boltTLDR
Moz released a dataset of 50,000 fan-out prompts across 1,000 subtopics, revealing that brand mentions cluster in entity and comparison queries, and that most fan-outs balance topical relevance with diversity.
The MOZ Blog published analysis from a year-long study that generated 50,000 AI-powered fan-out prompts—query variations that expand on core topics—to understand how brands appear across different search intents. The research found that while brand mentions occur naturally in entity and comparison-type queries, they're heavily influenced by the types of prompts tracked, and that an alignment score of 0.67 (on a 0-1 scale) represents the sweet spot between topical relevance and useful diversity. Moz is sharing the raw dataset and metrics freely, aiming to advance the conversation around query fan-out's impact on keyword research in the era of generative AI.