Glossary

Every term you'll come across in AI search, from GEO to query fan-out, and every term you'll see in AnswerOrbit. Each one has its own page, with an example and the related terms.

Looking for how we count our numbers? See What the numbers mean.

#

• 10x content: 10x content is a page so much better than what already ranks for a topic that it becomes the obvious one to read, link to and quote.
• 301 redirect: A 301 redirect sends people and bots from an old address to a new one, and tells search engines the move is permanent.
• 302 redirect: A 302 redirect sends people and bots to another address for now, and tells search engines to keep the original page.
• 304 Not Modified: 304 Not Modified is a server's reply saying a page hasn't changed since the visitor last fetched it, so the stored copy can be reused.
• 404 error: A 404 error is the reply a server gives when the page someone asked for doesn't exist at that address.
• 410 Gone: 410 Gone is a server's reply saying a page has been removed on purpose and won't be coming back.

A

• A2A (Agent2Agent): A2A is an open standard that lets AI agents from different companies find each other, hand each other tasks and pass back the results.
• Accelerated Mobile Pages (AMP): AMP is a way of building stripped-down web pages that load fast on phones, started by Google and once common on news sites.
• Accuracy prompt: An accuracy prompt asks an AI platform for facts about a brand, like prices or delivery areas, so you can check whether the answer gets them right.
• Accuracy score: An accuracy score is the share of what AI answers say about your brand that your own facts confirm.
• Ad impressions: An ad impression is one showing of an ad to someone, counted whether or not they click it.
• ADA website compliance: ADA website compliance means making your site usable by people with disabilities, in line with the Americans with Disabilities Act.
• AEO audit: An AEO audit is a full check of how AI platforms treat your brand: where you're named, what they say, what they read, and what blocks you.
• AEO content scorecard: An AEO content scorecard grades one page against the things that help AI platforms read and quote it, and lists what to fix.
• Agency mode: Agency mode is a way of running an AI search tool for many clients from one account, each with its own brands, prompts and reports.
• Agent Analytics: Agent analytics measures the AI bots and agents that visit your website, which pages they read, and the people AI assistants send you.
• Agent readiness: Agent readiness is how well your website works when the visitor is an AI agent, so it can find your facts and complete a task.
• Agentic browsing: Agentic browsing is when an AI agent opens web pages, clicks and fills in forms for a person, instead of the person doing it.
• Agentic Commerce Protocol (ACP): The Agentic Commerce Protocol is an open standard from OpenAI and Stripe that lets AI agents buy from a business for a person.
• Agentic marketing: Agentic marketing is handing marketing work to AI agents that plan and carry out the steps themselves. It also means marketing to agents.
• Agentic readability: Agentic readability is whether an AI agent can read a page and pull out the facts on it correctly, without a person's help.
• Agentic search: Agentic search is when an AI agent does the research for a person, running many searches, reading pages and comparing options on its own.
• Aggregated sentiment score: An aggregated sentiment score rolls the tone of every AI answer about your brand into one number, usually the share that's positive.
• AGI (Artificial General Intelligence): AGI is an AI that could do almost any thinking task as well as a person. It doesn't exist yet, and people disagree on what would count.
• AhrefsBot: AhrefsBot is the crawler from the SEO tool Ahrefs, which reads pages and links to build Ahrefs' index and the Yep search engine.
• AI agent: An AI agent is an AI system that carries out a task in several steps on its own, like researching products or booking something, not just answering a question.
• AI assistant: An AI assistant is the app people talk to, like ChatGPT, Gemini or Copilot, built around an AI model and its tools, such as web search.
• AI browser: An AI browser is a web browser with an AI assistant built in, which can answer questions about pages and browse sites for you.
• AI claims and brand facts: A claim is something an AI answer says about your brand, like a price or a delivery area. Brand facts are what's true, used to check each claim.
• AI copilot: A copilot is an AI tool that drafts or suggests and waits for a person to say yes, halfway between a chat assistant and an agent that acts alone.
• AI Marketer: An AI marketer is an AI assistant built into a marketing tool that reads your data, suggests work and can carry some of it out for you.
• AI Mode: AI Mode is a conversation inside Google Search where you ask questions and follow-ups and get a written answer with links, instead of a results page.
• AI Overviews: AI Overviews are the AI-written summaries Google shows above its usual results for some searches, with links to the pages they're based on.
• AI preferences: AI preferences are a proposed standard way for websites to say how AI may use their content, such as for training or for answers.
• AI referral: An AI referral is one visit to your website by a person who clicked a link in an AI assistant's answer, as opposed to a visit by a bot.
• AI referral traffic: The visits people make to your website after clicking a link in an AI assistant's answer.
• AI response: An AI response is the answer an AI platform writes to a question, including the brands it names and the pages it links to.
• AI search: AI search is searching by asking an AI platform a question and getting a written answer back, instead of a list of links.
• AI SEO: AI SEO is a loose name for getting your brand named in AI answers. It's also used for doing SEO work with AI tools.
• AI slop: AI slop is low-quality content churned out with AI in bulk: pages, posts and images made to fill space, with little care for whether they're right.
• AI visibility: AI visibility is how often AI platforms name your brand when people ask the questions your customers ask.
• AI washing: AI washing is claiming a product uses AI, or uses more of it than it does, to make it sound smarter or newer than it is.
• AI2Bot: AI2Bot is the web crawler of Ai2, the Allen Institute for AI, which collects pages to train its open language models.
• AIO: AIO has two meanings in AI search: short for Google's AI Overviews, and short for AI optimization, the work of getting named in AI answers.
• Alexa for Shopping: Alexa for Shopping is Amazon's AI shopping assistant, the new name for Rufus, which answers product questions inside Amazon's store.
• Alt text: Alt text is a short written description of an image, read out by screen readers and used by search engines and bots to understand it.
• Alternatives page: An alternatives page lists the options people consider instead of a product, such as "alternatives to Kestrel", often written by a rival.
• Amazonbot: Amazonbot is Amazon's web crawler. It reads pages to improve Amazon's products and services, and its data may be used to train Amazon's AI models.
• Amzn-SearchBot: Amzn-SearchBot is Amazon's search crawler. It reads pages so they can appear in search in Amazon products like Alexa, and isn't used for AI training.
• Amzn-User: Amzn-User is Amazon's on-demand fetcher. It visits a page when someone's request to Alexa needs fresh information, and isn't used for AI training.
• Anchor text: Anchor text is the clickable words of a link. Search engines read it as a short description of the page the link points to.
• Answer bubbles: Answer bubbles form when AI platforms give everyone much the same answer, so the same few brands keep getting named and others struggle to break in.
• Answer Economy: The answer economy is the idea that people now get their answers straight from AI, so brands compete to be named in the answer itself.
• Answer engine: An answer engine is an AI platform that replies to a question with a written answer, like ChatGPT, Perplexity or Google's AI Overviews.
• Answer Engine market share: Answer engine market share is how AI search use splits between platforms like ChatGPT, Gemini, Perplexity and Google's AI features.
• Answer inclusion rate: Answer inclusion rate is the share of AI answers for a topic or set of prompts that name your brand. AnswerOrbit calls it visibility.
• Answer layer: The answer layer is the AI-written reply between your pages and the people asking, which decides the brands they hear about.
• Answer variability: Answer variability is how AI platforms give different answers to the same question, even when nothing has changed. It's why one check proves little.
• Answer volatility: Answer volatility is how much your AI search numbers move from one week to the next when nothing real has changed. Know it before you report a win.
• Answer-first formatting: Answer-first formatting puts the direct answer at the top of a page or section, with the detail and reasoning after it.
• AP2 (Agent Payments Protocol): AP2 is an open protocol, started by Google, that lets an AI agent prove a person agreed to a purchase before it pays.
• Apple Intelligence: Apple Intelligence is Apple's set of AI features on iPhone, iPad and Mac, including a smarter Siri that can answer questions from the web.
• Applebot: Applebot is Apple's web crawler. It reads pages for search in Siri, Spotlight and Safari, and its data may help train Apple's AI models.
• Applebot-Extended: Applebot-Extended is a robots.txt name that tells Apple whether pages Applebot read may train its AI models. It doesn't crawl anything.
• Article spinning: Article spinning is rewriting one piece of content into many slightly different copies, by swapping words, to pass them off as new.
• Article syndication: Article syndication is when other sites republish an article that first ran somewhere else, with the original publisher's permission.
• Artificial intelligence (AI): Artificial intelligence is software that does things we usually think need a human mind, like understanding language or answering questions.
• Association: An association is a link AI platforms draw between your brand and a quality, like freshness or low prices, when they describe you.
• Atomic content: Atomic content is writing each section of a page so it makes sense on its own, without the reader needing what came before.
• Attention: Attention is the part of an AI language model that works out which words in a text matter most to each other while it reads and writes.
• Attribute: An attribute is a specific thing AI answers say about your brand, like "fast delivery" or "expensive for small teams".
• Attribute list: An attribute list sets out what you want AI platforms to say about your brand, with every item precise enough to check in an answer.
• Attribution gap: The attribution gap is the difference between how much AI answers shape what people buy and how much of it your analytics can show.
• Authentic brand mention: An authentic brand mention is a genuine, accurate mention of your brand on a site AI platforms read, earned honestly by the brand.
• Author markup: Author markup is structured data on a page that says who wrote it and links to more about them.
• Authority: In AI search, authority is whether the platforms see your brand as a credible name in your field, credible enough to name in their answers.
• Auto-generated content: Auto-generated content is text produced by a program rather than written page by page, from simple templates to AI writing tools.
• Automation: Automation is software that follows rules you set in advance. It does the same thing every time and doesn't decide anything for itself.
• Average position: Average position is where your brand comes, on average, among the brands an AI answer names, counted over the answers that name you.

