It's Koji Ikeda. I am the software analyst, one of the software analysts here at Bank of America. I'm absolutely thrilled to be closing out our tech conference with Zeta Global. We have three here. We have Matt Pfau, VP of Investor Relations, Neej Gore. You are the Chief Data Officer, and we have a special guest.
Chris Yurek-Bauley, EVP of Data Cloud.
Yep. Thanks so much for being here, guys. I guess we should start it off with, you know, maybe for those in the room that are unfamiliar with Zeta Global and those on the webcast that may be unfamiliar with Zeta Global, what is Zeta Global? What do you guys do? What is the opportunity that you're going after?
Sure. Maybe I'll start, and Matt, you can chime in.
Yeah.
Zeta Global is a marketing technology company, and we help enterprises, mostly mid-market to large Fortune 100 enterprises, acquire, grow, and retain their customers. These are mostly consumer-facing. There are two things that we do that are very different than the way they've traditionally been done in the market. One is that we believe that marketing is converging. In the last 10 years, CMOs have made significant investments in companies like Snowflake and Databricks. They've brought all their data into one place. The reason that's great is because you can apply the AI to one canonical view of the data. Zeta believes that because you've done that and you've made those investments, it is the best for the organization to be able to acquire new customers, to grow customer value, and to retain them from one singular platform.
That is also what we're seeing from CMOs around the U.S. and around the world. The second thing that is disruptive in the way we operate is that we bring identity-based marketing to the open web. You have all seen platforms like Facebook, Google, and Amazon do really well. In those platforms, you are always identified as a person. You are logged in. You action within the platform. You receive marketing. You make purchases. They optimize your workflow the next time they see you on the platform. In a very similar way, our system relies on identity-based marketing using a data set that Zeta has cultivated over several years. We own and operate large enterprises that service consumers and publishers across the internet. Through that, we are able to see a large swath of the U.S. population and global population.
We see about 90% of the U.S. adult population on our network every month in a similar way to how Facebook or Google see the population on their networks. We can use that to synthesize intent and identity around what someone is about to do next. We can use that to help our marketers, the enterprises that we service, be better with acquisition, growth, and retention. This model has traditionally never existed in MarTech and AdTech. I know in the investment community, you guys like to think along the lanes of MarTech and AdTech. If you look at this from the view of a customer of ours, the CMO, they're very much thinking about convergence and how they can do more out of one platform across all of those lanes.
You mentioned marketing and advertising. I think these are two demand environments that are hotly debated right now, you know, for better or worse.
Sure.
You guys have been growing pretty well. How are you seeing the demand environment today, specifically to Zeta Global?
Yeah. So I'll comment, and Matt, chime in, please. I think that there's two things that are happening that are divergent. We've all been hearing that consumer sentiment is all over the board, right, in surveys and published reports on consumer sentiment. That's led to people thinking that there's a pretty unstable macro right now. Generally, they think there's an unstable macro, then one of the first budgets that gets cut is marketing. That's pretty straightforward and makes sense. What we see actually is something different. Consumer behavior patterns haven't really shifted. Demand, even in categories like retail, are especially up. They haven't changed. Our customers are still engaging with us to drive more and more opportunity than they have historically. April was one of our strongest months. It was a month that was tepid with macro volatility from the external perspective.
In the last 90 days, we've closed three of our largest contracts in the history of Zeta. There is still tremendous opportunity this year. I think what you guys see from a macro perspective, what gets represented, is not really representative of the way that our customers are viewing the current landscape to operate.
Yeah. We also signed three of our largest deals ever in the past 90 days. Even if you look at it from a vertical perspective, some of the ones that people might think would be more impacted by tariffs or macro-sensitive, for example, auto, we've seen good momentum in that sector to start the year and have signed more business in there post-Liberation Day.
It seems like there's a, and you guys see it, the divergence between sentiment and actual transactions.
Activity.
Activity. Why do you think that's happening right now? I mean, why is consumer sentiment bad? I mean, I think it's clear, but why do people keep spending? You put out your economic index, right?
Yeah. Economic index. Yeah. So we put out the economic index as to how we see the view of the U.S. economy every month. I think sentiment is down because there's a lot of news that goes back and forth, right? For better or worse, one day it's X, the next day it's Y, the third day it's back to X. That makes consumers feel uneasy. Spending patterns have not shifted. The job market's strong, you know, relatively strong. Activity around consumers and their behaviors with merchants has continued to persist through this time. I think the underlying economic indicators that we track are actually still very healthy.
