What's Up with Tech?

From Channel-Centric To Consumer-Centric: The OS That Could Rewire Marketing

• Evan Kirstel

Share episode

Interested in being a guest? Email us at admin@evankirstel.com

What if your marketing stack could think with you and act for you? We sit down with Christian Monberg Chief Technology Officer & Head of Product from Zeta to unpack Athena, a super intelligent agent designed not as a chat add-on but as the operating system for modern marketing. Built on a consolidated data fabric and a supergraph grounded in identity, Athena translates insights into action—turning questions into live segments, optimizing campaigns for incremental lift, and giving teams software on demand without the usual tool sprawl.

We explore how Athena’s identity-first approach unlocks trustworthy measurement, why incremental gains at the margins drive real revenue, and how “impossible products” like Advisor run always-on strategies in the background. Christian shares a real beta story where natural-language analytics didn’t just produce a report—it generated targetable audiences ready to activate in one click. We also tackle the realities of enterprise adoption: pairing probabilistic agents with deterministic governance, standards like MCP that help agents talk to services, and the cultural shift from skeptical pilots to production-scale trust.

From replatforming and consolidation to generative UI and on-demand app creation, this conversation charts a path from channel-centric silos to true consumer-centric engagement. Expect practical insights on cutting onboarding time with agent-driven migrations, simulating strategies before you spend, and building the measurement rigor that proves impact. If you’re ready to move at the speed of business, reduce total cost of ownership, and give your team more room to experiment, this one delivers the blueprint.

If you found value, follow the show, share it with a teammate who’s battling tool sprawl, and leave a quick review so others can discover it.

PodMatch
PodMatch Automatically Matches Ideal Podcast Guests and Hosts For Interviews

Support the show

More at https://linktr.ee/EvanKirstel

SPEAKER_01:

Hey everyone. Fascinating chat today with Zeta, shaking up the marketing world again with their new AI agent, Athena. And we're going to dive into what that means with Christian. Christian, how are you?

SPEAKER_00:

Doing well, Levin. Thanks for having me.

SPEAKER_01:

Well, thanks for being here. Exciting times. You guys had a spectacular launch. And really excited for anyone new to Zeta. How do you describe the company these days and what you do for marketeers?

SPEAKER_00:

Sure. Well, real simply put, Zeta is a technology company. We provide AI-powered software to enterprises to help them drive more efficient and effective marketing strategies at a lower cost. Every marketer out there is trying to work on acquire, grow, and retain use cases. And the legacy marketing clouds have left them with a lot of gaps in the way their data connects or their identity connects. And the workflows are increasingly challenging because they've had to bring a lot of different technologies together. Zeta uses AI to simplify all of that and really focuses on business outcomes for them. And that's what we hold ourselves accountable for, too.

SPEAKER_01:

Fantastic. And you had a huge launch at uh at Zeta Live. It was quite a blockbuster event in New York City. What was the big idea behind behind launching Athena at Zeta? And what was the reception like?

SPEAKER_00:

It was a jaw-dropping moment. Very few times I get to be on stage talking to uh customers, prospects, investors, um, peers in the industry, and uh see such an exciting response from all of them. Athena was one big part of what we discussed at Zeta Live this year, and Athena is a super intelligent agent. Um, it turns what we call answers into action. And at Zeta, we've got a proprietary data asset with uh roughly 250 million Americans in it, a lot of data around them, a lot of signal. But as uh many know, data is quite abundant, and answers about that data are quite scarce. Now, there's a lot of agencies and consultants out there that will pour over data and make recommendations. Um, the difference is that these answers are built into the platform in such a way that Athena can take action on them immediately for you, quite literally in one click if you'd like. And this allows brands to not only know more about their customers and prospects, but it allows them to iterate and test much more rapidly. Um, it's a platform that moves at the speed of business rather than being a barrier to moving quickly.

SPEAKER_01:

Fantastic. And with so many marketing technology tools out there, maybe talk about what Athena allows marketers to do that they couldn't do before.

