Learn · Glossary
Plain definitions of the 35 terms you meet when you work on getting your company named by ChatGPT, Gemini, Claude and Perplexity. Each one says what it means, why it matters, and where to go next.
How often, how early and how well assistants like ChatGPT, Gemini, Claude or Perplexity bring up a company when people ask questions it ought to appear in.
Why it matters: Buyers increasingly ask an assistant for a shortlist before they visit any website. If you are not in the answer, you are not on the list.
How to measure AI visibility →A tool that replies to a question with a written answer instead of a list of links. ChatGPT, Gemini, Claude, Perplexity and Google's AI Overviews all work this way.
Why it matters: An answer engine usually names only a handful of companies, so the competition for each mention is tighter than for a page of search results.
What is answer engine optimization? →Also called: AEO
The work of making a company easy for answer engines to find, understand, trust and quote, so it is named in the answers buyers read.
Why it matters: It builds on SEO but aims at a different result: being quoted in the answer, not just ranking a page.
AEO explained →Also called: GEO
Another name for optimizing for AI-generated answers. It is often used for the same work as AEO, with more focus on generative assistants such as ChatGPT.
Why it matters: You will see AEO and GEO used for the same goal. What matters is the outcome: being named, accurately, when buyers ask.
AEO vs SEO vs GEO →Also called: SEO
Improving how pages show up in classic search results, such as Google's list of blue links.
Why it matters: Many AI assistants read the web to answer, so pages that search engines can find and understand still matter for AI answers.
How SEO and AEO fit together →Also called: LLM
The type of AI model behind assistants like ChatGPT, Gemini and Claude. It writes text by predicting likely words, based on what it learned in training and anything it reads at the time of the question.
Why it matters: An LLM can only describe your company well if clear, consistent information about you exists where it learns and where it searches.
Also called: Google AI Overview
The AI-written summary Google shows above its regular search results for some searches.
Why it matters: It answers the question on the results page itself, so being named inside the summary can matter as much as ranking below it.
A search where the person gets what they need on the results page or in an AI answer, without clicking through to any website.
Why it matters: Fewer clicks means the answer itself is the shop window. What the answer says about you matters more than ever.
Also called: Prompt
A question a real buyer might ask an AI assistant, such as “best CRM for a 20-person agency”. In AI visibility work these are the prompts you test.
Why it matters: Testing the questions buyers actually ask, instead of your own brand name, shows whether AI recommends you to people who do not know you yet.
Start with buyer questions →When an AI answer names your company anywhere in its reply.
Why it matters: A mention is the first step, but a passing mention is weaker than being recommended.
How to check if ChatGPT recommends you →When an AI answer suggests your company as a good choice for the question, not just names it.
Why it matters: Buyers act on recommendations. A company can be mentioned often and still rarely be recommended.
Why AI recommends your competitor →How early your company appears in an AI answer: first, in the top three, or further down.
Why it matters: People read the first names most closely, so moving up a list can matter as much as being on it.
Also called: Share of voice
Your share of all the company mentions across a set of buyer questions, compared with your competitors.
Why it matters: It turns many separate answers into one number you can compare with rivals and track over time.
Measuring share of voice →Also called: Source
A website an AI assistant shows or links to as a source for its answer. Perplexity, Google's AI Overviews and ChatGPT search often list their sources.
Why it matters: Citations show which websites shape the answers in your category, which tells you where your company needs to be described well.
Track what Perplexity cites →Also called: Run-to-run variation
The way an AI assistant can give a different answer, or a different list of companies, to the same question asked twice.
Why it matters: One answer is a snapshot, not a result. Reliable tracking asks each question more than once and reports a range.
Does ChatGPT give the same answer every time? →Whether what AI assistants say about your company is correct: what you sell, who you serve, your prices and your strengths.
Why it matters: Being named with the wrong facts can cost you more than not being named. Accuracy is worth checking as closely as visibility.
How we measure →Whether an AI answer describes your company in a positive, neutral or negative way.
Why it matters: Two companies can be mentioned equally often while one is described as “expensive” and the other as “easy to start”.
The large body of text an AI model learned from before it was released.