B

• Backlinks: Links from other websites to yours, which search engines read as a sign that your pages are worth visiting.
• Battlecard: A battlecard is a short reference document that tells your team how you compare with one competitor and what to say about it.
• Bi-encoder: A bi-encoder is a retrieval model that turns a question and each passage into numbers separately, so it can search millions of passages fast.
• Bing Webmaster Tools: Bing Webmaster Tools is Microsoft's free service for checking how Bing crawls, indexes and shows your site, including in its AI answers.
• Bingbot: Bingbot is Microsoft's web crawler. It reads pages for Bing's index, which also feeds Microsoft's AI search answers.
• Black hat SEO: Black hat SEO means tactics that break a search engine's rules to rank higher, such as cloaking, bought links or hidden text.
• BLUF (bottom line up front): BLUF, short for bottom line up front, means putting the answer or main point in the first sentences of a page or section.
• Borrowed authority: Borrowed authority is getting your brand mentioned on websites the AI platforms already trust, instead of relying on your own site alone.
• Bot: A bot is a program that visits websites on its own. AI companies run bots to train models, find pages to cite, and fetch pages for people.
• Bot directory: A bot directory is a public list of known bots, with who runs each one and what it does. Useful for spotting AI bots in your logs.
• Bot visits: Bot visits are the requests AI bots make to your website, which you can see in your server logs.
• Bot vs human readable content: The difference between what a person sees on a page in their browser and what a bot reads in the page's HTML.
• Bot-to-human ratio: The bot-to-human ratio is how many visits AI bots make to your site for every visit from a person.
• Bounce rate: Bounce rate is the share of visits where someone lands on your site and leaves without engaging with it.
• Brand: In AI search tracking, a brand is a company, product or name whose place in AI answers you follow, yours or a competitor's.
• Brand AI visibility index: A brand AI visibility index is a single score that blends several measures of how a brand shows up in AI answers into one number.
• Brand context: Brand context is the background a tool or AI system holds about your brand: what you sell, to whom, where, and at what price.
• Brand coverage: Brand coverage is the share of the prompts you track where AI answers name your brand at least once.
• Brand fidelity: Brand fidelity is how accurately AI answers describe your brand when they name it: your category, products, prices and what sets you apart.
• Brand mention: A brand mention is any AI answer that names your brand, whether or not it links to your website.
• Brand perception: Brand perception in AI search is how AI answers characterize your brand: the qualities, category and uses they tie to you.
• Brand placement: Brand placement is how early in an AI answer's text your brand first appears, near the top or deep in the answer.
• Brand position: Brand position is your rank among the brands named in one AI answer. Named first means position 1.
• Brand sentiment: Brand sentiment is the overall tone AI answers take about your brand across many answers: how many mentions are positive, neutral or negative.
• Brand SERP: A brand SERP is the search results page someone gets when they search for your brand name.
• Brand-owned citation: A brand-owned citation is a link in an AI answer that points to a website the brand itself controls.
• Branded content: Branded content is material a brand makes or pays for that people want to read, watch or hear for its own sake, like a guide or a video series.
• Branded keywords: Branded keywords are searches that include your brand's name or a product's name, like "Fernhill Coffee subscription".
• Branded prompt: A branded prompt is a question to an AI platform that names a brand, such as "Is Fernhill Coffee any good?" An unbranded prompt names none.
• Branded search lift: Branded search lift is a rise in the number of people searching for your brand by name, often a sign that more people are hearing about you.
• Brave Leo: Brave Leo is the AI assistant built into the Brave browser. It can answer with results from Brave Search, which runs its own web index.
• Bridge page: A bridge page is a thin page whose only job is to pass visitors on to somewhere else, usually through an affiliate link.
• Broad objection: A broad objection is a reason against your brand that AI answers give across many different questions, topics and platforms.
• Broken link: A broken link points to a page that no longer exists or can't be reached, so whoever follows it, person or bot, hits an error.
• Buy for Me: Buy for Me is an Amazon feature that lets shoppers buy from a brand's own website without leaving the Amazon app, with Amazon's AI placing the order.
• Buyer-journey prompts: Buyer-journey prompts are a set of questions to AI platforms that follow a customer from first learning about a need to choosing what to buy.