SPEAKER_00:

Sure. Um, so most importantly, Athena isn't just uh a voice interface to a marketing platform. It's designed to be the entire intelligent operating system of our platform. And we really take an uh operating system approach to thinking through Athena. Much like an operating system in the old days would have infrastructure and file systems and kernels and UI shells and package managers. Athena has all of those uh uh uh systems as well. That means one consolidated data fabric, one consolidated supergraph uh with an identity space that reaches into lots of different channels. Um, even our Athena SDK allows marketers to and uh and IT to quickly build applications on top of Athena. I'm really excited about that component specifically, this idea that for a long time we have been trying to change the way we work or the way we think to accommodate the limitations of software. As we move forward with Athena, um Athena will actually accommodate us and our needs. We'll have software on demand. So, example use cases that we shared during Zeta Live were uh several different, what we call impossible products at Zeta. Um, Advisor is an example. Advisor looks at your goals and it looks at AI-driven strategies that are perpetually running in the background, and it helps you understand where there's incremental improvements in all of your campaigns and will take action for you. In marketing, so much of the revenue is actually made on the margin. It's incrementally better at each little thing you do. And Athena is looking for that. And I want to discuss briefly that there Athena isn't just generative UI, it's not a parlor trick. Um, it is the culmination of data, system processes, and technology, predictive AI, so um machine learning workbench is what we call it, which has lots and lots of algorithms that are running in the background for predictive outcomes. And it also has generative UI, both in terms of um how we communicate through Athena and also how the uh understanding and insight from those predictive algorithms comes back to the consumer and they can understand it quickly.

SPEAKER_01:

Amazing. Wow, a super intelligent agent indeed. That's kind of uh mind-blowing. Can you walk us through an anecdote or a real-world example or story of Athena turning questions into marketing actions? How does it look like uh on the ground?

SPEAKER_00:

Yeah, sure. Um, I'll take you through uh an example and uh an early beta customer we have. Um so Zeta operates our own analytics environment. It's very important that all of that is tied together and quickly available for our customers. Uh FENA allows you to talk to the analytics environment just as you would natural language to any insight environment. What makes this specifically different in Zeta's use case is that it also gives you access to the signals and answers across our entire data cloud. Um so you can start to look at your trends, the insights that come from campaigns you are running and compare and contrast that to competitors out in the space. Um you can start to ask questions like who else behaves like this on the web? What are other great prospect audiences that we can we can go and find? And because the insights uh environment or the analytics environments integrated into the core of the platform, you don't just get a report, you actually get new segments that you can activate immediately. When we showed this to a um a beta customer that we had, the email I got back, which was wonderful, said, Can I send this to my CEO now? And we said, hold on a second, this is still in beta, you know, this isn't ready for prime time, but we'd love to work with your CEO and the rest of your team to help us co-develop uh the rest of this. And Zeta has always been really focused on our customers and very proud of the work we do with them through our customer advisory board, um, through our ongoing uh in-person meetings. And our uh as proud as I am and have some pride of authorship over Athena, as does the rest of the team at Zeta, we definitely would not be here without this amazing collaboration across the industry, different verticals with our customers. Uh the day before Zeta Lab this year, we had our biggest CAB event, customer advisory board event of the year. And we had uh 50 some odd companies represented at it. All of them got hands-on opportunities with Athena and the impossible products that sit underneath Athena.

SPEAKER_01:

Fantastic. So, one challenge the industry has is the number of tools, even for a small four-person business like mine. I've I've got a couple dozen tools, which is way too many. Of course, you've seen these market uh marketing technology landscapes with thousands, tens of thousands of tools. Um, how easy it is it for a marketing team to get started with Athena? What does onboarding look like? And uh what does it mean for consolidating uh some of the tools that they have in-house?

