Less Than Half of Mexican Brands Make GPT-4’s Top 100: See Who’s In!

Data from the study Positioning of the Top 100 Mexican Companies in Gen AI, conducted by marketing and communication agency Llorente & Cuenca (LLYC), shows that less than half of Mexico’s most recognized brands appear in the top 100 of GPT-4 when queried about three of eight key economic sectors (Entertainment, Tourism and Consumer Services, Industrial Manufacturing, and Financial Services and Holdings).

The study evaluated 12,000 assessments made by models GPT-4, Gemini 2.0, and Claude 3.7, comparing their brand recognition with the official information published by these companies.

Adolfo Corujo, CEO of marketing solutions at LLYC, explains that while there are well-positioned Mexican companies in reality, this is due to various factors that AI doesn’t consider when ranking, such as annual earnings or employee count, which are typically considered in rankings made by specialized media.

“In the case of AI,” he notes, “positioning is based on publicly available information, usually sourced from the media, social media interactions, and involves research conducted in other countries, allowing international brands to be better positioned in the eyes of consumers.”

New Brand Positioning in the Age of AI

ChatGPT was launched in November 2022, and its rapid adoption has significantly altered how the internet is used and understood. Previously, companies focused on positioning themselves through traditional media without considering how to do so through AI.

The study highlights a disconnect between the identity projected by Mexican brands and the perception held by major language models. While OpenAI’s models align closely with the official image of the companies, there are notable variations by dimension. For instance, aspects like social responsibility and corporate vision are less recognized compared to the basic definition of the company.

Corujo points out that national firms are associated with many concepts. “As humans, we make the mistake of associating a company with 15 themes instead of just three,” he says. “We talk about innovation, culture, social and corporate responsibility—a lot of things. But when a machine processes this information, it boils it down to three words, and the more concepts there are in a positioning, the more likely it is to get lost.”

Furthermore, he highlights structural biases in how brands are represented in AI. State-owned or industrial companies tend to be less visible or correctly classified by large language models (LLMs). This is linked to more restricted audiences and narratives marked by controversies that distort the perceived image. In contrast, brands that clearly communicate their values are more recognized.

The challenge for brands in Mexico facing AI is to adapt their narratives to this new medium, as there has not yet been a significant shift that would better position national brands; however, the expert predicts that in the next 12 to 15 months, a large percentage of companies will become competitive in their AI positioning strategy.

“In an era where artificial intelligence not only serves to consult information but also influences people’s perceptions,” concludes Corujo, “it’s crucial to foster a discussion that has not yet been addressed: how brand reputation is managed in the face of algorithms.”