Regardless of the macro environment, a marketer's job doesn't change. They need to acquire, grow, and retain. All of our products are returned on an investment base. As long as marketers are getting that return, they're going to continue to spend with us.
Yep. There's a great quote by the famous race car driver, Senna, where he said something to the effect of, "It's really hard to pass 15 cars on a sunny day, but we can certainly do it when it's raining." If you look at Zeta's track record and what happened during COVID, which was also a time of great macro uncertainty, we actually accelerated out of COVID. We took market share from our competitors. In those times, marketers looked to ROI delivering solutions more closely, and we are more aligned with those kinds of capabilities with them.
Glad to hear you're a Senna fan. I am too. He did it. I know he did that in a rainy race many, many years ago. Okay. We got the data guys from Zeta Global here. I think that's fantastic. I think the data aspect of Zeta Global is not only one of your key differentiators, but also one of the key debates out there in the investment community. Talk to us a little bit about what is the data that you guys have.
Sure.
How do you gather that data, and how do you make sure that the data quality is strong, but also compliant with the regulations that are out there?
Let's start with the last question first. We try to stay ahead of compliance in the U.S. Obviously, there was a short report last November. We held a full-day data summit in December where we went point by point and talked about our data asset. The reality is that we work with highly regulated industries like banking and telecom. We would never pass these long infosec processes if these customers felt there was something incorrect about the way we appropriated data and brought data into our environment. Truthfully, we were not as transparent as we could have been with the investment community. We tried to correct that from December moving forward. From a customer perspective, we had no disruptions from the short report. We feel like our business from that perspective has never been stronger than it is today.
I will say on the data, I'll go into it in some detail. We own two at-scale networks, and they interact with consumers every single day. One is called Discuss. Discuss is the largest commenting platform online. Imagine your favorite website. You scroll to the bottom, whether it is a commerce site or a publishing site, and you can leave a comment. You can upvote. You can downvote. You can respond to a poll. Discuss provides small to medium-sized publishers and some large with engagement tools that keep visitors on site. The value exchange there is that the publishers are generally able to monetize their traffic at higher rates, and Discuss gets access to the opted-in data that comes through those sites. On the flip side, we own Live Intent. Live Intent is the largest ad exchange serving the largest publishers in the world.
Think publishers like The New York Times, The Wall Street Journal, Groupon, Sam's Club. I use the word publisher broadly. Live Intent helps those publishers monetize their emails. When you get an email from that publisher, the ad units that are placed in those emails are being powered by Live Intent. Through that, we also get permission to use that type of data for graph building. We are not the only kind of company that uses open web data of these formats for graph building. Namely, Google does it, Facebook does it, Amazon does it, Zeta does it too. Those are our two types of data that we bring in from the two networks we bring in. We constitute three types of data from that. We constitute identity data, so the representation of a person.
We have about 200 million-245 million of those in the U.S. That's about 90% of the U.S. adult population. It is a mature at-scale graph. On top of that, we layer in signals. Signals are, what are you doing now, and what do you plan to do next? We use our LLMs internally to generate a view of your next best action that you plan to take. We also have identifiers. If we want to find you across different channels, let's say we want to send you a CTV ad, we want to send you an email, we want to send you an SMS, we want to send you a direct mail, how would we do that? That is the asset and how it comes together. That gets basically brought into our platform, and it helps our customers, again, know more about their customers.
It also helps them acquire customers at scale through our platform. They have to do it through our platform.
I've covered marketing and advertising technology vendors for a long time. And whenever I talk with customers and partners, they always tell me, "You're only good as your last ROI," essentially. So how do you maintain, or how do you think about maintaining that data asset lead for the next, call it, 5-10 years?
Yeah. So I think there's two things. The first is that these assets that I mentioned are growing. They're seeing more and more traffic that's in the U.S. and worldwide. There also are not a lot of these kinds of assets available. We get the question sometimes as to, like, "Well, why hasn't Salesforce gone out and replicated your model? Why hasn't Adobe replicated your model?" It's not so straightforward. You would need to have at-scale assets that were kind of integrated into your business operation, and you have to own and operate them to do what we're doing. This has been in the works for us for 10 years now, right? It's not an overnight thing.
