Glossary

The language of search and AI, in plain English

Direct definitions of the terms that decide whether your business gets found and recommended. No jargon: every entry answers the question in its first sentence.

SEO, AEO and GEO: the difference in one table

The three acronyms describe different jobs built on the same foundation. It is the most common confusion in the market, and this table settles it in thirty seconds.

	SEO	AEO	GEO
Where the brand shows up	In Google's list of links	In the answer box above the list	Inside the text the AI writes
Unit of competition	The keyword	The specific question	The whole question, in natural language
What the user does	Compares and clicks	Reads the answer, sometimes clicks	Receives 2 or 3 brands, pre-filtered
How it is measured	Position and organic traffic	Snippets and PAA boxes won	Presence and sentiment inside AI answers
Winning format	A complete, well-structured page	Short answer, list or table	Self-contained passage, with a source
Main lever	Authority and content	Structure and clarity	Entity identity and verifiable reputation
Typical timeline	3 to 6 months	Weeks to months	Weeks for access, months for citation

GEOGenerative Engine Optimization

GEO (Generative Engine Optimization) is the work of optimising a company's digital presence so that generative AI systems such as ChatGPT, Gemini, Perplexity, Claude and Google's AI answers find it, cite it and recommend it in their responses.

It is the equivalent of SEO for the era of ready-made answers. With SEO a business competes for a position among ten links and earns the click; with GEO it competes for the citation inside an answer that already arrives complete, naming two or three brands. Whoever gets cited wins the customer; whoever doesn't never learns they lost.

SEOSearch Engine Optimization

SEO (Search Engine Optimization) is the set of practices that gets a website into the top positions of search engines for the searches that bring customers.

It spans three layers: technical (speed, indexing, structure, structured data), content (the right pages for terms with buying intent) and authority (mentions and links from credible sources). With AI answers, SEO stopped being the finish line and became the path: the AI models themselves use search results as the source for deciding who to cite.

AEOAnswer Engine Optimization

AEO (Answer Engine Optimization) is optimising to own the direct answer: Google's featured snippet, the People Also Ask box and the reply a voice assistant reads aloud.

It sits between SEO and GEO. SEO competes for the position in the list, AEO competes for the answer box above the list, and GEO competes for the citation inside the text the AI writes. All three rest on the same foundation: clear, structured, sourced content that stays consistent across channels.

See also: SEO · GEO · Featured snippet

AI Overview

An AI Overview is the AI-generated summary Google shows at the top of results, above the blue links, answering the question right there.

It appears on roughly 45% of searches and cites a handful of linked sources. That changes the measuring stick: ranking fifth is worth little if the answer above already resolved the user's question. The goal becomes being one of the sources cited inside the summary.

See also: GEO · Zero-click search · AEO

Featured snippet

A featured snippet is the box Google shows at the top of results containing a passage lifted from a page that answers the search query directly.

It is earned, not bought: Google picks the page that answers most clearly and in the right shape (short paragraph, list or table). It is the training ground for AEO: whoever structures content well enough to win snippets is usually who AI cites too.

Zero-click search

A zero-click search is a search that ends without the user clicking any result, because the answer already appeared on the results page itself.

It used to be a traffic problem; it became an existence problem. If the question is resolved without a click, your brand has to be inside the answer, named and cited, not merely listed beneath it. That is the practical reason GEO stopped being optional.

llms.txt

llms.txt is a plain-text file at the root of a domain that introduces the business to language models: who it is, what it does, which pages matter and how to cite it.

It follows the same logic as robots.txt (what can be crawled) and sitemap.xml (what exists), but aimed at AI models rather than search crawlers. It is recent and still rarely adopted, which makes it a cheap differentiator for whoever publishes first.

robots.txt

robots.txt is the file at the root of a site that tells each automated bot what it may and may not crawl.

In GEO it is step zero: if GPTBot, PerplexityBot or ClaudeBot are blocked, no other effort matters. Watch the detail almost everyone gets wrong: beyond the indexing crawlers there are live-fetch agents (ChatGPT-User, Perplexity-User, Claude-User). Blocking those removes your brand from the answer at the exact moment it would have been cited.

Crawler

A crawler is the automated program that walks the web reading pages to feed a search engine or an AI model.

