SEO and GEO Glossary: search and AI visibility terms

The terms of search engine optimisation and of visibility in AI assistants, defined in plain words. Each definition links to the page where we explain how we apply it for businesses in Luxembourg.

A

AI Overviews

Summaries generated by Google's AI and shown above the standard results for some searches, with links to the pages used for the answer. Google states that no special markup is required: a page must be indexed and eligible to be shown with a snippet to appear there. How we approach GEO.

Alt attribute

Alternative text for an image, read by screen readers and used by search engines to understand what the image shows. It describes the image in one useful sentence, without a list of keywords. How we approach SEO.

Anchor text

The clickable text of a link. It tells the reader and the search engine what they will find on the destination page. A descriptive anchor (“SEO audit for an accounting firm”) is better than “click here”. How we approach SEO.

Authority

The trust a search engine places in a site, built in particular through links from reliable sites and the reputation of the brand. Authority scores from third-party tools (Domain Rating, Domain Authority) are those tools' own estimates: Google does not use them. How we approach SEO.

Average position

A page's average rank in search results for a query, calculated by Search Console across all its appearances. It is an average: a page can be third on desktop and twelfth on mobile, or vary depending on the searcher's town. How our audit works.

B

Backlink

A link pointing to your site from another site. It is one of the oldest popularity signals in SEO: an editorial link from a recognised site in your sector or region carries more weight than dozens of unrelated links. How we approach SEO.

Bingbot

The crawler of Microsoft's Bing search engine. Bing's index also feeds other search services, so being crawled properly by Bingbot extends your presence beyond Bing alone. How we approach SEO.

C

Canonical tag

An indication (rel="canonical") that designates, among several URLs with identical or very similar content, the version to index. It stops the search engine from spreading its signals across duplicates, for example a page reachable with and without URL parameters. How we approach SEO.

ChatGPT-User

OpenAI agent that visits a page at the request of a ChatGPT user, for example when asked to look at a website. It is not an automatic crawler: its visit shows that a conversation needed your page. How we approach GEO.

ClaudeBot, Claude-SearchBot and Claude-User

Anthropic's three agents: ClaudeBot collects public pages to train the models, Claude-SearchBot crawls the web to improve Claude's search results, and Claude-User reads a page at a user's request. Each one is controlled separately in the robots.txt file. How we approach GEO.

Content negotiation

An HTTP mechanism by which a client states the format it prefers (Accept header) and the server responds in that format at the same URL. A site can thus serve HTML to browsers and Markdown to AI agents that ask for it. How we approach GEO.

Core update

A major update to Google's ranking systems, announced several times a year. It does not target any particular site: it reassesses the quality of pages overall. A drop after a core update calls for work on quality, not a one-off technical fix. How our audit works.

Core Web Vitals

Three measures of real visitor experience defined by Google: LCP (time to display the largest visible element, good under 2.5 seconds), INP (responsiveness to interactions, good under 200 milliseconds) and CLS (visual stability, good under 0.1). They are shown in Search Console. How we approach SEO.

Crawl budget

The number of pages a search engine is willing to crawl on a site within a given time. Google specifies that the issue mainly concerns very large sites. For a site of a few hundred pages, the priority is rather to avoid useless pages, redirect chains and errors that waste visits. How we approach SEO.

Crawling

A robot's visit that downloads the pages of a site by following links and the sitemap. A page that has never been crawled cannot be indexed. Server log files show which pages robots actually visit, and how often. How we approach SEO.

CTR (click-through rate)

The share of a page's appearances in search results that turn into a click. Search Console gives it query by query. A low CTR on a good position often points to an unappealing title or description. How our audit works.

Customer reviews

Ratings and comments left on a Google Business Profile or on third-party platforms. Google states that the number of reviews and positive ratings contribute to a business's popularity in local results. Replying to reviews, positive and negative alike, is part of the work. How we approach SEO.

