Relevance evaluates direct model recommendations by executing binary yes-or-no probes across active LLM models, prompting the AI with targeted questions on whether it would recommend a brand for a particular entity or query. Repeated across multiple sampled trials, these evaluations aggregate into a Query Relevance Score that reflects the model's consistency and confidence in connecting the brand to commercial intent. This metric isolates categorical weaknesses by surfacing specific topics where competing brands dominate, highlighting strategic areas where content adjustments can improve direct AI endorsements.