If AI Can’t See Your Brand, Neither Can Consumers

283

Consumers are no longer typing into search bars. They are talking to ChatGPT. In 2024, over half of surveyed users preferred generative AI tools over traditional search engines for product advice and travel planning. This shift has not only disrupted how people discover brands, it is changing what visibility even means.

The old digital marketing battlegrounds like SEO, social media, and brand recall are being eclipsed by a new metric: Share of Model (SOM). Unlike Share of Search or Voice, SOM measures how frequently and favorably large language models (LLMs) such as GPT-4, Gemini, or Perplexity mention a brand in response to consumer prompts. If your brand does not show up, it might as well not exist.

Jellyfish, a marketing agency pioneering SOM analytics, has revealed wide disparities in how LLMs surface brand names. In the Italian detergent market, Ariel’s SOM ranged from 24 percent on LLaMA to under 1 percent on Gemini. Other brands were entirely absent from some models. Unlike search engines, LLMs offer no second page.

This is not just a tech quirk. It is a strategy gap. Some well-known brands are invisible to AI, while niche or newer ones thrive. Tesla and Rivian, for instance, score highly on LLM visibility due to content strategies focused on functionality and resolution. In contrast, legacy players like Lincoln, reliant on brand heritage, often go unnoticed by machines trained to solve problems rather than admire lineage.

To boost SOM, marketers must adapt. Content must address clear use cases, cite authoritative sources, and speak to AI’s goal of resolving user intent. Brands like Nike and The Ordinary excel in this area by offering structured, contextualized product information that AI can interpret and recommend.

Understanding LLM sentiment — what AI “likes” and associates with your brand opens new levers for positioning. Travel brands, for instance, are now tailoring content around concepts like privacy and variety after seeing how LLMs prioritize these traits.

In this new era, visibility is no longer just earned. It is engineered. The next generation of brand growth will depend on how well marketers can speak to the machines that speak to us.