GEO Glossary
Key terms and concepts in Generative Engine Optimization and AI visibility
RAG (Retrieval-Augmented Generation)
RAG is the architecture that powers ChatGPT, Claude, and Perplexity. When an AI generates a response, it retrieves relevant web sources from its training data and live web crawl, augments its response with citations from those sources, and generates a synthesized answer that appears authoritative.
GEO (Generative Engine Optimization)
GEO is the practice of optimizing content to rank in AI-powered search engines like ChatGPT, Claude, and Perplexity. Unlike traditional SEO which optimizes for Google's ranking algorithm, GEO focuses on ensuring your content is cited as a source when AI models generate responses to user queries.
N-Sampling Methodology
N-Sampling is a consistency scoring method that queries an AI model multiple times (typically 10x) for the same question to eliminate flukes and identify true ranking patterns. Single-query monitoring can miss inconsistencies that N-Sampling reveals.
Citation Decay (Source Authority Decay)
Citation Decay occurs when an AI cites outdated or stale competitor data. We identify these moments and flag them as strategic opportunities for your brand to replace outdated citations with fresh, authoritative content.
Citation Path Analysis
Citation Path Analysis is the forensic process of identifying the specific web sources used by an LLM to generate a response. It traces the exact citation chain from AI output back to source material, revealing why certain brands rank and others don't.
Share of Model (SoM)
Share of Model measures your brand's visibility across AI search engines. It calculates the percentage of relevant queries where your brand appears in AI responses, similar to Share of Voice in traditional marketing.
Brand Constitution
Brand Constitution is a verification protocol that prevents AI hallucinations by ensuring LLMs have access to accurate, up-to-date information about your brand. It acts as a "source of truth" that AI models can reference.
Answer-First Content
Answer-First Content is a content format that directly answers a question in the first paragraph. This format wins Google Featured Snippets and AI citations because it provides immediate value and is easily extractable by LLMs.
Key Takeaways
- •RAG architecture means AI responses are only as good as their source material
- •GEO optimizes for AI-powered search engines, not just traditional Google rankings
- •N-Sampling methodology eliminates flukes that single-query monitoring misses
- •Citation Decay creates opportunities to replace stale competitor citations