C

• Cached page: A cached page is a stored copy of a web page, kept by a search engine, browser or server so it doesn't have to be fetched again.
• Candidate set: A candidate set is the shortlist of passages an AI system gathers and considers before it writes an answer.
• Canonical tag: A canonical tag is a line of HTML, rel="canonical", that tells search engines which address is the main version of a page with copies.
• Canonical URL: The one address you tell search engines to treat as the main version of a page that's reachable at several addresses.
• Category: A category is the kind of product or service people put your brand in, like coffee subscriptions or oak furniture.
• CCBot: CCBot is Common Crawl's crawler. It builds a free, public archive of the web that many AI models have been trained on.
• ChatGPT: ChatGPT is OpenAI's AI assistant. It answers questions in conversation, can search the web, and can show products and ads.
• ChatGPT Atlas: ChatGPT Atlas was OpenAI's web browser with ChatGPT built in, released for Mac in October 2025 and closed in August 2026.
• ChatGPT-User: ChatGPT-User is the name OpenAI uses when ChatGPT visits a web page because someone asked it to, instead of crawling on its own.
• Chunk: A chunk is a short piece of a web page, often a section or a few paragraphs, that an AI system stores and retrieves on its own.
• Chunk optimization: Chunk optimization is shaping a page's headings, paragraphs and lists so the pieces an AI system pulls out are the ones you'd choose.
• Chunkability: Chunkability is how easily an AI system can lift a clear, complete answer out of a section of your page.
• Chunking: Chunking is how an AI or search system cuts a page into smaller pieces, so it can store them and find the right one for a question.
• Circular sourcing: Circular sourcing is when AI answers cite pages that were themselves written from AI answers, so one mistake can feed on itself.
• Citation and source: A source is a page an AI platform read to answer. A citation is a link to that page inside the answer.
• Citation categories: Citation categories sort the pages AI answers cite into kinds of source, such as your own site, competitors, press, reviews and forums.
• Citation coverage: Citation coverage is the share of AI answers that cite a given website at least once.
• Citation decay: Citation decay is how a page gets cited less and less by AI answers over time, as newer pages take its place.
• Citation domains: Citation domains are the websites AI answers cite, counted by website instead of by single page.
• Citation faithfulness: Citation faithfulness is whether a page an AI answer cites really says what the answer claims it says.
• Citation gap: A citation gap is a source AI answers use that names your competitors but not you, or a question where they're cited and you aren't.
• Citation hijacking: Citation hijacking is when an AI answer credits your content, data or idea to a different website.
• Citation pages: Citation pages are the individual web pages AI answers cite, each with its own address, counted page by page.
• Citation rank: Citation rank is your place among brands or websites, ordered by how often AI answers use their pages.
• Citation recall: Citation recall is the share of the statements in an AI answer that the sources it cites fully back up.
• Citation share: Citation share is the part of all the links in AI answers that point to one website or page.
• Citation supply chain: The citation supply chain is the set of websites, and kinds of pages, an AI platform tends to draw on when it answers questions about a topic.
• Citation velocity: Citation velocity is how fast a page or website gains, or loses, citations in AI answers over time.
• Cited domain: A cited domain is a website, counted as a whole, that an AI answer links to as one of its sources.
• Claude: Claude is Anthropic's AI assistant. It can search the web and cite the pages it used in its answers.
• Claude-SearchBot: Claude-SearchBot is Anthropic's crawler that reads web pages for the search index behind Claude's answers.
• Claude-User: Claude-User is the Anthropic bot that fetches a web page when someone's question to Claude needs it.
• ClaudeBot: ClaudeBot is Anthropic's crawler that collects web pages that may be used to train its AI models.
• Click depth: Click depth is how many clicks it takes to reach a page from your home page. Pages buried deep are harder for people and bots to find.
• Click economy: The click economy is the old deal of web search: search engines send visitors to websites, and websites earn from those visits.
• Client workspace: A client workspace is the separate space an agency keeps for each client in an AI visibility tool, with its own brands, prompts and access.
• Cloaking: Cloaking is showing search engines or bots different content from what people see, to rank for things the page doesn't deliver. It's spam.
• Cloudflare AI Crawl Control: Cloudflare AI Crawl Control is Cloudflare's tool for seeing which AI crawlers visit your site and allowing or blocking each one.
• Co-citation: When two pages or brands are cited by the same source, which ties them together in the eyes of search engines and AI.
• Co-occurrence: When two names or terms keep appearing close together in text, so they become linked in the minds of readers and models.
• Comet: Comet is Perplexity's web browser with an AI assistant built in, which can research, compare and carry out tasks for you.
• Commercial prompt: A commercial prompt is a question to an AI platform from someone getting ready to buy, like "What's the best coffee subscription for an office?"
• Comparative prompt: A comparative prompt asks an AI platform to weigh options against each other: two brands, two products or two ways of doing something.
• Comparison page: A comparison page weighs products or brands side by side, such as "Fernhill vs Kestrel" or "The 7 best coffee subscriptions."
• Competitive overlap: Competitive overlap is how often you and a given competitor are named in the same AI answers.
• Competitor: In AI visibility tracking, a competitor is a brand you measure yourself against in the answers, alongside your own.
• Competitor prompt: A competitor prompt names two or more brands and asks an AI platform to choose between them, like "Fernhill or Kestrel for an office?"
• Compound-job prompt: A compound-job prompt pairs the job someone needs done with the details that make it theirs, like budget, size or location.
• Comprehension problem: A comprehension problem is when AI platforms don't know something true about you, like a product you sell or a place you deliver.
• Computer-generated content: Computer-generated content is writing produced by software, today mostly AI tools, that's meant to read like it came from a person.
• Conductor model: The conductor model is a way to run AI search work where one team sets the direction and hands clear briefs to the teams who do the work.
• Connector: A connector links one tool to another so it can read data or take actions there, such as an AI assistant reading your analytics.
• Content brief: A content brief is a short plan for a page: who it's for, the questions it answers, the facts it must include and how you'll know it worked.
• Content delivery network (CDN): A CDN is a network of servers around the world that keeps copies of your site close to visitors, so pages load faster and the site stays up.
• Content fidelity: Content fidelity is how faithfully an AI answer passes on what your page says when it draws on it.
• Content freshness: How recently a page was published or updated, and whether readers and machines can tell.
• Content gap: A content gap is a subject your customers ask about, and your competitors cover, that your own website says nothing useful about.
• Content gap analysis: Content gap analysis finds the topics your competitors cover, and get found for, that your own website doesn't.
• Content hub: A content hub is a central page on one subject that gathers and links all your related content, so readers and search engines can find it together.
• Content licensing deal: A content licensing deal is an agreement in which an AI or tech company pays a publisher or platform for access to its content.
• Content refresh: A content refresh is updating an existing page, its facts, structure and date, so it stays accurate and keeps getting used in answers.
• Content relevance: Content relevance is how well a page matches what the person searching or asking actually wants to know.
• Content Signals: Content Signals are lines in robots.txt that say how your content may be used once it's read: for search, as AI input, or for AI training.
• Context window: A context window is how much text an AI model can take in at once: the question, the conversation so far and any pages it reads.
• Conversational search: Conversational search is searching by talking: asking a full question, then follow-ups, in one ongoing chat with an AI assistant.
• Conversion rate of AI traffic: The conversion rate of AI traffic is the share of visits from AI assistants that end in a sale, a sign-up or another goal you've set.
• Copilot Search: Copilot Search is Microsoft's AI search experience in Bing, which writes an answer and links the sources it used.
• Core update: A core update is a broad change Google makes to its search ranking systems, several times a year.
• Core Web Vitals: Core Web Vitals are Google's three measures of how a page feels to use: loading speed, responsiveness and visual stability.
• Cornerstone content: Cornerstone content is the small set of pages on your site that matter most to your business, the ones you most want people and search to find.
• Cosine similarity: Cosine similarity scores how close two pieces of text are in meaning, using the lists of numbers an AI model turns them into.
• Coverage: Coverage is how well your set of tracked prompts represents the questions your customers really ask, so you can trust the numbers it gives.
• Coverage topic: A coverage topic is a broad group of prompts that tracks whether you're part of the conversation across a whole category.
• Crawl budget: Crawl budget is how many pages a search engine can and wants to crawl on your site in a given time. It matters most for very large sites.
• Crawl-delay: Crawl-delay is a robots.txt rule asking a bot to wait a set number of seconds between requests to your site.
• Crawl-to-refer ratio: The crawl-to-refer ratio compares how many pages an AI company's bots read on a site with how many visitors it sends back.
• Crawlability: Crawlability is whether bots can reach your pages and read what's on them, which decides whether AI platforms can use those pages in answers.
• Crawlability for AI: Crawlability for AI is whether AI bots in particular can reach and read your pages, which can differ from what Googlebot sees.
• Crawler: A crawler is a bot that visits web pages and reads them, to build a search index, train an AI model, or fetch a page for someone.
• Crawler purpose category: A crawler's purpose category says what a bot reads your pages for: training AI models, search, or fetching a page for a person.
• Cross-encoder: A cross-encoder is a model that reads a question and a passage together to score how well the passage answers it.
• Custom corpus: A custom corpus is a chosen set of documents an AI system searches for its answers, in place of the open web or alongside it.
• Customer journey: The customer journey is the path someone takes from first noticing a need to buying and using a product, and every place they meet a brand along the way.

D

• Dark traffic: Dark traffic is visits that arrive with no record of where they came from, so your analytics files them as direct.
• Data cleansing: Data cleansing is removing broken, duplicate and misread records from a dataset before you count anything from it.
• DeepSeek: DeepSeek is a Chinese AI company that makes large language models and a chat assistant of the same name.
• Depth of mention: Depth of mention is how much an AI answer says about your brand when it names you, from a passing name in a list to a full recommendation.
• Depth topic: A depth topic is a small, focused group of prompts about one opportunity, set up so its numbers tell you whether you're winning that one thing.
• Detection rate: Detection rate is the share of AI answers where your brand shows up at all, either named in the text or cited as a source.
• Digital PR for AI: Digital PR for AI is getting your brand covered on the sites AI platforms already read and cite when they answer questions in your category.
• Discovery prompt: A discovery prompt is a question that names no brand and asks an AI platform who to pick, like "What's the best coffee subscription?"
• Discrete step: A discrete step is a task with one right answer that comes out the same every time, so a tool or an AI agent can do it for you.
• Dofollow link: A dofollow link is an ordinary link with no rel attribute asking search engines to ignore it, so it can pass ranking credit to the page it points to.
• Domain: A domain is a website's address, like fernhillcoffee.com. In AI search, it's the site behind a cited page, as opposed to the page itself.
• Domain authority: Domain authority is a score from an SEO tool that estimates how strong a website is, mostly from the links pointing to it.
• Domain citation: A domain citation is an AI answer using or linking to a page on your website. Counting them shows how often the platforms rely on your site.
• Domain coverage: Domain coverage is your website's share of the citations in AI answers, compared with your competitors' websites.
• Domain Rating (DR): Domain Rating (DR) is Ahrefs' score from 0 to 100 for how strong a website's backlink profile is compared with other sites.
• Domain segments: Domain segments split a website's data into parts, like product lines or countries, or combine several domains, so each team sees its own share.
• Domain structure: Domain structure is how a website's domain, subdomains and folders are organized, which shapes how people, search engines and AI read it.
• Doorway page: A doorway page is one of many near-identical pages made to rank for similar searches and funnel visitors somewhere else. Google treats it as spam.
• Duck.ai: Duck.ai is DuckDuckGo's private AI chat, where you can talk to models from several AI companies without your chats being stored or used for training.
• Duplicate content: Duplicate content is the same or nearly the same text showing up at more than one web address, on your own site or across sites.
• Dwell time: Dwell time is how long someone stays on a page they clicked from search results before going back to the results.
• Dynamic URL: A dynamic URL is a web address built on the fly from parameters, usually after a question mark, like ?colour=oak&sort=price.

E

• E-E-A-T: E-E-A-T stands for experience, expertise, authoritativeness and trustworthiness: the qualities Google looks for in content it can rely on.
• Earned citation: An earned citation is an AI answer citing an independent page that mentions you, like a review, a news story or a list article.
• Earned media: Earned media is coverage of your brand that someone else chose to publish, like reviews, news stories, list articles and forum posts.
• Editorial link: An editorial link is a link another site gives you because its writer chose to, without being paid, asked to swap or otherwise rewarded.
• Effective cutoff: The effective cutoff is the point where an AI model's knowledge starts to thin out, which is usually earlier than its official training cutoff.
• Ego bait: Ego bait is content that features or praises people or brands in your field, so they'll want to share it and link to it.
• Email outreach: Email outreach is writing to editors, writers and site owners to suggest your content, product or correction for something they publish.
• Embeddings: Embeddings are lists of numbers that capture what a piece of text means, so a system can tell how close a page is to a question.
• Entity: An entity is a distinct thing a search engine or AI model recognizes, like your company, a product or a person, with facts attached to it.
• Entity confusion: Entity confusion is when an AI platform mixes your brand up with another business, product or person, and tells people the wrong things about you.
• Entity disambiguation: Entity disambiguation is how search engines and AI models work out which of several same-named things you mean, and how you help them pick you.
• Entity establishment: Entity establishment is the work of getting search engines and AI models to recognize your brand as one clear thing, with the right facts attached.
• Entity salience: Entity salience is how central a brand or other entity is to a piece of text or a topic, as a search engine or AI model sees it.
• Entity-based SEO: Entity-based SEO is optimizing around the things search engines and AI recognize, like your brand and products, as well as the words people type.
• Entry page: An entry page is the first page someone sees on your site in a visit, wherever they came from. Analytics tools often call it the landing page.
• Evergreen content: Evergreen content is content that stays useful long after it's published, like how-to guides and explainers, as long as you keep it accurate.
• External link: An external link is a link that points from one website to a different website, as opposed to an internal link between pages of the same site.
• Extractability: Extractability is how easily an AI platform can lift a passage out of your page and use it in an answer, without the rest of the page around it.

