AI search, explained
The AEO & GEO Glossary
Every term you need to understand AI visibility, Answer Engine Optimization and Generative Engine Optimization — defined clearly, in plain English.
Technical
AI Crawler
An AI crawler is an automated bot operated by an AI company, such as OpenAI's GPTBot or PerplexityBot, that fetches web content to power training data or live retrieval for answer engines.
Fine-Tuning
Fine-tuning is the process of further training a pre-existing AI model on a focused dataset to specialize its behavior or knowledge, distinct from the retrieval that grounds most live answer-engine responses.
Grounding
Grounding is the practice of tying an AI model's generated answer to specific, verifiable external sources, typically via retrieval, so the response reflects real, current information rather than the model's unaided guesses.
Hallucination
A hallucination is when an AI model confidently generates false or fabricated information, such as inventing a product feature, a statistic, or a nonexistent fact about a brand, presented as if it were true.
Knowledge Graph
A knowledge graph is a structured network of entities, people, companies, products, and the relationships between them, that search and AI systems use to understand facts about the world and answer questions confidently.
llms.txt
llms.txt is a proposed plain-text file placed at a website's root that gives AI systems a concise, curated guide to the site, its purpose, key pages, and citable facts, so models can understand and reference it more accurately.
Retrievability
Retrievability is how easily an AI system can find, access, and pull your content into an answer, a prerequisite for citation, since content that can't be retrieved can never be quoted no matter how good it is.
Retrieval-Augmented Generation (RAG)
Retrieval-Augmented Generation (RAG) is an AI technique where a model retrieves relevant documents from an external source at query time and uses them to ground its generated answer, rather than relying only on knowledge stored in its parameters.
Schema Markup
Schema markup is standardized structured-data code, based on schema.org vocabulary, added to a web page so search and AI systems can unambiguously understand its content, such as that a page describes an Organization, Product, or FAQ.
Semantic Search
Semantic search is a retrieval approach that matches content by meaning rather than exact keywords, using vector representations of text so a query finds conceptually relevant passages even when the wording differs.
Structured Data
Structured data is information organized in a standardized, machine-readable format, most commonly schema.org markup on web pages, that lets search and AI systems reliably understand what your content means rather than inferring it from prose.
Training Data
Training data is the large corpus of text an AI model learns from during training, which shapes its parametric knowledge, including what it knows and believes about your brand before any live retrieval happens.
Vector Embedding
A vector embedding is a numerical representation of text that captures its meaning in a high-dimensional space, enabling AI systems to compare content by semantic similarity and retrieve the passages most relevant to a query.
Wikidata
Wikidata is a free, collaboratively edited knowledge base of structured facts that feeds Wikipedia, search knowledge graphs, and AI systems, making a well-maintained Wikidata entry a high-authority signal for how models understand your brand.
Metric
AI Referral Traffic
AI referral traffic is website visits that originate from AI engines like ChatGPT, Perplexity, or Google AI Overviews when a user clicks a cited link, typically low in volume but high in buyer intent.
AI Visibility
AI visibility is how often and how prominently your brand appears in answers from AI engines like ChatGPT, Perplexity, and Google AI Overviews, treated as a measurable signal across many prompts rather than a single lookup.
Brand Mention
A brand mention is any instance where an AI engine names your brand in its answer, the most basic unit of AI visibility, counted across prompts to compute your mention rate.
Citation in AI Answers
A citation in AI answers is when an AI engine links to or attributes a claim to your domain as a source within its generated response, signaling the model both retrieved and trusted your content.
Citation Rate
Citation rate is the share of AI answers in which your domain appears as a cited source, measuring how often engines not only know your brand but actively attribute information to your site.
Mention Rate
Mention rate is the share of AI answers in a tracked prompt set that name your brand anywhere in the text, whether or not your domain is cited, measuring how often the AI brings you up at all.
Prominence
Prominence measures how early and conspicuously your brand appears within an AI answer, since being named in the first sentence carries far more weight than a passing mention buried at the end.
Sentiment Analysis (AI Answers)
Sentiment analysis in AI answers assesses whether an engine describes your brand positively, neutrally, or negatively when it appears, revealing not just if you're mentioned but how favorably.
Share of Answer
Share of answer is the percentage of AI-generated answers to your category's buyer questions in which your brand is mentioned or cited, a direct measure of how much of the AI answer space you occupy versus competitors.
Share of Voice (AI)
Share of voice in AI search is the proportion of relevant AI answers in which your brand appears compared to competitors, showing how much of the AI conversation in your category you own.
Engine
Amazon Rufus
Amazon Rufus is Amazon's AI shopping assistant that answers product questions and makes recommendations directly within Amazon's storefront, making it a commerce-focused answer engine especially relevant to ecommerce and DTC brands.
ChatGPT Search
ChatGPT Search is OpenAI's web-connected mode within ChatGPT that answers questions using live retrieval and citations, making it one of the most influential answer engines buyers use to research and shortlist products.