D

Duplicate content

Identical or near-identical text found at several URLs, on your site or elsewhere. It is not a penalty in itself, but the search engine picks a single version to show. Product sheets copied from the manufacturer and town pages with interchangeable text are common examples. How we approach SEO.

E

E-E-A-T

Experience, Expertise, Authoritativeness, Trustworthiness: criteria from the guidelines Google gives its search quality raters. It is not a directly measured ranking factor, but a framework for judging whether a page deserves trust: who wrote it, based on what experience, with which sources. How we approach SEO.

Entity

Something a search engine or an AI can identify without ambiguity: a company, a person, a place, a product. Being recognised as an entity, with a name, an address, an activity and profiles that are consistent across the web, means being cited correctly rather than confused with a namesake. How we approach GEO.

F

Featured snippet

A block shown at the top of Google results that quotes a passage from a page to answer the question directly, with a link to the source. Short, clear passages that answer in their first sentence are the best placed to appear there. How we approach SEO.

G

GEO (Generative Engine Optimization)

All the actions that increase the chances that a generative answer engine (ChatGPT, Perplexity, Gemini, Google's AI Overviews) cites a business or relies on its pages to answer. GEO extends SEO: it works on AI crawler access, the business's identity, quotable passages and measurement in assistants. How we approach GEO.

Google Business Profile

Google's free business listing (formerly Google My Business), shown in Maps and in local results: category, opening hours, address or service area, photos, reviews, posts. For a local business, it is often the first thing a customer sees. How we approach SEO.

Google Search Console

Google's free tool for site owners: queries that generate impressions and clicks, indexed or excluded pages, errors, Core Web Vitals, manual actions. It is the reference data source for steering SEO work. How our audit works.

Google-Extended

A token declared in the robots.txt file to indicate whether a site's content may be used to train Google's Gemini models and to ground their answers. It is not a separate crawler: refusing it changes nothing for crawling by Googlebot or for ranking in Search. How we approach GEO.

Googlebot

Google Search's crawler. It mostly crawls pages with a smartphone agent, since Google indexes the mobile version of sites first. Its visit is the condition for any presence in Google's results, including AI Overviews. How we approach SEO.

GPTBot and OAI-SearchBot

Two separate OpenAI crawlers: GPTBot collects pages to train the models, OAI-SearchBot crawls the web for ChatGPT search. You can allow one and block the other in the robots.txt file. Blocking OAI-SearchBot removes the site from ChatGPT's search answers. How we approach GEO.

Grounding

The way an AI assistant bases its answer on documents retrieved at the time of the question rather than on its training memory alone. Grounding is what makes cited sources appear under the answer, and therefore links to the websites selected. How we approach GEO.

H

Hallucination

A false statement produced with confidence by a language model: a service you do not offer, a wrong address, an invented price. Clear, consistent and easily verifiable information on your site and elsewhere reduces the risk of an assistant getting your business wrong. How we approach GEO.

Heading tags (H1, H2, H3)

The heading levels of a page, from H1 (the main title, one per page) to H2, H3 and below. A logical hierarchy helps readers who skim the page, screen readers and search engines, which use it to split the content into passages. How we approach SEO.

Helpful content

A central principle of Google's guidance: content written first for people, offering original information, real experience or analysis, rather than text produced to rank. Google publishes a set of questions for assessing your own pages against this criterion. How we approach SEO.

Hreflang

An attribute that tells Google about the language or regional versions of the same page (for example fr-LU, de-LU, en-LU) so that it shows the right version to each searcher. In Luxembourg, where one business often serves its customers in three languages, it is a decisive setting. How we approach SEO.

HTTPS

The encrypted version of the web protocol, shown by the padlock in the browser. Google uses it as a light ranking signal and browsers warn visitors on unsecured pages: it is now a prerequisite. How we approach SEO.

I

Impressions

The number of times a page appeared in a user's search results, whether or not they clicked. In Search Console, a rise in impressions is often the first sign that a page is making progress, before the clicks follow. How our audit works.

