GenGrowth
GENGROWTH · GEO

/tools/geo-brief · Output ContentBrief v1.1

GEO Brief Builder

What this does not do

Reuses successful answers from the selected owned run without resampling. Without a run, no observed coverage is invented. AI answers determine required topics only; numbers and product claims must come from non-null facts. GEO has no SERP length baseline, so length stays unavailable and the page decision remains undecidable.

Questions about this

Where do the subtopics come from?
An owned visibility run supplies the actual headings and list lead-ins from its successful answers. Their cross-sample counts exclude failed calls. Without a run, the brief contains only frozen requirements and facts, or your explicitly labelled manual question; it does not invent observations or buy a replacement sample.
Can it write the numbers for me?
No, and that is deliberate. Anything specific comes from the fact table you verified in your knowledge base. A dimension you have not verified is printed as not verified, with the reason. The model that assembles the brief is told it may never state a fact that is not in that table.
Why is a sampled answer not a source of facts?
Because it is another company's model, answering once, and it can be wrong. It is good evidence about what a page has to cover — that is what it is used for — and no evidence at all about what is true of your product.
Does it write the page?
The Brief does not write or publish the page. It hands the same versioned contract to the shared Content Draft tool, whose sentences are constrained by the non-null fact table. After you publish, the GEO draft links to the page citability check.

The rest of the chain

A brief is worth writing when you know which questions you are missing, and worth publishing when the page it becomes can actually be read.