The AEO Glossary

The vocabulary behind answer engines and AI visibility — 20 terms, defined in plain English and cross-linked.

AEO (Answer Engine Optimization): The practice of improving how AI-powered answer engines — ChatGPT, Perplexity, Gemini, Claude, Copilot, Google AI Overviews — represent and recommend a brand. Where SEO optimizes pages to rank in a list of links, AEO optimizes entities and evidence so the engine names you inside the answer itself. It spans content structure, structured data, citations, and continuous measurement.; See also:AI Visibility Share of Voice Structured Data
AI Visibility: How often, how prominently, and how favorably a brand appears in AI-generated answers for the prompts that matter to it. AI visibility is measured across engines and prompts using metrics like mention frequency, share of voice, sentiment, and citation count — the AEO equivalent of rankings and impressions.; See also:Mention Frequency Share of Voice Sentiment Analysis
AI Overview: Google's AI-generated summary shown above traditional results for many queries. AI Overviews synthesize multiple sources into a direct answer with citations, which means a brand can be visible (or invisible) before a user ever sees a blue link. They are a major driver of zero-click search behavior.; See also:Zero-click Search Citation
Brand Mention: Any instance of an AI engine naming a brand, product, or domain in an answer — recommended, compared, or merely referenced. Mentions are the atomic unit of AI visibility measurement: they are counted, positioned (first vs. buried), and scored for sentiment and accuracy.; See also:Mention Frequency Sentiment Analysis
Citation: A source link an answer engine attaches to a claim in its response. Engines like Perplexity and Google AI Overviews cite pages they retrieved; being cited both drives referral traffic and signals that the engine treats your page as evidence. Earning citations from pages engines already trust is a core AEO tactic.; See also:RAG AI Overview
Entity Recognition: How language models identify and disambiguate real-world things — companies, products, people, places — in text. If your brand name is used inconsistently across the web, engines may fail to connect your product to the category it belongs to. Consistent naming and structured data strengthen entity recognition.; See also:Named Entity Knowledge Graph Structured Data
Hallucination: A confident but false statement generated by an AI model — a wrong price, a discontinued feature described as current, or a fabricated review. For brands, hallucinations are a reputational risk that must be monitored: correcting the source content and strengthening authoritative signals reduces their frequency.; See also:Training Data RAG
Knowledge Graph: A structured database of entities and the relationships between them, used by search and answer engines to ground responses in facts. Google's Knowledge Graph and Wikidata are the best known. Getting your organization correctly represented in public knowledge graphs anchors how engines describe you.; See also:Entity Recognition Structured Data
LLM (Large Language Model): The neural network underneath every answer engine — GPT-4/5 class models behind ChatGPT and Copilot, Gemini behind Google, Claude behind Anthropic products. LLMs generate answers from patterns learned in training data, optionally augmented with live retrieval. Understanding what an LLM knows about your brand is the starting point of AEO.; See also:Training Data RAG Prompt
Mention Frequency: The percentage of relevant prompts for which an engine mentions your brand, tracked over time and across engines. A presence rate of 40% means four in ten answers to your target prompts include you. It is the bluntest and most fundamental AI visibility metric — you cannot win an answer you do not appear in.; See also:Brand Mention AI Visibility
Named Entity: A specific, uniquely identifiable thing referenced in text — “AEOquest”, “Paris”, “iPhone 17”. Answer engines resolve named entities to records in knowledge graphs and training data. Brands are named entities competing to be retrieved when a category question is asked.; See also:Entity Recognition Knowledge Graph
Prompt: The question or instruction a user gives an AI engine — the AEO analogue of a search keyword. “Best CRM for a five-person startup” is a prompt with commercial intent. AEO programs track a portfolio of prompts across intents (informational, comparison, transactional) and measure who each engine names in response.; See also:Mention Frequency LLM
RAG (Retrieval-Augmented Generation): An architecture where the model retrieves live documents (web pages, indexes) and generates its answer from them, rather than relying on training data alone. Perplexity, Copilot, and AI Overviews are retrieval-heavy: for these engines, being retrievable and quotable right now matters as much as historical authority.; See also:Citation LLM Vector Search
Sentiment Analysis: Classifying whether an AI answer speaks about a brand positively, neutrally, or negatively. A brand can be highly visible and consistently framed as the expensive or legacy option — sentiment turns raw mention counts into a picture of how engines actually characterize you.; See also:Brand Mention AI Visibility
Structured Data: Machine-readable annotations embedded in web pages that state facts explicitly: this is a Product, its price is X, this page is an FAQ. Structured data removes ambiguity for crawlers and retrieval systems, making content easier to extract, attribute, and cite correctly.; See also:Schema.org Entity Recognition
Schema.org: The shared vocabulary for structured data, maintained by Google, Microsoft, and others. Types like Organization, Product, FAQPage, Article, and DefinedTerm let a site describe itself in terms every major engine understands. Schema.org markup in JSON-LD format is the standard implementation.; See also:Structured Data
Share of Voice (SOV): Your brand’s slice of all brand mentions across answers to a prompt set, relative to competitors. If engines mention some brand 100 times across your tracked prompts and 25 of those are you, your SOV is 25%. It reframes visibility as a competitive, zero-sum metric — the one executives usually care about most.; See also:AI Visibility Mention Frequency
Training Data: The text corpus a model learned from — web crawls, books, code, reference works — with a fixed cutoff date. What the corpus said about your brand shapes the model’s default beliefs; content published consistently and widely before the cutoff is what pure-LLM answers draw on when no retrieval happens.; See also:LLM Hallucination
Vector Search: Finding documents by semantic similarity: text is converted into numerical embeddings, and the engine retrieves passages whose meaning is closest to the prompt — regardless of exact keywords. Retrieval pipelines behind answer engines lean on vector search, which rewards clear, self-contained passages over keyword stuffing.; See also:RAG
Zero-click Search: A search that ends without any click to a website because the answer appeared directly on the results (or chat) surface. AI Overviews and chat assistants push more journeys toward zero-click, which is why brands measure presence inside answers instead of only counting sessions.; See also:AI Overview

What is AEO? The Complete Guide to AI Engine OptimizationThe pillar guide: definitions, the AEO vs SEO shift, and a 5-step framework.How to Improve Your Brand's AI VisibilityTen practical tactics, engine-specific tips, and how to measure ROI.

You might also like