Glossary · A–Z
SEO and AI search terms, explained briefly.
50 terms
50 terms · A–Z
All terms
AEO (Answer Engine Optimization)
AEO (answer engine optimization) is writing and structuring content so that search engines and AI assistants can lift a direct answer from it. The goal is not only that your page is found, but that the answer in it is picked up: in a featured snippet, in the People Also Ask block, by a voice assistant or in an AI answer with a source reference.
Read the explanationAgentic AI
Agentic AI is AI that does not just give an answer but carries out a task on its own in several steps: breaking a goal into parts, using tools — a search engine, a browser, an API, a CMS — checking intermediate results, adjusting and continuing until the task is done, with a human keeping oversight. The difference from a chatbot is the difference between 'explain how I do this' and 'do this for me'.
Read the explanationAI crawler
An AI crawler is a bot that fetches web pages on behalf of an AI system: to collect training data, to build a search index for an answer engine, or to read a page live when a user asks for it. Well-known examples are GPTBot, OAI-SearchBot and ChatGPT-User (OpenAI), ClaudeBot (Anthropic), PerplexityBot, Google-Extended (Google's opt-out for AI training), Applebot-Extended, Meta-ExternalAgent, Amazonbot, Bytespider (ByteDance) and CCBot (Common Crawl).
Read the explanationAI Mode (Google)
AI Mode is Google's conversational search mode: a separate tab next to the regular results where you ask a question in plain language and get an extended, Gemini-written answer with sources, after which you can follow up as if you were talking to an assistant. Google launched AI Mode in May 2025 in the United States and rolled it out over the following months to most countries and languages, including Dutch.
Read the explanationAI Overviews
AI Overviews are the AI-generated summaries at the top of Google search results: an answer of a few paragraphs, written by Gemini from a set of pages, with links to those sources beside or below it. Google introduced them in May 2024 in the United States and then rolled them out to more than two hundred countries and dozens of languages, including Dutch. They now appear on a large share of informational and comparison queries.
Read the explanationAnswer engine
An answer engine is a search system that answers a question directly with a generated, coherent answer instead of a list of links. ChatGPT with search, Perplexity, Google AI Mode and AI Overviews, Microsoft Copilot and Claude with web access are answer engines: they retrieve sources, have a language model write the answer and show the sources as citations. The user reads the answer and at most clicks through for depth.
Read the explanation
Brand mention
A brand mention is a reference to your brand name in text without a link: in an article, a review, a forum post, a podcast transcript or — increasingly important — in the answer of an AI engine. Where a backlink is a clickable reference, a mention is only the name. Yet it is a signal that search engines and language models use to determine how well known, trustworthy and relevant a brand is.
Read the explanation
Canonical URL
A canonical URL is the preferred version of a page that you declare to search engines when the same content is reachable at multiple addresses. You set it with a link tag in the head: <link rel="canonical" href="https://example.com/page/">. Search engines use it to consolidate the signals of all variants onto one address and show that version in the results.
Read the explanationCitation (AI source)
A citation is a source reference inside an AI answer: the link, footnote or source card that ChatGPT, Gemini, Perplexity or Google's AI Overviews show to indicate which web page a claim is based on. Where a classic search engine lists ten blue links, an answer engine writes a single answer and points to a handful of pages within it. Those references are the citations.
Read the explanationContent decay
Content decay is the gradual decline in rankings, traffic and citations of a page that once performed well, without anything having changed on the page itself. The content ages, competitors publish better or more recent versions, search intent shifts, or the results page changes shape. It is the natural erosion of content in an environment that does not stand still.
Read the explanationCore Web Vitals
Core Web Vitals is a set of three Google-defined metrics for the perceived speed and stability of a web page: Largest Contentful Paint (LCP), Interaction to Next Paint (INP) and Cumulative Layout Shift (CLS). They measure, respectively, how quickly the main content becomes visible, how quickly the page responds to interaction, and how much the layout shifts while loading.
Read the explanationCrawl budget
Crawl budget is the number of URLs a search engine wants to and can fetch on your site within a given period. Google derives it from two parts: crawl capacity (how many requests your server can handle without slowing down) and crawl demand (how interesting and how frequently changed your pages are). Together they determine how many pages Googlebot visits per day and which ones.
