AI search glossary

The words people use for getting found in ChatGPT, Perplexity, Gemini and Google's AI features, defined plainly, with what each one changes about how AI engines find and cite you.

Agentic search: Agentic search is a form of AI search in which the system acts as an agent: it plans a series of searches, visits and reads web pages, compares options, and can take actions such as filling forms or starting a purchase on the user's behalf, instead of returning a single list of links or one answer.
AI brand monitoring: AI brand monitoring is the ongoing practice of asking AI assistants and AI search engines about a brand and its category, then recording how often the brand appears, what the answers say about it, whether those claims are accurate, and which sources they cite.
AI citations: An AI citation is a link or source reference that an AI search engine attaches to its answer to show which web page supplied a piece of information, such as the numbered sources in Perplexity or the supporting links in Google AI Overviews.
AI rank tracking: AI rank tracking is the practice of asking AI engines the same set of prompts on a schedule and recording whether, where and how often a brand appears in the answers, so changes in AI visibility can be followed over time.
AI referral traffic: AI referral traffic is website visits that arrive from a link in an AI assistant or AI search answer, such as ChatGPT, Perplexity, Gemini, Copilot or Claude, identified in analytics by the referrer or by a tag like utm_source=chatgpt.com.
AI SEO: AI SEO is an umbrella term with two common meanings: optimizing a website so AI search engines and assistants find, cite and recommend it, and using AI tools to carry out search engine optimization work.
AI share of voice: AI share of voice is the portion of brand mentions or citations in AI-generated answers that go to one brand, compared with its competitors, across a defined set of prompts, engines and time period.
AI visibility: AI visibility is how often, and how favorably, AI assistants and AI search engines mention, cite or recommend a brand, product or site when people ask them questions the brand should be an answer to.
Answer engine optimization (AEO): Answer engine optimization (AEO) is the practice of structuring content so that an answer engine, such as a featured snippet, a voice assistant or an AI assistant like ChatGPT or Copilot, can take it as the direct answer to a question.
Brand mentions: A brand mention is any reference to a brand by name on a page, forum, review site, news article or AI-generated answer, whether or not it links to the brand's website.
Canonical tag: A canonical tag is a link element, written as rel="canonical" in a page's head, that tells search engines which URL is the preferred version of a page when the same or very similar content is reachable at several URLs.
Content freshness: Content freshness is how up to date a page's information is, and how clearly the page signals when it was published and last meaningfully updated, which matters most for queries where recent information is expected.
Conversational search: Conversational search is a way of searching where the user asks in natural language and keeps refining with follow-up questions, while the engine remembers the earlier turns and answers each one in context, usually with links to its sources.
Digital PR: Digital PR is public relations done online: earning coverage, mentions and links from news sites, trade publications, blogs, newsletters, podcasts and creators by giving them something worth reporting, rather than paying for the placement.
E-E-A-T: E-E-A-T stands for experience, expertise, authoritativeness and trustworthiness: the qualities Google's search quality raters use to judge whether a page is reliable and helpful, with trust named as the most important.
Entity SEO: Entity SEO is the practice of helping search engines and AI models recognise a brand, person, product or place as a distinct entity, with consistent facts and clear links to the sources that describe it, rather than as a string of keywords.
FAQ schema: FAQ schema is structured data using the schema.org FAQPage type to mark a page as a list of questions with answers, each question as a Question with an acceptedAnswer. Google used it for an FAQ rich result, which it stopped showing in May 2026.
Featured snippet: A featured snippet is a Google search result where a short extract from a page, chosen automatically by Google's systems, is shown first and above the page's title and link because Google judges that it answers the query well.
Generative engine optimization (GEO): Generative engine optimization (GEO) is the practice of making content more likely to be used, quoted and cited when a generative engine, such as ChatGPT, Perplexity or Google AI Overviews, writes an answer from several web sources.
IndexNow: IndexNow is an open protocol that lets a website notify participating search engines, with one HTTP request, that specific URLs have been added, updated or deleted, so they can be recrawled sooner instead of waiting to be discovered.
