The common category error
A fan-out can surface dozens of plausible questions. The mistake is treating the list as an information architecture. Questions can overlap, depend on the same evidence, or make sense only as parts of one decision. Publishing each variation creates repetition before it creates coverage.
A content plan needs a second layer of judgment: which user jobs are genuinely distinct, which page already owns them, and which evidence can be maintained over time. Without that layer, fan-out becomes a scaled-content prompt.
Four tests before a branch gets a URL
A separate page is justified only when the separation helps the reader and the site. Use four tests together; passing one is not enough.
- Different job: the reader is trying to reach a materially different outcome.
- Different proof: the answer needs its own dataset, test, expert evidence, or maintained reference.
- Different journey: the next action or conversion path is not the same as the parent page.
- Different lifecycle: the material changes on a different cadence or has a separate accountable owner.
What to do with the branches that remain
Keep supporting questions as sections on the strongest page when they serve the same promise. Turn unresolved proof into a research task. Merge wording variants into the clearest formulation. Reject branches that add no decision value or cannot be supported without speculation.
The result is usually smaller than the original map. That is a feature. A planning process should remove unnecessary work before it creates a publishing backlog.
A conservative default
If an existing page has one bounded gap, add a section. If its promise needs material expansion, extend it. If several pages compete, merge content only after checking intent equivalence. If evidence is missing, build an evidence asset. Create one new URL only when no existing page owns a distinct job and credible source material is ready.
That default does not guarantee a ranking outcome. It produces a cleaner, more reviewable reason for the page to exist.