A planning map, not a leak from a search system
On this site, fan-out means taking one declared user job and mapping the work needed to answer it responsibly. The branches are editorial prompts produced by a published rule set. They are not hidden Google searches, private model instructions, chain of thought, demand estimates, or ranking signals.
That boundary matters. A useful planning model can be inspected and challenged. A claim about secret platform behavior cannot. The framework therefore shows its inputs, rules, assumptions, and missing evidence instead of pretending to reconstruct an internal system.
What the map contains
A bounded map starts with the outcome a person needs. It then names the entities that must be resolved, the questions that determine the decision, the source roles those questions require, and the gaps that prevent a confident recommendation.
- Goal: the outcome the reader or practitioner must reach.
- Entities: the people, products, rules, pages, or systems that need an explicit identity.
- Questions: the smallest set of issues that could change the decision.
- Source roles: the kind of evidence needed, such as a current specification, firsthand test, or page inventory.
- Page decision: extend a page, add a section, merge content, build evidence first, create one URL, or take no action.
Why the page decision comes last
A list of related phrases is easy to produce. It is also a poor reason to create URLs. The framework checks whether an existing page already owns the job and whether the evidence is strong enough before a new page is allowed into the plan.
This changes the default. A branch is usually a section, a research task, or an evidence requirement. It becomes a separate page only when it represents a distinct user job with a durable purpose, its own proof, and a useful place in the site architecture.
Where the model stops
The tool does not inspect a site, Search Console property, live search result, or AI answer unless the user supplies that evidence in a separate review. It does not estimate traffic, predict citations, or judge factual accuracy. Its output is a structured decision prompt, not a finished strategy.
Use the map to decide what to verify next. Use real URLs, primary sources, firsthand material, and measured demand to decide whether the recommendation survives contact with the actual site.