F

• FAQPage schema: FAQPage schema is structured data that marks up a page of questions and answers, so machines can read each question and its answer as a pair.
• Featured snippet: A featured snippet is a box at the top of Google's results that shows a passage from one page as the answer, with the link underneath.
• Fetchable, chosen, extractable: The three tests a page passes before an AI answer cites it: the platform can read it, picks it, and can lift a passage out of it.
• Fine-tuning: Fine-tuning is extra training that adapts an AI model that's already been trained to a narrower job, like a company's support questions.
• First-mention rate: First-mention rate is how often your brand is the first one named in AI answers that name any brand at all.
• Foundation model: A foundation model is a large AI model trained on a huge amount of general data, which other products and assistants are then built on.
• Frontier model: A frontier model is one of the newest and most capable AI models available at a given time. The label moves on as new models come out.

G

• GA4 AI Assistants channel: The AI Assistant channel is a default channel in Google Analytics 4 that groups visits from AI assistants like ChatGPT, Gemini and Copilot.
• Gated content: Gated content is anything on your site people can only read after they hand over details, usually an email address, through a form.
• Gateway page: A gateway page, or doorway page, is a thin page made only to rank for a search and pass visitors on somewhere else. Google treats it as spam.
• Gemini: Gemini is Google's AI assistant app and the family of AI models behind it. It's separate from AI Overviews and AI Mode in Google Search.
• Gemini Notebook: Gemini Notebook, formerly NotebookLM, is Google's research tool that answers questions from the sources a person adds, such as web pages and files.
• Generative AI: Generative AI is AI that makes new content, such as text, images or code, in answer to a request, instead of just sorting or labelling what exists.
• Generative AI performance report: The Generative AI performance report in Google Search Console shows how often links to your site were shown in AI Overviews and AI Mode.
• Generative engine: A generative engine is a search tool that writes its own answer from the sources it finds, like ChatGPT search, Perplexity or Google AI Mode.
• Generative features opt-out: The Search generative AI control in Google Search Console lets you keep your site out of AI Overviews, AI Mode and Discover's AI features.
• Generative prompt: A generative prompt asks an AI platform to make something, like an email, a plan or a menu, instead of finding information or recommending a brand.
• GEO and AEO: GEO (generative engine optimization) and AEO (answer engine optimization) are two names for getting your brand named in AI answers.
• GEO spam: GEO spam is trying to game AI answers with fake, mass-produced or misleading content, like invented reviews or pages written only for AI bots.
• Google Ads integration: A Google Ads integration brings the keywords you pay for in Google Ads into another tool, so you can compare paid search with AI visibility.
• Google Alerts: Google Alerts is a free Google service that emails you when new pages appear in Google's results for words you choose, like your brand name.
• Google algorithm: The Google algorithm is the set of systems Google uses to pick and order the results for a search, which it changes many times a year.
• Google Analytics: Google Analytics is Google's free tool for measuring who visits your website, where they came from and what they did there.
• Google Autocomplete: Google Autocomplete is the list of predicted searches Google shows as you type, based on what people really search for.
• Google bombing: Google bombing is a prank in which many sites link to a page with the same odd words, to make it rank on Google for a phrase it has nothing to do with.
• Google Business Profile: Google Business Profile is Google's free listing for local businesses, showing your address, hours, photos and reviews in Search and Maps.
• Google Caffeine: Google Caffeine was the new web index Google finished in 2010, built to add pages faster so search results could be fresher.
• Google Dance: Google Dance is SEO slang for the way a page's Google rankings jump around, especially when it's new or Google is updating its systems.
• Google Hummingbird: Hummingbird was a major rework of Google's ranking systems in 2013 that helped Google understand what a whole search means, beyond its single words.
• Google Knowledge Graph: The Google Knowledge Graph is Google's store of facts about people, places, companies and things, and how they connect.
• Google Knowledge Panel: A Google Knowledge Panel is the box of facts Google shows for a well-known person, company or thing, drawn from its Knowledge Graph.
• Google Panda: Google Panda was a 2011 Google system that pushed thin, low-quality pages down in search results. It's been part of Google's core ranking since 2015.
• Google penalty: A Google penalty is when Google ranks a site lower or drops it from results for breaking its spam policies, by a manual action or by its systems.
• Google Penguin: Google Penguin was a 2012 Google system against link spam. Since 2016 it's part of Google's core ranking and works in real time.
• Google Pigeon: Google Pigeon is the name SEO writers gave to a 2014 change to Google's local results, which tied them more closely to its web ranking.
• Google Sandbox: The Google Sandbox is the SEO theory that Google holds new websites back from ranking well for their first months. Google has never confirmed it.
• Google Search Console: Google Search Console is Google's free tool for site owners to see how their site does in Google Search and fix what stops it showing up.
• Google Top Heavy update: The Top Heavy update was a 2012 Google change that ranked pages lower when ads filled the top of the page and pushed the content out of sight.
• Google Webmaster Guidelines: Google Webmaster Guidelines were Google's rules for getting a site into its results. They're now called Google Search Essentials.
• Google Webmaster Tools: Google Webmaster Tools was the old name for Google Search Console, Google's free tool for watching how your site does in its search.
• Google-Agent: Google-Agent is the user agent of Google's hosted AI agents when they visit websites and take actions for a person, like filling a form.
• Google-Extended: Google-Extended is a robots.txt token that controls whether Google may use your pages to train Gemini models and ground Gemini answers.
• Google-GeminiNotebook: Google-GeminiNotebook is the fetcher Google uses to read a web page when someone adds it as a source in Gemini Notebook, formerly NotebookLM.
• Googlebot: Google's main web crawler, which reads pages for Google Search, including AI Overviews and AI Mode.
• GoogleOther: GoogleOther is Google's general-purpose crawler, used by its product teams for things like research crawls. It isn't tied to Google Search.
• GPTBot: GPTBot is OpenAI's crawler for collecting pages that may be used to train its AI models. It's separate from the bot that powers ChatGPT search.
• Grey hat SEO: Grey hat SEO is using tactics that aren't clearly against search engines' rules but bend them, sitting between white hat and black hat SEO.
• Grok: Grok is the AI assistant built into X, also available on its own at grok.com. It can search both X posts and the wider web before answering.
• Grounding: Grounding is when an AI platform bases its answer on sources it reads at the moment you ask, like web pages, instead of only on what it remembers.
• Grounding page: A grounding page is one page on your site that states the key facts about your brand plainly, for AI platforms and people to check against.
• Guest blogging: Guest blogging is writing an article for someone else's website, to reach their readers and, often, to earn a link back to your own site.
• Guestographic: A guestographic is an infographic you make and offer to other websites to publish, with a short article and a link back to you.

H

• Hallucinated claims about your brand: Hallucinated claims about your brand are false statements AI platforms make about you, like a wrong price, a feature you don't have or a closed shop.
• Hallucination: A hallucination is when an AI model states something false as if it were fact, often fluently and with no source behind it.
• HTTP 402: HTTP 402 Payment Required is a web status code set aside for payments. It has no agreed use, and some schemes for paid AI crawling are trying it out.

I

• Implied query: An implied query is the question an AI platform works out you meant, and searches for, from the words you actually typed.
• Impression-only reporting: Impression-only reporting tells you your page was shown somewhere, like an AI answer, but not whether anyone clicked, read it or bought.
• Impressions: Impressions count how many times something was shown to people. In AI search, it usually means an estimate of how many people saw an answer naming you.
• Inbound link: An inbound link is a link from another website to yours. It's another name for a backlink.
• Incumbent advantage: Incumbent advantage is the head start well-known brands get in AI answers, because the pages the platforms learn from and read already talk about them.
• Index bloat: Index bloat is when a search engine has indexed lots of your site's low-value pages, like filter results or empty tag pages, that nobody needs to find.
• Indexability: Indexability is whether a search engine can add a page to its index. A page that can't be indexed can't rank, and can't be linked in Google's AI answers.
• Indexing: Indexing is when a search engine stores a page it has crawled, so it can find that page again and show it, or hand it to an AI answer.
• Indirect prompt injection: Indirect prompt injection is hidden instructions on a web page that an AI model follows when it reads the page to answer someone.
• Inference: Inference is running a trained AI model to get an output. Every answer an AI platform gives you is one inference.
• Information gain: Information gain is what a page adds that other pages on the same subject don't: new data, first-hand experience or an original view.
• Informational prompt: An informational prompt is a question to an AI platform from someone who wants to learn or understand something, before they're ready to buy.
• Informational query: An informational query is a search from someone who wants to learn something, like "how to descale a coffee machine", and isn't shopping yet.
• Inline citation: An inline citation is a link to a source placed inside the text of an AI answer, next to the sentence it backs up.
• Inline domain: An inline domain is the website behind a link placed inside an AI answer's text, as opposed to one only listed among its sources.
• Intent: Intent is what someone is trying to do when they ask a question: learn something, compare options, buy, or reach a particular site.
• Internal link: An internal link is a link from one page of your website to another page on the same site, like a guide linking to a product page.
• Interstitial ad: An interstitial ad is a full-screen ad or pop-up that covers a page's content until the visitor closes it or waits it out.
• Inverted pyramid: The inverted pyramid is a way of writing that puts the answer first, then the most important details, then the background.