Indexing

Recording a crawled page in the search engine's index, the condition for it to appear in results. A page can be crawled without being indexed, for example if it is judged to be a duplicate or too thin. The “Pages” report in Search Console gives the reasons. How we approach SEO.

IndexNow

An open protocol that lets a site notify participating search engines (including Bing) immediately that a page has been created, modified or deleted. Google does not take part: for Google, the sitemap and Search Console remain the channels to use. How we approach SEO.

Internal linking

All the links between the pages of the same site. It guides visitors to useful pages, helps robots discover every page and indicates which pages matter most. How we approach SEO.

J

JSON-LD

The format Google recommends for structured data: a block of code placed in the page, separate from the visible text, that describes the business, its services or its FAQ using the schema.org vocabulary. How we approach GEO.

K

Keyword

A term or phrase typed by users into a search engine. The job is less about repeating a keyword than about fully covering the need it expresses, with your customers' real vocabulary, in their language. How we approach SEO.

Keyword cannibalisation

A situation where several pages of the same site target the same search and compete with each other. The search engine hesitates between them and none ranks as well as a single, complete page would. The fix is to merge, redirect or differentiate the pages. How we approach SEO.

Knowledge Graph and knowledge panel

Google's base of facts about entities (companies, people, places) and their relationships. When Google recognises an entity, it can show a knowledge panel to the right of the results. A clear and consistent identity across the web helps you get recognised. How we approach GEO.

L

Link building

The work of earning external links to a site: partnerships, press relations, reference content, reputable directories. Google prohibits buying links intended to manipulate rankings; commercial links must carry the sponsored or nofollow attribute. How we approach SEO.

LLM (large language model)

An AI model trained on very large volumes of text to understand and produce language: it is the engine behind ChatGPT, Claude, Gemini or Mistral. On its own, it answers from memory; connected to a web search, it can cite recent sources. How we approach GEO.

llms.txt

A text file placed at the root of a site that presents the site and its main pages in Markdown for language models. Proposed in 2024 on llmstxt.org, it is a convention, not a standard adopted by the major search engines: it replaces neither the sitemap nor the robots.txt file. How we approach GEO.

Local citations

Mentions of a business's name, address and phone number in directories, on the websites of professional chambers or in local media. Their consistency with the Google Business Profile strengthens the search engine's confidence in that information. How we approach SEO.

Local pack

A block of local results with a map and a few businesses, shown by Google for searches with local intent (“accountant Luxembourg”, “plumber Esch-sur-Alzette”). Google says it ranks these results on three criteria: relevance, distance and popularity. How we approach SEO.

Local SEO

Optimising a business for searches tied to a place: the Google map, the local pack, “trade + town” searches. It relies on the Google Business Profile, consistent contact details, reviews and pages about the area actually served. How we approach SEO.

Long tail

Long, precise searches, each typed rarely but numerous overall (“accountant for a self-employed cross-border worker in Esch”). Less contested than generic searches, they bring visitors whose need is already clear. How we approach SEO.

M

Manual action

Penalty applied by a human reviewer at Google when a site breaches the spam policies (bought links, content generated to manipulate rankings, hidden text). It is reported in Search Console, together with the steps for requesting a review once the problem has been fixed. How our audit works.

Markdown

A lightweight text format without layout, which language models read easily. Offering a Markdown copy of each page, referenced in the HTML, gives AI assistants clean text, free of menus and code. How we approach GEO.

Meta description

A summary of a page declared in its code, which Google often shows under the title in search results. It does not directly influence ranking, but a precise and appealing description improves the click-through rate. How we approach SEO.

Mobile-first indexing

Google's way of working, which crawls and evaluates the mobile version of a site first. If content is missing on mobile, it is ignored, even if it appears on the desktop version. How we approach SEO.

N

NAP

Name, Address, Phone: a business's name, address and phone number. Keeping them exactly consistent, down to the last comma, across the website, the Google Business Profile and directories is a foundation of local SEO. How we approach SEO.

Nofollow, sponsored and ugc

Attributes that qualify a link: sponsored for a paid or commercial link, ugc for a link posted by users (comments, forums), nofollow when you do not want to endorse the destination. Google treats them as hints. How we approach SEO.

