The Brands Winning the Next Decade Are Building Their Moat Right Now

The conversation I keep having with marketing leaders right now tends to start in the same place, no matter the company or the category. How do you actually reinvent the way your team operates around AI? Not just layer it onto existing campaigns, but lead with it, build toward outcomes that compound, and create a team that works differently because of what AI makes possible. 

Most are moving fast on that question. Very few are asking what I think is the more important one: what do your agents actually know about your customer?

AI won’t fix a broken customer data foundation

A year ago, most marketing leaders were talking about experimentation. Today, they’re talking about accountability. Budgets are under pressure and boards want to understand where AI is actually moving the needle, not just where it’s been deployed. 

At the same time, adoption is accelerating.

Across ecommerce and retail, 89% of retailers are actively using or assessing AI projects, but only 25% of AI initiatives deliver the expected ROI.

That gap exists because agents are only as good as what they know about the customer. Without a complete, current picture of who she is, even the most sophisticated AI decisioning gets it wrong in ways that are costly. A retargeting ad runs for days after she already purchased. A discount goes to someone who never needed one. Three channels reach the same person with three contradictory messages, none of which reflect what actually just happened. Those aren't technology failures. They're context failures, and most marketing stacks were never built to solve for them.

Knowing and doing have always been two different systems

The last decade produced extraordinary investment in understanding customers. CDPs, data warehouses, identity resolution, first-party data strategy all pushed the industry to get very good at building a detailed picture of who the customer is. But the architecture was never designed to close the loop in real time. Understanding lived in one system and execution lived in five others. Each channel made its own decisions with no shared view of what the others were doing, and no single system deciding what was right for this customer right now.

The more decisions we ask AI agents to make, the more dangerous disconnected context becomes. 

Context is the competitive moat

That's the gap agentic marketing finally closes, and the opportunity on the other side of it is significant. McKinsey estimates agentic commerce could redirect $3 to $5 trillion in global retail spend by 2030. The brands that capture that shift won't simply be the ones who deployed agents earliest. They'll be the ones who gave their agents something real to work from.

Consider what that looks like when it's actually working. A customer browses a high-margin product, leaves, and comes back two days later. The agent already knows her channel preference, her purchase history, and that she's never needed a discount to convert. So instead of a generic retargeting ad with 15% off, she gets a well-timed reminder on the channel she actually uses, with no margin given away unnecessarily. She buys. That outcome feeds back into her profile and sharpens the next decision automatically.

The moat isn't the agent, the moat is what the agent knows.

This is the era we've been building toward

Marketers have always had extraordinary instincts about how to build brands and tell stories that resonate with people. For years, though, our ambition outpaced our infrastructure. We knew more about our customers than our systems could ever act on. The understanding we worked so hard to build rarely made it to the right channel, for the right customer, at the right time. So the best thinking in the room kept getting swallowed by execution.

That's exactly what our acquisition of Blueshift was built to solve. Their team spent a decade on the decisioning and execution side of that problem, at real scale, for real ecommerce and retail brands. When you put their AI decisioning engine next to BlueConic's real-time customer profiles, the gap finally closes. Signals feed decisions, outcomes sharpen profiles, and the operational work that used to absorb all your team’s time now runs automatically, and gets smarter with every interaction.

What marketers get back

What energizes me most about this is not the efficiency gain, real as that is. It’s what marketers get back when AI is finally doing the coordination effectively: the space to do the work only humans can. The creative instinct, the brand judgment, the message that stops someone mid-scroll because it feels like it was written just for them. Agentic marketing doesn't replace that. It protects your team’s headspace to focus on exactly what drew them to marketing in the first place. Your team sets the strategy, and the system makes sure it lands on the right channel, at exactly the right moment, without anyone having to manually orchestrate it.

The brands building that foundation now aren't just going to run better campaigns. They're going to build relationships with their customers that compound over time, ones where the brand actually feels like it gets them. That's what great marketing has always been trying to deliver. Now there's finally a system built to deliver it.