Conference Transcript - ZETA 2025-08-13

Brian Schwartz, Analyst — Oppenheimer

Hello, everyone. I want to thank you for dialing in to this session of day three of Oppenheimer's Technology Conference. I'm thrilled to have with us part of the leadership team from Zeta Global. Zeta's a regular here at the Oppenhe tech conference. And with us, we have Nij Gord, he's the chief data officer, and Matt Fo, who is the senior vice president of IR. So welcome to both of you. Maybe I'll just call it a jump ball, whether it needs your Matt wants to take the first one. You know, we may have new listeners in the audience that are unfamiliar with the Zeta Global story. So can we maybe just start from a high level, 20,000 feet view? And can you please share the company background and the problems that Zeta Global is solving for your customers?

Nij Gord, Other

Matt, you want me to take this one? Zeta is a marketing technology company, and we primarily service the enterprise marketer that focuses on B2C marketing, so consumer marketing. Those marketers, since the beginning of time, have had some very consistent challenges. Their challenges are to acquire new customers, to grow customer value, and to retain those customers for longer. So if you think of the biggest brands in the world, anyone from T-Mobile to American Express to even TJ Maxx to other retail stores and outlets, they have the exact same types of challenges. Zeta is unique in the market for two very specific reasons. The first is that we allow enterprises to both acquire, grow, and retain all within one platform. And historically, this has all been done in point solutions. When the world moved to consolidating data for the purposes of AI into Databricks and Snowflake and other, our platform had a tailwind behind it because marketers wanted to do more with single source solutions, again, across acquire, grow and retain. And the second thing we do exceptionally well is we have our own data asset that helps enterprises see intelligence outside of their four walls. So what can we tell them about their business and their opportunities that they wouldn't already know themselves? This helps with acquire, this helps with grow and helps with retain. And it really positions us differently in the MarTech landscape than any other provider out there. And those have been two big sources of our growth. Obviously, we are also a pioneer in AI and marketing technology. We like saying we've been doing it for seven years, not seven months. Now it'd be longer than seven months at this point for those that had started it with the chat GPT boom. But even my background, I joined Zeta eight years ago, and my company was very focused on AI. Zeta acquired our company, and then that same infrastructure we brought here, and that has been a real boon for us in the company in terms of our ability to adopt new AI technology, incorporate it into client workflows, including our own workflows, and then bring the benefits of that to ROI that we deliver for customers.

Brian Schwartz, Analyst — Oppenheimer

Thank you for that intro. That was really good. Why don't we bring Matt into the discussion here so um man maybe we can do a look back uh in terms of the first half business highlights you just record just uh reported uh earnings last week your 2q results looked really good for the business you exceed your guidance you raised your annual targets so tell us what is working well for the business in the first half of 2025 yeah thanks brian i think there's a few things that

Matt Pfau, Head of Investor Relations

highlight and then Nij can chime in with anything I missed. I think, first of all, our agency business is performing quite well. That's both within the holdcos where we're gaining more traction with the large agency holding companies, as well as we've seen good success with independents this year as well. Going into the year, we had about three independents that were platformed. We've well more than doubled that number this year. And so that's an area where we're seeing really good traction. We also this year have put more emphasis on our one Zeta initiative, which is showing good signs of early success. We've had a lot of nice cross sell this, this year. We called out a few of those customers on our, our last earnings call. That's still early, still getting started, but good signs of initial success. And I'd also call out that the live intent integration is going quite well. And that's certainly reflected in the scaled customer account where we've had good success cross-selling a lot of live intent customers with Zeta products. And then the pipeline of deals with the programmatic email product that we've derived from the live intent assets is showing good promise as well. Nish, anything I missed?

Nij Gord, Other

No, I think that's right. Again, you know, to reiterate something you said, our ability to provide solutions to marketers that allow them to acquire, grow and retain in one platform has become more important than ever before. And those are some of our largest customers. Those are some of our happiest customers. And they gave us the greatest opportunity for growth as well. So, you know, we're seeing really growth on all of those vectors, but the one Zeta model is really coming into light here.

Brian Schwartz, Analyst — Oppenheimer

Nisha, I wanted to ask you about the data advantage of your platform, you know, versus the competition. You did talk about the comprehensiveness advantage of the platform and the all-in-one. But, you know, every marketing software company that I follow says they have the best data. So share with us how Zeta Global is differentiated versus those other suppliers from a data capabilities perspective.

