how AI Mode draws on it, and what its public API can and cannot show.

llms.txt: llms.txt is a proposed Markdown file at a website's root that gives AI models a short summary of the site and links to its key pages. The format, who publishes one, what Google says about it and the common misconceptions.

Prompt tracking: Prompt tracking means running a fixed set of questions through AI engines on a schedule and recording which brands each answer names and which pages it cites. How it works, why one run proves little, and what our scans show.

Query fan-out: Query fan-out is when an AI search system splits one question into several related searches, runs them at once and merges the results into one answer. How Google and OpenAI describe it, a worked example and common misconceptions.

Retrieval-augmented generation (RAG): Retrieval-augmented generation (RAG) is how an AI model looks up documents first and then writes its answer from them. Where it came from, how answer engines use it, why it decides which pages can be cited, and common misconceptions.

Search intent: Search intent is the goal behind a query, meaning what the person wants to do rather than the words they typed. The classic categories, how Google's raters apply them, and why intent shapes whether and how AI answers appear.
Structured data (schema markup): Structured data is standardized code, usually schema.org written in JSON-LD, that labels what a page contains. How it works, what it does and does not do for Google's AI Overviews and AI Mode, and common myths.

Topical authority: Topical authority is the idea that a site known for covering a subject in depth is trusted more on it. What Google has actually said, including its 2023 topic authority system for news, and how the idea applies to AI answers.

Zero-click search: A zero-click search ends without the searcher clicking any result. How it happens, what named studies from SparkToro and Pew Research Center measured, and why AI answers make it the usual case rather than the exception.