Web Mentions

The four indicators

• Sites talking about you: how many distinct domains cite the brand.
• Already picked up by AI: how many of those sites ChatGPT, Perplexity or Gemini already cite in your tracked prompts.
• High-authority sources: the share of citations coming from solid sites.
• Dominant tone: shown only when the data actually exists — never a number invented to fill the box.

The most actionable indicator

Already picked up by AI is the one that sets your priorities. Landing a mention on a site the models already cite is worth far more than ten mentions on sites they ignore. The Who cites you list puts those sites first: they are your next PR targets, in that order.

A verdict that goes both ways

• Good source: the site really does cover your client. Its results are marked « confirmed » by default, instead of going back through « to verify » every week.
• Not my client: the domain is permanently excluded from all of External presence, cache purged straight away.
• Undo reverses either one.
• A deliberate limit: a term you have banned still drops the content, even from a confirmed source. The verdict says the site talks about your client, not that everything it publishes should be accepted.
• A summary line recalls how many sources are confirmed and how many are excluded.

When the data is missing

The content analysis API does not cover every brand. When it does not, the page falls back on Google search results rather than staying empty — and it tells you which of the two sources it is showing, so you always know what you are reading.