Skip to main content
Visibility is the share of AI answers that mention your brand. It’s the first number to look at because it answers the most basic question — are you even in the conversation when buyers ask AI about your space?

​
How it’s calculated

Visibility = (responses that mention your brand ÷ total responses) × 100
For each prompt, 3RD checks every AI response for your brand. A response where your brand appears counts as “visible.” Each prompt is scored individually, and the overall score is the average across all prompts — every prompt counts equally, regardless of how many models answered it.
A mention is a detected occurrence of your brand in the answer. Detection is robust to spelling variations, capitalization, possessives, and suffixes, with an AI-based fallback for ambiguous phrasings — so mentions get caught even when the model words them loosely.

​
How to read it

Visibility is always relative to a filter. The same brand can be at 70% for branded prompts and 8% for unbranded “best X” prompts. Always check which model, market, topic, and prompt type you’re looking at.
  • Trend matters more than the absolute number. Watch the direction after you ship changes.
  • Segment it. Low visibility on Must-Win prompts is far more urgent than on Foundational prompts.
  • Compare against competitors on the Leaderboard, not against an arbitrary target.

​
How to improve it

Visibility is downstream of being useful and citable for the questions buyers ask. The levers, in order of leverage:
1

Cover the prompts you're absent from

Publish content that directly answers the questions where you have no page. See Content.
2

Earn citations on sources AI already trusts

Get featured on the third-party pages models cite for your category. See Authority & Outreach.
3

Make your own pages extractable

Ensure models can cleanly read and quote your content. See Crawlability.
Don’t chase 100%. On competitive unbranded prompts, a healthy, rising visibility that’s at or above your nearest competitor is the real goal.