Read the explanation
Doorway page
A doorway page is a page that exists only to rank for a specific search query or a variant of it and then sends the visitor on to another, generic destination. Classic examples are series of pages such as 'plumber Utrecht', 'plumber Amersfoort' and 'plumber Zeist' that are identical apart from the place name and all lead to the same contact form.
Read the explanation
E-E-A-T
E-E-A-T is a framework from Google's Search Quality Rater Guidelines that stands for Experience, Expertise, Authoritativeness and Trustworthiness. It describes how human raters assess the quality of a page and its creator, and by extension which characteristics Google's ranking systems try to reward. The first E (Experience) was added to the older E-A-T in December 2022.
Read the explanationEntity
An entity is a uniquely identifiable thing or concept: a person, brand, product, place, organisation or idea that a search engine or language model can recognise independently of the exact words used. 'Google', 'Amsterdam', 'Shopify' and 'core web vitals' are entities. The word 'jaguar' is not an entity but a string, because it could be an animal, a car brand or an operating system.
Read the explanation
Faceted navigation
Faceted navigation is the filter system in webshops and catalogues that lets visitors refine a product list by attributes such as brand, price, colour, size and material, and sort the results. Every combination of filters usually produces its own URL, such as /shoes?brand=nike&colour=black&size=43. For visitors it is indispensable; for search engines it is one of the biggest sources of duplicates and wasted crawl budget.
Read the explanationFAQ schema
FAQ schema is structured data of the FAQPage type (schema.org) that marks up the questions and answers on a page in a machine-readable way. Each question becomes a Question with an acceptedAnswer, usually as JSON-LD in the head. Search engines and AI systems can then see which question belongs to which answer without guessing.
Read the explanationFeatured snippet
A featured snippet is a highlighted answer box at the top of Google's search results showing an excerpt from a web page — a paragraph, list, table or video — with the source URL beneath it. Google picks the excerpt automatically from a page that already ranks highly, because it directly answers the searcher's question. The spot is therefore also called 'position zero'.
Read the explanation
GEO (Generative Engine Optimization)
GEO (generative engine optimization) is optimising your content and your brand so that generative AI systems — ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews and AI Mode — mention and cite you when someone asks a question you have the answer to. Where SEO is about a position in a list of links, GEO is about being part of one composed answer.
Read the explanationGoogle Business Profile
Google Business Profile (formerly Google My Business) is the free business listing that determines how your business appears in Google Search and Google Maps: name, category, address or service area, opening hours, phone number, website, photos, services, products, posts, questions and answers, and reviews. The profile is the basis of the local map block and of the knowledge panel that appears to the right of the search results when someone searches for your business name.
Read the explanationGoogle Merchant Center
Google Merchant Center is the free platform through which webshops supply their product data to Google, so products can appear in Google Shopping, in the free product listings in the search results and the Shopping tab, in Google Images, YouTube and Maps, and as a source of product information in AI Overviews and AI Mode. You connect a product feed to it, or let Google fetch the products from your site itself.
Read the explanationGPTBot
GPTBot is OpenAI's web crawler that fetches publicly accessible web pages to collect training data for future models such as those behind ChatGPT. The bot identifies itself with the user agent 'GPTBot' and has been active since August 2023. OpenAI publishes the IP ranges GPTBot crawls from, so you can distinguish real visits from bots impersonating it.
Read the explanation
Hallucination (AI)
A hallucination is an answer from a language model that sounds convincing but is factually wrong or fabricated: a non-existent product, an incorrect price, a quote that was never said, a source that does not exist. The model does not 'lie' deliberately; it predicts the most likely next words and has no built-in sense of true or false. Without a reliable source it fills the gaps with whatever sounds plausible.
Read the explanationHreflang
Hreflang is an HTML attribute that tells search engines which language and country versions of a page belong together, so a Dutch-speaking visitor sees the Dutch version and a Belgian visitor the Belgian one. You place it as a link tag in the head (<link rel="alternate" hreflang="nl-be" href="…">), in the HTTP header or in the XML sitemap. Every version lists all the other versions and itself.
Read the explanation
Indexing
Indexing is the process by which a search engine analyses, understands and adds a crawled page to its index: the database of pages from which search results are assembled. A page that is not indexed cannot rank, however good it is. Crawling (fetching) and indexing (including) are two separate steps; Google crawls far more pages than it indexes.