Our goal is to continue to keep the graphs healthy, continue to provide publishers with disproportionate value so that the graph stays up to date, expand into geographies where we're currently not. There is a lot more we could be doing in Western Europe. There is a lot more we could be doing in Latin America. We are looking at those things right now. As new assets come up that we think are interesting, if there is an opportunity to acquire them into our graph, we will take those types of actions as well.
As you move more internationally, are the data privacy and regulations different?
They are.
Yeah?
Yeah.
How do we think about that?
Yeah. They'll be different. The types of capabilities that we can launch will be different in different geographies. You're as good as your delta between you and your next best competitor. We feel like there are models that are going to look a lot like the U.S., and there are going to be models that are going to look very different than the U.S., but they can still be significantly better than what's offered in those markets currently. Germany is a good example of a place where you're going to find something that's very different, but we can still build to a level that's even better than what's currently available there.
You mentioned two primary assets, Discuss and Live Intent as kind of your, what's the right word? Not demand, the generators of your data.
Scale. We also license some data. We partner with the credit bureaus. We bring in that data to validate and verify our data sets as well. There is some licensed data. About 80% of our data is coming from our own network, and then about 20% would be topped off from partners.
Are these two, Discuss and Live Intent, enough to drive 10 years of data differentiation, or are you guys going to come out with new ways to get data?
I'd say they're enough to drive differentiation because they are at scale and growing, and they have a maturity level right now that is very, very large. I think, as I mentioned, if there are assets that come and become available that we think are interesting that plug into our model and help us service enterprises, we will certainly take a look at those along the way.
Our identity graph at the core is what the unlock is because it connects all these disparate data points into a unified view of consumers, helping brands see really outside their four walls, right? Brands really typically only see what the interactions are within their four walls. We're able to, through identity, connect whatever signal we see from our networks or from our partners or from any of the channels that we interact with and be able to bring that view. On top, the LLMs and the AI that we have then drive those answers that help drive the marketing decisions that help bring more profitable customers for our clients.
Okay. Okay. So you guys kind of have, I think I wrote in a note before, like a three-pronged approach to attacking the market, CDP, marketing automation, and execution, right? How do we think about the differentiation of Zeta within those three prongs? What is the key hook that makes you guys different?
We have two sales motions today. One is that you're starting, or I say historically we've had two sales motions. You start with Zeta by buying media. Media means you're looking for new customers. And that's usually a share shift from vendors like The Trade Desk or Quantcast. It's usually in concert with the agency that you're working with. The agency will come to Zeta and will become a subordinate of the agency to help drive ROI. You could also come to us and you could say, "I want a CDP or I want a marketing automation suite." That's typically called owned media. In that lane, we're competing with Salesforce, Adobe, and there's a litany of smaller players that we see from time to time. Our new model is called One Zeta, and it's much more aligned with where the modern CMO is going.
The modern CMO is asking the question, "How do I acquire, grow, and retain from one place? How do I use my one set of data to inform all of those decisions?" I mean, if you could tell me, and I'm sure in your own experience, the amount of times that I have an American Express card, the amount of times I get marketed an American Express card, again, is like ridiculous right now, right? Because they do not understand. Their acquisition group does not understand that I'm already a customer on the growth and retention side. This is a problem for virtually every company that is not doing it the right way. We can help solve that problem. That is the kind of problem that we are looking to solve right now for enterprises.
The One Zeta model we're very excited about because there is tremendous revenue leverage that gets created when one of our customers moves from one use case to a second use case. Historically, we've always talked about channel expansion. I'm using Zeta for email, and now I'm using Zeta for SMS, and now I'm using Zeta for website personalization. That's not use case expansion. That's channel expansion. Use cases, I'm using Zeta to grow, and now I'm using them to acquire. Only 15% of our customers today are using us for multiple use cases. The revenue leverage you can create from moving from grow to acquire can be 5x to 10x. An example of that would be look at T-Mobile as an example. T-Mobile probably spends, this is just my guess, but somewhere between $30 million-$50 million on software every year.
Their media budget is about $3 billion a year. Your ability to tackle use cases is going to give you disproportionate revenue leverage. This is a key area of focus. We brought in a Chief Growth Officer. He was the partner at McKinsey responsible for their marketing practice. His name is Ed C. He is leading this One Zeta approach right now.
Besides the fact we can do multiple use cases, even if you look within each one of those use cases, we're doing multiple channels within there, which some of those competitors that Neej mentioned won't do. For example, we can acquire customers through email, display and video, social, CTV. A lot of the other competitors are just doing one or two of those channels.
