In short
- What happens
- The asked question disappears and sub-queries take its place. You do not see them; the engine does.
- Where you drop out
- At one sub-query, usually a comparative one. The rest of the answer simply carries on without you.
- What it means for measuring
- A measurement that only asks the main question measures the tip. Coverage across the sub-queries is the real measure.
What does such a split look like in practice?
Take a question like "what is a good solution for a small business starting out". It contains at least four sub-queries: what is this actually, what does it cost, what are the options, and what to watch for when you are small. The engine fetches something for each, and you can be perfectly present on three of the four.
The fourth is then the one carrying the recommendation, and it appears nowhere in your material. That is the pattern: the hole is rarely at the question you expected, because you had already answered that one.
Do all sub-queries weigh the same?
No, and that difference is measurable. We weight a sub-query by the intent behind it, with these weights — the same ones used in our scan, so the figure in the report and the figure in the dashboard cannot drift apart.
What stands out in that table: the intent nearly every fan-out starts with is the lightest of the six. Anyone building their entire content plan on explanatory questions is building on the cheapest half of the journey.
| Intent | Weight | Where it shows up in a fan-out |
|---|---|---|
| info | 0.55 | The first sub-queries: what is it, how does it work. Nearly every fan-out starts here and nearly nobody buys here. |
| commercial | 1.00 | The sub-query asking whether it is allowed, whether it is safe, whether it fits your case. The highest weight of the six. |
| compare | 0.92 | "X versus Y", "alternatives to". This is where the longlist becomes a shortlist, and where you drop off without noticing. |
| transactional | 1.00 | Price, quote, request, appointment. As heavy as commercial: whoever types this is done orienting. |
| local | 0.80 | With a place name in it. A lower weight than purchase-ready, yet the only intent where distance decides the answer. |
| navigational | 0.50 | Someone is looking for a name they already know. The lowest weight, because the choice was made before the question. |
Can I see which sub-queries were asked?
Not directly — the engine does not show them. What you can do is infer them: ask the same thing five ways and see which sources keep surfacing. Sources returning in four of five phrasings answer a sub-query that lies beneath all five.
What do I do with a hole in my coverage?
Make a page of it, rather than devoting a paragraph to it in an existing one. A sub-query deserving its own answer deserves its own address — then an engine can point at it without having to carry the rest of your story along.
Start at the hole with the highest weight, not the biggest hole. Four missing explanatory sub-queries together weigh less than one missing comparative one, and that is exactly the order most content plans do not follow.
Frequently asked questions
Do all engines do this?
Not all to the same degree, and not on every question. A short factual question is often simply answered; an open question with a trade-off in it is almost always split. Rule of thumb: the more the question asks for a recommendation, the more happens under the hood.
So must I create a page for every conceivable sub-query?
No, and that would be the agent-page mistake: thirty near-identical pages crowding each other out. What works is one page per sub-query that genuinely recurs, with an answer not found elsewhere. A sub-query you already answer in an existing section needs no second address.
Read on
Free, no account and no card.