Glossary
SEO, GEO and AEO terms, defined plainly
No recycled textbook phrasing. Every definition below that describes how one of our own tools checks something is written to match that tool's actual code, not a generic idea of what the term means.
SEO
Search Engine Optimization. The practice of structuring a page so a traditional search engine — Google, Bing — can crawl it, understand what it covers, and rank it for a matching query. The target is a listing on a results page that a person then clicks.
GEO (Generative Engine Optimization)
The same underlying work as SEO — clean structure, structured data, a crawlable site — aimed at a different destination: getting a generative AI system to read, cite or quote a page inside an answer it writes, rather than just linking to it. Our AI Visibility Checker scores a site against specific signals — which AI crawlers may fetch it, structured data, on-page basics, llms.txt, a sitemap — rather than reporting a single, unexplained "GEO score."
AEO (Answer Engine Optimization)
Optimizing content to be the thing an answer engine — ChatGPT, Perplexity, an AI Overview — states directly as the answer, rather than one of several ranked links. In practice it overlaps heavily with GEO: FAQPage schema, direct answers near the top of a page, and clear headings serve both goals at once, which is why our AI Visibility Checker does not try to score the two separately.
AI crawler
An automated fetcher operated by an AI company — to gather training data, to build a live search index, or to browse a page a user asked about — as distinct from a general-purpose search crawler like Googlebot. Our AI Visibility Checker checks a site's robots.txt against ten named AI crawlers, including OpenAI's GPTBot and ChatGPT-User, Anthropic's ClaudeBot, PerplexityBot and Google-Extended, and reports which ones may fetch the homepage.
robots.txt
A plain-text file at a site's root that tells crawlers, in named User-agent groups, which paths they may or may not fetch — the format is standardized in RFC 9309. It can block or allow one specific AI crawler by its own user-agent token without touching any other crawler, and it can declare a Sitemap: line pointing at the site's XML sitemap. You can see this site's own file at /robots.txt.
llms.txt
An emerging, informal convention: a plain-markdown file at /llms.txt listing a site's most important pages with one-line summaries, meant as a curated map for AI systems that would otherwise have to guess which pages matter. No standards body governs it and no engine treats it as authoritative yet, which is why our AI Visibility Checker only ever flags a missing one as a warning, never a failure. See this site's own file at /llms.txt.
XML sitemap
An XML file — usually at /sitemap.xml — listing a site's URLs so a crawler can discover pages without following every link on every page. It should be declared in robots.txt with a Sitemap: line; our AI Visibility Checker checks both whether robots.txt declares one and whether /sitemap.xml itself responds, and flags the gap as a fix if the file exists but isn't declared where crawlers actually look for it.
JSON-LD / structured data
A <script type="application/ld+json"> block embedding facts about a page as machine-readable JSON, using the schema.org vocabulary, so a crawler can read what a page is directly instead of inferring it from prose. Our Schema Markup Generator builds a valid block; the AI Visibility Checker parses every JSON-LD block on a homepage independently, so one malformed block can't hide the valid ones next to it, and reports every type it finds.
Schema.org
The shared vocabulary of types — Organization, WebSite, LocalBusiness, Person, FAQPage, Article, and hundreds more — that a JSON-LD block declares its type from. It is maintained jointly by Google, Microsoft, Yahoo and Yandex, which is why the same markup is readable by all of their crawlers instead of needing a separate format per engine.
FAQPage schema
A schema.org type that marks up a genuine list of visible question-and-answer pairs on a page, letting an answer engine lift a specific pair directly into a result. It should only ever describe content actually rendered on the page — marking up hidden or absent questions is a structured-data violation, not a shortcut. Our FAQ Schema Generator builds a valid block from your own questions and answers.
Canonical URL
The URL declared in a page's <link rel="canonical"> tag as the authoritative version of that content, used whenever the same or near-duplicate content is reachable at more than one address — with and without a trailing slash, with tracking parameters, across http and https. Search and AI engines index and cite the canonical URL, not necessarily the one a visitor actually landed on.
Core Web Vitals (LCP, CLS, INP)
Google's three field metrics for how a page feels to actually use: Largest Contentful Paint (how long the biggest visible element takes to render), Cumulative Layout Shift (how much content jumps around while loading), and Interaction to Next Paint (how quickly the page responds to a click or tap). Our Website Speed Test reports all three.
UTM parameters
Query-string parameters — utm_source, utm_medium, utm_campaign, and optionally utm_term/utm_content — appended to a link so an analytics tool can attribute the resulting visit to a specific channel and campaign instead of lumping it in as generic referral traffic. Build your own with the UTM Builder.
ROI vs ROAS
Return on Investment is (revenue − cost) ÷ cost — it accounts for every cost in a campaign, including margin. Return on Ad Spend is revenue ÷ ad spend alone — a narrower ratio that only looks at media cost. A campaign can report an impressive ROAS and still lose money once product cost and overhead are counted, which is the gap our Marketing ROI Calculator shows side by side.
Meta description
The text in a page's <meta name="description"> tag — a one- or two-sentence summary a search engine may show under the title in results, and an AI system may quote when it needs a quick description of what a page covers. It has no direct effect on ranking, but a missing one is one of the on-page basics our AI Visibility Checker flags, because it's free context an engine would otherwise have to guess at.
Open Graph tags
A set of <meta property="og:..."> tags — og:title, og:description, og:image — originally built for social-media share cards, now also read by many AI and social crawlers to understand a page's context without fetching and parsing the whole document. Open Graph uses the property attribute, not name, which is the one detail that trips up a hand-written meta tag.
Alt text
The alt attribute on an <img> tag, describing what an image shows for a screen reader and for any crawler that cannot run a vision model over every image on a page. Our AI Visibility Checker treats a homepage with under 80% alt-text coverage across its images as a signal worth fixing; only a non-empty value counts, since an empty alt="" gives an engine nothing to read even though it's a valid pattern for a purely decorative image.
E-E-A-T
Google's Experience, Expertise, Authoritativeness and Trustworthiness framework, used in its search quality rater guidelines to judge content. It is not a directly measurable ranking factor — it's a lens: does a page show it was written by someone with real, checkable experience of the subject, on a site that discloses who is actually accountable for it. Every tool page on this site points its byline at a real, named author page for exactly that reason.
Knowledge Graph
A search engine's own database of entities — people, organizations, places — and the verified facts and relationships between them, used to disambiguate what an entity actually is before it appears in an AI answer. Organization and Person schema markup is one of the inputs an engine draws on to build or correct an entry, though inclusion is never guaranteed just because the markup exists.
noindex
A directive — in a page's <meta name="robots"> tag or its X-Robots-Tag HTTP header — telling every search and AI engine not to index the page at all. It differs from the milder noai directive, which only asks AI systems not to train on or reuse a page's content; a noai page can still be indexed and cited, but a noindex page has nothing to retrieve or cite in the first place. Our AI Visibility Checker reports it as the single most severe finding it can surface, even though — like noai — it deliberately leaves it out of the numeric score, since opting out of indexing is sometimes a deliberate choice rather than a mistake.
