In this episode, we interviewed Mat Zucker - Chief Marketing Officer of Prophet, the global strategy and creative firm whose clients include CVS Health, Salesforce, Google and Netflix
Mat brings nearly 30 years in marketing, including executive creative roles at OgilvyOne, R/GA and Razorfish. He runs brand, communications and demand generation across Prophet's 15 offices, wrote Bronze Seeks Silver, and hosts two podcasts, Rising and Cidiot
His core argument: the AI novelty phase is ending, and now that anyone can look polished, buyers pick the firm that gives them belief - working proof, human expertise and taste
Key takeaways:
The novelty phase is ending - Mat sees people becoming choosier about accuracy, utility and security, happy to use AI as a synthesiser but not to hand it their judgement
Prophet is its own client zero - AI is improving how the firm captures, qualifies and triages inbound business, while AI-generated avatars were cut from its marketing videos because the business is built on human expertise and authentic relationships
Polish now signals software, not expertise - trust looks back at what you have done, belief looks ahead, so Mat's advice is to let buyers experience a working model, prototype or live simulation early on
Taste is the skill to keep sharp - in a world of generated content, the advantage goes to people who can tell meh from great, not just good from slop
🔗 Connect with Mat
Who are you and what do you do?
I’m Chief Marketing Officer at Prophet, a global strategy and creative firm. I lead the firm’s brand strategy, marketing operations, communications and PR, digital and content strategy, thought leadership, and demand generation efforts globally.
Prophet's 2026 AI-Powered Consumer Report found something uncomfortable: adoption went from 45% to 73% in two years, but enthusiasm fell and belief that AI would become central to daily decisions dropped 30%. What did you see in that data that the headline number hides?
The headline tells a story of reach, but the underlying data tells a story of maturity and boundaries. Chosen adoption will become clearer as people become more familiar and pick favorites. The drop in daily decisions shows the trust issue - appreciating AI as a synthesizer but not as ceding judgment as a decision-maker. The novelty phase is ending and now we’re becoming choosier about accuracy, utility, and security.
Trust is backward-looking. It answers, “Have you done this before?” Belief is future-back. It answers: “Can you take me somewhere new?”
You are the CMO of a firm that sells AI transformation, which means you have to do it to yourself first. Walk us through one thing you have put into Prophet's own marketing operation in the last year that genuinely worked, and one you killed
Yes, we are our own client zero, testing tools on ourselves before bringing them to market. I’ve been looking for juicy areas of our process to reimagine AI-enabled. One that’s working well is in our upfront inbound process to capture, qualify and triage incoming business opportunities. We’re improving both the volume and vetting of them - in terms of quality and speed as well as the support to those who will pursue the lead. What I killed? I have squashed AI-generated synthetic avatars in our marketing videos. While video automation certainly has its place and I’ve seen others do a good job with this format, using AI-generated spokespeople felt antithetical to professional services. Our business is built on human expertise and authentic relationships. Perhaps there’s a role for more video automation further down the relationship cycle and supercharge our experts behind the scenes, but it doesn’t feel right to impersonate them up front.
"Rethinking Marketing Maturity in the Age of GenAI" argues teams can now skip maturity stages instead of climbing them. For a reader running a ten-person consultancy rather than a global marketing function, which stage should they be trying to skip?
For many organizations, I would skip both ends of the maturity spectrum. There’s little incentive to do “Ad-Hoc” when the ease of deployment can get you to “Experimenting.” And then, at the other end, you can be “Strategic” realizing high impact without needing to go all the way to “Sophisticated” on most levers.
Yes, we are our own client zero, testing tools on ourselves before bringing them to market.
You have said trust is no longer enough and that buyers now need belief - conviction you can deliver an outcome there is no evidence for yet. In a market where anyone can generate a polished deck in ten minutes, what actually produces belief?
Polished decks used to signify effort and expertise. Today, they just signal access to software. As baseline polish is commoditized, belief can’t be bought with credentials or legacy case studies. Trust is backward-looking. It answers, “Have you done this before?” Belief is future-back. It answers: “Can you take me somewhere new?” The way to show up is tangible, forward-facing proof. Instead of a static case study from three years ago, maybe bring a working model, an interactive prototype or a live simulation. You don’t tell the buyer what’s possible but let them experience it illustrated and in action early on.
You spent years as a creative director. What is the one craft skill you would tell a 45-year-old consultant to keep sharp even as the tools get better at it?
Creative direction teaches and refines your taste level. We learn it through mentors, through art classes, through portfolio school, through award shows, through peer conversations. Especially now in a world of generated content, you need to be able to distinguish not just good from slop, but meh from great. You must identify - and advocate for - great.
You have hosted Cidiot since 2018, about city people learning rural life, and you describe the approach as leaving your biases at the door and being willing to look foolish. Most of our readers are senior people who are beginners at AI. What does the Cidiot method look like applied to that?
My bias moving upstate was that “city knowledge is the real knowledge.” The AI equivalent for senior folks is “the tools I mastered are the tools that matter.” The Cidiot method is just noticing that bias, setting it down, even unlearning it. You don’t need to become an engineer, you need to be willing to be the least-informed person in a conversation with a chatbot, the same way I was the least-informed person last week at a livestock auction. And be as curious as a goat.
Where can readers find you and follow your work?
Start with Prophet and follow on LinkedIn. Then if you want to get more personal, my Substack is Sugar & Spice.
AI Central Voices is where the AI Central team sits down with the founders, executives, and builders shaping AI - going behind the scenes of how they operate, what they're betting on, and where the industry goes next.
Want to be featured, or have an event you'd like us to cover? Reach out at [email protected]
