Research across 14,140 controlled runs, 12 brands and five generative AI platforms, written up on Search Engine Land on 15 October, found that UK activewear brand TALA appeared in none of the athleisure or sportswear recommendations tested. The AI models placed TALA correctly in athleisure when asked about it by name. But Google's Knowledge Graph (the database Google uses to understand who's who) described TALA as a financial services company, because a fintech shares the name. The authors are careful: the data shows the two together, and doesn't prove the mix-up caused the gap.
New Balance makes the point from the other side. It had the strongest Knowledge Graph result in the study and still got zero athleisure recommendations, because Google and the AI models see it as a footwear brand. Being known matters less than being known for the right thing.
Under the headline, this is about identity. AI tools recommend brands they're confident about, and that confidence comes from consistent signals: what your site says, what other sites say about you, what Google's own records say. A short name, an acronym, a rebrand or a shared name all blur those signals.
Our SEO rule is to be the answer, wherever people ask the question. Before AI can recommend you, it has to know which company you are and what you're for. That's foundational work, and the boring stuff just became strategic.