A visibility figure is only meaningful relative to what was asked. Twenty questions you would obviously win produce a wonderful number and no information. Twenty a competitor's marketing team would write produce the opposite. The question set is not setup: it is the instrument.
Ours is generated from the site's own Search Console queries with branded searches excluded, then confirmed by a human before anything runs. A question is only attributed to Search Console when its wording genuinely overlaps a real query, because 'grounded in real demand' is a claim that has to be true rather than decorative.
The rules that keep it honest
- Never name yourselfA question naming them measures whether the engine can describe a brand it was just told about. That is not the same as being recommended, and it is much easier to score well on.
- Ask what a buyer typesReal purchase questions, in the reader's words, not category descriptions written by a marketer.
- Keep the set fixedWeek-over-week movement is only interpretable if the questions did not move. Additions are dated, so a jump caused by new questions is never read as a jump caused by your work.
- Enough questions to survive one odd answerBelow about ten, a single engine having a strange day moves the number more than anything you did that month.
- Exclude branded demandPeople already searching your name are not the audience this measures, and leaving them in inflates every figure in the same flattering direction.
Who approves the questions?
You do. They are proposed, and nothing runs until you have read the list. A generated question set nobody checked is a measurement nobody can defend.
Can I add questions later?
Yes, and the week they entered is recorded so that a change in the number can be attributed to the change in the set rather than to the market.
Try it on your own site, free
No card, and the free tier does not expire — the first verdict takes about six weeks to arrive.