Brands Flex. Marketing Attracts. AI Decides.

Broadly speaking, marketing AI comes in two buckets: tools that help organizations find and keep customers, and tools that show consumers only the few brands AI decides are right for the prompt.

For a little over 2 years, I have been on a mission to understand the latter. Not that the tools for efficiency and productivity aren’t important -- they surely are -- but it is the intermediating, probabilistic role of AI for brand recommendations that fascinates me. What makes AI decide one brand is more worthy than another, when both operate in the same category, with similar histories and similar products? And of course, the vast majority of brands are not recommended -- and either don’t know it, or don’t know what to do about it. Why? Is it because some brands have better marketing? Or do the brands that get chosen have characteristics that matter more to machines?

Brand AI Report is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

I’ve been so obsessed with those questions that I spent the last year building tools and methodologies to answer them. Others are building systems around AI visibility and other marketing goals, like content, personalization, and ranking. Those have real merit, of course. The danger, though, is confusing marketing with the brand that AI is evaluating. SEO and GEO may get a brand into one of the many sources AI considers, but without independent recognition, for example, it is unlikely to be enough for AI to recommend it in responses to category prompts that don’t name the brand directly.

Rather, it is the evidence around an organization -- its claim-fact pairings, demonstrated values, and relevant associations -- that can be the deciding factor. And there are other characteristics – the recency and density of quality press coverage, historical accolades, leaders with citable pedigrees, and partners with strong credibility, to name a few – that can influence an AI decision today and for years to come.

So, marketing to AI systems is important for business growth, just as brand-building is, but resist the temptation to manage them the same way. You can have a website full of your brand-building credentials and press coverage, and for prompts about your brand specifically, that should help you be well described in the response. But being recommended in a category prompt? That is the steeper climb. Some brands inherit the characteristics AI trusts -- through decades of press coverage, for example. Others have to build them up deliberately.

That is what this series is about. I’ll be looking at how brand and marketing are intertwined in each of the 7 tiers of sources that AI may draw on to synthesize a recommendation. Marketing doesn’t have the same effect in each tier, but we will look at what organizations can do to keep their brand strong and clear in the media that matters to machines.

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