AI Visibility Glossary

The foundation every website needs before AEO or GEO can work. SEO covers the signals search engines use to understand, index, and rank a page: technical health, content structure, heading hierarchy, schema markup, and page speed.

CitedScore runs an SEO score as the base layer because AI engines and search engines share the same structural requirements. A site with broken SEO gives AI platforms nothing solid to build on.

The practice of structuring content so AI-powered answer engines can extract, understand, and cite it in generated responses. Where SEO optimizes for ranking position, AEO optimizes for extraction: clear question-and-answer structure, FAQ schema, llms.txt, and content depth that lets an AI platform pull a precise, quotable answer.

CitedScore measures AEO Readiness as the second score in every report. It tells you how citation-ready your site is before the live prompts run.

The practice of building brand visibility in AI-generated answers. Where SEO targets a position in a list, GEO targets a mention in a response. GEO works at the brand and entity level: how consistently your name appears across third-party sources, whether AI platforms have encountered you enough to trust you, and whether your on-site signals confirm what those external sources say.

CitedScore's GEO Citation Score measures this with live prompts across ChatGPT, Google AI Overviews, Gemini, and Perplexity.

A ghost citation occurs when an AI engine uses a website's content to power its response, links the domain as a source, and never names the brand in the answer text.

The measured prevalence, with sources and dates, lives on the /sources page. CitedScore classifies every result so you know whether you are getting credit or just providing the raw material.

A mention is when an AI engine names your brand directly in the text of a generated answer. The reader encounters your name. This is the form of AI visibility that moves buying decisions because the buyer gets a recommendation by name, not just a source link in the footnotes.

The fix for a missing mention is different from the fix for a missing citation. That is why CitedScore classifies them separately and scores each lever in your report.

A citation is when an AI engine includes your domain as a source link in its response. If the brand name does not appear alongside the link, it is a ghost citation.

Citations without mentions have limited commercial value because the reader has no brand name to remember or act on. It happens often enough to have its own name: the ghost citation.

The composite score CitedScore assigns after running live prompts across all four AI platforms. It combines mention rate, citation rate, and Mentioned vs Cited classification into a single number weighted by each platform's share of AI-driven buyer interactions.

Google AI Overviews and ChatGPT receive higher weight because they reach more buyers. The score tracks business impact rather than raw appearances, and it changes as your signals improve. The exact weights and constants are published at /how-it-works.

Machine-readable code embedded in a web page that tells search engines and AI platforms exactly what the page contains. JSON-LD is the standard format. Schema types relevant to AI citation include Organization, LocalBusiness, FAQPage, Article, and Person.

Without schema, AI platforms infer what your site is about from context. With schema, they have verified, structured signals to extract and cite. CitedScore records which schema types each sampled page carries and scores your homepage schema in the AEO Readiness layer of every report.

A plain-text file placed at the root of a domain (/llms.txt) that gives AI language models explicit guidance about what the site is, what it does, and what content is available for use in AI responses. It is the AI-native counterpart to robots.txt, the file that tells crawlers what they may access.

An llms.txt file signals that a site is AI-aware and has structured its content deliberately for AI consumption. CitedScore checks for llms.txt presence and quality in the AEO Readiness layer.

The process by which AI engines verify a brand by cross-referencing it across multiple independent sources: the brand's own website, third-party directories, review platforms, media coverage, and community discussions. An entity that appears in one place is easy to overlook. An entity that appears consistently across authoritative sources becomes citation-worthy.

Entity corroboration is the primary mechanism behind AI mentions. It is why brand footprint across the web matters as much as on-site structure, and why CitedScore audits both levers.

A search that ends without the user clicking any result because the search engine or AI engine resolved the query on the results page itself. In the US, most Google searches already end this way, and only a fraction ever reach the open web; the current measured figures, with sources and dates, are on the /sources page.

The traffic that used to flow through those clicks now flows through AI recommendations. The businesses named in those recommendations capture the audience that used to arrive through search. The businesses that do not appear in those recommendations are invisible to that audience.

A buyer persona is a structured profile of a specific type of customer, built from the vocabulary, intent, and phrasing patterns that type of buyer actually uses when asking AI engines questions. Different buyer types ask the same underlying question in different ways, and AI engines respond to those differences. A business can be visible to one buyer type and completely absent for another.

Generic AI visibility tools run the same prompts against every domain. CitedScore generates four buyer personas from your site before running a single prompt, one for each buying stage: First Thought, Exploring Options, Evaluating Options, and Ready to Decide. Each persona reflects a real buyer type for your business, in the vocabulary of that stage. The prompts for each span both registers, short conversational and long structured, and one question per persona is asked both ways to measure the divergence, because the two formats produce dramatically different AI responses.

The per-persona results appear in the GEO section of your report. They tell you which buyer types find you, which platforms surface you for them, and where you're invisible to the audience most likely to convert.

The Perception Gap is the distance between three things: how you describe your own business, how your site presents your business to AI engines, and how AI engines actually describe you in generated answers. When the three diverge, the gap is where visibility and revenue leak.

A business whose site presents it as a premium brand for experienced buyers but whose AI appearances describe it as a generic budget option has a Perception Gap that no amount of ad spend resolves. The problem is structural: the signals AI reads are not telling the story the business intends. The fix is a change in what the site communicates and how the brand appears across the web.

CitedScore generates the Perception Gap panel in every report. It compares the Company Profile extracted from your site against the patterns in your live AI results across all four platforms. The gap between how your site signals your identity and how AI describes you is where the audit focuses first.

A 0–100 score is precise, but it does not tell you what it means. The CitedScore Maturity Ladder is a five-rung scale that places your brand according to your own measured GEO evidence: how often you are named, how often you are cited, and where you land when you are named. It is how CitedScore turns a measured score into a named stage of AI visibility.

The five rungs, evidence-first: 1. Invisible (not appearing in AI answers for your category), 2. Overlooked (named occasionally, rarely leading the answer), 3. Mentioned (consistently named across AI answers), 4. Recommended (named early and used as a source), and 5. Cited Authority (the default citation for your category).

Your composite score sets the candidate rung. Evidence caps can only lower it from there, never raise it: a strong Technical SEO score cannot buy a rung your live GEO results do not support. Nothing flows from the ladder back into the score. It reads your Technical, AEO, and GEO scores and reports where they place you; every report also shows the specific, evidence-backed gate to the next rung.