CitationLab
Resources

Glossary for AI visibility and AI search

The AI visibility industry has its own vocabulary. Here you will find short, precise explanations of all important terms.

120+
terms
Ongoing
Updated
Norwegian + English
coverage

All terms explained

Short and precise definitions you can use right away.

AEO

Answer Engine Optimization

Optimizing content to be cited in AI-generated answers

Read more

GEO

Generative Engine Optimization

Strategies for visibility in generative AI systems

LLMO

Large Language Model Optimization

Adapting content and data for LLM visibility

AI Visibility Score

Overall score for visibility across AI engines

Citation

Reference to a specific URL in an AI-generated answer

Share of Voice

Your share of AI answers compared to competitors

AI-hallusinasjon

AI Hallucination

When AI generates incorrect or fabricated information presented as fact

RAG

Retrieval-Augmented Generation

AI technique that retrieves external source data before answering

Chunk

Text segment that AI uses as context window when generating answers

E-E-A-T

Experience, Expertise, Authoritativeness, Trustworthiness

Google's quality signals for content and publisher trust

Prompt

Question or instruction sent to an AI engine

LLM

Large Language Model

Large language model like ChatGPT, Gemini, Claude

AI-synlighet

AI Visibility

The degree to which your brand is mentioned, recommended and cited in AI-powered search tools

Read more

Entitetsautoritet

Entity Authority

The weight your brand carries in AI models, built through structured data and credible sources

Share of Model

SoM

Brand's share of AI recommendations compared to competitors — the AI equivalent of market share

AI-sitering

AI Citation

When an AI model uses your content as a source and links to it in its answer

AI-omtale

AI Mention

When an AI model mentions your brand without linking to your website

Parametrisk hukommelse

Parametric Memory

Knowledge baked into a language model's weights during training, as opposed to real-time retrieval

AI Overviews

Google's AI-generated summaries displayed above traditional search results

AI-referert trafikk

AI Referral Traffic

Visitors who come to your website via links in AI-generated answers

Knowledge Graph

Structured database of entities and relationships that AI uses to understand the world

Chunk-optimalisering

Structuring content into segments that AI can extract and cite effectively

Entitet

Entity

A uniquely identifiable thing (person, company, concept) that AI can recognize and discuss

AI-crawler

Automated program that fetches content from websites for use in AI models

Siteringsrate

Citation Rate

The share of AI answers where your content is used as a source — key metric for AI visibility

Sentimentanalyse

Sentiment Analysis

Measuring whether AI mentions are positive, negative or neutral

Schema.org

Standardized vocabulary for structured data that helps AI understand your content

Null-klikk-søk

Zero-Click Search

Search where the user gets the answer directly without clicking to a website

Konversasjonelt AI-søk

Conversational AI Search

Multi-step search where the user has a conversation with AI, as opposed to single queries

CAVIS

Conversational AI Visibility Simulation

CitationLab's framework for measuring visibility in multi-step AI conversations

Read more

Fact Density

The density of verifiable facts in your content — higher fact density increases the chance of AI citation

Perception Match

Degree of alignment between desired brand position and how AI actually describes you

Prompt Engineering

The technique of formulating questions to AI for optimal answers — important for testing AI visibility

Treningsdata

Training Data

Data used to train an AI model — your content must exist here for parametric visibility

Kildemangfold

Source Diversity

Spread of brand mentions across various credible sources — strengthens AI's trust in you

Fanalyzer

Fan-out Analysis

Analysis tool in CitationLab for discovering topic expansions and content opportunities

Chunkalyzer

Analysis tool that shows how AI models break down and process your content

Prompt-styring

Prompt Management

Systematic management of search queries to monitor AI visibility over time

Periodesammenligning

Period Comparison

Comparison of AI visibility metrics between two time periods

AI Mode

Google AI Mode

Google's conversational search model where the entire experience is AI-driven, not just a box at the top. Users can ask follow-up questions.

Answer-first

Content format where the direct answer comes first (40–60 words), before elaboration. The format AI models cite best.

ChatGPT

OpenAI's AI assistant and the most-used AI platform in the world. Recommends businesses and services on direct request.

Core Web Vitals

Google's three core user-experience metrics: LCP (load time), INP (responsiveness) and CLS (visual stability). Direct ranking factors.

CTR

Click Through Rate

The share of users who click a link compared to those who see it. Measured in Google Search Console.

DeepSeek

Chinese AI model launched in 2025, known for high performance at low cost. Should be included in AI visibility monitoring.

Google Business Profile

Google's free business profile in Search and Maps. One of the most important verification sources AI models use to confirm a business is real.

Indeksering

Indexing

The process where search engines store and categorize content from websites. A basic prerequisite for any AI visibility.

Søkeintensjon

Search Intent

What the user actually wants to achieve with a search. Four types: informational, navigational, commercial investigation, and transactional.

