Santiago Gonzalez on Untangling Creative Ops, AI in Marketing, and the Myth of Personalization – ML Stories

In this edition of ML Stories, we spoke with Santiago Gonzalez, former Senior Manager of Creative Operations & Marketing at Keurig Dr Pepper. With a background that spans documentary filmmaking and corporate marketing, Santiago brings a refreshingly pragmatic and brutally honest lens to creative operations. From scaling complex workflows to building smarter infrastructure for modern brands, his approach blends strategic oversight with a keen sense of what slows teams down—and what sets them free.
We caught up with Santiago to dig into the realities of operating creative at scale, the evolving role of AI, and what happens when personalization goes too far.
What keeps you up at night in creative operations?
When we think about creative—not just operationally, but from a tech standpoint—the thing that keeps me up is that everybody’s still reacting. No one’s really creating the technology. You end up building on top of a system that gets more and more tangled. Like the transit systems of New York or London: no one’s ever gone back to fix the base. It’s just stacked infrastructure.
What role should AI play in brand marketing?
Let AI do a lot of that lower funnel work. Variant generation, data scrubbing, metadata tagging—that stuff can be done with AI. We should let AI do all of our hygienic marketing and creative work and save our really talented teams for the high-level stuff. Brand relaunch? Throw the best at it. New market? Get your best brains in a room and go.
What’s a belief you’ve changed your mind about?
Everyone always said personalization is the golden ticket. But what I’m seeing out in the real world is people starting to distrust things that feel too specific. It’s not like the Coke bottles where people searched for their name. Now it’s like—how do they know it’s me? That kind of extreme personalization might actually push customers away.
Do marketing teams have enough visibility into what’s working?
I think there’s plenty of visibility. I don’t think there’s enough understanding. You can AB test till the cows come home, but we haven’t established a widely agreed-upon way to interpret the results. One test might resonate with one demographic, another with a different one—so which wins? At the end of the day, you should let data decide for you. It’s very inhuman—and that’s what it should be.
What makes a marketing team great?
The best teams I’ve been on weren’t just built around the job titles—they were built around the people. I’ve worked with editors who were in punk bands, project managers with law degrees, designers who performed slam poetry. That kind of range gives you options you don’t even realize you’ll need. It creates better ideas, better collaboration, and it makes the team feel like they’ve got skin in the game.
