GLOSSARY / THE VOCABULARY
Answer engines, defined.
22 terms behind AI search, in plain English - what each one means, how it’s measured, and where people usually get it wrong.
FOUNDATIONS
Answer Engine Optimization (AEO)
The practice of optimizing content so AI answer engines mention, cite, and describe a brand accurately.
Generative Engine Optimization (GEO)
The research term for optimizing content visibility inside generative AI outputs, coined in academic literature.
Answer engine
A system that answers user queries directly with a synthesized response instead of a list of links.
LLM SEO
An informal term for optimizing content so large language models retrieve, trust, and repeat it accurately.
Zero-click search
A search interaction where the user gets their answer without clicking through to any website.
METRICS
Share of voice (AI search)
The percentage of AI-generated answers in a topic or category that mention a given brand.
Answer position
Where in an AI-generated response a brand is mentioned, from first-named to buried in a list.
Brand mention rate
The percentage of tracked prompts in which an AI answer engine names a brand at all.
Citation share
The percentage of a brand’s own URLs that AI answer engines actually cite as sources.
Sentiment (AI answers)
Whether an AI-generated mention of a brand reads as positive, neutral, or negative in tone.
Visibility score
A composite index combining mention rate, position, and citation data into one overall AI-visibility number.
MECHANICS
Retrieval
The step where an AI system fetches candidate documents or passages before generating an answer.
Grounding
Anchoring a generated answer in retrieved or verifiable source material instead of the model’s memorized parameters.
Retrieval-augmented generation (RAG)
An architecture that retrieves external documents at query time and feeds them to a language model as context for its answer.
Prompt fan-out
One user query expanding into multiple retrieval searches or sub-queries behind the scenes before the answer is composed.
Hallucination
When an AI answer states something false or unsupported with the same confidence as a verified fact.
Model-recalled citation
A source a model names from memory when answering, not necessarily a page it retrieved and read in that session.
Entity consistency
Whether a brand’s name, category, and attributes are described the same way across different AI answers and engines.
TECHNICAL
llms.txt
A proposed root-level text file that lists a site’s key pages for AI systems to read, similar in spirit to robots.txt.
AI crawler
An automated bot that fetches web pages on behalf of an AI company, for either training data collection or live retrieval.
Schema markup for AI
Structured data (schema.org markup) that gives answer engines explicit, machine-readable facts about a page instead of requiring inference.
AI Overviews
Google’s AI-generated summary shown above traditional search results for eligible queries, citing a handful of source links.
Know the words. Now see your numbers.
Run a free scan and find out how AI answer engines describe your brand.