Glossary

The GEO & AI Search Glossary

Clear, quotable definitions for the language of Generative Engine Optimization — the terms behind how AI answer engines like ChatGPT, Perplexity, and Google AI Overviews find, rank, and cite sources.

Last updated July 2026

This glossary defines the core terms of Generative Engine Optimization (GEO) and AI search. Each entry is written to stand on its own — a definition-first sentence you can quote directly. Jump to a term or read straight through.

Generative Engine Optimization (GEO): Generative Engine Optimization (GEO) is the practice of optimizing content and brand signals so AI answer engines cite, quote, and recommend you in their generated answers. It is the generative-AI counterpart to SEO.
Answer Engine Optimization (AEO): Answer Engine Optimization (AEO) is the practice of optimizing content to be the direct answer an engine returns — in AI chat, voice assistants, or featured snippets. It is a near-synonym of GEO that emphasizes the answer surface.
Large Language Model Optimization (LLMO): Large Language Model Optimization (LLMO) is optimizing content so large language models surface and cite it. It overlaps almost entirely with GEO; the term simply foregrounds the underlying model rather than the search product.
Answer engine: An answer engine is a system that responds to a query with a synthesized, direct answer instead of a list of links — for example ChatGPT, Perplexity, or Google AI Overviews. Answer engines usually cite a small set of sources.
AI Overviews: AI Overviews are Google's AI-generated answer summaries shown above the traditional search results, with links to cited sources. They are one of the highest-stakes GEO surfaces because of Google's reach.
AI Mode: AI Mode is Google's dedicated conversational search experience — a full-page generative answer interface, distinct from the AI Overview shown atop classic results. It relies more heavily on live retrieval and follow-up questions.
Citation: A citation is an explicit source attribution an answer engine attaches to a claim, usually as a link or footnote. Earning citations is the central goal of GEO — it is how a brand appears in, and is credited by, an AI answer.
Mention rate: Mention rate is the share of sampled AI answers to a set of tracked prompts in which a brand appears. It is the headline GEO visibility metric, expressed as a probability rather than a fixed rank.
Share of voice: Share of voice is how often a brand is cited relative to its competitors for the same set of prompts. It converts raw mentions into a competitive standing within a category.
llms.txt: llms.txt is a proposed plain-text file, placed at a site's root, that lists the content a site owner wants AI systems to prioritize. Adoption is limited — Google has said Search does not use it — so treat it as an optional supplement, not a ranking factor.
GPTBot: GPTBot is OpenAI's web crawler that collects publicly available content to help train future models. Allowing it in robots.txt lets your content inform those models; blocking it opts your site out of training crawls.
OAI-SearchBot: OAI-SearchBot is OpenAI's crawler for ChatGPT's search feature. Unlike GPTBot, which gathers training data, OAI-SearchBot indexes pages so they can be retrieved and cited in live ChatGPT answers.
ChatGPT-User: ChatGPT-User is the OpenAI agent that fetches a specific page on demand, the moment a user asks ChatGPT about a URL or triggers browsing. It represents a real user's request rather than a bulk crawl.
ClaudeBot: ClaudeBot is Anthropic's web crawler, which gathers content used by Claude. As with other AI crawlers, your robots.txt controls whether it may access your site.
PerplexityBot: PerplexityBot is the crawler Perplexity uses to index pages for its answer engine. Because Perplexity leans heavily on live retrieval and citations, being crawlable by PerplexityBot is a prerequisite for appearing in its answers.
Google-Extended: Google-Extended is a robots.txt token that controls whether your content may be used to train and ground Google's generative AI, such as Gemini and AI Overviews. Disallowing it does not affect your normal Google Search ranking.
Crawler (AI crawler): A crawler, or bot, is an automated program that fetches web pages. AI crawlers — GPTBot, OAI-SearchBot, PerplexityBot, ClaudeBot, and others — feed content to model training, AI search indexes, or on-demand retrieval; if they cannot reach a page, it cannot be cited.
Retrieval-Augmented Generation (RAG): Retrieval-Augmented Generation (RAG) is a technique in which an AI model retrieves relevant documents at query time and generates its answer from them. RAG is why fresh, crawlable, quotable web content can be cited even if it was published after the model was trained.
Grounding: Grounding is tying a model's generated answer to specific retrieved sources, so claims can be attributed and verified. Well-grounded answers cite sources; GEO works by making your content the material an engine grounds on.
Hallucination: A hallucination is a confident but false or fabricated statement produced by an AI model. Grounding an answer in retrieved, authoritative sources — the material GEO helps you supply — is a primary defense against hallucinations.
Entity: An entity is a distinct, identifiable thing — a company, person, product, or place — that engines recognize and reason about. Consistent naming, structured data, and corroboration help engines resolve your brand to the correct entity.
Knowledge graph: A knowledge graph is a structured network of entities and the relationships between them. Search and AI engines use knowledge graphs to understand who or what a source is about; clear entity signals help you be represented accurately.
Structured data (JSON-LD): Structured data is machine-readable markup — usually schema.org in JSON-LD format — that labels the meaning of a page's content. It helps engines resolve entities and extract answers, making it a core GEO tactic.
Prompt sampling: Prompt sampling is asking an AI engine the same question multiple times and aggregating the results, because answers vary between runs. It turns non-deterministic responses into a stable mention rate with a confidence range.
Visibility score: A visibility score is a composite metric summarizing how present a brand is across AI answers — typically blending mention rate, answer position, and share of voice into a single trackable number.
Zero-click search: A zero-click search is a query answered directly on the results surface — by an AI Overview, snippet, or chat answer — without the user clicking through to a website. Zero-click behavior makes being cited in the answer itself increasingly important.
E-E-A-T: E-E-A-T stands for Experience, Expertise, Authoritativeness, and Trustworthiness — Google's framework for content quality. The same signals that build E-E-A-T (authorship, citations, corroboration) also make content more likely to be trusted and cited by AI engines.
Freshness: Freshness is how recent and up-to-date content is. For answer engines that retrieve the live web, clearly dated, recently updated pages are preferred over stale ones, making freshness a practical GEO signal.

Turn these terms into visibility

Run a free AI-visibility check and see exactly where AI answer engines mention your brand today.

Run a free check