JavaScript SEO: JavaScript SEO is the practice of making sure pages that rely on JavaScript to load or build their content can still be crawled, rendered and indexed by search engines, usually by serving the important content in the initial HTML through server-side rendering or pre-rendering.
JSON-LD: JSON-LD (JSON for Linking Data) is a W3C standard for describing things and the links between them in plain JSON. On websites it is placed in a script tag so search engines and other machines can read facts about the page, such as who publishes it and what product it sells.
Knowledge graph: A knowledge graph is a structured database that stores facts as entities (people, companies, places, products, ideas) connected by labelled relationships, so software can look up what something is and how it relates to other things.
Knowledge panel: A knowledge panel is the information box Google shows in search results for people, organisations, places and things, generated automatically from its Knowledge Graph and sources across the web to give a quick snapshot of the entity.
LLM hallucination: An LLM hallucination is output from a large language model that contains false or misleading information presented as fact, such as an invented citation, a wrong price, or a feature a product does not have.
LLM SEO (LLMO): LLM SEO, also called LLMO or large language model optimization, is the practice of influencing how large language models such as GPT, Claude and Gemini understand, describe and recommend a brand, both through what they learn in training and what they retrieve when answering.
llms.txt: llms.txt is a proposed standard, introduced by Jeremy Howard in September 2024, for a Markdown file at /llms.txt that gives language models and AI agents a short summary of a site and a curated list of links to its most useful pages.
Organization schema: Organization schema is structured data using the schema.org Organization type to describe a company, nonprofit or other organization: its official name, website, logo, contact details, identifiers, and the other web pages (through sameAs) that describe the same organization.
People Also Ask: People Also Ask is a Google search feature that shows a group of questions related to the original search; expanding a question reveals a short answer taken from a web page, shown as a featured snippet with a link.
Product schema: Product schema is structured data using the schema.org Product type, usually with a nested Offer, to describe an item for sale: its name, image, brand, identifiers such as GTIN and SKU, price, currency, availability and reviews.
Query fan-out: Query fan-out is the technique AI search engines use to answer one question by splitting it into subtopics and running many related searches at once, then writing an answer from the combined results.
Reddit SEO: Reddit SEO is the practice of making a brand, product or expert useful and visible in Reddit discussions, because those threads rank in search engines and are read by AI engines when they answer questions about products and services.
Retrieval-augmented generation (RAG): Retrieval-augmented generation (RAG) is a method in which a language model first retrieves relevant documents from an external source, such as a search index or a document database, and then generates its answer using those documents as context instead of relying only on what it memorized in training.
Schema markup (structured data): Schema markup is structured data added to a web page, using the shared schema.org vocabulary, that labels what the page is about (a business, a product, an article, an event) in a format search engines can read directly instead of inferring it from the text.
Semantic SEO: Semantic SEO is the practice of creating content around the meaning and intent behind searches, covering a topic's related concepts, entities and questions, so that search engines that understand language can match the page to queries it never repeats word for word.
Topical authority: Topical authority is the degree to which a site is recognised as an expert, go-to source on a particular subject, shown by deep coverage of that subject, original work, and citations and recommendations from others.
User agent: A user agent is the software that makes a web request on someone's behalf, such as a browser or a crawler, and the User-Agent string is the HTTP request header in which that software names itself, its version and often its platform.
Wikidata: Wikidata is a free, collaborative, multilingual knowledge base run under the Wikimedia umbrella that stores structured facts about entities as items and statements, released under a CC0 public domain dedication so anyone can reuse them.
XML sitemap: An XML sitemap is a file, usually at /sitemap.xml, that lists the URLs on a website a search engine should crawl, optionally with the date each page last changed, following the protocol published at sitemaps.org.
Zero-click search: A zero-click search is a search that ends on the results page without the user clicking through to any website, because the answer was shown directly in a snippet, panel, AI summary or other search feature.