Of 2582 brands rated across 67 categories, 25% sit above indifference, 74.7% are indifferent and 0.3% below.
Across 67 categories and 2,582 brands, the BrandInsightAI macro dataset maps how AI models judge commercial standing, revealing widespread indifference and stark category divides.
Of 2,582 brands rated across 67 categories in the BrandInsightAI macro dataset, 25% sit above indifference, 74.7% are indifferent, and 0.3% fall below.
Universities has the largest share of brands above indifference at 96.6%, with 29 brands evaluated and 24 named across responses.
Betting and Gaming records the smallest share of brands above indifference at 0%, containing 40 brands and 28 named entities.
Figure 1Share of each category’s brands above indifference
Above indifference: Desire above two-thirds of its maximum. The 20 highest of 67 categories.Data table
Emotional magnetism is the feature the most categories are divided on, appearing as contested in 17 of 67 categories in the BrandInsightAI macro dataset.
Fragrance and Beauty leads the index for Emotional magnetism with an index score of 138 against the macro dataset baseline.
Values and integrity is contested in 11 categories, while Emotional bond is contested in 10 categories.
Figure 2The feature each category is most divided on
Number of categories whose brands are most spread on each feature, measured against how spread that feature usually is.Data table
With 74.7% of brands sitting in the indifferent tier across 2,582 evaluated entities, marketing teams should analyse how AI models perceive emotional magnetism, which is contested across 17 categories.
Given that Gemini rates brands 4.5 points above peers while ChatGPT sits at minus 1.7 points, SEO strategies must monitor cross model variances rather than relying on a single AI output.
Since Apple leads Consumer Electronics and Appliances at 84.9% of maximum Desire, brands trailing category leaders should examine feature indexing to align with high performing benchmarks.
Method & limitations
Dataset findings are descriptive and correlational, reflecting stated model evaluations rather than direct consumer transactions.
Model prompts use the 31 statements of the Science of Desire glossary, which may not capture every commercial nuance.
How every figure on this page is counted, floored and
reviewed: our methodology.
Where next
Every figure in this piece comes out of the same warehouse these views render live. Read on from whichever question you have next.