JSON-LD

The recommended format for schema markup. Placed in the head tag of a page and gives AI structured context about the content.

Lokal SEO

Local SEO

Search engine optimization for geographically specific searches. Important for AI visibility because AI uses GBP and local directories as verification sources.

Mention Rate

The share of AI responses where your brand is mentioned — without necessarily being cited as a source. Broader than Citation Rate.

Metabeskrivelse

Meta Description

A short description (140–155 characters) shown beneath the page title in search results. Affects CTR.

Organisk søk

Organic Search

Search results shown in Google without payment. The starting point for SEO and AEO strategy.

People Also Ask

PAA

A box in Google search results with related questions. Optimizing for PAA gives extra visibility and builds topical authority.

Perplexity

AI-driven search engine with over 100 million weekly searches. The second most important AI search platform after ChatGPT, with a strong recency bias.

Prompt-optimalisering

Prompt Optimization

Identifying and optimizing content for the specific questions users ask AI models. AEO's answer to keyword research.

Robots.txt

A text file that tells search engines and AI crawlers what they may fetch. Misconfiguration is a common cause of invisibility in AI search.

SEO

Search Engine Optimization

The process of improving a website's visibility in organic search results. The foundation all AEO is built on.

SERP

Search Engine Results Page

The page Google shows after a search. Contains organic links, AI Overviews, featured snippets, PAA, and knowledge panels.

Sitemap

An XML file that lists important pages on the site and helps search engines discover and index content. Should be submitted to Google Search Console.

Teknisk SEO

Technical SEO

Optimization of the site's technical infrastructure: speed, indexing, mobile, schema, and Core Web Vitals. Includes robots.txt access for AI crawlers.

Topical Authority

The degree of subject-matter authority a site has within a particular topic. One of the strongest drivers of AI citations.

Wikidata

An open, machine-readable knowledge database used by Google Knowledge Graph and many AI models as a verification source.

Wikipedia

The single most important source for AI models' knowledge about entities. An article about your company or industry is one of the strongest authority signals available.

UCP

Universal Commerce Protocol

Open standard, led by Google, that lets AI agents discover, buy and manage orders across stores with one integration. Powers Universal Cart and checkout in Google AI Mode and Gemini.

ACP

Agentic Commerce Protocol

OpenAI and Stripe's standard that powers Instant Checkout in ChatGPT, letting users buy directly in the conversation. Distinct from Google's UCP.

ChatGPT Shopping

ChatGPT's product discovery and shopping experience. Shopify and Etsy catalogs connect automatically; others apply via the ChatGPT Merchant portal and a product feed.

AI-handelsagent

AI Shopping Agent

An AI that discovers, compares and buys products on a user's behalf via standards like UCP or ACP. Requires machine-readable product data to act.

LLM SEO

Optimizing content so large language models retrieve, understand and cite it. Overlaps with GEO and LLMO; focuses on entities, chunks and authority.

Gemini

Google Gemini

Google's AI assistant and model family, built into Search (AI Mode and AI Overviews), Android and Workspace. A key surface to monitor for AI visibility.

Claude

Anthropic Claude

Anthropic's AI assistant, known for long-context reasoning and increasingly used for research and work. Cites web sources when it browses.

Grok

xAI Grok

xAI's AI assistant built into X (Twitter), with real-time access to posts. Relevant for brands with an active social presence.

Microsoft Copilot

Microsoft's AI assistant across Bing, Windows and Microsoft 365. Built on OpenAI models and surfaces cited web sources.

SGE

Search Generative Experience

Google's original name for AI answers in Search, now rolled into AI Overviews and AI Mode. Still seen in older documentation.

Bing

Microsoft's search engine and the web index behind Copilot and, in part, ChatGPT search. Bing Webmaster Tools controls how it crawls you.

Generativ AI

Generative AI

AI that creates new text, images or code from training data, rather than just classifying. The technology behind ChatGPT, Gemini and AI Overviews.

Grounding

Tying an AI model's answer to verifiable external sources instead of relying on parametric memory. Reduces hallucinations and produces citations.

Query Fan-out

When an AI search splits one question into several sub-queries, retrieves sources for each, and synthesizes one answer. Being cited means covering the sub-questions.

Embeddings

Vector Embeddings

Numeric representations of text meaning that let AI compare content by semantic similarity. The basis for retrieval in RAG and AI search.

Vektordatabase

Vector Database

A database that stores embeddings and finds the most relevant content by meaning rather than keywords. Powers retrieval in AI search and RAG.

Semantisk søk

Semantic Search

Search that matches on meaning and intent rather than exact keywords, using embeddings. The retrieval model AI engines rely on.

Hybrid søk

Hybrid Search

Combining keyword search (BM25) with semantic vector search for better retrieval. Common in modern AI search pipelines.

