Get your AI audit

Glossary

GEO terms, in plain words.

The words used in generative engine optimisation and AI search, defined briefly and neutrally.

The short answer

This GEO glossary defines the terms used in generative engine optimisation and AI search: how engines such as ChatGPT, Gemini and Google AI Overviews build answers, how brands get cited in them, and how visibility is measured. Each definition is short and neutral, so you can use it in a brief, a board paper or a conversation with an agency.

AI search has brought a new vocabulary with it. Some terms come from search, some from machine learning and some from how agencies like ours measure the work. Where a term describes how we measure, we say so.

Generative engine optimisation (GEO)
Generative engine optimisation is the practice of making a brand easier for AI engines to find, trust, name and cite in their answers. It covers the pages a brand publishes, the sources engines rely on and how accurately the brand is described. The US spelling is generative engine optimization.
Answer engine optimisation (AEO)
Answer engine optimisation is a closely related term for structuring content so that AI search features and voice assistants can lift a direct answer from it. The terms are often used interchangeably; AEO tends to focus on the format of answers, GEO on being named and cited.
LLM (large language model)
A large language model is an AI model trained on large amounts of text to understand and generate language. LLMs power chat assistants such as ChatGPT, Gemini and Claude, often combined with live web search.
AI Overviews
AI Overviews are AI-generated summaries that Google shows above some search results, with links to the sources used. They are one of the 8 engines we track.
AI Mode
AI Mode is Google's conversational search experience, in which a user asks questions and follow-ups and receives an AI-generated answer with links instead of a standard list of results.
Retrieval-augmented generation (RAG)
Retrieval-augmented generation is a method in which an AI system first retrieves relevant documents, for example through a web search, and then writes its answer from them. It is why current, readable pages can change what an engine says.
Citation
A citation is a source an AI engine links to or names as the basis for its answer. Being cited is different from being mentioned: a brand can be named without its own pages being the source.
Citation rate
Citation rate is the share of AI answers that cite a brand's own pages as a source. It shows whether engines rely on what the brand publishes or on third parties.
Visibility score
Visibility score is the share of tracked questions, across all the engines measured, where the brand appears in the answer. It is the headline measure of AI visibility in our reporting.
Share of voice
Share of voice is a brand's share of all brand mentions in the answers for its category, measured against the rivals the engines name.
Brand sentiment
Brand sentiment is the tone in which AI engines describe a brand, scored on every stored answer. It shows whether the brand is described positively, neutrally or with caveats.
Prompt
A prompt is the question or instruction a person types into an AI engine. In GEO tracking, a prompt is one buyer question run on each engine and recorded with its answer and date.
Prompt universe
A prompt universe is the full set of questions a brand's buyers are likely to ask AI engines. Ours is built from SEO data, the brand's own enquiries, a category map and a live engine baseline, then ranked by demand and winnability.
Question cluster
A question cluster is a group of closely related prompts that share one intent, such as questions comparing two neighbourhoods or two products. Work is usually planned and measured cluster by cluster.
Entity
An entity is a distinct thing, such as a company, person, product or place, that search engines and AI systems recognise and connect to facts. Clear, consistent entity data helps engines describe a brand correctly.
Knowledge graph
A knowledge graph is a database of entities and the relationships between them, used by search engines and AI systems to answer factual questions. Google's Knowledge Graph is the best-known example.
Wikidata
Wikidata is a free, collaboratively edited knowledge base run by the Wikimedia Foundation that stores structured facts about entities. Search engines and AI systems use it as a reference source.
Structured data (schema markup)
Structured data, or schema markup, is code added to a web page, usually using the Schema.org vocabulary, that labels facts such as an organisation's name, a product's details or a page's questions and answers. It helps machines read the page accurately.
llms.txt
llms.txt is a proposed convention: a plain-text Markdown file at a website's root that points AI systems to the site's most useful pages. It is not an official standard, and support among engines varies.
E-E-A-T
E-E-A-T stands for experience, expertise, authoritativeness and trustworthiness. The term comes from Google's search quality rater guidelines and describes what Google looks for in reliable content.
Drift
Drift is when what AI engines say about a brand changes or falls out of date, often after a launch, a price change or new results. Drift alerts flag these changes so wrong facts are caught early.
Week-1 baseline
The Week-1 baseline is the first full measurement of a brand's AI visibility, taken before any fixes ship. Every later result is reported against it.
Readiness scorecard
A readiness scorecard rates how prepared a brand is to be named in AI answers. Ours scores 8 factors out of 10 each, for a total out of 80: brand recognition, trust and credibility, entity accuracy, technical readiness, share of answer, location authority, discovery and comparisons.
Feeder network
A feeder network is the set of sites and channels that send buyers and AI engines towards a brand. It includes owned channels, such as a blog, guide site, video channel, newsletter or WhatsApp channel, and earned ones, such as creators, 'best X in Y' lists and newsletters.
Owned vs rented growth
Rented growth comes from ads, sponsored posts and paid placements, and stops when the budget stops. Owned growth comes from assets a brand controls, such as its pages, reviews, community and channels, which keep working and compound.
Digital PR
Digital PR is earning coverage, mentions and links in online media, lists and publications. In GEO it matters because AI engines often draw on trusted media when they describe a brand.
Hallucination
A hallucination is a statement an AI model produces that sounds plausible but is false or unsupported, such as a wrong price, date or figure. Accurate, current, well-sourced pages leave fewer gaps for hallucinations to fill.

Questions

What is the difference between GEO and AEO?

They overlap and are often used interchangeably. AEO usually means formatting content so engines can lift a direct answer; GEO is broader, covering whether a brand is named, cited and described accurately in AI answers.

What is the difference between a mention and a citation in AI answers?

A mention is when the engine names the brand in its answer. A citation is when it links to or names a source as the basis for the answer. A brand can be mentioned without its own pages being cited.

Does llms.txt help a brand appear in AI answers?

llms.txt is a proposed convention, not an official standard, and support among engines varies. It can do no harm, but clear, accurate, well-structured pages and trusted third-party sources matter far more.

Own your growth

See what AI says about you.

One Week-1 measurement across eight AI engines: who they name, who they name instead of you, and the sources they trust.