SPEAKER_00:

Yeah, a great, great question. And Evan, in um a business like yours, and especially how technology forward you are, it's fairly easy to adopt tens or dozens of new tools and make it work yourself. And there's frustrations that come with it, but there's also huge advantages in terms of productivity and new insights for you that you can navigate easily as a small team. In enterprise software, the same problem remains, but it gets infinitely more complex between training, connectivity, governance, security. Um, and teams have been doing this, but it's left a very fragmented environment where earlier I said technology working at the speed of business. Technology is nowhere close to working at the speed of business, much less human uh creativity and ingenuity. So, what we see in the enterprise space is uh pretty massive replatforming. Um, organizations are tired of being held to the constraints of moving data from one place to another. And I'll give you an example. I was talking to uh a large enterprise, one of the biggest retailers in uh America, in North America, and um they had just to send an email, they had to lock down the process 14 hours before the email was scheduled to send because they had so many downstream dependencies. So consolidation is something that a lot of the analysts are writing about, like Forrester and others. And while this tends to ebb and flow in cycles, we're definitely in a replacement cycle right now towards uh lowering the total cost of ownership and simplifying the complexity of marketing, which is uh something we've leaned into very uh uh adamantly at Zeta, that we want to simplify the complexities of sophisticated marketing. That word sophisticated is operative. A lot of marketers are doing things kind of the same way they did them a long time ago, maybe slightly new to newer tools, but they're looking for similar outcomes as well. Um, we think that they can change the way they interact with platforms and uh mechanisms like Athena will forever change the way that uh our customers work with marketing software and further beyond the way they work with any software at all. There's entirely new modalities and affordances of software that are coming to light right now. It's a very, very uh busy time, but exciting time to be in technology.

SPEAKER_01:

Indeed, the most exciting time of certainly my 35-year career. Uh, let's talk big picture. How does Athena change the way brands think about data, you know, their data, about campaigns, about customer engagement? I mean, this is really rethinking everything.

SPEAKER_00:

Yeah. Um, so uh just go back to another uh customer story. I was visiting with one um one of the top three publishers globally. Um, and a new gentleman had joined the team, and he was looking through their entire stack uh from Soup Nuts stuff way outside of the purview of Zeta. And it was very clear that they were taking a legacy approach that wasn't going to help them drive more loyalty, higher LTV, and they're optimizing for the wrong local variables. So it's pretty common in enterprises for teams to have channel-centric strategies, and you've experienced this. You've been um sitting in the United Lounge or in the terminal, just trying to figure out when your plane's gonna take off. You just want an update that says it's gonna take off in 35 minutes or in two hours, like when's the delay gonna be over? And you're looking through text and email and in the app, and then you get an email that says, hey, enjoy 20% off if you book now to Mexico, which is the wrong message for the time. It's not consumer-centric at all. This gentleman went on to say that the approach he wants to mirror, and it's something we agree with, is that inside of highly complex enterprises, uh, organizations that have stakeholders that want a different share of voice with the customer, they should all basically put their message into, for lack of a better term, an auction. And technology should exist to figure out the right message at the right time on the right channel. When that, when it's time to go, it can look at this auction and decide exactly what to say to the customer that's right for them, not right just for the business. Now, this all needs to be managed by business margin, revenue goals, all of those things that are real constraints. But I believe AI can do that in a very uh effective and sophisticated way. The industry saw some challenges with real-time uh generative AI, although those hurdles are being solved for on nearly a daily basis right now. But I believe this time next year, it won't just be one customer that's asking for that, that the entire industry is going to move towards a much more consumer-centric approach to marketing, which we've all been promising and talking about for decades. Uh, we're just very far from getting there. Or uh the industry, I should say, is very far from getting there. Zeta's been focused on this really since 2017.

SPEAKER_01:

Amazing. And you said that now and over the next months uh is the time for AI to go from pilot to production. Uh what's pushing that beyond amazing technology and the access to LLMs and all the goodness that we experience? What's pushing that shift?

