Fan-out coverage

When an AI answers a prompt, it silently breaks it into sub-queries and researches each one. Fan-out coverage measures how much of that hidden research you participate in. The Fan-out tab in the Gaps area shows it as a short summary per topic; the full list lives on the Query Fan-out page.

How coverage is counted

A sub-query is covered when you appeared, cited or mentioned, in at least one answer that ran it in the selected period. Everything else counts as not covered: topics the AI researches around your market where you do not exist.

Sub-queries are counted once per topic. Spellings that differ only in capitals or spaces, such as "Best SEO tool" and "best seo tool", count as one sub-query. When the same sub-query comes up for several prompts of a topic, it counts once, and it is covered when you appeared in any answer of that topic that ran it.

The overview tile sums the distinct not covered sub-queries across your topics for the period.

The Fan-out tab

The tab has one row per topic:

• The coverage count, for example "61 of 148 covered".
• Up to five not covered sub-queries, the ones the AI ran most often.
• See all: opens the Query Fan-out page filtered to that topic, the same period and the not covered sub-queries only. The "+N" chip after the five sub-queries opens the same view.

A topic where every sub-query is covered says so instead of listing chips. The Query Fan-out page counts with the same rule, so its numbers match the tab.

The Fan-out tab is available on Pro and Agency plans.

Why not covered sub-queries are gold

They are pre-validated content ideas. Each one is a question the AI demonstrably researches when answering your customers, which is stronger demand evidence than most keyword tools provide. Covering a sub-query in a title, a heading or a dedicated section closes a specific hole in the AI's picture of your market. When a Search Console property or a Bing Webmaster Tools site is mapped, the Query Fan-out page also shows which of them have search demand behind them.

Fan-out data comes mostly from Gemini, which reliably emits sub-queries, while ChatGPT typically echoes the prompt and Claude and Perplexity expose few.