James Dooley is joined by Sergey Lucktinov to break down how query fan-out works inside large language models and how SEO strategies must adapt. The discussion explains why optimisation has shifted away from individual keywords towards intent, entities, and topic coverage. Sergey outlines how LLMs expand searches into multiple background queries, why complexity increases fan-out, and how different models handle it. The conversation also covers semantic SEO, attribute selection, content depth, and the balance between being comprehensive and staying focused. This episode is essential viewing for anyone looking to optimise content for ChatGPT, Gemini, Perplexity, and other AI-driven search experiences.