What does the audience survey actually do?
The survey calibrates how your guidance is framed, not which gaps the analysis finds. Skipping it does not degrade your results.
What the survey changes, and what it does not
The analysis will find the same gaps either way. Your audience helps us prioritise which ones matter most for your readers, and how to phrase the fixes.
The 12 core concepts that ContentGrapher identifies as missing or underdeveloped on your page are determined by the page content, not by what you enter in the survey. Those concepts appear regardless. What the survey changes is how the pipeline frames the guidance for those gaps: which ones are flagged as essential, and how specifically the writing brief prescribes the fix.
Where the survey fits in the flow
Phase 1 runs first, producing the concept map and boundary classifications from your page. The survey prompt then appears. After you fill it in, or choose to skip it, Phase 2 runs using your audience input to frame its output.
The structural recommendations in Phase 2c, which determine whether your page should be split or kept as one, do not use the survey at all. They operate on the concept map and content context from Phase 1 only. Phase 2 produces the writing brief and explanation framework, both of which are audience-calibrated when you specify your reader. Phase 2 requires an analysis credit. Phase 1 is free.
What changes in practice
The same gap gets a different prescription depending on who is reading. Here are two examples from the same analysis, applied to the same gap, with and without audience input.
No audience specified: "Structural completeness is the central diagnostic criterion but the current content treats it as self-evident. Add a sentence that defines it operationally, what makes a piece of content structurally complete versus incomplete, so readers have a working definition they can apply to their own content."
Audience specified (expert content strategist, task: identify LLM retrieval gaps): "Tighten the definition of structural completeness so it reads as a benchmark, not a buzzword. The current content uses the term but doesn't give strategists a clear mental model of what 'complete' looks like versus 'incomplete' at the structural level. Add a concrete threshold or dimension."
Same gap. The no-audience version asks for a working definition. The fully specified version asks for a benchmark with a concrete threshold. Both are correct for their respective readers.
What happens if you skip it
If you skip the survey, the pipeline uses inferred audience context derived from your page content to frame the output. You receive a complete, accurate analysis. Nothing is missing. The only difference is that the framing will be calibrated to the audience ContentGrapher inferred, rather than the one you specified.