Ahrefs analyzed over 1 billion data points in three months to understand AI search optimization. The study reveals that YouTube mentions are the strongest predictor of AI visibility, with a correlation of 0.737, surpassing traditional SEO factors like Domain Rating and backlinks. This is significant because both Google and OpenAI rely heavily on YouTube content for training their AI models. For queries processed by AI, there's an 86% agreement in conclusions between AI Mode and AI Overviews, but they cite different sources. AI Mode responses are notably longer and more detailed. Content length has little impact on AI citations, with a correlation of just 0.04. Interestingly, over half of the citations in AI Overviews come from pages under 1,000 words. Google still drives 345 times more traffic than AI platforms like ChatGPT and Gemini, though ChatGPT is responsible for over 80% of AI-related site traffic. Freshness of content is critical; 79% of blog lists cited by ChatGPT were updated this year. Meanwhile, Wikipedia, homepages, and educational content dominate ChatGPT's top citations, making many valuable sources inaccessible to marketers. The findings suggest that marketers need to adapt their content strategies. Emphasizing recent updates and engaging with short-form content could yield better visibility in AI-driven results. As the study shows, the nature of the queries—especially in the B2B space—affects the types of sources that gain traction. The landscape is shifting, moving towards a focus on immediacy and relevance rather than just traditional SEO metrics.