Reranking

A second-pass step that reorders retrieved results by relevance before the AI writes its answer. Strong content can win the rerank and get cited.

Token

The smallest unit of text an LLM processes — roughly three-quarters of a word. Models have token limits that cap how much context they can read.

Kontekstvindu

Context Window

The maximum amount of text (in tokens) an AI model can consider at once. Determines how much of your content fits into a single answer.

Inferens

Inference

The moment an AI model generates an answer from a prompt, as opposed to training. Each inference draws on parametric memory plus any retrieved sources.

Foundation Model

A large model trained on broad data that serves as the base for many applications (e.g. GPT, Gemini, Claude). Fine-tuned or prompted for specific tasks.

Fine-tuning

Further training of a foundation model on specific data to specialize its behavior. Distinct from prompting or retrieval.

RLHF

Reinforcement Learning from Human Feedback

Training technique that aligns model answers with human preferences. Shapes which sources and phrasings a model favors.

Temperature

A setting that controls how random or deterministic an AI's output is. Lower values give more consistent, repeatable answers.

Answer Engine

A tool that answers questions directly instead of returning a list of links — ChatGPT, Perplexity, AI Overviews. The target of AEO.

Kildeattribusjon

Source Attribution

How an AI engine credits and links the sources behind an answer. Strong attribution is what turns visibility into traffic.

Siteringsverdighet

Citability

How easy your content is for AI to quote — clear claims, facts, structure and authority. The core objective of GEO.

Innholdsferskhet

Content Freshness

How recent and up-to-date content is. AI engines like Perplexity weight fresh sources heavily (recency bias).

Information Gain

How much new, unique information a page adds beyond what already exists. High information gain increases the chance of being cited.

Passage Ranking

When search engines rank and surface a single relevant passage from a page rather than the whole page. Rewards well-structured, self-contained sections.

Topical Map

A structured plan covering every subtopic of a theme to build topical authority. Helps AI see you as a comprehensive source.

Digital PR

Earning brand mentions and links on authoritative third-party sites. One of the strongest off-site levers for entity authority and AI citations.

Brand Authority

The overall trust and recognition a brand has across the web. AI models weigh it when deciding whom to recommend.

Co-citation

When your brand is mentioned alongside known authorities in the same context. Signals to AI that you belong in that set.

Retrieval

The step where an AI system fetches relevant external documents before answering. The "R" in RAG and the gateway to being cited.

Prompt Injection

An attack that hides instructions in content to manipulate an AI model's behavior. A security concern for AI-facing pages and agents.

NER

Named Entity Recognition

AI technique that detects and classifies entities (people, brands, places) in text. The first step toward linking your brand to a knowledge graph.

Entity Linking

Entity Disambiguation

Connecting a mention to the correct entity in a knowledge base (e.g. the right "Apple"). Clean, consistent data helps AI link to you correctly.

Knowledge Panel

The info box Google shows about an entity, drawn from its Knowledge Graph. A strong public signal that AI recognizes your brand as a real entity.

Canonical URL

The tag that tells search engines which version of a page is the original. Prevents duplicate-content confusion for crawlers and AI.

Hreflang

Markup that tells search engines which language and region a page targets. Helps AI serve the right localized version of your content.

llms.txt

A proposed standard file that gives AI models a curated, markdown overview of your most important content. An emerging lever for AI visibility.

llms-full.txt

An extended llms.txt that includes the full text of key pages so AI can ingest them directly. Companion to llms.txt.

GPTBot

OpenAI's crawler that gathers web content for training and ChatGPT. Allowing it in robots.txt is required to be eligible for citation.

ClaudeBot

Anthropic's web crawler for Claude. Control its access in robots.txt to decide whether Claude can use your content.

PerplexityBot

Perplexity's crawler that indexes pages for its answer engine. Blocking it removes you from Perplexity's citations.

Google-Extended

A robots.txt token that controls whether Google may use your content for Gemini and AI training, separate from normal Search indexing.

Crawlbarhet

Crawlability

How easily bots and AI crawlers can access and read your pages. Blocked or JS-heavy pages are invisible to AI.

AI-agent

Agentic AI

An AI that plans and takes multi-step actions toward a goal — searching, comparing, even buying — with little human input. The next layer beyond chat.

MCP

Model Context Protocol

An open standard for connecting AI assistants to external tools and data sources. Lets agents act on live information.

System Prompt

The hidden instructions that set an AI assistant's role and rules before the user's prompt. Shapes tone, guardrails and source preferences.

Multimodal AI

AI that understands and generates across text, images, audio and video. Means your images and video can also be read and cited by AI.

Zero-shot / Few-shot

Prompting an AI with no examples (zero-shot) or a handful (few-shot) to guide its answer. Affects how reliably it represents your brand.

Want to learn more about AI visibility?

Go deeper on the blog or start a free trial of CitationLab.