The "What Does
That Even Mean?" Glossary.

The practice of structuring and writing content so it is more likely to be selected as a direct answer by AI tools, voice assistants, or any system that synthesizes information rather than just listing links. AEO focuses on clarity, question-and-answer formatting, direct responses to specific queries, and schema markup. It overlaps significantly with GEO and SEO but puts the answer itself — not the ranking — at the center.

An automated program that reads web pages on behalf of an AI company or AI-powered search product. Examples include GPTBot and OAI-SearchBot (OpenAI), PerplexityBot (Perplexity), and ClaudeBot (Anthropic). Like traditional search crawlers, these can be selectively allowed or blocked using robots.txt rules. Whether you allow or block them affects whether your content may be used in AI training or AI-generated answers.

Google's name for the AI-generated summary blocks that appear at the top of some search results. AI Overviews synthesize information from multiple web pages into a direct answer with cited sources. They are distinct from featured snippets — AI Overviews generate new text, while featured snippets display an excerpt directly from a page. Whether a page gets cited in AI Overviews depends on content quality, trustworthiness, and clarity.

Any system that responds to queries with synthesized answers rather than a ranked list of links. ChatGPT, Perplexity, Google AI Overviews, and Bing Copilot all function as answer engines. The key difference from a search engine: an answer engine produces a response, not a list of results. For website owners, the implication is that being referenced or cited by an answer engine matters — not just ranking on page one.

References included by an AI system in a generated answer, attributing information to a specific source. A citation usually includes the source name and a link back to the original page. Being cited is one of the clearest indicators that an AI system has used your content as a source. Citations are not guaranteed — they depend on content quality, specificity, and the particular system's behavior — but pages with clear, direct, authoritative answers are more likely to earn them.

How a webpage is organized — including its headings, paragraphs, lists, tables, summary blocks, and FAQ sections. Clear content structure helps both human readers and AI systems navigate and understand a page. Good structure means: one H1 per page, logical H2/H3 hierarchy, short paragraphs with direct answers, and supplemental elements like tables or FAQs where they add value. Structure is one of the highest-leverage improvements most pages can make.

An approach to search optimization that focuses on establishing clear, consistent facts about a person, organization, product, or concept — rather than targeting keywords alone. Search engines and AI systems use entity relationships to understand context and authority. Strong entity signals include consistent name/address/phone data, Wikipedia mentions, Wikidata entries, and Knowledge Graph presence. Entity SEO matters more as AI systems rely increasingly on structured knowledge rather than keyword matching.

A short excerpt from a web page that Google surfaces at the top of some search results — above the regular ranked links — to directly answer a query. Unlike AI Overviews, featured snippets display text taken directly from your page rather than AI-generated text. Featured snippets are earned by writing clear, direct answers to specific questions, with proper heading structure and concise explanations.

The practice of optimizing content specifically for AI systems that generate new text in response to queries, rather than retrieving and ranking existing pages. GEO focuses on making content clear, authoritative, specific, and easily parseable by large language models. It's closely related to AEO — the distinction is mostly that GEO emphasizes the generative aspect of modern AI tools, while AEO emphasizes direct answers.

A type of search experience where AI generates a custom answer to each query, drawing on indexed web content. Unlike traditional search — which finds and ranks existing pages — generative search synthesizes information from multiple sources into a new response. Google AI Overviews, Perplexity, and Bing Copilot are examples. Generative search is one of the main reasons traditional link-click traffic patterns are changing.

The process of connecting an AI model's output to specific source documents or factual data, rather than relying entirely on what the model learned during training. Grounded AI responses are anchored to real sources — which is why they include citations. Grounding reduces hallucination risk because the model is constrained to what the retrieved documents actually say. RAG is one common grounding technique.

A Google framework for evaluating whether content genuinely helps people, as opposed to content created primarily to rank in search. Google's helpful content guidance emphasizes: content written by real people with genuine expertise and experience, for a specific audience, that fully satisfies the reason someone searched. In the LLMAO context, helpful content is also the foundation of AI visibility — AI systems tend to cite content that is clear, specific, and actually answers questions.

A structured database of facts and relationships between real-world entities — people, places, organizations, products, concepts. Google's Knowledge Graph powers features like knowledge panels and many AI-assisted search results. AI systems use similar graph structures to understand how concepts relate. For SEO, establishing a clear entity footprint — consistent facts about who you are and what you do — is how you build a presence in knowledge graphs.

A type of AI model trained on large amounts of text data to understand and generate human language. GPT-4, Claude, Gemini, and Llama are all large language models. LLMs power most modern AI assistants and AI search tools. Understanding that LLMs work by pattern-matching and prediction — not by understanding in the human sense — helps explain why content clarity, structure, and specificity matter so much for AI visibility.

The broader practice of making website content easier for large language models to read, parse, cite, and summarize. It overlaps with AEO and GEO but refers more specifically to the technical and content decisions that affect how LLMs process your pages. This includes content structure, heading clarity, schema markup, internal linking, and the specificity of answers. LLMAO.ca is largely built around this concept.

A proposed community convention for providing machine-readable, AI-friendly documentation at the root of a domain. It is written in plain Markdown and serves as a curated index or table of contents specifically for LLM crawlers, AI search assistants, and IDE code assistants to digest key site content efficiently.

A method used by AI systems to supplement their built-in knowledge with information retrieved from external sources at the moment a query is made. Instead of relying only on training data, a RAG system searches for relevant content, retrieves it, and uses it to generate a more accurate, up-to-date answer. Perplexity, Google AI Overviews, and Bing Copilot all use RAG-like approaches. This is why keeping your content fresh and crawlable matters — it affects whether your pages get retrieved and used.

Code added to a web page — usually in JSON-LD format, inside a <script> tag — that tells search engines and AI systems what the content means, not just what it says. Schema markup uses vocabulary from Schema.org to label content types like Article, FAQ, Product, Person, Organization, and more. It makes it easier for machines to extract structured facts from a page. Google uses it for rich results; AI systems use it to better understand content context.

HTML written using elements that describe the meaning of content, not just its appearance. Examples: using <article>, <nav>, <main>, <section>, <h1>–<h6>, <blockquote>, and <aside> appropriately — rather than stacking everything in <div> and <span> elements. Semantic HTML helps screen readers, search crawlers, and AI parsers understand the structure and purpose of different parts of a page. It is foundational to good accessibility and good AI parsability.

Information formatted in a standardized, machine-readable way so that systems can easily parse and use it. On the web, structured data usually refers to Schema.org markup added to pages in JSON-LD, Microdata, or RDFa format. Google and other systems use structured data to understand facts, relationships, and content types — and to power features like rich search results, knowledge panels, and AI answer citations. JSON-LD is the recommended format.