Glossaire AEO / GEO / LLMO
Ce glossaire rassemble les termes essentiels pour comprendre et maîtriser la visibilité sur les IA conversationnelles. Que vous découvriez l'AEO ou que vous cherchiez à approfondir un concept technique, chaque définition est conçue pour être claire, actionnable et directement utilisable.
A
AEO (Answer Engine Optimization)
AEO (Answer Engine Optimization) is a strategy for optimizing content to be cited directly by AI assistants and voice search. Unlike SEO which targets search result rankings, AEO aims to position your content as the source of answers provided by ChatGPT, Claude, Gemini, and other conversational AIs.
Voir la définition StratégieAEO Audit (Answer Engine Optimization)
An AEO audit analyzes your brand's visibility in AI answer engine responses (ChatGPT, Claude, Perplexity, Gemini). It identifies if you're cited, how you're presented, and improvement opportunities.
Voir la définition RéglementationAI Act (EU Regulation)
The AI Act is the European regulation on artificial intelligence, adopted in 2024. It classifies AI systems by risk and imposes obligations on developers and deployers of AI.
Voir la définition MétriqueAI Citation
An AI citation is when an AI assistant references your brand, website, or content as a source. Citations are the ultimate AEO goal - being explicitly mentioned and recommended by AI.
Voir la définition RéglementationAI Compliance
AI compliance refers to adhering to regulations governing artificial intelligence use, including the AI Act, GDPR for data, and industry-specific rules. Essential for businesses using or being referenced by AI.
Voir la définition TechniqueAI Indexing
AI indexing is the process by which your content becomes accessible to AI systems - either through training data, RAG databases, or real-time web access. Different from traditional search indexing.
Voir la définition StratégieAI Optimization
AI optimization encompasses all strategies aimed at improving a brand's visibility in artificial intelligence responses. It includes AEO, GEO, and LLMO.
Voir la définition PlateformeAI Overviews (Google SGE)
AI Overviews are Google's AI-generated summaries displayed at the top of search results. Part of the Search Generative Experience (SGE), they synthesize information from multiple sources.
Voir la définition StratégieAI SEO Audit
An AI SEO audit combines traditional SEO analysis with AI visibility evaluation. It measures your presence on Google AND on ChatGPT, Claude, Perplexity, Gemini.
Voir la définition MétriqueAI Share of Voice
AI Share of Voice measures how often your brand is mentioned by AI assistants compared to competitors. A key AEO metric for tracking visibility across ChatGPT, Claude, Perplexity, and Gemini.
Voir la définition ContenuAnswer Capsule
An Answer Capsule is a concise, standalone text block designed to be directly extracted by AI assistants. It provides a complete answer in 2-3 sentences, optimized for AI citation.
Voir la définition ConceptAnswer Engine
An Answer Engine is a search system providing direct answers rather than links. Perplexity, ChatGPT, and Google AI Overviews are answer engines. AEO specifically optimizes for these systems.
Voir la définitionB
Backlink
A backlink is a hyperlink from another website pointing to your site. Backlinks are a key ranking factor for SEO and influence how AIs perceive your content's authority.
Voir la définition OutilBing Webmaster Tools
Bing Webmaster Tools is Microsoft's platform for managing your site's presence in Bing search. Important for AEO as Bing powers Microsoft Copilot's web search capabilities.
Voir la définitionC
ChatGPT
ChatGPT is OpenAI's conversational AI assistant, powered by the GPT-4 model. Launched in November 2022, it revolutionized public access to AI and popularized the use of conversational AI assistants.
Voir la définition PlateformeChatGPT Audit
A ChatGPT audit specifically analyzes how your brand appears in ChatGPT (OpenAI) responses. It evaluates mentions, sentiment, information accuracy, and competitive positioning.
Voir la définition PlateformeClaude (Anthropic)
Claude is the conversational AI developed by Anthropic, founded by former OpenAI researchers. Claude stands out for its large context window (200K tokens), safety focus, and nuanced response quality.
Voir la définition PlateformeClaude Audit (Anthropic)
A Claude audit analyzes your visibility on Anthropic's AI assistant. Claude stands out for its extended context window (200K tokens) and safety-focused approach, attracting a professional audience.
Voir la définition TechniqueCrawl (Web Indexing)
Crawling is the automated process where bots explore the web to discover and index content. For AEO, being crawlable by AI data sources (Common Crawl, etc.) is essential to be included in training data.
Voir la définitionD
Deepfake
A deepfake is AI-generated synthetic media (video, audio, image) that realistically depicts someone saying or doing something they never did. Deepfakes pose brand reputation risks.
Voir la définition SEODomain Authority
Domain Authority is a metric (usually 0-100) predicting how well a website will rank in search results. In the AI era, domain authority influences which sources AIs consider trustworthy.