J

• JavaScript SEO: JavaScript SEO is making sure search engines and AI bots can read the pages on a site that uses JavaScript to build its content.
• JSON-LD: The format most websites use to add structured data to a page, as a block of JSON in the page's HTML.
• Judgment work: Judgment work is the part of a job that needs a person to decide, because it depends on context, experience or taste that isn't written down.

K

• Keyword cannibalization: Keyword cannibalization is when several pages on your site go after the same search, so they compete with each other and none does as well as one would.
• Keyword clustering: Keyword clustering is grouping searches that mean the same thing, so one page can target the whole group instead of one phrase at a time.
• Keyword density: Keyword density is how often a word or phrase appears on a page, as a share of all the words on it.
• Keyword difficulty: Keyword difficulty is an SEO tool's estimate of how hard it would be to rank on Google's first page for a search, usually on a scale of 0 to 100.
• Keyword list: A keyword list is a saved group of related words and phrases you follow together, to see demand and results for one subject.
• Keyword ranking: A keyword ranking is where your page appears in a search engine's results for a given search, such as position 3 on Google.
• Keyword stemming: Keyword stemming is when a search engine treats different forms of a word, like roast, roasted and roasting, as the same word.
• Keyword stuffing: Filling a page with the same words over and over to rank for them, which search engines treat as spam.
• Keywords: Keywords are the words and phrases people type into search engines, and the ones you choose to make your pages findable for.
• Knowledge base: A knowledge base is a collection of documents an AI system treats as true and draws on. For a brand, it's the record of what's true about you.
• Knowledge decay: Knowledge decay is when AI platforms keep repeating facts about you that used to be true, like an old price or a product you've retired.
• Knowledge distillation: Knowledge distillation is training a small AI model to copy a larger one, so it gives similar answers faster and for less money.
• Knowledge graph: A knowledge graph is a database of things, like companies, people and products, and how they're connected, used by search engines and AI.
• Knowledge panel: A knowledge panel is the box of facts Google shows beside its results for a well-known company, person or thing.

L

• Landing page: A landing page is the page a visitor first arrives on, often one built for a single campaign with one thing for the visitor to do.
• Large language model (LLM): A large language model is an AI model trained on huge amounts of text to predict words, which lets it write answers. ChatGPT and Gemini run on them.
• Late chunking: Late chunking is a way of splitting a document for AI search that reads the whole document first, so each piece keeps its surrounding context.
• Latent intent: Latent intent is what someone really wants when they ask a question, beyond the words they typed. AI platforms often try to answer it too.
• Latent Semantic Analysis (LSA): Latent Semantic Analysis is an older way for computers to find which words and documents are related, by spotting words that tend to appear together.
• Link bait: Link bait is content made to attract links and mentions from other sites, like original research, free tools or a strong opinion.
• Link building: Link building is the work of getting other websites to link to yours, to bring visitors and help your pages rank.
• Link citations: Link citations are the clickable links to sources in an AI answer. They show which pages the answer relied on and can send visitors to them.
• Link equity: Link equity is the ranking value a link passes from one page to another, which depends on the linking page's strength, relevance and the link itself.
• Link exchange: A link exchange is when two websites agree to swap links. Done at scale to boost rankings, Google treats it as link spam.
• Link farm: A link farm is a group of websites that exist mainly to link to each other or to paying sites, to push up rankings. Google treats it as spam.
• Link juice: Link juice is SEO slang for the ranking value a link passes from one page to another. It means the same as link equity.
• Link popularity: Link popularity is how many links point to a website or page, and how good they are. It was one of the first things search engines ranked by.
• Link profile: A link profile is the full set of links pointing to your website: how many there are, which sites they come from and what they say.
• Link reclamation: Link reclamation is getting back links to your site that broke or disappeared, and asking sites that mention you to add a link.
• Link rot: Link rot is the way links on the web break over time, as the pages they point to move, change or disappear.
• Link scheme: A link scheme is any arrangement to create links to a site to raise its search rankings, such as buying, swapping or mass-producing links.
• Link spam: Link spam is links placed to manipulate search rankings, from comment and forum links to bought links. Google's policies forbid it.
• Link text: Link text, or anchor text, is the visible, clickable words of a link. It tells readers and search engines what the linked page is about.
• Link velocity: Link velocity is how fast a website gains (or loses) links from other sites over a period, such as new linking sites per month.
• Listicle leverage: Listicle leverage means getting your brand onto the "best of" list articles that AI platforms read when they answer questions about your category.
• LLM crawler: An LLM crawler is a bot an AI company sends to read web pages, to train its language models or to find pages for its answers.
• LLM-as-a-judge: LLM-as-a-judge means using one AI model to grade or label another model's output, such as whether an answer is positive about a brand.
• LLMO (Large Language Model Optimization): LLMO, or large language model optimization, is work to get your brand named and described correctly in the answers AI models write.
• llms-full.txt: llms-full.txt is a file some websites publish with the full text of their key pages in one Markdown file, for AI tools to read at once.
• llms.txt: A proposed Markdown file at the root of a website that lists its most useful pages for AI tools to read.
• Local business schema: Local business schema is structured data that describes a local business, its address, hours and contact details, in a form machines can read.
• Local citation: A local citation is a listing or page online that names your business with its address and phone number, like a directory entry or a news story.
• Local pack: The local pack is the map and short list of nearby businesses Google shows for searches with a local intent.
• Local search marketing: Local search marketing is everything a business does to be found by people searching nearby, from local SEO to local ads and reviews.
• Local SEO: Local SEO is the work of getting a business found in searches with a place in them, like "florist near me" or "coffee shop in Leeds."
• Log file analysis: Log file analysis is reading your server's records of every request to see how search engine and AI bots actually crawl your site.
• Long-tail keyword: A long-tail keyword is a specific search phrase that few people search each month, but which says clearly what the person wants.
• Long-tail relevance: Long-tail relevance is being the right answer for specific, detailed questions, where fewer people ask but those who do are close to deciding.

M

• Machine learning: Machine learning is how computers learn to do tasks from examples, without rules written by a person. Today's AI platforms are built with it.
• Machine-readable content: Machine-readable content is text a bot can read and understand without guessing: in the page's HTML, in clear structure, with no key facts hidden in images.
• Manual action: A manual action is a penalty a person at Google applies when a site breaks its spam policies, lowering or removing pages from search.
• Market perception prompt: A market perception prompt asks an AI platform how to choose in a category, like "What should I look for in a coffee subscription?", without naming brands.
• Marketing Engineer: A Marketing Engineer is someone in marketing who builds their own automations and AI workflows, joining marketing know-how with technical skill.
• max-snippet: max-snippet is a robots meta rule that limits how many characters of your page Google may quote, in search results and in AI Overviews and AI Mode.
• MCP (Model Context Protocol): MCP is an open standard that lets AI assistants like Claude, ChatGPT and Cursor connect to other tools and use them on your behalf.
• MCP server: An MCP server is a connection point a tool offers so AI assistants can read its data or use it, through the Model Context Protocol.
• Mention frequency: Mention frequency is how many times AI answers name your brand, counting every repeat of the name inside one answer.
• Mention refresh: A mention refresh is getting an existing mention of your brand updated, because it's wrong, out of date or describes you poorly.
• Meta AI: Meta AI is Meta's AI assistant, built into WhatsApp, Instagram, Facebook and Messenger, and available on its own at meta.ai.
• Meta description: A meta description is a short summary of a page, written in its HTML, that search engines may show under the title in their results.
• Meta keywords: The meta keywords tag lists a page's keywords in its HTML. Google has ignored it for ranking since at least 2009.
• Meta redirect: A meta redirect, or meta refresh, is an HTML tag that sends the browser to another address, at once or after a delay.
• Meta robots tag: The meta robots tag is a line of HTML that tells search engines whether to index a page, follow its links and show snippets from it.
• Meta tags: Meta tags are lines of HTML in a page's head that describe the page to browsers, search engines and social networks, without showing on the page.
• meta-externalagent: meta-externalagent is Meta's web crawler for training its AI models and for indexing content for its products. You can block it in robots.txt.
• meta-externalfetcher: meta-externalfetcher is Meta's bot that fetches a single link when a person asks for it. Meta says it may not follow robots.txt rules.
• meta-webindexer: meta-webindexer is Meta's crawler for Meta AI's search results. Letting it read your site helps Meta AI find your pages when it answers.
• Microsoft Copilot: Microsoft Copilot is Microsoft's AI assistant, available on the web, in Windows, in the Edge browser and in Microsoft's apps. It searches the web with Bing.
• Mirror site: A mirror site is a full copy of a website hosted at another address, used for downloads, backups or access in other regions.
• MistralAI-User: MistralAI-User is the bot Mistral AI uses to visit a web page when someone asks its assistant a question that needs it.
• Mixed-use crawler: A mixed-use crawler is a bot that reads your pages for more than one purpose, such as search and AI, so blocking it affects both.
• Mixture of Experts (MoE): Mixture of Experts is a way of building AI models where each word is handled by only a few specialised parts of the model.
• Mobile-first indexing: Mobile-first indexing means Google reads and ranks the mobile version of your pages, so what's missing on a phone is missing for Google.
• Model: A model is the trained AI system that writes an answer. A platform like ChatGPT is the app around it, and may use several models.
• Model collapse: Model collapse is when AI models get worse because they're trained on text written by other AI models, losing the rare and varied parts.
• Multimodal AI: Multimodal AI is AI that works with more than text, like images, audio and video, so it can read a photo or answer a question about a chart.