Nij Gord, Other

It's a great question, and it's an interesting one because I would make the claim that most marketing technology companies don't have any data of their own, right? They have data from their customers that they do things with. You know, when I look at data assets, I think about scale and durability. So, you know, how much data do you have? How durable has it been over time? The second thing I think about is governance. Is it consented? Where is it coming from? And the third thing I think about is actionability. How closely is it tied into your ability to create a marketing outcome? We have made key investments over the last eight years and even before that to acquire data around identity, signals, and identifiers that are sewn directly into our platform. These aren't licensed data sets. The core of our data is our own. And that data is consented in our framework. And we are able to bring that into intelligence that our brands can take advantage of directly through the ZMP. So that model of allowing enterprises to see outside of their four walls becomes incredibly powerful. And we can do that from meeting number one, even before they've onboarded any of their data into our environment, we can start showing them value around acquire, grow, and retain just using the assets that Zeta has collected. And the closest model I would say this is too is not other marketing technologies or even advertising technologies. It's really close to the walled gardens because they also have scale, they have governance and they have actionability. So that model has been very successful for companies like Facebook and Google. We are the first and really only company to bring that same model into the

Brian Schwartz, Analyst — Oppenheimer

MarTech landscape. Thank you. You brought up AI. So if I think about Zeta Global, they were early evangelists for AI. So again, the differentiation question, please share with us how Zeta Global's AI is differentiated from, say, Agent 4s or those other MarTech suppliers in the market?

Nij Gord, Other

So I think the key thing right now is the first is that I'd say we're 12 to 18 months ahead. You know, AI and an integrated platform that delivers against AI has been our ethos since I've been here for eight years. We aren't a loosely coupled set of technologies. We're one central technology that AI can perform operations on. So just having that infrastructure in place puts us in a better spot to start. But we started to experiment with GPT's egenic workflows very, very early on. And so we have production use cases where you can come into the system, you can build end-to-end workflows, you can launch those workflows into activation, and then the AIs can also take advantage of optimization. So there's a little bit of a lead there. Obviously, technology changes very quickly, and we are blessed to have that lead, but leads can go away. The other thing I'd say that is even potentially more important, and we're going to double down on what we have today and the lead we have, but the data that can feed our AI models is world class. So you probably have heard the expression, your AI is only going to be and your outcome is only going to be as good as your data. The fact that we can synthesize data around identity, around intelligence that exists outside of the enterprise's four walls and use our same AI systems around workflows, around agents to then bring that into customer outcomes has been a big differentiator. And that's one that's very hard to replicate. The types of assets we have consolidated and brought under the Zeta umbrella are not readily available. The way we've done it is not easy. It's taken years, and that's going to give us a competitive advantage for years to come. But I'd say even from a functional perspective, if you have seen a demo of our platform, and I encourage everyone to see a demo at Zeta Live or in an investor meeting, if you have one set up with us, you will see is that AI is sewn into virtually every part of our operating environment. There's a, you know, a good example of this is a famous retailer that has RFP'd like 10 times in the last 10 years. They're very large. And they've never switched from their original provider, you know, their marketing automation stack, because the switching cost is just too high. you know think of all the campaigns you have to migrate the audiences the data model now with our new systems and with zeta's new system called compass it really cuts the onboarding time and work by about two-thirds or sometimes even more so this might be the year right i think we think they're going to rfp soon again it's created such a great opportunity for enterprises to take advantage of new capabilities and get in front of the curve if they've been laggards and really our AI is a powerhouse to enable them to do that. Great. Let's shift to

Brian Schwartz, Analyst — Oppenheimer

the monetization path for AI, for Zeta Global. So you do have an agentic platform, got all this AI functionality. How should investors think about a framework of the AI monetization path for the

Nij Gord, Other

business? Yeah, it's a good question. And I know it's one that comes up very frequently. In the today world, I think that we're thinking of monetization primarily from a perspective of utilization. So with AI, you're able to make decisions that are easier, faster, smarter. And that encourages you to put more campaigns into effect and really activate your marketing with greater efficacy and with higher frequency. That leads to more monetization. We've seen the utilization of our platform increase significantly from clients that have adopted our AI methodologies and tactics and tools much faster than the ones that haven't. So in our current world, we democratize the AI. We make it available because we feel like utilization is going to be the end point for monetization. There may be a future world where that changes, but in the current world, that's the rubric in which we're operating. We've seen some good success around.

