Every company in America has an AI budget now. Most of them have a slide that says "transformation." A healthy number have a vendor who promised them half their workflow would run itself by Q3.

Matt Egol has watched this movie before. He watched it with CRM. He watched it with ERP. He watched it with Salesforce and Adobe and whatever the shiny thing was the year your CFO bought it. And he will tell you, cheerfully and without a trace of cynicism, how it ends:

"If you don't change the work, if you don't change behavior, nothing changes."

Matt is the founder and CEO of Journey Spark Consulting and a strategic advisor to AgenticAdvertising.org. He spent 15 years as a partner at PwC, came up through Booz & Company alongside Randall Rothenberg, and studied economics at Dartmouth, where he made Phi Beta Kappa, before an MBA at Columbia. On paper, he is the guy you hire to install the future. In person, he is the guy who tells you the future is the easy part.

Technology is the easy part. Culture is the hard part. That's the whole thesis, and it sounds like something embroidered on a pillow in a consultant's waiting room, right up until he walks you through what actually happens when a company ignores it.

Matt starts with the line every MBA has heard: culture eats strategy for breakfast. His point is that AI didn't repeal it. AI made the breakfast bigger.

"Adoption is not just about getting people to learn new skills. It's about changing the way work is done in an organization. That's not new with AI. It's just that AI is placing more stress on organizations, 'cause the magnitude of change we're demanding is larger."

And before you go blaming the rank and file, he stops you. "Cultures are not to blame here. The issue is the leaders, and how leaders engage and evolve things."

That's a sentence worth reading twice, because it is not what the vendor deck says. The vendor deck says your people are the bottleneck. Matt says your people are the mirror.

Here is the story Matt and his partner Randy Rothenberg wrote up, and it is the best ten minutes in the episode.

A global entertainment company, about 3,500 people, had been working AI into its marketing organization: content, image sizing, translations, campaigns. The vision was big and, frankly, reasonable. Agentic workflows would let the company support more product launches in more countries with less human time, because updating a product on the site or expanding into a new country could eat thousands of hours of human work.

So leadership set a target: more than 50% of the workflow would be automated.

Then the people said no.

Not out loud, necessarily. They just didn't use it. They were, in Matt's words, "attached with craft to their old tools." They were "very proud of what they could do." They were worried about the quality of the work. And they were worried about the thing nobody puts on the slide: whether they were automating themselves out of a job.

"There were a whole host of reasons why pushing a mandate and introducing the technology would not get you the results," Matt says. Same story as CRM. Same story as ERP. "It's just that it's much larger scale a challenge with what companies are trying to do with AI right now."

So how did they get from "no" to "yes"?

That's below.

Matt has a line that I'm going to steal for the rest of my life, and so will you.

The industry, he argues, has spent years pointing data and analytics at exactly two things: targeting and procurement. Who to hit and how cheaply to hit them. The real opportunity, he says, is how the creative actually gets made.

"When we look at creativity versus crap, both spelled with AI, C-R-E-A-I, or C-R-A-I-P..."

CREAI versus CRAIP. Print it on a mug. Hand it to every agency pitching you 4,000 AI-generated variants by Friday.

His point is not that machines should stay out of the creative process. It's that you can use them to "spit it out and target people," or you can use them "as part of the creative process and optimize content." One of those is a strategy. The other is a content farm with a brand guideline.

And to do the good version, he says, you need a culture of experimentation, a culture of collaboration, people who are open-minded and innovative. Which brings us right back to the part nobody can buy.

I asked Matt the obvious question: if creativity and analytics work best together, why has this industry kept them in separate buildings for fifty years?

His answer is the art-and-science thing, and a little bit of tribal pride. "In the marketing world, a lot of people have prided themselves on their craft and their creativity, and they see it as the suits versus the creatives." The creative people, he says, "have pride in being the creative people, and they've resisted the use of analytics in support of their work, thinking it's a craft, it's not a science. It's actually both."

He would know. His mom is an artist. She works in printing, painting and collage, a style he calls mixed media layerism, and Matt uses her artwork on his book cover, his website and his podcast intro. She made a series of deconstructed customer journeys that he still uses in his own work.

So when he says he admires people who combine art and science, he isn't describing a LinkedIn persona. He's describing his kitchen table.

His path into consulting is even better. He started college thinking he'd be a philosophy professor. Then his philosophy professor told him, "You care too much about why people really do what they do," and sent him to psychology. His psychology professor sent him to economics, because he cared about business too. Economics sent him to consulting, with a focus on culture and change.

Three professors in a row told him he was in the wrong room. He ended up building a career out of the hallway between them.

Keep reading free with your email: how the entertainment company got from "no" to "yes," the one number Matt says predicts results twice as well as NPS, and his before-you-buy-another-AI-tool checklist, in his own words.