SPEAKER_00:

Um, there's a few components to it. Um, technology, like you hit you hit on, is a big one. And um, you know, we've all of us that are deep in this industry have been inventing the technology for a long time and co-inventing it, and it becomes a little bit of the zeitgeist, and then we get uh better reference architectures, better patterns uh that we can all adhere to. Um, MCP is a good example of that. My team had already built our version of MCP before it came out, which is Model Complex Protocol, a way of um um agents talking to services. Uh, when that came out and we had standards, it actually allows all of us to move much more quickly. Uh the second piece is culturally, and there's two pieces. Um, there's the organizations that are building agentix solutions. And you know, I can turn back the clock many years. There were people in our product and engineering team that were a little bit slow to adopt it at first. Uh, they had 20 years of coding the way they coded, and they're very good at it. They weren't ready to uh change process immediately. So that took for us um probably a couple of quarters to really turn the corner where everybody was a believer. Um, at Zeta, we actually used um uh a Flack channel where people shared stories, and that was the make it or break it moment. One person would finally have the aha moment. It was a lightning in a bottle kind of event, and they would go through and write their story of what they were able to accomplish using AI in a way that would have uh been much harder, taking longer before. Um, so you have to have an organizational strategy workflows, however you do it in your organization swim lanes, where you're tracking productivity across BAU, like HR and sales or customer success, as well as engineering and product tasks, including QA and observability and alerting, monitoring, and bug regressions. Once those are in place and you have a consistent, predictable uh environment where probabilistic agents are paired perfectly with a deterministic environment, which um in ERP space and then kind of a CRM and data space, it has to be deterministic. Then you can build trust with your customers. And that has taken a little bit of time to get there. Um almost everybody uh in the enterprise world has used something like ChatGPT. And it's been an awesome companion. Um, it's been something you can brainstorm with, something you can try new language out with, something that will help you organize your uh your thoughts. But putting it into a runtime environment where you need to depend on it for uh real-time outcomes took a minute. And we had to cross that trust barrier. Um, Zeta has seen that happen. And I think there's other industries that have seen that happen as well, where people start to believe in this AI and how hardened it's been. Now, those organizations have to transform their operations. And we've gone through many different iterations of this over the last 20 years of my technology career, anyways. Um today, when I share those two stories, I'm sitting down with organizations that are thinking about redefining a 20 or 30, 30-year-old process. It's daunting, but it's driven by the board, it's driven by the executive team. Top-down um goals are very, very important, but the only way that adoption is going to be found is through bottoms-up adoption. And so my team spends a lot of time understanding the person that's hands-on keyboard, the person that's going to be using it every single day to help um really get to the marrow of what would prevent them from using any AI feature. Through time-staking, detailed work bottoms up, we've finally seen a shift from skeptical curiosity to ardent adoption. And that's happening one by one right now. And like so many things in life, once you get enough momentum and enough people moving, it really turns into a snowball effect.

SPEAKER_01:

Fantastic. Uh, amazing time for those who did not have the opportunity to attend Zeta Live and see, you know, you and Tom Brady and Serena Williams. Um, what were uh some of the takeaways from you know the roadmap for Zeta's marketing cloud, what where it's headed, and maybe you can give me a peek into uh the roadmap.

SPEAKER_00:

Yeah, absolutely. Um, so first of all, my name doesn't belong in the same list as Tom Brady and Serena Williams. They were phenomenal. Um, you know, I don't think I've ever been starstruck in my life until I met Tom. Um so Zeta's organization has all in on Athena. And the Athena operating system is something we're continuing to evolve, and it really is our future. We'll talk more about it as the year um moves forward. Um, that Athena operating system gives us access to a few things. First, the supergraph that Zeta manages is rooted in identity. It's a graph of graphs, and it allows us to move from identifier to identifier to identifier very easily. Um, a lot of your listeners probably aren't in marketing, but for those that are, um, identity becomes the unlock for all measurement. And a measurement sets the proof standard. As an industry and many industries, we need to be thinking about the proof that things are working. Um, so for marketing, that's measurement outcomes from campaigns, seeing that it drives revenue. In the AI space, that's evals and observability and making sure that we have the proof these agents are working uh effectively. We need to have new decision standards, and these decision standards um change because we have more intelligence at our grasp. Um, humans have been making decisions, and there's a fine line between having too much information and too little. AI helps us synthesize and find the needle in the haystack of what signals are going to boost our decisions the right way. And there's a new UI standard. I spoke earlier briefly about generative UI. Um, and I'm so excited for uh what we're working on, which is kind of like vibe coding for the enterprise, but it's built on top of this uh, you know, OS. Um, and uh anybody in technology that's listening would understand when you have a deterministic data foundation, when you that data foundation is enriched by Zeta's data, so everything's understood through an identity space, and you build on top of that data services, APIs, MCPs, HOA, when you've got a UI kit that has clearly defined components, uh, when you've got a trust layer that has security, governance, and appliance built into it, you have the building blocks to generate whatever application you need in that moment. So if a CFO wants an application or his team wants an application to look at warehousing by geography combined with sales by geography, combined with shipment times, and they want to understand the insights of what's driving demand in those areas and what they can do to drive more demand, they can build that app quite literally in minutes. Um, and that is an entire paradigm shift. Possibly the biggest thing to happen to software since we moved to the cloud, where uh all users are no longer um slowed down by the limitations of their technology, but really uh technologies like Athena will set let all consumers free and achieve their creativity they want to bring to work, the um the compliance they want to bring to work, the data-driven decisions they want to bring to work. If that can happen in moments instead of days, weeks, or months, I think this is going to drive businesses to achieve profitability, outcomes, growths, brand loyalty at a level never seen before.

SPEAKER_01:

Amazing. Uh, what an opportunity. Uh looking ahead, it's hard to believe we're almost at 2026. Where is the year gone? Um, what what do you think will surprise marketeers most about how AI reshapes their work uh in the coming year?

SPEAKER_00:

Great question. So um I suspect uh marketers uh optimistically, I think marketers have an opportunity to do a lot more testing and experimentation in their work at a level they've never seen before, which will be rewarding. It'll require collaboration. Um, that collaboration is sorely missing from a lot of enterprises since most of us work from home these days. Um I think they'll be surprised by the number of tools that come to market. Um, Zeta has one we call it simulator. We share it at Zeta Live as well. It allows you to simulate lots of different approaches before a single dollar is spent. When you do that, you get a lot more dexterity in your brand. You start to understand uh opportunities and even new objectives that you had never really considered before. So the optimistic uh part of me as a technologist thinks that brands and marketers are going to experience uh liberation because the technology is available and is no longer blocking them. The pessimist to me worries about compliance and governance of all of these distributed tools that haven't been brought together yet. Um we believe in data federation that data needs to live in lots of places. It's never going to be in one data warehouse as much as people try to bring everything in a data one data warehouse. Um that means you need to have really rigorous governance and security protocols in place to be able to access data where it lives, bring that insight back or that materialized view back and with a platform to make sense of it. And so until this replacement cycle really takes foot for every organization and they get the executive mandate and courage to do this work, um, I think we have a long ways to go. One of our initiatives going back two years was to decrease time to value. We launched something internally we call Line, which is um agent-driven and it creates a job card for every task to be done to migrate onto our platform. Um, and every task, every job card gets an agent. Um, when we deployed it early without many agents helping it, we immediately saw a 30% reduction in migration, which is really important. And so all businesses are gonna have to think about this replacement cycle, what it means to consolidate all of these different tools. So there isn't that problem with um with security governance, compliance, and the tools they need to create to help with that migration. Agents will be an unlock for almost all of them.

SPEAKER_01:

Wow. Well, incredible insights, incredible opportunity. It's like we're living through the renaissance in marketing and branding, and uh amazing to see you leading the way. Congratulations on all the success, Christian. Evan, thanks for having me. It's been a joy. Thank you. And thanks everyone for listening, watching, sharing the episode. And be sure to check out our TV show, techimpact.tv, now in Bloomberg and Fox Business. Thanks, Christian. Thanks, everyone. Thank you.