Claude
Claude is Anthropic's family of AI assistant models, used both directly by millions of users and embedded in products and workflows, making it another engine whose answers can mention or recommend brands in a category.
DeepSeek
DeepSeek is a family of open, cost-efficient AI models from a Chinese lab that has gained rapid adoption, and it appears both as a consumer assistant and as an affordable engine powering AI features and tooling.
Google AI Mode
Google AI Mode is a conversational, AI-first search experience within Google that returns fully generated, multi-step answers to complex queries, going beyond AI Overviews toward a chat-style search interface.
Google AI Overviews
Google AI Overviews are AI-generated summaries that appear at the top of Google search results, synthesizing an answer from multiple web sources and often citing them, frequently resolving a query without the user clicking through to any site.
Google Gemini
Google Gemini is Google's family of AI models powering its Gemini assistant and features across Google's ecosystem, including contributions to AI Overviews and AI Mode, making it a key engine for AI visibility in Google's surfaces.
Grok
Grok is xAI's AI assistant, integrated tightly with the X platform and its real-time data, that answers user questions and can surface brand recommendations, making it another engine relevant to AI visibility.
Meta AI
Meta AI is Meta's assistant built into WhatsApp, Instagram, Messenger, and Facebook, reaching an enormous audience and answering user questions, which makes it a high-reach engine for brand mentions in AI answers.
Microsoft Copilot
Microsoft Copilot is Microsoft's AI assistant, integrated across Windows, Bing, and Microsoft 365, that answers questions using AI models with web grounding, serving as another answer engine where brands compete for inclusion.
Perplexity
Perplexity is an AI answer engine that combines live web search with language models to produce concise, citation-rich answers, and its transparent sourcing makes it a benchmark for measuring AI citation visibility.
Core Concept
Answer Engine
An answer engine is a system that returns a single synthesized answer to a user's question rather than a list of links, drawing on AI models and often live web retrieval; examples include ChatGPT, Perplexity, and Google AI Overviews.
Answer Engine Optimization (AEO)
Answer Engine Optimization (AEO) is the practice of structuring and publishing content so that AI answer engines like ChatGPT, Perplexity, and Google AI Overviews mention, cite, or recommend your brand directly in their generated answers.
Consideration Set
A consideration set is the shortlist of brands a buyer, or an AI engine acting on their behalf, treats as viable options for a category, and appearing in it is the practical goal of AI visibility.
E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness)
E-E-A-T stands for Experience, Expertise, Authoritativeness, and Trustworthiness, Google's framework for content quality that also strongly influences whether AI engines treat your content as credible enough to cite.
Featured Snippet
A featured snippet is a highlighted answer box at the top of Google's results that quotes a concise answer directly from a web page, an early form of zero-click, answer-first surface and a precursor to today's AI Overviews.
Generative Engine Optimization (GEO)
Generative Engine Optimization (GEO) is the practice of optimizing content to be surfaced, quoted, and cited by generative AI engines, so your brand appears inside the AI-generated responses users read instead of a traditional search results page.
Knowledge Panel
A knowledge panel is the information box that appears in search results summarizing key facts about an entity, drawn from a knowledge graph, and its presence signals that search and AI systems recognize your brand as a well-defined entity.
Large Language Model Optimization (LLMO)
Large Language Model Optimization (LLMO) is the practice of shaping your online presence so large language models learn, remember, and correctly represent your brand when generating answers, whether from training data or live retrieval.
Non-Deterministic Answers
Non-deterministic answers refers to the fact that AI engines can return different responses to the same prompt on different runs, which is why AI visibility must be measured as rates across many prompts over time rather than from a single lookup.
Search Engine Results Page (SERP)
A SERP, or search engine results page, is the page a search engine returns for a query, traditionally a list of ranked links now increasingly topped by AI-generated features like AI Overviews that answer the query directly.
Topical Authority
Topical authority is the depth and breadth of credible coverage a brand or site has on a subject area, signaling to search and AI systems that it is a go-to source, which makes it more likely to be mentioned and cited for that topic.
Zero-Click Search
Zero-click search is any search where the user gets their answer directly on the results surface, such as an AI Overview or featured snippet, and never clicks through to a website, meaning the source site earns visibility but no traffic.
Technique
Entity SEO
Entity SEO is the practice of making your brand an unambiguous, well-defined entity that search and AI systems recognize, using consistent naming, structured data, and authoritative references so models reliably associate the right facts with you.
Prompt Engineering
Prompt engineering is the practice of crafting the inputs given to an AI model to elicit the most accurate and useful outputs, and in AI visibility it underpins designing the buyer-intent prompts used to test how engines represent a brand.
Prompt Tracking
Prompt tracking is the practice of repeatedly running a fixed set of buyer-intent questions through AI engines and recording how your brand appears each time, the AI-era equivalent of keyword rank tracking.
See how AI engines rank you today
Get a free, personalized AI visibility report for your brand — see where ChatGPT, Perplexity and Google AI mention you, and exactly what to fix.
Free report · no credit card · from just $19/mo when you upgrade.