Okay. Okay. Sticking with the data theme, can't have that without AI. You guys are attacking it in your own way with having your own AI strategy, but also customers use you for AI. Let's tackle those both. What is your AI strategy? How is that potentially driving more monetization? And then how are your customers using your data for AI?
I'll give an example of something we're doing internally that's representative of how we think of AI. This is for our internal uses. The Gap, and I've been in marketing for 25+ years now. The Gap has probably RFPed for their marketing automation stack like five times in the last 10 years, which shows you they're not happy with their current solution. They've never switched, right? The switching cost of moving from their current solution to something new would just be too high. That's because they have thousands of email templates to move over. They probably have hundreds of marketing workflows to move over. There is a tremendous amount of work and switching costs that's required to change from what you're currently doing to a new system.
That's contrasted by this idea that we're right in the middle of a replacement cycle where CMOs are looking for new solutions. For example, we have a tool called Compass internally, and it reduces your onboarding time by about two-thirds because we use generative AI to take all of the heavy lifting off of things like rebuilding templates, moving experiential workflows, moving your data model over from your previous system to the Zeta system. This has been in beta for us for probably six months. It's moving into production. We were already using it internally. Now it's going to be exposed to customers directly. It's a huge step forward in accelerating the replacement cycle. I think if you're thinking about changing, this is going to be a lever that really pushes you to do that.
Whereas in a past world, you may have said, "This is going to be too much work for our team." That is one. On the customer side, we look at generative in three lanes. We look at this idea of productivity gains. Can you help a marketer do more with the time that they have? We look at prescience. Can you help a marketer understand data and visualizations in a way that they could not do without a data analyst previously? We look at personalization. Can you actually get to the reality of one-to-one personalization, one message for one person, one channel for one person? This is where generative is going to go. We have tools that are all oriented towards this. Yesterday, we had a really prominent announcement. I am not sure if you guys saw it, but we announced something called the Answers Framework.
I'll contrast this with something I heard at the Snowflake conference this week in San Francisco. The Answers Framework that we've launched is the intersection between intelligence and action. When you can make it easy for a marketer to create intelligence, interpret intelligence, and then act on it into a marketing workflow, you've really solved a lot of their problem. Zuckerberg talked about how in Facebook, I think it was in The New York Times, he had written about how you're going to tell us your outcome and your budget, and our agents are going to do the rest for you and deliver against that. Now, that's in one channel, but that's alluding to the same idea where you can make it very easy for a marketer to take action on something that we think they should do. That's where our systems are pointed.
Most of the other capabilities that exist, and at Snowflake's conference, they have a new agentic framework through Cortex. They're talking about the interpretation of data. You can ask questions of your data, which I think is great. It's a great start. Moving that question into something that you can actually do with it is still very challenging for the marketer. That is the piece that you really have to master. That is the piece that we're unlocking with the Answers Framework. It's being released in some of our tools as of yesterday when we announced it. A litany of tools within our platform will get that between now and Zeta Live and into the future. Zeta Live is our conference in October.
I would also add that as we're getting more data, first-party and third-party into our universe, that data is all feeding the models and the AI that helps reduce, as Neej said, the time to impactful value for our clients. That drives consumption, right? The more that you shorten that time to impactful or ROI-positive impact, the more they're doing with us, the more they expand use cases and channels, and that drives the consumption up.
Is it okay if I get technical on the tech stack for you guys and data? I think with data and AI, lots of buzzwords get thrown out here and there. I wanted to dig in on what the stack looks like. Is there anything specific about the data stack, the technology stack that makes you guys differentiate it? Is it the database structure or the search functionality within it, the plugins, the APIs, whatever it may be?
Let you take this first.
Sure.
Go as technical as you want, right?
Sure.
It's happy to understand. Would love to understand that.
I think one of the main advantages Zeta was built on modern frameworks. Foundations like Snowflake that are sort of modern database systems where, as opposed to sort of the older models where a lot of the bigger legacy competitors are built on. It is a modern data platform. On top, we have our own AI framework and agentic framework that is plugging into all the leading LLM models and then some of our own internal ones that are training specifically. Obviously, the fuel for all of this is our data cloud and all the data that we feed in that completes the view of each profile of customer or prospect that feeds into the data.