N

• Name matching: Name matching is how a tool decides which brand an AI answer means, counting your short names and spellings as you and leaving out other businesses.
• NAP: NAP is short for name, address and phone number, the basic details of a local business that should match everywhere it's listed online.
• Natural language understanding (NLU): Natural language understanding is the part of AI that works out what a piece of human language means: its intent, the things it names and its tone.
• Natural link: A natural link is one a site owner chose to give you because your page was worth pointing to, with no payment or deal behind it.
• Navigational prompt: A navigational prompt is a question to an AI platform from someone trying to reach a particular brand, website or page, like "Fernhill Coffee login".
• Navigational query: A navigational query is a search from someone trying to reach a particular website or page, like "Fernhill Coffee login" or "Kestrel Roasters."
• Negative SEO: Negative SEO is an attempt to hurt a competitor's search rankings, for example by pointing spammy links at their site or copying their content.
• Network benchmarking: Network benchmarking compares your page or brand numbers with a large pool of other sites, so you can tell whether a result is good or poor.
• New mention: A new mention is getting your brand added to a page the AI platforms already use for your customers' questions, where it names competitors but not you.
• NLWeb: NLWeb is an open project, started by Microsoft, that lets a website answer questions in plain language, for people and for AI agents.
• noai: noai is an unofficial tag a website can add to say it doesn't want its content used for AI. No standard backs it, so most AI bots ignore it.
• noarchive: noarchive is a robots meta rule that asked search engines not to keep a cached copy of a page. Google no longer uses it.
• nofollow: nofollow is a link attribute, rel="nofollow", that asks search engines not to count a link as an endorsement of the page it points to.
• Noindex tag: noindex is a rule that tells search engines not to show a page in their results. At Google, that also keeps it out of AI Overviews and AI Mode.
• Non-branded prompt: A non-branded prompt is a question to an AI platform that names no brand, such as "What's the best coffee subscription?"
• Non-determinism: Non-determinism means an AI platform can give different answers to the same question, even when nothing else has changed.
• Non-parametric memory: Non-parametric memory is the knowledge an AI model looks up while it answers, such as web pages it searches for, as opposed to what it learned in training.
• noopener: noopener is a link attribute, rel="noopener", that stops a page opened in a new tab from controlling the page that opened it.
• noreferrer: noreferrer is a link attribute, rel="noreferrer", that stops the browser telling the next site which page the visitor came from.
• nosnippet: nosnippet tells Google not to show text from a page in its results, and not to use the page as direct input for AI Overviews and AI Mode.
• Not provided in Google Analytics: "(not provided)" is the label Google Analytics shows in place of the search terms Google no longer passes on to websites.
• Notability: Notability is Wikipedia's test for whether a subject gets its own article: has it had significant coverage in independent, reliable sources?

O

• OAI-AdsBot: OAI-AdsBot is OpenAI's bot for checking the pages advertisers submit as ads on ChatGPT. It only visits those pages.
• OAI-SearchBot: OAI-SearchBot is OpenAI's search bot. It reads web pages so ChatGPT can find them and show them in its search answers.
• Objection: An objection is a reason an AI platform gives for not choosing your brand, such as a high price or limited delivery.
• Off-page SEO: Off-site, or off-page, means everything said about your brand on websites you don't run: reviews, press, directories, forums and list articles.
• On-page SEO: On-site, or on-page, means the content and setup of the website you run yourself: your product pages, guides, FAQs and how bots can read them.
• Open Graph meta tags: Open Graph meta tags control how a link to your page looks when it's shared: its title, description and image in the preview card.
• Opportunity: An opportunity is a place where AI answers leave you out, or rank you low, even though you have good reason to be named there.
• Organic search results: Organic search results are the unpaid results a search engine ranks by relevance, as opposed to the ads placed around them.
• Organic traffic: Organic traffic is the visits your site gets from unpaid search results, as opposed to ads, email, social or other sources.
• Organization schema: Organization schema is structured data on your site that tells search engines and AI platforms your company's name, logo, site and profiles.
• Original research: Original research is data you collect and publish yourself, such as a survey or a study, that nobody else has.
• Orphan page: An orphaned page is a page on your site that no other page links to, so crawlers and visitors struggle to find it.
• Outbound link: An outbound link is a link on your website that points to a page on another website.

P

• Page: A page is one web address, such as a single article or product page, as opposed to the whole website it sits on.
• Page health: Page health is how well a page loads and works, judged by measures like loading speed and layout stability, for people and for bots.
• Page speed: Page speed is how quickly a web page loads and becomes usable, usually measured on a phone with Google's Core Web Vitals.
• PageRank: PageRank is Google's original way of judging a page by the links pointing to it, weighing each link by how important its own page is.
• Pages versus domains: Counting sources by domain shows which websites AI platforms rely on. Counting by page shows the exact articles to work on.
• Paid link: A paid link is a link you get in return for money, products or a service. Search engines expect it to be marked as sponsored.
• Parametric memory: Parametric memory is what an AI model learned in training and holds in its own weights, as opposed to what it looks up when it answers.
• Parasite SEO: Parasite SEO is publishing your content on someone else's well-ranked website to borrow its standing in search results and AI answers.
• Passage retrieval: Passage retrieval is when a search system picks out one section of a page that answers a question, instead of treating the whole page as one.
• Pay-per-crawl: Pay-per-crawl is charging AI companies a fee each time their crawler reads a page on your site.
• Pay-per-inference: Pay-per-inference is the idea of paying a website each time its content is used to write an AI answer, instead of each time it's crawled.
• People Also Ask: People Also Ask is the Google search feature that lists related questions, each opening to a short answer from a web page.
• Perception problem: A perception problem is when AI answers paint your brand wrongly because the pages they read tell the wrong story, while the product itself is fine.
• Perplexity: Perplexity is an AI answer engine that searches the web for each question and shows numbered sources next to its answer.
• Perplexity instant checkout: Perplexity instant checkout lets a shopper buy some products inside a Perplexity answer, without going to the shop's own site.
• Perplexity Search: Perplexity Search is Perplexity's web search, offered to developers as an API that returns ranked web results instead of a written answer.
• Perplexity-User: Perplexity-User is the fetcher Perplexity sends to read a page when someone asks it a question, as opposed to crawling the web ahead of time.
• PerplexityBot: PerplexityBot is Perplexity's crawler. It reads web pages so Perplexity can find them and link them in its answers, and it isn't used for model training.
• Personas: In AI search tracking, a persona is a kind of customer you ask a question as, so you see the answer that person would get.
• Pickup lag: Pickup lag is how long it takes for a change to a web page to show up in what the AI platforms say.
• Pillar page: A long page that covers a whole topic and links to shorter pages that go into each part of it in detail.
• Pitch workspace: A pitch workspace tracks a prospect's AI visibility for a short time, so an agency can show them where they stand before they sign.
• Platform: A platform is an AI product people ask questions, like ChatGPT, Perplexity or Google AI Overviews, whose answers you track.
• Pogo-sticking: Pogo-sticking is when a searcher clicks a result, comes straight back to the results page, and tries another one.
• Position: Position is where your brand is named in an AI answer, counting the first brand named as 1.
• Position bias: Position bias is a language model's habit of using information at the start or end of what it reads better than information in the middle.
• Position-Adjusted Word Count (PAWC): Position-Adjusted Word Count is a research measure of how much of an AI answer comes from one source, giving more weight to sentences near the top.
• PR gap: A PR gap is where the sites AI platforms read talk about your competitors and leave you out, or describe you worse than them.
• Presence rate: Presence rate is the share of AI answers that name your brand at least once. In AnswerOrbit it's called visibility.
• Primary keyword: A primary keyword is the main search term a page is written to rank for, which shapes its title, headings and opening.
• Private blog network (PBN): A private blog network is a group of websites run by one owner to link to another site and push it up the search results.
• Probability-based engine: A probability-based engine writes its answers by predicting likely words, so the same question can get a different answer each time.
• Product schema: Product schema is structured data on a product page that states the product's name, price, availability and reviews in a form machines can read.
• Product tile: A product tile is a card in an AI answer that shows one product, with its picture, price and a link to buy it.
• Prominence: Prominence is how strongly an AI answer puts you forward when it names you: how early you come and how much space you get.
• Prompt: A prompt is a question someone asks an AI platform, in their own words. In AI search it takes the place the keyword had in classic search.
• Prompt clustering: Prompt clustering is grouping prompts that ask the same thing in different words, so you can measure a subject as a whole.
• Prompt engineering: Prompt engineering is writing the input to an AI model so it gives the output you want. In AI search tracking, it means writing prompts the way customers ask.
• Prompt generator: A prompt generator writes a set of prompts for you to track, from your topics, markets and the kind of questions you want to cover.
• Prompt group: A prompt group is a set of prompts you track and read together, usually about one subject, worded different ways on purpose.
• Prompt injection: Prompt injection is text written to make an AI model ignore its instructions and follow someone else's, typed into the chat or hidden in what it reads.
• Prompt research: Prompt research is finding the questions your customers ask AI platforms, so you know which ones to track and which to answer.
• Prompt research reports: A prompt research report compares the questions people ask AI platforms with the prompts you track, and lists the ones you're missing.
• Prompt run: A prompt run is one ask of one prompt on one AI platform. Every visibility number is built from runs, so their count tells you how solid a number is.
• Prompt space: Prompt space is every way people could ask an AI platform about your category. You track a sample of it, so the sample needs to be fair.
• Prompt type: Prompt type is the kind of question a prompt is, such as learning, comparing or buying, so you can read each kind on its own.
• Prompt volume: Prompt volume is an estimate of how often people ask AI platforms a question or use a phrase, the AI search version of keyword search volume.

