Skip to main content

Presence Rate

Understand how consistently a brand or domain appears in generative answers.

Written by Apolline Vanneste
Updated yesterday

🎯 What the Presence Rate captures

The Presence Rate measures how often a brand or domain appears across individual LLM-generated search results.

The Presence Rate is expressed as a percentage from 0 to 100, reflecting the share of results in which:

  • a brand is mentioned in the answer,

  • a domain appears as a link,

  • or both, depending on the selected view.


🧮 How the Presence Rate is calculated

For each analysed result:

  • the presence of a brand mention and/or a domain link is detected

  • each occurrence is counted, including multiple brand mentions within the same answer

The Presence Rate represents the proportion of results where the brand or domain appears, relative to the total number of eligible analysed results.

Results are considered eligible when the answer contains at least one brand mention or domain link. Results where no brands or domains appear are excluded from the calculation, as they offer no opportunity for visibility.

Presence is measured at the result level, not at the prompt level, meaning each result is counted independently. For example, if the same prompt is analysed across five different engines, it represents five distinct results.


🔍 How to read the Presence Rate

The Presence Rate is a consistency signal.

  • A high Presence Rate indicates that a brand or domain appears regularly across engines and prompts

  • A low Presence Rate suggests sporadic or limited exposure

This KPI does not indicate how prominently a brand or domain appears, only whether it appears.


🔗 How this KPI fits with other signals

The Presence Rate works closely with:

  • Visibility Score, to understand how presence translates into prominence

  • Sentiment Score, to assess whether frequent mentions are framed positively, neutrally or negatively

Together, these signals help contextualise exposure, impact and perception in generative answers.


To go further

Did this answer your question?