For most of my career, marketing has been about understanding people. What they want, what gets their attention, what makes them trust one brand over another, and ultimately what makes them click, buy, come back or tell somebody else about it.
We have spent millions of dollars getting better at that. Search engines helped people find us. Social media helped us reach them. Data helped us target them. Reviews helped them make decisions. Somewhere along the way, we built an entire industry around trying to put the right product in front of the right person at the exact moment they were ready to buy.
I think we’re approaching a pretty significant change, though, because soon the person may not be the one making the decision.
Imagine I’m headed to Miami for three days for a basketball tournament. I’ve got games at two different gyms, meetings scattered around the city and probably a dinner or two I haven’t planned yet. Instead of opening Google, searching for hotels, checking how far they are from each gym and comparing prices, I tell an AI agent exactly what I need. Find me the best hotel under $500 a night, keep me within 20 minutes of both gyms, give me a good place for a meeting, a decent gym of its own and no ridiculous resort fee.
The agent searches, compares the options and recommends one. Eventually, I probably won’t even need the recommendation. I’ll tell it to book whatever it thinks is best. Now think about that from the hotel’s side. They spent years figuring out how to get my attention, but this time I never saw their ad, visited their website or scrolled through their Instagram account. The decision about whether they even made it in front of me happened somewhere else.
They didn’t really market to me. They had to convince the agent.
That’s a completely different problem. For years, companies obsessed over where they ranked on Google. Then we worried about where we showed up on Amazon. Then Instagram, TikTok, YouTube and every other platform inserted itself somewhere between a company and its customer. AI agents could become the biggest intermediary yet because they aren’t simply helping us find things. Increasingly, they’re going to help us choose them.
And I don’t think most companies are prepared for that distinction. If an AI is choosing my hotel, what matters to it? Price certainly matters. So do location, availability and reviews. But what about cancellation policies, customer service history, amenities, loyalty benefits, or whether the hotel’s pricing and product information are structured clearly enough to be compared against competitors?
I’ve spent a fair amount of my free time over the last couple of years trying to understand where all of this is actually going. I’ve taken three AI courses through Harvard, including work around machine learning, recommendation engines, AI agents and how AI changes management and negotiation. In one of those courses, I actually built a recommendation engine, which made me think differently about this entire problem. Once you understand how a machine can evaluate information, weigh variables and narrow a set of choices, you start looking differently at what happens when that machine sits between a business and its customer.
I see the same question coming in sports. If an athlete we represent is looking for a training facility in a city, a recovery product, a financial advisor or even a brand to work with, an AI agent could eventually help narrow those choices before a person ever gets involved. Imagine an athlete’s AI evaluating endorsement opportunities based on contract economics, audience overlap, category conflicts, brand reputation and previous athlete partnerships, then deciding which three opportunities are actually worth putting in front of them.
AI doesn’t have to make the final decision to change marketing. It only has to decide who makes the shortlist.
That’s where this gets really interesting to me. The challenge for a company trying to reach an athlete may eventually become the same challenge as the hotel trying to reach me. Before you can convince the person, you first have to survive the filter.
That means we may be entering a world where brands have two customers: the human who ultimately consumes the product and the machine helping decide which products, services or opportunities that human ever sees. Those two customers may value completely different things.
Humans are emotional. We like stories, brands and beautiful packaging. We can be influenced by nostalgia, celebrity, status and a hundred other things that have absolutely nothing to do with whether one product is objectively better than another. A machine doesn’t necessarily shop that way. It can evaluate hundreds of options against a set of criteria in seconds without getting tired, distracted or impressed because somebody famous was in the commercial.
That doesn’t mean branding suddenly becomes irrelevant. I actually think the opposite could happen. A strong brand may become one of the signals an agent uses because its human has already demonstrated a preference for it. Trust still matters. Reputation still matters. The difference is that businesses may also have to make sure those things are understandable to the technology making the first cut.
So companies are going to have to start asking a question we haven’t really had to ask before: Are we easy for machines to understand, evaluate and recommend?
Twenty-five years ago, businesses learned they needed a website. Then they learned they needed to be searchable. Then mobile. Then social. Each time, companies initially treated the new behavior like another marketing channel before eventually realizing it changed how people bought things.
I don’t think AI agents are another channel. They could fundamentally change the path between a business and its customer, including who decides which brands even get considered.
Your next customer might not be a person at all.
