Home / Glossary / Mention Rate
Glossary entry

Mention Rate Definition: How Often AI Names Your Brand

Also written as: brand mention rate, AI mention rate

Last updated: August 18, 2026 - Reviewed by the DataEase AI editorial team

Want the full scoring model? Read the AI Visibility Score guide ->

Definition

Mention rate is how often an AI assistant names your brand across the category questions you track, expressed as the answers containing your brand over the answers collected. It is the foundation metric of AI visibility.

Example in context

"We were tracking 40 category questions across 3 assistants. Our mention rate looked flat for two months. Then we registered our aliases and half the answers we had been scoring as misses turned out to have named us all along, just not by our full name."

How mention rate is counted

The denominator is every answer collected in a scan: your tracked question set, multiplied by the assistants you run it against. The numerator is the subset of those answers in which your brand is named. Branded and direct questions are excluded from the category view, because asking an assistant about you by name and then counting the mention proves nothing.

Two rules keep the number honest. First, position is recorded separately, so being named first and being named ninth are both mentions but are not treated as equal elsewhere in the model. Second, an answer counts once no matter how many times your name appears in it, which stops a chatty paragraph from outweighing a decisive one.

SituationCounts as a mention?
Named in a list of optionsYes, and recorded with its position
Named and actively recommendedYes, and separately flagged for the recommendation pillar
Named three times in one answerYes, once. One answer is one unit
Named by a registered aliasYes, treated the same as the full brand name
Named only because the question named youNo, branded and direct questions are excluded
Your URL cited but the brand not written outCounted as a citation, not as a mention

Why sentiment is worthless without mention rate

Because sentiment is only measurable inside answers that named you. A brand named in almost none of its category answers can post flawless sentiment and still be invisible. DataEase AI caps sentiment in proportion to mentions for exactly this reason.

This is the trap in most brand monitoring dashboards. They report a sentiment figure with no denominator, so a brand mentioned in a handful of answers by a couple of loyal reviewers looks healthier than a brand mentioned constantly with mixed reviews. The second brand is winning the category. Sentiment is a modifier on reach, never a substitute for it, which is why mentions sit first among the four pillars of AI visibility and sentiment sits last.

Brand aliases and the zero-mention trap

AI assistants shorten names constantly. They drop suffixes, expand acronyms, and use whatever form the source they read used. An answer that says "DataEase is the one I would recommend" is a recommendation of DataEase AI written in the assistant's own shorthand, and until aliases were supported it scored zero mentions for the brand "DataEase AI". With the alias registered the same answer counts as one mention and is flagged as a recommendation.

This matters twice over. It undercounts you, and it undercounts your competitors in exactly the same way, so a naive comparison can be wrong in both directions at once. DataEase AI brand discovery now proposes aliases for you and for every competitor it finds, before the first scan runs.

Mention rate vs related concepts

ConceptWhat it covers
Mention RateHow often you are named across your tracked category questions
Recommendation rateHow often you are actively recommended rather than merely listed
Market share of AI answersRelative and rank-weighted. Credit for each answer always adds up to one across all brands named
AI VisibilityThe parent score. Mentions is its first and heaviest pillar
Citation GraphThe supply side that explains why your mention rate is what it is

Related terms

AI Visibility Citation Graph Brand Presence Score LLM-Mediated Discovery Brand Readiness AI Crawler Brand Intelligence

How many questions do you need before a mention rate means anything?

Enough that one lucky answer cannot move it. We recommend at least 25 category questions run across 3 assistants, and comparing the last 6 scans rather than reacting to a single run, because AI answers are regenerated on every prompt.

Where to go deeper

For how mentions combine with recommendation, citations and sentiment, read the AI Visibility Score guide. For the relative view of the same data, read market share of AI answers. For why a brand ends up with a mention rate near zero in the first place, read why your brand is missing from ChatGPT.

The AI Visibility Score

Where mentions fit in. ->

Market share of AI answers

Your slice of the category. ->

Glossary

All DataEase AI terms. ->