Each platform has its own, with its own name: Googlebot, Bingbot, GPTBot (OpenAI), ClaudeBot (Anthropic), PerplexityBot. A site that requires JavaScript to show its content serves empty pages to many of them, which is why server-side rendering stopped being a technical detail and became a visibility prerequisite.

SSRServer-Side Rendering

SSR (server-side rendering) is the technique where the server delivers the page with all its content already in the HTML, instead of sending an empty shell for the browser to fill in with JavaScript.

It matters for GEO because most AI crawlers do not execute JavaScript. A client-rendered site can look perfect to a human visitor and be completely invisible to ChatGPT, which loads the same URL and finds a blank page.

Structured dataschema.org / JSON-LD

Structured data is markup, invisible to the reader, that declares in machine-readable form what each piece of information on a page means: this is a business, that is an address, that other thing is a frequently asked question.

It is the difference between a machine reading "50 Main Street" as arbitrary text and understanding it as a business address. The standard is schema.org and the recommended format is JSON-LD. For AI, the decisive block is business identity: legal name, registration number, address, phone and links to official profiles.

sameAs

sameAs is the structured-data property that lists an entity's official profiles (Instagram, LinkedIn, Facebook, Google Business Profile), declaring that they all belong to the same company.

It is what resolves identity. Without sameAs, your site, your social profiles and your directory listings are unrelated fragments, and the AI has no basis for joining them or for telling them apart from similarly named businesses. It is the highest-leverage, lowest-effort fix in all of GEO, and the most frequently skipped.

See also: Structured data · NAP · Entity

Entity

An entity, in search vocabulary, is a real-world thing the system recognises as unique and distinct (a company, a person, a place, a product) and about which it accumulates facts from many sources.

Search engines and AI models do not recommend texts: they recommend entities. The question that decides everything is whether the system can assert "this business exists, it is this one, and not that other one with a similar name". Entity resolution is the precondition for citation.

See also: sameAs · NAP · Structured data

NAPName, Address, Phone

NAP is a business's name, address and phone, and the golden rule of local search is that it must appear identical on every channel where the business exists.

Disagreement between channels, such as two different phone numbers or an old address in a directory, is the fastest way to lose the trust of a system whose job is cross-referencing sources. Your website has to be the canonical source that settles the rest.

See also: Google Business Profile · Map pack · Entity

Google Business Profile

A Google Business Profile is the free listing a business maintains with Google, feeding its panel in search results and on Google Maps.

For a local business it is the most valuable digital asset after the website itself, and sometimes more. Primary category, reviews, response rate, photos and hours decide who appears in the map pack. AI assistants also consult it when answering "near me" questions.

Map packlocal pack

The map pack is the block of three businesses Google shows alongside a map when a search has local intent, such as "dentist near me".

Three slots for an entire city, sitting above the organic results. Winning one is usually worth more than the first organic position, because local search carries immediate buying intent.

See also: Google Business Profile · NAP

E-E-A-TExperience, Expertise, Authoritativeness, Trustworthiness

E-E-A-T is the set of criteria Google uses to assess content quality: lived experience, the expertise of whoever wrote it, recognised authority, and the trustworthiness of the source.

It is not a number inside the algorithm: it is the yardstick guiding quality raters, and it became more relevant with AI, which has to decide who to trust before citing. In practice it translates into verifiable things: an identifiable author with a track record, sourced data, real reviews and consistent information.

FAQPage

FAQPage is the structured-data type that marks a block of questions and answers on a page, telling machines which text is the question and which is the answer.

It is the highest-return format in AEO and GEO, with one condition: the questions and answers must be visible to the reader, not hidden in markup. Each answer has to work on its own, away from the page, because that is how it will be quoted.

See also: AEO · Structured data · Featured snippet

hreflang

hreflang is the markup that tells search engines what language and region a page was made for, and where the equivalent versions live.

For anyone serving more than one market, it is what separates "generic English" from "the US market". Declaring en-US on a page built for the United States is an explicit targeting signal. Watch the common mistake: hreflang pointing at a URL that redirects is discarded, and the whole pair is invalidated.

Want to know where your business sits on all three maps?

The Visibility X-Ray measures all three fronts at once: where you rank on Google, whether you own the direct answer, and whether AI cites your brand.

Free diagnosis