Why it matters: What the model learned in training shapes its default view of your category, which is why consistent public information about you matters over time.
The date after which an AI model's training data stops. Events after the cutoff are unknown to the model unless it searches the web.
Why it matters: Recent launches, prices or rebrands may be missing from answers that rely on training data alone.
Also called: Grounding, retrieval
When an AI assistant searches the web while answering and uses the pages it finds. Often called grounding or retrieval.
Why it matters: When an assistant searches, your current pages and the sites that mention you can shape today's answer, not just what it learned in the past.
How buyers use AI to shortlist vendors →Also called: AI bot
A program an AI company uses to read web pages, either to train its models or to fetch pages when answering a question.
Why it matters: If your website blocks the crawlers that fetch pages for answers, assistants may not be able to read your site when buyers ask.
Which AI crawlers to allow →A small file on your site, at yourdomain.com/robots.txt, that tells crawlers where they may and may not go.
Why it matters: It is the main switch for allowing or blocking AI crawlers. One wrong line can hide your whole site from an assistant.
How to check your robots.txt →A suggested file, found at yourdomain.com/llms.txt, that offers AI tools a brief overview of a site and points to its most useful pages. It is a community idea, not an official standard.
Why it matters: Adding one costs little. So far, the big AI companies have not confirmed they read it when choosing which companies to name.
Does llms.txt help? →A clearly identifiable thing, such as a company, product or person, that search engines and AI models can recognise and connect to facts about it.
Why it matters: When your company is a clear entity, with the same name, description and links everywhere, AI is less likely to confuse you with someone else.
A database of entities and the facts that connect them, such as which company makes which product. Google runs the best-known one.
Why it matters: Facts in knowledge graphs feed many answers. Consistent, structured information about your company helps it be represented correctly.
Also called: Off-site mentions
What other websites say about your company: reviews, lists, comparisons, news and directories.
Why it matters: AI assistants often lean on what independent sites say. Being described well on the sites that rank for your category can matter as much as your own pages.
Why rivals get named instead →Also called: Structured data
Code added to a web page that labels its content for machines, such as the company name, a product's price, or a list of questions and answers.
Why it matters: It makes your facts unambiguous to search engines and AI, and it is one of the few things you fully control.
Run the free Schema Check →The most common format for adding schema markup: a small block of structured code placed in the page.
Why it matters: Google recommends this format, and it keeps the markup separate from the visible text, so it is easy to add and check.
Schema markup for AI search →Also called: FAQPage
Schema markup that labels a list of questions and answers on a page.
Why it matters: It presents your answers as clear question and answer pairs, the same shape AI assistants use when they reply.
Which schema types matter →Writing that answers the question directly in the first one or two sentences, then adds detail.
Why it matters: AI assistants look for a clear, quotable answer. Burying it under a long introduction makes your page harder to use.
Mentiq writes answer-first pages →Also called: Vs page
A page that compares your company with a named alternative or explains how to choose between options.
Why it matters: Buyers often ask AI to compare options. A fair, factual comparison page gives assistants something to quote.
Write a comparison page with Mentiq →How completely and consistently a website covers a subject, so it becomes a trusted source on it.
Why it matters: Several connected, accurate pages on one subject usually earn more trust than one page on many subjects.
Also called: Experience, expertise, authoritativeness, trust
Google's way of describing what makes content trustworthy: first-hand experience, real expertise, a recognised reputation and accurate, honest information.
Why it matters: The same qualities make a page safer for an AI assistant to quote: named authors, real examples, and facts that can be checked.
When an AI model states something false as if it were fact, such as a feature a product does not have or a price that is wrong.
Why it matters: It can happen about your company. Checking brand accuracy regularly and publishing clear facts reduces the room for mistakes.
How we check answers →Also called: Hype
A bold statement with no proof behind it, such as “the best”, “industry-leading” or “guaranteed results”.
Why it matters: Vague superlatives give AI nothing it can verify, and they cost trust with careful buyers. Specific, provable facts travel further.
Mentiq flags hype claims →No term matches that. Try a shorter word.
See whether AI names your company today, then fix what it gets wrong.