Branded vs. unbranded queries in AI search — and how to track them

Branded queries in AI search occur when users specifically name a brand or product, indicating existing awareness and intent to research or compare. These queries help measure how well a brand’s messaging or facts resonate with audiences already familiar with it. For example, a prompt like 'Are Nikes good shoes for marathon runners?' directly involves the brand Nike, showing where the company stands in AI-generated answers for its own name or products.

Unbranded queries describe categories, problems, or use cases without naming any brand, such as 'What’s the best shoe for marathon runners?' These queries are critical for reaching new audiences who haven’t yet decided on a brand. AI engines must decide whether to include a brand in answers to unbranded prompts, making visibility here a measure of discovery and potential growth.

Google’s Search Console introduced a branded queries filter in November 2025 to separate branded and unbranded search terms, highlighting the need for similar tracking in AI search. Unlike traditional search, AI search uses longer, conversational prompts, requiring precise categorization to avoid misleading results. Mixing branded and unbranded prompts can skew visibility metrics, leading to misinformed strategies.

The distinction between branded and unbranded prompts is essential because they measure different aspects of answer engine optimization (AEO). Branded prompts assess messaging accuracy, while unbranded prompts reflect how often a brand is discussed across the web. Combining these scores can overstate new audience discovery, making separate tracking necessary for accurate insights and actionable adjustments.