Brian Schwartz, Analyst — Oppenheimer

and then um last question on ai is just uh it's topical but the pushback um you know specifically that ai it's gonna be the end of software you know because of uh ai and chat gpt and i think someone pointed out earlier hey mark benioff said that we were gonna have the end of software 25 years ago and and here we are we still have software so um but clearly it's topical chat gpt 5.0 came out last week, hit most of the software stocks that I'm covering. What is your perspective in terms of generative AI and the impact of software? Do you think it is going to be the

Nij Gord, Other

end of software? I don't. I think it's going to be more like your teammate on Turbo. That's how I would describe it. Again, I think the jobs to be done for enterprise marketing are complex. They operate across channels. They require supervision. I think AI is going to make us all better, right? It's going to make us smarter. It's going to make us faster. And it's going to make the work easier. Consumers over time have demanded more. They're going to want better experiences. But, you know, AI is a tool that helps us get there. And it offloads the busy work so that we can think about more strategic work that's required for our clients. I don't think it's the end of software. I think it's really going to be an accelerant for marketers and their ability to get, you know, the job done and provide great results to the brands they work for and the consumers that they are responsible to. So that's my personal point of view on it.

Brian Schwartz, Analyst — Oppenheimer

Thank you for sharing it as the chief data officer. Appreciate that. Okay, let's take a step back and, again, look at kind of the market here. And let's talk about the moat and, you know, Zeta Global's right to win in your market. So, you know, why are you winning? The market you compete in has been around for a long time here. So it's crowded. You've got these mega platform companies like Salesforce and Adobe, IBM. You've got all these point providers like Braze or Interval out there. So what gives Zeta Global the right to win? You know, how is how is your business differentiated, you know, against those larger providers as well as those point providers?

Nij Gord, Other

So I'll start and Matt, maybe you can chime in after. But I think this goes back to what I mentioned at the outset, which is that, you know, marketers are not looking for more tools. They're looking for better performance with the tools they have. And as part of that, if you think again about the marketer jobs to be done, acquire new customers, grow customers at scale, and retain them for longer, Zeta is a platform that operates across all of those vectors. It can use a canonicalized set of data from the brands that we work with plus our data, and it can bring new intelligence to bear. So this is just a very disruptive operating model in a, you know, technology area where things haven't changed a lot over the last 10 or 15 years, right? The jobs to be done haven't changed, but now we've finally approached them with an entirely new vector. And again, doing more with less is a key theme of what's happening today in the marketing ecosystem. It covers AI, but it also covers the solutions that you choose to partner with. And the fact that you can acquire, grow, and retain through Zeta, your intelligence can lead you across all of those vectors. We help you understand and see opportunities that you don't see yourself. These are all difference makers in our ability to not just help our customers win, but they create a big moat around our business, which has been reflected in our growth and the adoption of

Matt Pfau, Head of Investor Relations

our technologies. And it's also a flywheel effect too, Brian. The more you use Zeta across multiple use cases, the higher return on investment you're going to get. And that ties into the one Zeta initiative we have. So as Nish mentioned before, the clients that are using us for multiple use cases have a materially higher NPS score than those that are only using us for one use case.

Brian Schwartz, Analyst — Oppenheimer

Let's shift over to some of the growth vectors here for the business. I think Matt pointed out your agency business is outperforming here today. So maybe let's start there first. So, you know, if I think about the agency business, it's been around for a very long time. So, you know, why are you doing so well there? Why are agencies increasingly doing more business with Zeta Global?

Nij Gord, Other

Yeah, I can start. So I think let's dimensionalize agencies into two groups to start. So there's independent agencies. Typically, those customers of Zeta and partners of Zeta want to use all of our technology, right? They're never going to be able to make the level of technology investment that we can make. We've given them the ability to co-brand our tools and white label our tools, and they want to compete with the holdcos. And in many cases, the independent agencies are also getting into use cases outside of Acquire. They're getting into Grow and Retain as well. So they use our technologies in a very similar manner to how an enterprise would. And we give them this tool set that's, again, hard to replicate. The same reasons the enterprise selects us is the same reason the independent agencies will also be selecting us. On the other side, with the large holding companies, where we are an activation partner for them, There's a number of things happening. You know, first of all, our technologies are set up for scale, so we can give them efficiencies in delivery and operation and fulfillment that they may not be able to achieve themselves. The second thing is that we've made considerable investments in the tech to actually deploy digital marketing with our data baked in. So when you work with Zeta on an activation program, you know, the Zeta data asset comes with to help power those solutions. But at the end of the day, this comes back in all cases for independence and for the hold codes to performance. You know, we deliver industry leading performance with our technology and our data assets under one umbrella. And that's what's leading to growth and better outcomes for the business. So all of those together materialize in the growth rates you've seen. But certainly, the performance is an important part of how this all materializes as well.

Brian Schwartz, Analyst — Oppenheimer

When we shift to the publisher cloud to live intent, so you've had this business now for a few quarters. How is the integration going? And are you achieving the synergies that you're hoping for when you acquired the company? Has anything surprised you about the technology or the company now that you have it as your business and you're running it?