Q

• Query deserves freshness (QDF): Query deserves freshness is the idea that some searches need recent results, so Google favours newer pages when a topic is in the news.
• Query fan-out: Query fan-out is when an AI platform turns one question into several web searches of its own before it writes the answer.
• Query fan-out estimator: A query fan-out estimator guesses which web searches an AI platform will run for a question, so you can write content that covers them.

R

• RAG (retrieval-augmented generation): RAG is how most AI search works: the platform fetches pages first, then writes its answer from what they say.
• Rank and citation relationship: The rank-citation relationship is the link between ranking well in Google and being cited in AI answers, real but looser than it was.
• Re-indexing: Re-indexing is when a search engine or AI platform fetches a page again and replaces its stored copy with the current one.
• Reality problem: A reality problem is when AI answers describe your brand correctly, and what they describe puts people off.
• Recommendation poisoning: Recommendation poisoning is planting or rigging list articles, reviews and comparisons so AI platforms recommend one brand over others.
• Referrer string: A referrer string is the address a browser passes to your site to say which site sent a visitor, like chatgpt.com.
• Region: A region is the country, and sometimes city, a prompt is asked from. AI platforms answer differently depending on where the person is.
• Relevance engineering: Relevance engineering is shaping your content so AI systems can find it, read it and judge it the best match for a question.
• Reranking: Reranking is a second, closer look at the passages a search found, putting the best matches first before the AI model reads them.
• Retired crawler tokens: Retired crawler tokens are bot names an AI company no longer uses or lists. robots.txt rules that name them do nothing.
• Retrieval: Retrieval is the step where an AI platform fetches web pages or passages to answer a question, on top of what its model learned in training.
• Retrieval poisoning: Retrieval poisoning is planting pages built to be fetched for certain questions so an AI answer repeats something false.
• Reverse mention: A reverse mention is a deal between two brands to name each other in their own content, so each shows up where the other is read.
• Review-site influence: Review-site influence is how much the reviews on sites like Trustpilot or G2 shape what AI answers say about a brand.
• RLHF (Reinforcement Learning from Human Feedback): RLHF trains an AI model on people's ratings of its answers, so it learns to give the kind of reply people prefer.
• robots.txt: A plain text file at the root of your website that tells crawlers, AI bots among them, which pages they may read.
• RSL (Really Simple Licensing): RSL is an open standard that lets publishers state, in a file machines can read, how AI companies may use their content and on what terms.

S

• SAGE: SAGE is a four-stage loop for improving how AI platforms present your brand: Setup, Analyze, Generate, Engineer.
• sameAs: sameAs is a structured data property that links your brand on your website to its official profiles elsewhere, like Wikipedia or LinkedIn.
• Scaled content abuse: Scaled content abuse is Google's name for publishing lots of pages mainly to game search rankings, whether written by people or by AI.
• Schema evidence problem: The schema evidence problem is that nobody has shown structured data makes AI answers cite you more, even though it's often sold as if it does.
• Schema markup and Schema.org: The shared vocabulary of types and properties that structured data uses to describe things on the web.
• Search algorithm: A search algorithm is the set of rules a search engine uses to decide which pages to show for a search, and in what order.
• Search engine poisoning: Search engine poisoning is when attackers push harmful pages up the search results so people click them and get scammed or infected.
• Search engine results pages (SERPs): A SERP is the page a search engine shows after a search: the list of links, plus ads, AI answers and other boxes.
• Search gap: A search gap is a search AI platforms run while answering your customers' questions that none of your pages answer.
• Search intent: Search intent is the reason behind a search: what the person wants to learn, compare, buy or reach when they type it.
• Search results: Search results are the pages, links and answers a search engine shows in reply to a search.
• Search term: A search term is the word or phrase a person types into a search engine. In AI search, the question they ask plays the same part.
• Search variation: Search variation is when an AI platform runs different searches and reads different pages each time it answers the same question.
• Search visibility: Search visibility estimates how much of the clicks from search results your site gets for a set of keywords, based on its rankings.
• Search volume: Search volume is how many times people search a keyword in a month, on average. It's the usual way to size demand in SEO.
• Secondary keywords: Secondary keywords are the related phrases a page targets alongside its main keyword, like synonyms, subtopics and longer versions.
• Secure Sockets Layer (SSL): SSL is the old protocol for encrypting the connection between a browser and a website. TLS replaced it, but the name stuck.
• Seed keywords: Seed keywords are the short, broad words you start keyword research from, which tools then grow into longer lists of phrases.
• Self-reported attribution (SRA): Self-reported attribution means asking customers how they found you, which picks up AI recommendations your analytics can't see.
• Semantic chunking: Semantic chunking splits a page into pieces where the subject changes, so an AI system can read each piece by itself.
• Semantic HTML hierarchy: Semantic HTML hierarchy is a page's outline in its code: one main heading, sections under it and subsections under those, in order.
• Semantic search: Semantic search finds content by what it means, so a page can match a question even when it uses different words.
• Sentiment: Sentiment is the tone an AI answer takes when it names your brand: positive, neutral or negative.
• Sentiment drift: Sentiment drift is a slow change in the tone AI answers take about your brand, usually traced to a source that changed or a new one.
• Sentiment prompt: A sentiment prompt asks an AI platform to judge a brand, like "Is Fernhill Coffee any good?", so you can read what it thinks.
• Sentiment score: A sentiment score is one number for how positively AI answers talk about a brand, usually its share of positive mentions.
• SEO audit: An SEO audit is a check of a website for the problems that hold it back in search, from pages bots can't read to weak or missing content.
• SEO gap: An SEO gap is a technical problem that stops AI platforms reaching, reading or using your pages, like a blocked bot or hidden text.
• SEO silo: An SEO silo is a way of grouping a site's pages by topic, with internal links that mostly stay inside each group.
• SERP features: SERP features are the extra boxes on a search results page besides the plain links: AI Overviews, maps, snippets, videos, products and more.
• Server-side rendering: Server-side rendering builds a page's full HTML on the server, so bots get all the text without running any scripts.
• SGE (Search Generative Experience): SGE was Google's 2023 test of AI-written answers at the top of search results. It became AI Overviews in 2024.
• Share of sources: Share of sources is the part of all the pages AI answers used that came from one website, such as yours or a competitor's.
• Share of voice: Share of voice is your share of all the mentions of the brands you track in AI answers.
• Share of voice rank: Share of voice rank is your place among the brands you track, ordered by their share of the mentions in AI answers.
• Shopping: Shopping, in AI search, is when people ask an AI platform what to buy and it answers with products, prices and links to shops.
• Shopping mode: Shopping mode is when an AI platform answers a buying question with product cards, prices and shop links, in place of plain text.
• Shopping mode rate: Shopping mode rate is the share of your prompts, or of their answers, where an AI platform showed products in place of plain text.
• Shopping visibility score: Shopping visibility score is the share of AI shopping answers that show at least one of your products.
• Short-tail keywords: Short-tail keywords are short, broad search phrases of one to three words that many people search, like "coffee beans."
• Signed agents: Signed agents are AI bots that sign each request with a cryptographic key, so a website can check who really sent it.
• Site reputation abuse: Site reputation abuse is publishing third-party content on a trusted site mainly to borrow its search ranking, against Google's rules.
• Sitelinks: Sitelinks are the extra links to other pages of a site that Google sometimes shows under a search result, often for brand searches.
• Sitemaps: A sitemap is a file that lists the pages on your site you want search engines to find, usually in XML at /sitemap.xml.
• Sitewide link: A sitewide link is a link repeated across all the pages of a site, usually because it sits in the menu, sidebar or footer.
• Skills: Skills are packaged instructions an AI assistant can load to do one kind of task well, like writing in a set format or using a tool.
• Slopsquatting: Slopsquatting is registering a name that AI models tend to make up, such as a software package or a web address, to catch the people who trust the answer.
• Social citation: A social citation is when an AI answer cites a community or social site, such as a forum thread, a Reddit post or a YouTube video.
• Source gap: A source gap is a page AI platforms read for answers that name your competitors and not you.
• Source visibility: Source visibility is how often a website or page is used across AI answers over time, which shows which sites the platforms lean on.
• Sources block: A sources block is the list of links an AI answer shows apart from its text, such as at the end or in a side panel.
• Spamdexing: Spamdexing is tricking a search engine into ranking pages higher than they deserve, through tricks like keyword stuffing, hidden text and link schemes.
• Specific objection: A specific objection is a reason against your brand that AI answers give only for certain questions, so a targeted fix can deal with it.
• srcset: srcset is an HTML attribute that offers the browser several sizes of an image, so each screen downloads one that fits.
• Statistics page: A statistics page collects figures on one subject, each with its source, and AI answers often cite it when they need a number.
• Strategic text sequence: A strategic text sequence is a string of text written to push an AI model into recommending a product, a form of manipulation shown in research.
• Strength: In AI search, a strength is a quality AI answers already tie to your brand, accurately and in your favour, like fresh beans or fast delivery.
• Structured data: Code added to a web page that tells machines what the page is about, such as a product, its price and its reviews.
• Subdomain: A subdomain is a separate section of a domain with its own name in front, like shop.yourbrand.com or blog.yourbrand.com.
• Synthesis: Synthesis is the step where an AI platform writes one answer from the several sources it has read, merging and shortening what they say.
• Synthetic data: Synthetic data is text or other data made by AI models and used to train other models, alongside or in place of data written by people.
• Synthetic query: A synthetic query is a web search an AI platform writes and runs by itself while answering someone. No person typed it.

