“This won’t differentiate us.”

Be sure that AI associates you with your category first. Then differentiate.

“This is too slow.”

Think foundation, not campaign. Building it takes months. Maintaining it will become routine.

“This is like SEO, right?”

Make sure your ship floats first. SEO comes later.

“No one cares about all these details”

AI probably won’t recommend a brand it can’t verify over one that it can. Your details are what it verifies.

The hands-on practice of Brand AI can feel like a stream of contradictions; brand strategy’s Best Practices bumping up against AI’s processing realities. When I survey a brand, I am doing 2 things: mapping which of its claims to the category can be verified and finding the most unassailable sources that help to illustrate its identity. That means understanding the company’s position and business, how well its product and service descriptions connect it to a category, the currency and cadence of the category’s ecosystem, and what 3rd party evidence corroborates that connection. It’s a lot to absorb in a short period of time, especially when the brand is in multiple, if not many, categories.

Those steps fit so nicely onto slides and webpages – easy to follow. But in practice, it’s another story – at least, a topic for another article. Suffice it to say that brand identity, products, and services are not always current on an owned website, let alone on a valued partner’s site the brand doesn’t control. And the work can have the effect of “stirring the pot”. Under the microscope, how the brand talks about itself and how important non-owned sources talk about the brand can lead to revelations and decisions to correct unknown inconsistencies.

What the work builds is category authority, and it comes from multiple citations to facts that corroborate the brand’s claims in often very dry and undifferentiating sources; everything from government filings to e-commerce storefronts to industry associations. The goal is to make it clear to AI exactly what the brand does, at what level, and according to what trusted sources. It is a lot of research and opening links that bots can’t – paywalls, database search results, etc. – to piece together the strongest possible argument that a brand belongs in the category where it competes. The holy grail: a brand so universally associated with its category that it practically defines it, to become “the anchor”, the reference that AI compares all others to when asked for recommendations.

The undercurrent to mapping authority is producing it, ongoing. Not big productions, not continuous publishing, simply making high-quality associations that deepen a consensus of belonging. It is one thing to connect the dots of a brand’s credibility across AI sources. It is another to keep them connected as the business reacts and evolves. Because the potential value is so great – being recommended to prospective customers by their AI systems – the process of building and maintaining it should really be a shared responsibility, and as routine as tracking expenses.

Often, it seems, these sources of authority are only known to the departments or solo experts who deal with them, in arcane journals or product spec sheets, for example, that a researcher would probably have to stumble on. So, people should be incentivized to routinely compile and organize the less obvious, more esoteric citations that could associate the brand with its category, a flow of citable facts to the keeper of the brand’s “authority map”.

This gap-finding and research process uses traditional search and AI tools, but it is not automatic. Building a Brand AI foundation the first time can feel like a triathlon (I imagine). But once it is built, it is a relatively inexpensive source of ongoing brand value. If there is a dusty basement full of important documents in the House of AI, this is it. And neither AI nor anyone else will see the brand’s “big picture” of identity until you build this frame.

The identity. The counterintuitive news is that associating a brand with its category requires abundant public corroboration of resolute sameness. It is the price of consideration by AI for inclusion in responses to prompts about the category, and so, consideration by the customer. The much better news is that for brands that make the cut, AI will probably characterize them by what sets each one apart. In AI, this is where the brand’s unique identity shines, synthesized from various sources of owned-media, 3rd party lists, product reviews and service descriptions.

In my view, the purpose of a brand’s facts and citations is to get its poetic position in front of a customer using AI. And because AI considers claims it can verify over those it cannot, brand strategy can begin to identify which citable facts best underpin its value proposition. Everything from leadership qualifications (3rd party experience, published credentials) to published 3rd party tests of materials used in production can validate the value proposition. Verified results, public filings, or accolades in one category can also make an exploration into a related category more reasonable to AI. When a brand is widely acknowledged to have expertise in one area, it can move into a tangential one more credibly.

But it can be tricky. For example, I was recently researching the watch category and was surprised when a well-known brand was not in the response to “What are the best brands for a watch at around $1500-2000?”. Asked why not, the AI ventured a guess: “Those [named] brands have unambiguous category association, whereas [the unnamed brand] is a racing brand, a smartwatch brand, an entry-luxury brand and an F1 sponsor at once, and diffuse positioning may show up as diffuse sources.” AI systems don’t know how they produce a response, but it is worth asking whether a brand that competes in four categories can be an authority in any one of them without committing to continuous foundation building for AI systems used by customers – very possibly worth it, but it’s a decision to be all-in.

All of which is to say: Building a deep foundation of facts, authoritative associations and reviews from sources that cover a category is key to AI systems associating the brand with that category. Only then can the purpose and identity of the brand be in the mind of the customer using AI when it’s time to make a decision.