Glossary: AI search and Citerank terms
Plain-English definitions of the terms used across AI search and these docs, each linked to the guide where it matters.
Search is moving from links to answers, and the vocabulary is new for most teams. These are the terms you will meet in Citerank and in AI search generally. Each links to the guide where it matters most.
- @id #
- A unique identifier (usually a URL with a fragment, such as https://example.com/#organization) that links the same entity across schema blocks and pages. Guide →
- Accessibility tree #
- The structured view of a page (roles, names, headings) that assistive technology and AI agents read instead of the visual page. Guide →
- ACP and UCP #
- Agent commerce protocols that let AI agents browse catalogues and check out on a shopper's behalf. Guide →
- AI agent #
- Software that browses and acts for a person, such as comparing products, filling a quote form or booking a slot. Guide →
- AI engine check #
- Citerank's unit for asking the AI engines one prompt or keyword once. Plans include a monthly number of checks. Guide →
- AI Mode #
- Google's conversational search mode, where answers are built by AI from multiple searches. Guide →
- AI Overview #
- The AI-written answer Google shows above the search results for some queries, with links to the sources it drew on. Guide →
- Answer engine #
- A search product that answers a question directly, such as ChatGPT search, Perplexity, Gemini or Google AI Overviews, rather than listing links.
- Citation #
- A link to your page that an AI engine shows as a source for its answer. Guide →
- Citation position #
- Your place in the list of sources an AI engine cites. Position 1 is the first source shown. Guide →
- Common Crawl #
- A large open archive of the web that is used in many AI model training sets. Guide →
- Credits #
- Citerank's unit for paying for runs that are not engine checks, about one cent of provider cost each. Packs add credits that never expire. Guide →
- E-E-A-T #
- Experience, expertise, authoritativeness and trust: the qualities Google's guidelines use to judge content, shown through on-page and schema signals. Guide →
- Entity #
- A distinct thing AI systems can identify: your organisation, a person, a product, a place. Guide →
- Fan-out query #
- One of the searches an AI engine runs behind the scenes to answer a question. Pages that answer fan-out queries are the ones that get cited. Guide →
- GPTBot, ClaudeBot, PerplexityBot #
- Crawlers run by OpenAI, Anthropic and Perplexity. Some collect training data; others (such as OAI-SearchBot and Claude-SearchBot) fetch pages to answer people. Guide →
- Grounding #
- When an AI model looks up current web sources (for example through Google Search) before answering, instead of answering from memory. Guide →
- Information gain #
- How much new, useful content a page adds compared with the pages that already rank for the topic. Guide →
- JSON-LD #
- The format for adding schema.org structured data to a page, inside a script tag in the head. Guide →
- Knowledge Graph #
- Google's database of entities and the facts about them, behind knowledge panels and much of AI search's understanding of brands. Guide →
- llms.txt #
- A Markdown file at a site's root that gives AI tools a short, curated map of the site. Guide →
- MCP #
- Model Context Protocol, the open standard that lets AI applications connect to tools and data. Guide →
- Mention #
- Your brand named in an AI answer, with or without a link. Guide →
- Passage #
- A self-contained section of a page that an AI engine can quote. Engines retrieve and cite passages, not whole pages. Guide →
- potentialAction #
- Schema.org markup describing an action a page supports, such as search, order or reserve. Guide →
- Retrieval #
- The step where an AI engine selects passages from candidate pages before writing its answer. Guide →
- sameAs #
- A schema property listing your official profiles elsewhere (Wikidata, LinkedIn, Google Business Profile) so AI systems connect them to you. Guide →
- Schema.org #
- The shared vocabulary of types (Organization, Product, Service, FAQPage) used for structured data. Guide →
- Shadow gap #
- A query where Google shows an AI Overview and your page ranks in the top 10, but the AI Overview does not cite you. Guide →
- The share of all citations (or mentions) across a set of queries that go to you rather than to others. Guide →
- Speakable #
- A schema property marking the passages of a page best suited to be read aloud or quoted. Guide →
- WebMCP #
- Markup and browser APIs that describe a page's actions so an AI agent in the browser can perform them. Guide →
- Wikidata #
- The open knowledge base of entities that many AI systems ground facts against. Guide →
PreviousCommon problems: limits, checks, plans and pages Citerank cannot readNextPlaybook: your first AI visibility audit of a site
Something here does not match what you see in the app? Tell us and we will fix the guide.