The B2B search glossary, defined.

The practice of structuring a brand's content and off-site signals so that AI systems like ChatGPT, Perplexity, Gemini and Google AI Overviews cite and.

The practice of structuring content so it wins the single, direct answer that an answer engine or AI assistant returns to a question, rather than one position.

Google's AI-generated summaries that appear at the top of search results, answering a query directly and citing a small set of sources instead of only listing.

An AI system that answers a query by synthesising a single response from multiple retrieved sources, rather than returning a ranked list of links for the user.

The architecture most AI search systems use to answer a query: they retrieve relevant passages from an external source such as the web, then generate an answer.

A reference to a source that an AI system names or links inside a generated answer, marking that source as one of the pages the model drew on to produce its response.

The difference between how often AI systems cite one brand versus its competitors for the same set of buyer questions, exposing who owns the AI answer in a.

A GEO metric that measures how often a brand is named or cited in AI-generated answers across a defined set of buyer prompts, relative to its competitors.

A self-contained block of content, usually the opening sentences of a page or section, that states a complete, quotable answer to a single question, formatted.

The share of claims on a page that are paired with a verifiable proof point, such as a specific statistic, a cited source, or a named example, rather than left.

A plain-text file placed at a website's root that gives AI systems a curated, Markdown-formatted map of the site's most important content, so models can find.

OpenAI's crawler for its search product: it fetches and indexes web pages so they can be retrieved and cited when ChatGPT search answers a user's query.

The crawler Perplexity uses to discover and index web pages so they can be retrieved and cited as sources in its AI-generated answers.

Two different OpenAI crawlers: GPTBot collects data used to train models, while OAI-SearchBot builds the live search index that lets ChatGPT retrieve and cite.

A large public dataset of crawled web pages, gathered by its crawler CCBot, that many AI models and tools use as a training and reference corpus.

Structured data added to a page's code, using the Schema.

A Schema.

The process by which a search or AI system decides which real-world thing a name refers to, linking a brand, person or product to a single, disambiguated.

A structured network of entities, people, companies, products, places, and the relationships between them, that a search or AI system uses to understand and.

The depth and breadth of a brand's content across a subject area, which signals to search and AI systems that it is a credible source on that topic and makes.

E-E-A-T stands for Experience, Expertise, Authoritativeness and Trustworthiness: Google's framework for assessing content quality, which has now become a.

The rate at which a page loses its AI citations as it ages without being updated, reflecting that recency is now a hard ranking signal in AI search rather than.

The practice of building a brand's reputation across third-party sites, communities and publications so that AI systems encounter consistent, trusted signals.

References to a brand by name across the web that do not include a hyperlink, and which AI systems still read as trust and authority signals when deciding whom.

When an AI system states something false or misleading about a brand, such as wrong features, wrong category or invented facts, because it lacks clear.

A content structure that puts the direct answer to a question first, at the top of a page or section, before any context or build-up, so an AI system can.

The practice of measuring AI visibility by running a fixed set of buyer prompts across AI engines on a cadence and recording which brands get cited, since.

A curated, fixed set of buyer-intent prompts that a brand runs across AI engines on a regular cadence to measure its citation visibility consistently over time.

The visits a website receives from AI assistants and answer engines, such as ChatGPT, Perplexity and Google AI Overviews, when a user follows a cited source.

A backlink-analysis technique that finds the domains linking to several of your competitors but not to you, surfacing the specific sites most likely to link to.

A 0-to-100 score, popularised by Ahrefs, that estimates the strength of a website's backlink profile relative to other sites, used as a rough proxy for overall.

The number of unique websites that link to a given site, a backlink metric that counts distinct linking sources rather than the total number of individual links.

The number of URLs a search engine's crawler will fetch on a site within a given window, set by crawl rate and crawl demand.

A plain-text file at a site's root that tells crawlers which paths they may or may not request. It controls crawling, not indexing.

Instructions such as Allow and Disallow in robots.txt that tell crawlers which URL paths to fetch or skip, matched per user-agent.

Examining a server's raw access logs to see exactly which URLs each crawler requested, when, and what status code they received.

A page with no internal links pointing to it, leaving crawlers and users unable to reach it through normal navigation.

The process by which a search engine stores and organizes a crawled page so it can be retrieved and shown in results.

A directive, via meta tag or HTTP header, that tells search engines not to store a page in their index even though they may crawl it.

The process of telling search engines which URL is the preferred version of a set of identical or near-identical pages, consolidating signals onto one.

A link element in a page's head that names the preferred URL for that content, used to consolidate duplicates onto one indexable version.

Substantively identical or very similar content available at more than one URL, which splits ranking signals and forces engines to pick one version.

A condition where a site has many low-value, thin, or duplicate URLs in the index, diluting crawl attention and the site's perceived quality.

The process of turning a page's HTML, CSS, and JavaScript into the final DOM that a browser displays and a crawler reads.

An approach where the server sends minimal HTML and the browser builds the page content with JavaScript after load.

An approach where the server builds the full HTML of a page before sending it, so content is present in the initial response.

The process where JavaScript attaches interactivity to server-rendered HTML in the browser, turning static markup into an interactive app.

The discipline of ensuring JavaScript-heavy sites are crawlable, renderable, and indexable, so content built by scripts is visible to engines.

The way a site's pages are organized and linked, determining how crawlers and users navigate from the homepage to any given page.

Links from one page of a site to another on the same site, used to aid discovery, distribute authority, and signal topical relationships.

A machine-readable file listing a site's important URLs to help search engines discover and prioritize them for crawling.

A secondary navigation trail showing a page's position in the site hierarchy, typically rendered as Home / Section / Page.

Filtering controls (by price, color, size, etc.) that generate URL combinations, often creating large numbers of crawlable low-value pages.

The format and organization of a site's URLs, including paths, slugs, and parameters, which affects clarity, crawlability, and consolidation.

A set of Google metrics measuring real-world loading, interactivity, and visual stability: LCP, INP, and CLS.

A Core Web Vital measuring how long it takes for the largest visible content element to render, used as a loading-speed proxy.

A Core Web Vital measuring how much a page's visible elements move around unexpectedly during loading, scoring visual stability.

A Core Web Vital measuring a page's overall responsiveness to user interactions, which replaced First Input Delay in 2024.

The time between a request and the first byte of the server's response, a measure of server responsiveness and a component of load speed.

Three-digit codes a server returns with each response, signaling success (200), redirection (3xx), client error (4xx), or server error (5xx).

A permanent redirect status code that sends users and crawlers from an old URL to a new one and passes ranking signals to the target.

A temporary redirect status code that points to a different URL while signaling the original should remain indexed.

A sequence of two or more redirects between the requested URL and the final destination, adding latency and risking signal loss.

A page that shows a not-found or empty message to users but returns a 200 success status, confusing crawlers about whether it exists.