T

• Tag: A tag is a label you add to prompts to group them across topics, like "price" or "comparison," so you can filter your numbers by it.
• Taxonomy SEO: Taxonomy SEO is organizing a site's content into categories and tags that make sense to people and search engines, so pages are easy to find.
• Technical SEO: Technical SEO is the work that lets search engines and AI bots reach, read and index your pages: access, speed, structure and code.
• Temperature: Temperature is a setting that controls how much randomness an AI model uses as it writes, one reason the same prompt gets different answers.
• Terminology war: The terminology war is the industry argument over what to call the work of getting named in AI answers: GEO, AEO, LLMO, AI SEO and more.
• TF-IDF: TF-IDF is a formula that scores how important a word is to a page: frequent on that page, rare across others.
• Theme: A theme is a topic AI answers keep bringing up when they describe brands in your industry, like price or delivery.
• Thin content: Thin content is a page with little of value to a reader: too short, copied, auto-generated or written only to rank.
• Third-party signals: Third-party signals are what independent sites say about your brand, such as reviews, press, forums and directories, which AI answers lean on.
• Third-party validation: Third-party validation is independent sources backing up a specific claim you make, which makes AI answers far more likely to repeat it.
• Tiered link building: Tiered link building means building links to the pages that link to you, in layers, to pass more strength to your site. It's link spam.
• Title tag: A title tag is the HTML element that names a page. Search engines often use it as the clickable headline in results.
• Token: A token is the small piece of text an AI model reads and writes, a word or part of one. Models count length, limits and cost in tokens.
• TollBit: TollBit is a company that helps websites see, control and charge for visits from AI bots, and publishes data on how AI bots use the web.
• Top citation domains: Top citation domains are the websites AI answers cite most for your prompts, ranked, so you can see which few sites shape your answers.
• Top citation pages: Top citation pages are the single web pages AI answers cite most for your prompts, ranked, which tells you exactly which articles to work on.
• Top-level domain (TLD): A top-level domain is the last part of a domain name, after the final dot, such as .com, .uk or .coffee.
• Topic: In AI visibility tracking, a topic is a group of prompts about one subject, like office coffee, so you can see which subjects you're named for.
• Topic cluster: A topic cluster is a set of pages on one subject, built around a main pillar page and linked to each other, so a site covers the subject fully.
• Topical authority: Topical authority is being treated as a reliable source on a whole subject because you cover it in depth, which helps you get read and cited.
• Topical relevance: Topical relevance is how closely a page or link matches the subject at hand. Links and mentions from pages on your topic count for more.
• Training data: Training data is the text an AI model learned from before it was released. What it says about your brand is fixed until the next model.
• Training data cutoff: The training data cutoff is the date an AI model's training data ends. Anything about your brand after it, the model only knows by searching.
• Transactional prompt: A transactional prompt is a question to an AI platform from someone ready to buy or act, like "Order a bag of decaf beans under £10".
• Transactional query: A transactional query is a search from someone ready to act, usually to buy, like "order coffee beans online."
• Transformer: The transformer is the design behind today's AI language models. Each word in a passage is compared with every other word to work out meaning.
• Transport Layer Security (TLS): TLS is the protocol that encrypts the connection between a browser and a website. It's the S in HTTPS and the successor to SSL.
• Trust gap: A trust gap is when AI answers name your brand but don't use or cite your own pages, so the platforms know you but don't trust your site as a source.
• Trust problem: A trust problem is when AI platforms know you do something well but don't recommend you for it, because too few outside sources back you up.
• TrustRank: TrustRank is a method for separating good sites from spam by starting from a few trusted sites and following their links outward.

U

• UCP (Universal Commerce Protocol): UCP, the Universal Commerce Protocol, is an open standard that lets AI agents find products, fill a cart and check out with a shop.
• UGC link attribute: rel="ugc" marks a link as user-generated content, like a link in a comment or forum post, so search engines treat it with care.
• Unbranded prompt: An unbranded prompt is a question to an AI platform that names no brand, like "What's the best coffee subscription?" Also called non-branded.
• Universal search: Universal search is Google mixing images, videos, news, maps and products into one results page, instead of showing only web pages.
• Unnatural links: Unnatural links are links made to game search rankings, such as bought, swapped or mass-produced links, instead of links a site chose to give.
• URL Rating (UR): URL Rating is Ahrefs' score from 0 to 100 for how strong one page's backlinks are. It's a tool's estimate, which Google doesn't use.
• URL slug: A URL slug is the last part of a web address, the words that name one page, like cold-brew-recipe in /blog/cold-brew-recipe.
• User agent: A user agent is the name a bot or browser gives with every request to your site, like GPTBot. robots.txt rules are written against it.
• User intent: User intent is what a person is really trying to get done when they search or ask, which can go beyond the words they use.
• utm_source=chatgpt.com: utm_source=chatgpt.com is the tag ChatGPT adds to the links it sends people to, so your analytics can tell those visits came from ChatGPT.

V

• Validation prompt: A validation prompt checks whether an AI platform knows a brand does a specific thing, like "Does Fernhill Coffee offer decaf?"
• Vector database: A vector database stores text as lists of numbers that capture meaning, so an AI system can find passages that mean the same as a question.
• Vertical search: Vertical search is search focused on one kind of content or one field, like products, jobs, travel, maps or videos.
• Visibility rank: Visibility rank is your place among the brands AI answers name, ordered by how often each one is named.
• Visibility score: A visibility score sums up how often AI answers name your brand across the prompts you track.
• Voice search: Voice search is asking a search engine or assistant a question out loud, through a phone, smart speaker, car or AI app.

W

• Watchlist: A watchlist is a set of websites or pages you pick to follow in an AI visibility tool, to see how often AI answers use them over time.
• Web Bot Auth: Web Bot Auth is a proposed standard that lets bots prove who they are by signing each request, so a site can tell real AI bots from fakes.
• WebMCP: WebMCP is a draft web standard that lets a website offer AI agents ready-made actions, like search the catalogue or check delivery, in the browser.
• Website authority: Website authority is an SEO tool's estimate of how strong a whole site is, mostly from the links pointing to it.
• Website hit: A website hit is any single request to a web server, for a page, an image, a script or a font. One page view can be dozens of hits.
• Website structure: Website structure is how a site's pages are organized and linked to each other, from the homepage down to each article and product.
• Webspam: Webspam is content or links made to trick search engines into ranking pages higher than they deserve, at readers' expense.
• White-hat SEO: White-hat SEO is improving search rankings by methods search engines allow: useful content, a sound site and links people choose to give.
• White-label client report: A white-label report is a report an agency sends its client under the agency's own name and logo, with the tool behind it left out.
• Wikidata: Wikidata is the free, open database of facts behind Wikipedia, which search engines and AI systems often draw on to understand entities like brands.
• Wikipedia presence: Wikipedia presence is whether your brand has a Wikipedia article, or is mentioned in others, which AI models read in training and in search.
• Workspace: A workspace is the shared account your team works in, holding your brands, members, access and billing. In some tools it means one brand.

X

• X-Robots-Tag: X-Robots-Tag is an HTTP header that gives crawlers rules like noindex for a file, including PDFs and images that can't hold a meta tag.

Y

• YMYL pages: YMYL pages, short for Your Money or Your Life, cover topics where bad information could hurt someone's health, finances or safety.

Z

• Zero-click rate: The zero-click rate is the share of searches or AI answers that end without a click to any website.
• Zero-click search: A zero-click search is one where the person gets what they need from the results or the AI answer itself, and doesn't click through to any site.