Relevance scoring measures how consistently AI models recommend a brand for specific entities or search queries by executing repeated binary classification prompts (asking "Would you recommend {brand} for someone interested in {entity}?") across multiple platforms. By probing models like Google Gemini, OpenAI GPT, and Anthropic Claude across multi-sample runs, the system calculates a Query Relevance Score (QRS) percentage based on affirmative responses, enabling granular relevance-gap analysis for targeted optimization.