What's A Brand To AI?

A preamble about brand, for this audience, is like explaining water to fish. Let’s jump ahead. Take all that you know about brands -- for humans -- and consider what they are to machines. More specifically, how does AI decide which brands are good enough to include in its short, reasoned responses to prompts about an infinite number of categories?

This is a sharp turn -- easy to miss. Searching with AI sounds a lot like searching with a browser. As we know, searching-by-browser produces pages of links to webpages that are often optimized to be found. Searching-by-AI returns just a few links, maybe in a short list, often explained with some confident text. Sometimes, it returns no links at all -- just an answer. And the brands on AI’s short list? They’re not there because of SEO.

Sure, if a category lacks more-credible sources that define it, brand websites will do. And in those situations, the top results from a traditional search-by-browser can become a legit source (among others). But that’s a fragile position. Competitors with more structured credibility, facts, and consensus as category players in more authoritative sources -- combined with what they promote on their own websites -- may appear in AI search, crowding out others that don’t.

So, what exactly is a brand to a machine? To humans, a brand is a structure of values, history, context, identity, and vision (apologies to brand strategists for any omissions). To AI, a brand must have three dimensions to qualify: where, what, and “says who.” The surface is the trusted sources (where) AI finds brands and their category. Below the surface is the scaffolding of credibility (what) -- the brand’s claims, paired with and validated by facts (”says who”): third-party acknowledgement for innovation, earnings reports, staff pedigrees, testimony, transcripts, reviews, and rankings.

Think of it as a series of roads and maps, always under construction, improving. AI is the driver. It likes routes that are paved, clearly marked, and verified by other maps. The brand is the engineer, designing the straightaways and turns. The claims on a brand’s website that are paired with third-party citations are the road signs, just for AI. The authority of the sources behind those citations is the pavement -- smooth and certain, or rough and full of holes. Consensus across citations and sources is what reassures the driver that the road is safe. If a competitor’s road is smoother, better marked, and safer, AI drives there instead. And you, the passenger, will never see the road not taken.

So, how is a CMO supposed to manage their brand in AI responses when it all seems so foggy and indirect? Build roads and maps for AI that are better than the competition’s. It starts with measuring how AI describes a category and the sources that define it, across different prompts and models. Measure the volume of facts that specifically support how the brand talks about itself -- its claims -- and the citable facts that prove them true. Measure how the competition is described, how often, and where. And link to as many third-party citations as possible that substantiate your unique value, identity, and vision -- the accolades of products, personnel, and partners, of category leadership, organizational heft, geographic focus, and community recognition -- all the brand aspects that your competitors don’t have and AI will trust.

These external sources and internal scaffolding are what help brands -- young and legendary alike -- get recommended by AI. Only when you have your brand road, maps, and signs that AI trusts can marketing get traction and build the speed to compete.

This post first appeared in the Brand AI Report, my newsletter on how AI describes and recommends brands.