Read the explanationInternal linking
Internal linking is the system of links between pages on the same website: from the navigation and footer to the contextual links inside the text of an article or product description. It determines how visitors and crawlers move through your site, which pages are considered important and how search engines understand the relationships between topics.
Read the explanation
Keyword cannibalization
Cannibalisation (keyword cannibalization) is the situation in which two or more pages on the same website compete for the same search query, undermining each other's ranking, clicks and authority. Google has to choose which page to show, often alternates between them, and gives neither the position a single strong page could have reached.
Read the explanationKeyword cluster
A keyword cluster is a group of keywords and questions that share the same search intent and should therefore be answered by one page. 'Shopify SEO', 'SEO for Shopify store', 'improve Shopify visibility' and 'Shopify SEO tips' together form one cluster: whoever types these terms wants the same thing, and Google shows largely the same results for them.
Read the explanation
LLM (large language model)
An LLM (large language model) is an AI model trained on a very large amount of text so that it can understand, summarise, translate and generate language. ChatGPT, Gemini, Claude and the models behind Perplexity and Google AI Overviews are all LLMs or systems built on them. The model keeps predicting the most likely next piece of text, and does that so well that coherent answers come out.
Read the explanationLLMO (LLM Optimization)
LLMO (large language model optimization) is increasing the odds that large language models — ChatGPT, Gemini, Claude, Perplexity — know, correctly describe and recommend your brand when someone asks a question in your category. Where SEO wants a page higher in a list, LLMO wants the model to hold you in memory: when it talks about your subject, you are part of the conversation.
Read the explanationllms.txt
llms.txt is a proposed text file at the root of your website (/llms.txt) that gives language models a compact, Markdown-formatted summary of what your site is and which pages matter most. The proposal dates from September 2024 and comes from Jeremy Howard (Answer.AI); it is meant as a counterpart to robots.txt: not 'what may you crawl' but 'this is what you should know and where to find it'.
Read the explanationLocal SEO
Local SEO is optimising a business's online visibility for searches with a geographic intent: 'dentist Haarlem', 'bike repair near me' or 'best coffee Rotterdam centre'. The goal is appearing in Google's local map block (the local pack), in Google Maps, in the regular results for local questions and, increasingly, in AI answers that recommend a business for a place.
Read the explanationLong-tail keyword
A long-tail keyword is a specific, usually longer search query with low search volume but a clear intent, such as 'rent electric cargo bike for three children Utrecht' instead of 'cargo bike'. The name refers to the long tail of the demand curve: a few short terms receive enormous numbers of searches, but the sum of all specific variants is larger than that head.
Read the explanation
Orphan page
An orphan page is a page on your website that no other page on your site links to. The page exists and is reachable through its direct URL, but a visitor or crawler who starts clicking from the homepage never gets there. Often the page is still in the XML sitemap, so Google knows about it but treats it as unimportant.
Read the explanation
People Also Ask
People Also Ask (PAA) is the block in Google's search results with expandable follow-up questions related to the original query. Behind each question sits a short answer from a web page, with a link to the source. When you open a question, Google loads new questions beneath it, so the block can grow endlessly.
Read the explanationProduct feed
A product feed is a structured file or data stream (XML, CSV, TSV or via API) in which a webshop supplies its products with all relevant attributes to platforms such as Google Merchant Center, Meta, Microsoft, Bol, Amazon and price comparison sites. Each row is a product; each column an attribute: ID, title, description, link, image, price, availability, brand, GTIN, category, colour, size, shipping costs.
Read the explanationProgrammatic SEO
Programmatic SEO is the creation of landing pages at scale from a template and a structured dataset, so you have a dedicated page for hundreds or thousands of similar search intents. Think 'office space rent in {city}', '{product} versus {product}' or '{service} for {industry}'. Every page follows the same structure but is filled with its own data, text and internal links.
Read the explanationPrompt sampling
Prompt sampling is a measurement method for AI visibility in which you ask a fixed set of questions (prompts) repeatedly to several AI engines and record the answers systematically: which brands are named, which sources are cited, in what order and in what context. Because a language model gives a slightly different answer each time it is asked the same question, one answer is not a measurement; a sample of answers is.
