Direct recommendation queries evaluate whether an AI model actively endorses a brand for specific topics by prompting it with questions like "Would you recommend {brand} for someone interested in {entity}?" constrained strictly to a binary yes-or-no output. By executing these structured prompts across multiple models over repeated sample batches, the system computes a percentage-based Query Relevance Score (QRS) that pinpoints topical gaps where AI platforms consistently hesitate or refuse to recommend the business.