Many GEO recommendations sound plausible but rest on technically unsound assumptions. Anyone who understands how AI bots actually access content and different content formats can assess measures accurately instead of pouring effort into ineffective optimisations. Myth check: analysis of common GEO recommendations with concrete examples — and why they fail in practice. How AI bots really crawl: how LLM crawlers handle different content formats such as images, videos and documents, what they can process and where the real technical barriers are. Which measures still pay into AI visibility after the myth check — and how editorial teams should prioritise their effort accordingly.