Read the explanation
RAG (Retrieval-Augmented Generation)
RAG (retrieval-augmented generation) is a technique in which a language model first retrieves relevant documents — from a search index, a database or the live web — before generating an answer, and passes them to the model as context. The answer is then assembled not only from the model's memory but from current, verifiable sources the model can cite alongside the answer.
Read the explanationRendering
Rendering is the process by which the raw code of a web page (HTML, CSS and JavaScript) is turned into the actual content and layout a visitor or crawler gets to see. For SEO it comes down to one question: is the content already in the HTML the server sends, or is it only built in the browser by JavaScript?
Read the explanation
Search intent
Search intent is the goal behind a query: what someone wants to achieve the moment they type a question into Google or put it to an AI assistant. The classic model has four types — informational (know something), navigational (go somewhere), commercial (compare) and transactional (buy or request) — and that intent decides which type of page can rank and which source an AI answer cites.
Read the explanationSemantic search
Semantic search is a search method that understands the meaning and intent behind a query instead of only matching the literal words. Someone searching 'how do I get my webshop higher in ChatGPT' gets results about generative engine optimization, even if those pages never use those exact words. The search engine compares the meaning of the question with the meaning of pages.
Read the explanationSERP (search engine results page)
A SERP (search engine results page) is the page Google, Bing or another search engine shows after you enter a query. It used to be a list of ten blue links; today it is a composed page with organic results, ads, a featured snippet, People Also Ask, images, videos, a local map pack, shopping results and, increasingly, an AI answer at the top.
Read the explanationShare of voice
Share of voice is the share of all visibility in a market that your brand owns. The term comes from advertising — your slice of total media spend in a category — but in search it means your share of rankings and estimated clicks on a fixed set of keywords, and in AI answers how often you are mentioned versus your competitors for the same questions. It is one number that says: of all the times someone searches in this market, how often do they see you?
Read the explanationStructured data (schema.org)
Structured data is machine-readable information you add to a web page to tell search engines and AI systems explicitly what is on it: that this is a product with this price and stock, an article by this author on this date, an organisation with this address, or an FAQ with these questions. The vocabulary used is schema.org; the common format is JSON-LD, a block of JSON in a script tag in the head or body.
Read the explanation
Thin content
Thin content is content that offers the visitor little or no value: pages with a few sentences and nothing else, automatically generated or copied text, rewritten versions of what already exists elsewhere, affiliate pages without an opinion of their own, and doorway pages that exist only to rank for a keyword variant. It is not about word count, but about whether the page adds anything to what is already available.
Read the explanationTopical authority
Topical authority is the degree to which a website is seen by search engines and AI engines as a reliable, complete source on one subject. A site with topical authority on 'solar panels' does not only answer the main question but also the sub-questions about yield, subsidies, installation, maintenance and payback time, and connects those pages logically.
Read the explanationTTFB (Time to First Byte)
TTFB (Time to First Byte) is the time between the moment a browser or crawler requests a page and the moment the first byte of the response arrives. The measurement includes the DNS lookup, establishing the connection (TCP and TLS), the time the server needs to build the page and the return trip across the network. Google considers a TTFB under 800 milliseconds good and above 1.8 seconds poor.
Read the explanation
Vector embedding
A vector embedding is a numerical representation of the meaning of a piece of text (a word, sentence, paragraph or entire page) in the form of a long series of numbers — a vector with hundreds to thousands of dimensions. Texts with similar meaning get vectors that lie close together; 'cheap running shoes' and 'affordable running trainers' sit right next to each other in that space, even though they share no words.
Read the explanation
Zero-click search
A zero-click search is a search that ends without the user clicking a result, because the answer is already on the results page: in a featured snippet, a knowledge panel, an AI Overview, a weather forecast, a calculator or a local map pack. The searcher gets what they were looking for; the websites that supplied the information get no visit.
Read the explanation
Missing a term? Let us know and we will add it if it keeps coming up.
In practice
How Traze applies this
Traze measures your visibility in Google and AI answers and adapts your content automatically. See how that works, or check for free where you stand today.
FAQ · glossary
About this glossary
From definition to result
Knowing what the terms mean is step one.
Traze measures your visibility in Google and AI answers and adapts your content automatically. Start with a free check.