AI Ranks Brands Differently for Every Audience

To understand why audience segments are such a key aspect of AI performance, consider this example: When a suburban parent asks an LLM to recommend a value retailer, Costco usually shows up. But when a younger, trend-driven shopper asks the same LLM the same question, Costco is nowhere to be seen. Despite being near-identical, the prompts produce two answers with nothing in common. 

The Audience Data Hiding Inside Costco's Score 

Costco is one of the leaders in the Value Retail product category, but its aggregate AI Presence score suggests it appears in only 55% of AI responses about value retailers.

AI Presence: How frequently AI systems include a brand in responses about key topics

This is because the superstore appears often to some audiences of shoppers yet barely gets surfaced to others. Bluefish’s State of Enterprise Brands in AI report found that Costco appears in 84% of AI responses for Suburban Family Shoppers, compared with just 7% for Trend-Driven Shoppers. That’s a 77-point swing driven entirely by who’s prompting the AI.  

LLMs tailor answers to the individual user, surfacing different brands depending on what they know about them based on previous conversations. An aggregate category score combines distinct audience segments into a single number, masking meaningful differences. A brand reading only the aggregate score would never see the more granular data about their highest-value customers, or understand competitive standing among different audiences.  

Why Aggregate Category Scores Don’t Go Deep Enough 

This divergence by audience segment holds across verticals. Sephora's AI Presence in the Health and Beauty category ranges from 93% with Trend-Driven Shoppers down to 70% with Value Seekers. In each case, a category-level score would flatten rich data like this into a potentially misleading figure.  

An aggregate ranking is wrong for nearly every real customer; Read on its own, it offers very little insight into where a brand loses or earns AI representation. 

This is why audience-level measurement is the foundation of agentic marketing. It captures the actual experiences that individual customers are having with brands while using AI, not a statistical average of them. For enterprise marketers, the discipline is to treat the category score as a starting point, then go deeper to see where decisions are really made: By measuring AI presence by audience, product, and topic. 

Don't Let AI Define Your Brand 

You’ve worked hard to build your brand, don't let AI define it for you. The Bluefish Agentic Marketing Platform (AMP) gives enterprise marketers deep insight into how AI represents brands at the audience, product, and topic level, because an AEO strategy is only as precise as the data backing it up. 

Connect with our team to see how AI represents your brands to the customers that matter.