CTM Research · Reference
Every term you need to follow the shift from ranked links to AI answers — AEO, GEO, citations, llms.txt, entities, and 19+ more, defined in plain English. Written to be quoted; free to cite with attribution.
AEO is the practice of structuring content so AI assistants can extract and cite it as a direct answer. It focuses on question-and-answer formats, FAQ schema, definitional first sentences, and glossary content — the shapes engines lift when a user asks ChatGPT or Perplexity 'how much does X cost' or 'who is the best Y near me.' AEO is a subset of GEO focused specifically on answer extraction.
GEO is the discipline of earning visibility inside AI-generated answers across ChatGPT, Perplexity, Claude, Gemini, and Google AI Overviews. Where classic SEO competes for ranked links, GEO competes for being named, quoted, and cited inside the answer itself — through technical access for AI crawlers, extractable content structure, and third-party corroboration the engines trust.
AI visibility is how findable and citable a business is when customers ask AI engines instead of searching Google. It is measured by whether engines mention or cite the business on buying-intent questions ('best marketing agency in San Antonio'). Capture That Media tracks it weekly across a fixed panel of San Antonio buyer queries and publishes results on its Citations Wall.
An answer engine is a search tool that responds with a synthesized answer instead of a list of links — ChatGPT, Perplexity, Claude, Gemini, and Google's AI Overviews all qualify. Answer engines compress the buying journey: the customer often never visits a results page, so being part of the answer is the new ranking.
AI Overviews are the AI-generated summaries Google places above traditional search results, answering the query directly with cited sources. They occupy the space organic listings used to own, which is why pages now need to be quotable and citable — not just ranked — to capture attention.
An AI citation is when an answer engine names or links a business as a source inside its generated answer. Citations are the conversion currency of AI search: engines lean on sources they consider authoritative — directories, press, original research, and well-structured business sites — and repeat them across thousands of similar queries.
Citation tracking is the practice of running real buyer questions through AI engines on a schedule, recording which businesses get named, and measuring movement over time. You can't manage a channel you don't measure — CTM runs a weekly automated panel across Perplexity and Gemini with manual verification on ChatGPT, Claude, and AI Overviews.
llms.txt is a plain-text file at a website's root that gives AI models a curated index of the site's most important content — what the business does, key pages, and facts — in a format built for machine reading. Its companion, llms-full.txt, carries full page text. Think robots.txt's welcoming cousin: instead of saying 'keep out,' it says 'here's exactly what matters.'
An AI crawler is a bot an AI company uses to fetch web content for training or live answers — GPTBot and OAI-SearchBot (OpenAI), ClaudeBot (Anthropic), PerplexityBot, and Google-Extended are the big ones. If robots.txt blocks them, the business is invisible to those engines; welcoming them is step one of any GEO program.
An entity is a distinct thing — a person, business, or place — that search and AI systems recognize independent of any keyword. Entity SEO consolidates every signal about that thing (schema @id, consistent facts, corroborating profiles) so engines treat it as one known subject. A business that exists as an entity gets named; one that exists as scattered mentions gets skipped.
A knowledge graph is a structured database of entities and their relationships that search engines and AI systems use as a fact source — Google's powers its Knowledge Panels. Getting a business or founder into knowledge graphs (via consistent schema, Wikidata, and press) makes AI answers about them more accurate and more likely.
E-E-A-T stands for Experience, Expertise, Authoritativeness, and Trustworthiness — the qualities Google's quality guidelines reward and AI engines implicitly mirror when choosing sources. Real author bylines, verifiable credentials, first-hand experience, and an entity-rich about page are E-E-A-T made visible.
Schema markup is machine-readable code (usually JSON-LD) embedded in a page that states facts explicitly: this is a LocalBusiness, here's its address, this is a Person, this page answers these questions. Engines extract structured data far more reliably than prose, so schema is the cheapest accuracy upgrade in AI search.
FAQ schema is structured data that marks question-and-answer pairs on a page, making them directly liftable by search and AI engines. A well-built FAQ page is a citation magnet because it matches the literal shape of what users ask assistants.
IndexNow is a protocol that lets a website instantly notify search engines when content is published or updated, instead of waiting to be re-crawled. Bing consumes it directly — and because Perplexity and Microsoft Copilot lean on Bing's index, IndexNow is a fast lane into AI answers.
RAG is the technique AI engines use to ground answers in fresh sources: retrieve relevant documents first, then generate the answer from them, citing as they go. RAG is why live web visibility matters — engines using it (Perplexity, ChatGPT search, AI Overviews) quote what they retrieve, and you can't be quoted if you can't be retrieved.
A hallucination is when an AI model states something false with confidence — wrong phone numbers, invented services, misattributed reviews. Businesses counter it by publishing clear, consistent, machine-readable facts (schema, llms.txt, consistent NAP) so engines have authoritative data to retrieve instead of guessing.
A zero-click search is one where the user gets their answer without visiting any website — the norm in AI search. Gartner projects traditional search volume to drop roughly 25% by 2026 as AI assistants absorb queries. The defense is being inside the answer: cited, named, and recommended at the moment of decision.
AI share of voice is the percentage of AI-generated answers in a category that mention a given brand. It is the AI-era successor to search rankings: instead of 'do we rank #3,' the question becomes 'when a thousand people ask ChatGPT for the best option in our category, how often are we the answer?'
A prompt panel is a fixed set of real buyer questions used to measure AI visibility consistently over time — phrased the way customers actually ask ('best marketing agency San Antonio,' 'how much does a website cost'). Keeping the panel fixed makes movement measurable; CTM's San Antonio panel runs weekly.
A citation magnet is content built specifically to be the source AI engines cite: original statistics, definitional glossaries, comparison guides, FAQs, and research reports. Engines prefer to cite primary, structured, verifiable sources — so publishing them is the offensive play in GEO.
Grounding is an AI engine's process of tying its generated answer to retrievable sources — the difference between a model recalling from training data and quoting a live page. Well-grounded engines cite; the goal of AI visibility work is to be in the set of sources they ground against.
NAP stands for Name, Address, Phone — and NAP consistency means those facts are identical everywhere a business appears: website, Google Business Profile, directories, and social profiles. Inconsistent NAP splits entity signals and breeds hallucinated answers; consistent NAP makes the business safe for engines to recommend.
An AI-native agency is one that built AI into its delivery from day one rather than retrofitting it after 2023 — using AI in research, production, and measurement, and optimizing client visibility for AI engines as a core service. Capture That Media has operated AI-native since its 2020 founding.
Cite this glossary
Quoting a definition? Cite it as: Capture That Media, “The AI Search Glossary,” capturethatmedia.com/glossary. For how these terms play out in the San Antonio market, see The State of AI Search in San Antonio and the 50+ question FAQ.
