AI SEO Glossary: The Terms You Actually Need to Know

AI SEO comes with its own vocabulary, and a lot of it overlaps with classical SEO. This reference glossary defines the terms you'll actually use in real engagements, with examples and the surfaces each applies to.

AI SEO disciplines

AI SEO. The umbrella discipline of earning visibility across AI-powered search engines and assistants. Covers Google AI Overviews, ChatGPT, Perplexity, Claude, Gemini, and Bing Copilot. See our what is an AI SEO company guide.

Generative Engine Optimization (GEO). Optimizing for AI features built into search engines. Most commonly Google AI Overviews, Bing Copilot, and SGE. See our GEO guide.

Answer Engine Optimization (AEO). Optimizing for standalone AI answer engines like Perplexity, You.com, and Arc Search. See our AEO guide.

LLM SEO. Shaping what large language models say about your brand, at both training-corpus and live-retrieval layers. See our LLM SEO guide.

AI Search Optimization. The unified discipline spanning GEO, AEO, and LLM SEO. Used when work covers every AI search surface in one program. See cluster page.

Entity SEO. Building the brand, product, and topic associations AI engines use to decide who to cite. See our entity SEO guide.

AI search engines and surfaces

Google AI Overviews. Google's AI-written answer shown at the top of search results on supported queries. Cites 2-5 sources.

SGE (Search Generative Experience). Google's broader generative search initiative. AI Overviews are one feature of SGE.

ChatGPT. OpenAI's conversational AI assistant. Pulls from training corpus and (with browsing) live web.

Claude. Anthropic's conversational AI assistant. Known for careful, well-cited responses. Common in US professional research.

Gemini. Google's conversational AI assistant. Closely integrated with Google Search.

Perplexity. AI answer engine that writes responses with a ranked source list. Pure citation-driven.

Bing Copilot. Microsoft's AI assistant inside Bing. Uses Bing's index for retrieval.

llms.txt. A text file at the root of a domain offering a curated map of priority content to AI engines. Not a directive. A positive signal.

Technical terms

Crawler / bot user-agent. The identity an AI engine's crawler uses when fetching pages. Examples: GPTBot, ClaudeBot, PerplexityBot, Google-Extended.

Schema markup. Structured data (usually JSON-LD) embedded in pages to help engines understand entities, relationships, and content type.

sameAs. A schema field linking your brand entity to its profiles on other platforms. LinkedIn, Crunchbase, Wikidata, Wikipedia, etc.

Knowledge graph. A structured database of entities and their relationships. Google maintains one. AI engines lean on Wikidata and Wikipedia as inputs.

Retrieval-augmented generation (RAG). The technique most AI engines use. Retrieve relevant pages, then have a language model write an answer using the retrieved content.

Training corpus. The body of public text a language model learned from at training time. Influences answers even when the model is not browsing.

Content and structure terms

Quotability. The degree to which a page can be extracted from and cited cleanly. Direct answers, short paragraphs, defined entities, and structured content all improve quotability.

Direct answer. A clear, factual response to a page's primary question, usually placed in the first 100-150 words. The highest-leverage on-page AI SEO signal.

Defined entity. An entity (product, person, method) introduced with a short definition on first mention. Helps AI engines attribute information correctly.

Topical cluster. A coherent set of related pages on the same topic, linked together. Signals topical depth and authority.

YMYL (Your Money Your Life). Google's term for topics that can affect health, finance, safety, or major life decisions. AI engines apply stricter trust standards to YMYL.

Measurement terms

Citation share. The percent of citations on a given prompt that include your brand. The headline AI SEO metric.

Prompt-level visibility. Whether your brand appears in the answer or citation set for a specific prompt. Tracked across a fixed prompt set monthly.

Branded prompt accuracy. When asked about your brand specifically, how accurately does the model describe you. Important for reputation and LLM SEO.

Source rank. When you appear in a citation set, how high you rank within it (first cited, third cited, etc.).

Cross-engine share-of-voice. Your citation share averaged across the major AI engines for a fixed prompt set. Useful for executive reporting.

For more depth on any of these, see our blog or services overview.

Vocabulary is a tool. Knowing these terms makes it faster to talk with strategists, vendors, and internal teams about real work . and faster to spot when someone is using buzzwords to obscure rather than clarify.