I would say modern bottom stack, proprietary and plugged into all the leading LLM AI on top and then fueling them all the amount of data, the unique data that we have that we can feed this all through.
I would also add that our systems are built for streaming data, and a lot of legacy systems are built for batch data. That is important because if you want to have first-mover advantage to talk to a consumer when they demonstrate intent, you need to be able to stream that data in. You need to be able to understand what it means and then take action on it immediately, right? Streaming data at the scale that we operate is non-trivial. You need to understand data pipelines and have built out robust data pipelines requiring years of development to do that thing.
Our identity also, again, with the streaming and having the identity at the core, we're able to stitch together what's happening across all your channel investments. As you're running campaigns on CTV or social, all that stuff is coming in, getting stitched into the profile in real time. You can react based on those and adjust your tactics. These are the foundations of getting to autonomous marketing.
You mentioned two things there that I actually wanted to ask about, which were streaming data and then the customer journey across different channels. I guess specifically on streaming data, what type of technology do you use for streaming data? Is it Kafka? Some sort of version of Kafka?
Yeah, we have a version of Kafka that's been configured for streaming queues. There's probably 10 different database and database structures that are used at Zeta. As an example, Snowflake's really great for some things, but it's not great for streaming. You need to have other infrastructures in place that play nice with those kinds of technologies. Maybe Snowflake would tell you differently, but that's my opinion and our opinion at Zeta. Yeah.
On the customer journey side, I remember when we did a demo with you guys in New York about a year ago, I'd like to say. One of the things that I thought was very interesting about that demo was the ability to follow the customer across different channels, which is not something other vendors can easily do from what I gather. What is it about the tech stack that enables you guys to be able to see?
Yeah, so we have a consolidated view of identity across our channels and external channels. When you take an action on a website, when you click on an ad on the internet, when you respond to an email, when you maybe receive a direct mail announcement, all of that can be stitched back to your profile. I do not like the word follow because it sounds creepy, but we can use that to inform the way our models will work so that you can receive better marketing. At the end of the day, marketing comes down to two things. Where is someone addressable and where are they likely to respond? These are the two things you have to master if you want to be a good marketer. Zeta is exceptional at understanding both of those vectors. Yeah.
Maybe to round out the conversation here in the last few minutes, it sounds like you guys have a differentiated data asset. You have the ability to execute campaigns, and you have marketing capabilities too, which are three categories that feel very, very large. There are players in there that are large. Why is it so difficult for a vendor within one of those categories to do what you do?
I think there's two things. First of all, the marketing clouds have, in my view, lost some of their focus. You may have seen Salesforce, and they're reporting on their marketing cloud growth at like 4%, whereas we're growing significantly faster than that because we've made investments that have set us up for this new modern age of marketing. Getting back to where I started, I think that where the CMO is today is they believe in convergence, meaning acquire, grow, and retain happening in one place because that's the way they've organized their data. It is hard for a marketing company to, if they're doing acquisition, to immediately get into grow and retain. Like maybe tomorrow, Salesforce will go out and buy, I'm sorry, maybe Trade Desk will go out and buy Brays, but I don't see that happening this week.
You would have to make a transformational change to your business strategy to do it. That's number one. The second thing is building the kind of data asset that we have built over the years in the way that we've built it is hard to replicate. Not impossible, but it would take a company a significant amount of time, especially in today's privacy landscape, to do this the right way. We feel like we have a significant advantage there as well. Both of those things are contributing to the growth that you're seeing in our quarterly updates.
Is the two assets that you have that can drive your data or the two main ones Discuss and Live Intent, are those very difficult to replace? Meaning.
Yeah.
If someone wanted to get a data and asset, they would have to find another place.
Yeah, the technology itself is maybe not as difficult. Anything can be built these days, but the network they've created would be very hard to replicate. It's got tremendous scale.
Okay. I guess in the last minute here, as we think about the data asset for Zeta Global, what's the key investment priority for you guys to keep data differentiated for you guys over the next couple of years?
I think it has less to do with differentiating on the data side. It has more to do with differentiating on the outcome side. It has more to do with making sure that the generative capabilities we build can play nice with our data to generate better outcomes for customers, which inevitably is all they care about. They want higher ROI, and you need to have both of these components to be able to deliver against that.
Awesome. Guys, we're out of time. Thank you so much for being here. We appreciate it. Happy to close out our tech conference with Zeta Global. Thank you so much.