Noindex

A directive (robots meta tag or X-Robots-Tag HTTP header) asking search engines not to show a page in their results. It does not prevent crawling; conversely, robots.txt blocks crawling but does not guarantee that a page stays out of the index. How we approach SEO.

P

PerplexityBot

Perplexity's crawler, used to surface and cite websites in its answers. Perplexity also uses the Perplexity-User agent when a user asks a question that requires visiting a page. How we approach GEO.

Q

Quotable passage

A paragraph that answers a precise question on its own, with the subject named, the fact and its source, without depending on the rest of the page. It is the format answer engines most readily extract to build and source their answers. How we approach GEO.

R

RAG (retrieval-augmented generation)

A technique in which an AI assistant searches for relevant documents before writing its answer, then relies on them. This is how the web searches of ChatGPT, Perplexity or Gemini work: what they cite depends on what they were able to retrieve. How we approach GEO.

Redirects (301 and 302)

Automatic forwarding from one URL to another. A 301 is permanent: it passes the old page's signals to the new one and is required during a redesign or a change of URL. A 302 is temporary. Redirect chains slow down both visitors and robots. How we approach SEO.

Rich results

Enhanced displays in Google's results (review stars, breadcrumbs, event details) made possible by structured data. Google decides whether to show them: the markup makes them possible, it does not guarantee them. How we approach SEO.

Robots.txt

A file placed at the root of a site that tells robots what they may crawl, robot by robot. Its format is standardised by RFC 9309. It controls crawling, not indexing: a blocked page can remain indexed if other sites link to it. How we approach GEO.

S

SEA

Search Engine Advertising: advertising on search engines, such as Google Ads, shown in placements labelled as sponsored. It is paid per click and stops when the budget stops; it does not influence organic rankings. How we approach SEO.

Search intent

What the searcher really wants behind the query: to find information, compare, find a provider, go somewhere, buy. A page that answers a different intent from that of the query ranks poorly, even when well optimised. How we approach SEO.

SEO (search engine optimisation)

Search Engine Optimization: work that improves a site's visibility in the unpaid results of search engines. It combines technical work (crawling, indexing, speed), content (answering real searches) and popularity (links, reputation). How we approach SEO.

SERP

Search Engine Results Page: a search engine's results page. It is no longer limited to ten blue links: AI Overviews, the local pack, featured snippets, related questions, videos and ads share the space, which changes the real value of each position. How we approach SEO.

Structured data

Markup, usually in JSON-LD format using the schema.org vocabulary, that explicitly describes the content of a page: business, service, FAQ, review, event. It helps search engines interpret the page and makes certain enhanced displays possible, without guaranteeing them. How we approach GEO.

T

Title tag

The page title declared in the HTML code, often used as the clickable title in search results. It must be unique, describe the content of the page and put first what the searcher is looking for. How we approach SEO.

Topic cluster (semantic cocoon)

Organising a site's pages into thematic groups: a parent page covering the main subject and child pages detailing each aspect, connected by logical internal links. Readers move around naturally and the search engine understands the full extent of your expertise. How we approach SEO.

X

XML sitemap

A file listing the URLs of a site to be crawled, with their update date. It is declared in Search Console and in the robots.txt file. A single file can contain up to 50,000 URLs; beyond that, a sitemap index is used. How we approach SEO.

Y

YMYL

Your Money or Your Life: topics that can affect people's health, safety or financial situation (finance, law, health). Google's guidelines require a higher level of trustworthiness for them: accounting firms, lawyers and health professionals are directly concerned. How we approach SEO.

Z

Zero-click search

A search that ends without a click to a website, because the answer appears directly on the results page or in an AI assistant. Hence the value of being cited in these answers, and of measuring visibility beyond clicks alone. How we approach GEO.

Sources

Definitions based on the following official documentation:

Last updated: 18 October 2026.

Glossary of SEO, local SEO and GEO

A