Glossary
Key terms in AI visibility, GEO (Generative Engine Optimization), and LLM-powered search — defined and maintained by the AI Visibility Barometer.
20 terms · Last updated Q2 2026
GEO — Generative Engine Optimization
The practice of optimising a company's content and digital presence to be cited by LLM-powered search engines (ChatGPT, Perplexity, Gemini, Claude). GEO is distinct from SEO: the target is a citation in a generated answer, not a position in a ranked list of links.
# geoAEO — Answer Engine Optimization
A broader term for optimising content so that AI-powered answer engines surface it in direct responses. Often used interchangeably with GEO. AEO emphasises the shift from search engines (returning links) to answer engines (returning synthesised answers).
# aeoAI Visibility
The degree to which a company, brand, or product is cited, mentioned, or recommended by LLM-powered search engines when users ask relevant category queries. High AI visibility means the company appears consistently across multiple LLMs and prompts.
# ai-visibilityAI Visibility Score
A composite metric (0–100) measuring AI visibility. In the AI Visibility Barometer, it is calculated as: citation presence rate (50%) + average rank score (30%) + share of voice (20%). The score is reproducible and methodology-first.
# ai-visibility-scoreLLM Citation
An instance where a large language model names, recommends, or references a specific company, product, or source in a generated response. LLM citations drive direct brand discovery — users receive a recommendation without clicking a search result.
# llm-citationCitation Presence Rate
The share of (LLM × prompt) combinations in which a company is cited at least once. A citation presence rate of 60% means the company appears in 60% of tested LLM-prompt pairs. Used as the primary weight (50%) in the AI Visibility Score.
# citation-presence-rateRAG — Retrieval-Augmented Generation
An LLM architecture where the model retrieves external documents or web content at query time before generating a response. RAG-based systems (including Perplexity and ChatGPT with browsing) are more sensitive to real-time content signals than pure generative models.
# ragAI Overviews (Google)
Google's LLM-generated summaries displayed at the top of search results pages, replacing or supplementing traditional blue links for many queries. AI Overviews draw from indexed web content and are a distinct channel from conversational LLMs — but share GEO principles.
# ai-overviewsLLM — Large Language Model
A neural network trained on large text datasets to generate and understand natural language. Examples: GPT-4o (OpenAI), Gemini (Google), Claude (Anthropic). LLMs power both conversational AI tools and AI-powered search engines.
# llmZero-click Search
A search interaction where the user receives their answer directly in the search interface — without clicking through to a website. LLM-powered search dramatically increases zero-click rates, making AI visibility (rather than web traffic) the key commercial metric.
# zero-clickStructured Data / Schema.org
Machine-readable metadata embedded in web pages using vocabulary from schema.org. Key schemas for AI visibility include FAQPage, DefinedTermSet, Dataset, Article, and Organization. Structured data helps LLMs identify, interpret, and cite content accurately.
# structured-datallms.txt
A plain-text file placed at the root of a website (e.g. /llms.txt) listing the site's pages with short descriptions, intended for LLM crawlers. Analogous to robots.txt for AI bots. Adopted as an emerging standard by GEO practitioners.
# llms-txtFAQPage Schema
A schema.org markup type that structures question-and-answer content in a machine-readable format. FAQPage schema is one of the strongest GEO signals: LLMs frequently extract and cite Q&A content marked up with this schema in their generated responses.
# faqpage-schemaTopical Authority
The degree to which a website or publisher is recognised by LLMs and search engines as a credible, comprehensive source on a given topic. High topical authority increases the probability of LLM citation. It is built through consistent, expert, well-sourced content on a focused subject area.
# topical-authorityConversational Search
A search paradigm where users interact with a search engine using natural language questions rather than keyword strings. LLM-powered tools (ChatGPT, Perplexity, Gemini) have accelerated the shift to conversational search, fundamentally changing how brands are discovered.
# conversational-searchPrompt-based Search
The use of natural language prompts — rather than keyword queries — to retrieve information from LLM-powered search engines. GEO practitioners optimise content to match the intent and language patterns of typical prompts in their category.
# prompt-engineering-searchEntity Recognition (LLM)
The ability of an LLM to identify and consistently reference a company, person, or product as a distinct entity. Strong entity recognition — built through consistent naming, structured data, and cross-source corroboration — increases citation frequency and accuracy.
# entity-recognitionCitation Rank
The position at which a company is cited within an LLM response. Position 1 (first mention) carries the highest visibility and recall. The AI Visibility Barometer uses average citation rank as a 30% weight in the AI Visibility Score.
# citation-rankGEO vs SEO
SEO (Search Engine Optimisation) targets ranked positions in traditional search results pages (Google, Bing). GEO (Generative Engine Optimisation) targets citations in LLM-generated answers. The two disciplines share technical foundations (structured data, content quality) but differ in measurement, distribution, and conversion mechanics.
# geoai-vs-seo