Research

Research shapes real search-demand data plus a language model into around 32 candidate prompts — at least 80% unbranded, organised into 6–8 clusters — and feeds them into your suggestion queue. It also carries a Google Search Console striking distance block.

What it does

It helps you find questions worth tracking that you haven't thought of yet, grounded in real demand rather than guesswork, and points you at queries where you are close to breaking through in search.

How to use it

Open Research:

• Generate candidates and send them to the suggestion queue, or straight to tracking.
• Review the striking-distance list: queries ranked 8–20 in Search Console with impressions above zero, showing the top 12 by impressions. These are the near-misses where a small push could earn ranking.

How it's computed

A fail-open relevance filter trims off-topic candidates — if the filter can't run, it errs toward keeping candidates rather than dropping good ones silently.

Per-prompt volumes use a long-tail decay model and are shown as honest ≈ asks/month, never unit-precise, because AI-search demand can't be measured to the unit. See data honesty.

The striking-distance tile deliberately has no sparkline: it is defined over a 28-day aggregate, and there is no honest daily series to draw — so rather than fake one, it draws none.

Limits

Striking distance requires a connected Search Console. Volume estimates require DataForSEO credentials; without them, volumes fall back to honest bands rather than invented numbers.

Related

• Suggestion queue — where generated candidates land for review.
• Fan-outs — another source of real, trackable demand.
• Google Search Console — the source of the striking-distance block.