AI can’t build your brand for you

AI is flattening the purchase funnel. Consumers are going to hand their buying decisions to agents. The whole journey from awareness to purchase collapses into one AI-mediated moment, and brand-building becomes something we used to do.

That’s the trendy argument I’ve been hearing in boardrooms and on conference stages. And that narrative is wrong.

It’s built on half-truths and an arrogant, tech-driven dismissal of the extraordinary value of brand-building, and the brands that buy into it are going to pay for it. 

The funnel didn’t flatten. The stakes got higher.

Ask an AI assistant for the best running shoe. You’ll get a list. HOKA is probably on it, and you could make a good case that it’s one of the best performance shoes out there. Nike’s on it too. Which one does the person buy? In most cases, Nike. The AI didn’t push it harder, and it wasn’t cheaper. They just know Nike. They have a feeling about Nike that’s been building for years through advertising, athletes and moments in culture, and a ranked list from a chatbot isn’t going to undo that. 

That feeling is brand affinity, and trust comes with it. But people in our industry still talk about both like soft metrics, even though what I’m seeing more and more is that when a purchase decision is run through AI, those two tip the scales. Consumers are getting more skeptical, and for good reason. There are suddenly more products that look like game changers until you click through to a site that was clearly spun up by AI. Then you start wondering if it’s safe to share your credit card details or if the product will ever actually show up. A brand you already trust doesn’t have that problem.

AI sits on top of the funnel and filters what’s already there. If your brand isn’t in someone’s head before they open the chat window, the model has nothing to work with. And the models pick up on what people do: the more a brand gets talked about, reviewed and written about, the more often it shows up in the answer. Upper-funnel advertising is what feeds it. Walled gardens are taking advantage of these dynamics. 

Visit a sneaker site, any sneaker site. Within a few hours, you’ll likely be followed around the internet by brands like Hoka, Nike, Adidas, New Balance, Brooks and On. One site visit gets packaged up and sold to the whole category. The platforms call that precision targeting. What you actually see is a wall of ads that all look alike, all chasing the same low-intent click.  

Brands are paying for a spot in the same crowded hallway as everyone else, and AI is accelerating that. 

The brands that win will own the moment before the algorithm

I’m not saying walk away from performance marketing. I’m saying get the order right. Affinity has to exist before someone opens the chat window or types into the search bar or walks into a store. The brands that win have already earned a spot in the consumer’s mind by the time any of that happens, and the way you earn it is with creative, emotion-evoking, mindset-driven advertising that reaches people when they’re open to hearing from you. A brand can build decades of loyalty through real engagement. We all know this to be true. 

Mindset matters because it’s the one thing the walled gardens can’t commoditize. Knowing who someone is gets you a user ID and a retargeting pixel. Knowing what headspace they’re in right now, whether they’re relaxed, curious, planning or distracted, is a different kind of signal, and it’s where brands can actually separate themselves.

We’ve spent years outsourcing brand-building to algorithms, and we got what you’d expect: a lot of brands that look and sound the same, competing on price, while the platform takes its cut no matter who wins.

The bill is coming due

There’s a bigger problem under all of this. The walled garden model is built to retarget and take credit. Someone spends years seeing a brand’s campaigns, across dozens of impressions and very different moments in their life. At the end they click a retargeted ad, and the platform books the sale. The lower funnel gets the credit, and the upper funnel gets cut in next year’s plan. Despite consumer skepticism, AI agents could start making more purchases for people, at which point the gap between what built the sale and what gets credit for it gets wider.

The good news is that the signals are changing, too. An impression carries a lot more than a single intent signal: the content around it, the environment, the emotional tone, the moment. AI is starting to read all of that. 

But for now, AI can amplify a brand that already exists, but somebody still has to build its reputation.

This op-ed represents the views and opinions of the author and not of The Current, a division of The Trade Desk, or The Trade Desk. The appearance of the op-ed on The Current does not constitute an endorsement by The Current or The Trade Desk.