Voir la définitionE
E-E-A-T (Google)
E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) is Google's framework for evaluating content quality. In the AI era, E-E-A-T signals influence how LLMs perceive and cite your content.
Voir la définition TechniqueEmbedding
An embedding is a numerical representation of text in a multidimensional vector. Embeddings allow AIs to measure semantic similarity between texts and are fundamental to RAG search.
Voir la définitionF
Featured Snippet
A Featured Snippet is an enhanced result displayed at the top of Google results (position 0), directly answering the user's query. Featured Snippets are precursors to AI answers and share AEO optimization strategies.
Voir la définition TechniqueFine-tuning
Fine-tuning is additional training of an LLM on specialized data to adapt it for a specific task or domain. Fine-tuning allows customizing a model's behavior without training from scratch.
Voir la définitionG
GEO (Generative Engine Optimization)
GEO (Generative Engine Optimization) is the optimization of content specifically for generative AI systems that synthesize information to produce unique responses. GEO focuses on how AIs understand, process, and reformulate your content in their answers.
Voir la définition StratégieGEO Audit (Generative Engine Optimization)
A GEO audit evaluates your visibility specifically on generative search engines that synthesize responses (Google AI Overview, Perplexity). It analyzes how your content is used to generate answers.
Voir la définition RéglementationGPAI (General Purpose AI)
GPAI (General Purpose AI) is a regulatory term from the EU AI Act for AI systems capable of performing various tasks (like GPT-4, Claude). GPAI providers have specific compliance obligations.
Voir la définition PlateformeGemini (Google)
Gemini is Google's multimodal AI model, integrated into Google Search, Workspace, and as a standalone assistant. Gemini represents Google's response to ChatGPT and directly impacts search results.
Voir la définition PlateformeGemini Audit (Google)
A Gemini audit analyzes your visibility on Google's multimodal AI, integrated into Google Search via AI Overviews. Gemini directly impacts your visibility in the Google ecosystem.
Voir la définitionL
LLM (Large Language Model)
An LLM (Large Language Model) is an AI model trained on massive amounts of text to understand and generate human language. Examples include GPT-4, Claude, Gemini, Llama. LLMs power modern AI assistants.
Voir la définition StratégieLLMO (Large Language Model Optimization)
LLMO (Large Language Model Optimization) is a technical approach to optimize content for large language models (LLMs) like GPT-4, Claude, or Llama. LLMO focuses on how LLMs process text during training and inference.
Voir la définitionM
Machine Learning
Machine Learning is a branch of AI where systems learn from data to improve performance without being explicitly programmed.
Voir la définition ia-techniqueMultimodal (Multimodal AI)
A multimodal AI can process and generate multiple data types: text, images, audio, video. Models like GPT-4V, Gemini, and Claude 3 are multimodal.
Voir la définitionP
Perplexity AI
Perplexity AI is an AI-powered search engine that combines real-time web search with LLM response generation. Unlike ChatGPT, Perplexity always cites its sources, making it a key platform for AEO.
Voir la définition PlateformePerplexity Audit
A Perplexity audit analyzes your visibility on the AI search engine that systematically cites its sources. Perplexity generates direct traffic as users click on cited sources.
Voir la définition SEOPosition Zero
Position Zero refers to the Featured Snippet placement above organic search results. In the AI era, it represents direct answer placements that bypass traditional rankings.
Voir la définition ConceptPrompt
A prompt is the text instruction or question given to an AI to get a response. Prompt quality directly influences response quality. In AEO, understanding prompts helps optimize content for how users query AIs.
Voir la définitionS
Schema Markup
Schema markup is structured data code added to web pages helping search engines and AIs understand content. Using Schema.org vocabulary improves how AI interprets your information.
Voir la définition RéglementationSystemic Risk (AI)
Systemic risk in AI refers to risks that could affect society at large - misinformation, manipulation, discrimination at scale. The AI Act has specific provisions for high-risk AI systems.
Voir la définitionT
Temperature (AI Parameter)
Temperature is a parameter controlling LLM response randomness. A low temperature (0-0.3) produces deterministic, factual responses. A high temperature (0.7-1) generates more creative, varied responses.
Voir la définition TechniqueToken
A token is the basic text unit processed by LLMs. A token can be a word, part of a word, or punctuation. Context window limits are measured in tokens, typically around 4 characters per token in English.
Voir la définition TechniqueTraining Data
Training data is the massive dataset used to teach LLMs language understanding and generation. Includes web pages, books, articles that shape AI knowledge and biases.
Voir la définitionAuditez votre visibilité IA
Découvrez comment votre marque apparaît sur ChatGPT, Claude, Gemini et Perplexity.
Commencer gratuitement