Running a query multiple times can reveal much more than a single request. Daniel Hinckley emphasizes this point by illustrating how he repeated the query "AI SEO Agencies" through Gemini 3 Flash, a model designed to mimic Google’s AI capabilities. After 124 iterations, he identified 124 unique query fan-outs and categorized them into 25 groups. The most common expansions included agency services (43 occurrences), reviews and rankings (38), and case studies (11). This approach highlights that the AI model seeks to provide comprehensive answers, focusing not just on identifying the best agencies but also on trust, proof, differentiation, and technology. Hinckley notes that many SEO teams optimize only for the main query, overlooking the broader context the AI considers important. By understanding the fan-outs, teams can identify gaps in their content and positioning. The AI’s aggressive fan-out behavior suggests a more exploratory nature in Gemini 3 Flash compared to earlier models, stressing the need for marketers to adapt their strategies. He proposes that presenting these query fan-outs on various platforms may enhance credibility and reach, as the AI often validates information across multiple sources. The conversation also touches on the differences in user intent, particularly between those looking for agencies and those researching on behalf of agencies. This distinction is crucial, as the AI might not fully grasp these nuances. The discussion concludes with an acknowledgment of the importance of aligning on-page claims with off-page evidence to ensure that AI models will recommend a brand. Mapping fan-outs is just the first step in a more complex SEO strategy that requires thorough alignment and validation.