Fan-Out Mapper
AI search answers one question by running many smaller searches. This maps them for any keyword, from the searches Gemini and ChatGPT actually ran plus real Google searches, groups them by what the searcher wants, and checks which ones your page answers.
Add searches you saw an AI engine run
Your maps
saved in this browser, they reopen at no cost
Sub-query map
Page blueprint
built from the fan-out
Where the sub-queries come from
Seen in Gemini and seen in ChatGPT are the searches each engine ran while answering the keyword, read from Gemini's search grounding (run twice, because its searches vary) and ChatGPT's web search. Added by you marks searches you saw an engine run and pasted in. Google searches are related searches with monthly volume from DataForSEO and Google Autocomplete. They are merged, near-duplicates are dropped, and up to 40 are kept, favouring ones that were really searched, stay on topic and differ from each other; the next 40 show as grey context dots. Branded searches are set aside, because a page cannot rank on another company's name. No engine publishes the exact fan-out it uses for a given person, so treat this as evidence of it, not a copy. Coverage compares each sub-query's words with your page's headings and passages: it shows what is there, not how well it is written.