Nij Gord, Other

Yeah, good question. So first of all, we were always very impressed with their data asset. You know, it's an identity-based data asset. They work with the world's biggest publishers and brands. So having that in-house and unlocking all of the value of that data asset and merging it with our data asset has been very powerful. That happened very early. If you may recall, we were already partnered with them to some degree before we made the acquisition. So we were somewhat familiar with it, but then we acquired more of the data after the acquisition and brought it into our world. So that's been a big advantage for us. The second thing is on the fulfillment side. Live Intent essentially gives our platform an entirely new channel. You know, whether you want to call it inbox advertising or direct to inbox, it extends our ability to find individuals where they want to be reached in premium environments. So that model has been merged in with our deployment architecture. It's being used in conjunction with existing Zeta clients. And Zeta technology has also been extended to live intent customers. And that's around marketing automation. That's around identity resolution in the bid stream. So there's been a lot of synergy on both sides happening, but we're really seeing that flywheel around adoption from both sides and really adoption of new technology from where they didn't have it previously take hold. And we'll see more of that in the coming months and year.

Brian Schwartz, Analyst — Oppenheimer

Then maybe it's a question for Matt. Is it fair to assume that the live intent, the publisher cloud business, has accelerated in Q2 versus Q1, since that was the second quarter, since that was one of the drivers of the upside.

Matt Pfau, Head of Investor Relations

Yeah, Brian, we haven't broken out anything specifically from a revenue perspective for, you know, the publisher cloud per se, except for we did give the live intent revenue, which was, you know, related to the acquired customers from live intent and doesn't include some of the revenue synergies that we've gotten subsequently. But like I said previously, I think if you look at the customer count and the customer additions that we've had over the past several quarters, they've been very strong. And partly a factor of that has been our ability to cross-sell live intent customers and get them above the 100K threshold to become a scaled customer. So that's a piece there. And then, you know, I think the pipeline we're seeing with the, you know, Zeta Direct, the email product that we've developed with the live intent assets is very exciting as well. So it has been going, you know, very well post-acquisition, and the live intent numbers that we've broken up don't necessarily tell the whole story.

Brian Schwartz, Analyst — Oppenheimer

Okay, that's a good intro into my next question. I wanted to talk about the install base, spending the back-to-the-base motion. You get the key metric, which is your scaled customer RPO. That has continued to grow at a double-digit rate in 2Q and for a long time. And it's actually up more than 50% since the IPO four years ago. And we're talking big numbers. You know, that scaled customer RPO is now over half a billion dollars. So how should we think about penetration within that scaled customer cohort? Whether we think about consumption or from a product or maybe even a marketing IT budget basis, how long of a runway is there to continue to grow the scale to customer output?

Matt Pfau, Head of Investor Relations

Do you want to give big picture what you see within customers?

Nij Gord, Other

So I would say at the highest level, we are underpenetrated in a major way right now with the OneZeta model. We started OneZeta about two years ago. as a test and the basic idea was if we have customers that are using one two three channels with us right channels would be things like ctv or display or audio or email would they want to be doing one two or three use cases with us instead use cases are a major point of leverage on revenue and use cases again meaning acquire grow retain uh what we found and and matt mentioned this earlier, is that the customers that move from one use case to two to three end up being some of our happiest customers from an NPS perspective. So we know that there is an amazing opportunity here. Initially, we started with 10 customers that we wanted to get to $25 million in revenue by the end of 25. This was just our initial kind of goal. Now, what we found along the way is that the one Zeta model applies not just to our largest enterprises, but it applies across the board to Zeta's customer set. The majority of our customers want to be doing more with us. And so moving them from one use case to two to three requires internal collaboration, which we're working through and we're learning more about. It also requires a different way to manage the customer and a different type of customer relationship, which we've also been building out. So I'd say that the future for that model of moving from one to two to three use cases is very, very bright. You know, we have a low percentage of customers that currently do that with us, but I think we have a high percentage of customers that are either interested in doing that with us or will be interested in doing that with us just because of everything I've mentioned. So really amazing growth vector for us just within our own install base that we've identified, and we're seeing a lot of good success around in the early days here.

Brian Schwartz, Analyst — Oppenheimer

If we look back again in the first half with that scale cost margin, does anything stand out or interesting in terms of either from a use case or a product perspective, you know, that's been driving that growth in the first half of the year for that metric?

Nij Gord, Other

I think, you know, Brian. Sorry, Nish, go ahead. Oh, I was going to say one thing. I think that, you know, if you think of a large enterprise and you think about how they spend money, they could spend, you know, up to maybe five, six, seven million dollars in annual commitments for their marketing automation stack. that same enterprise probably spends hundreds of millions in media. So the opportunity to capture both marketing automation, CDP, and media give us the most leverage from a use case perspective. And some of the large jumps we've seen in ARPU are for customers that are now doing both marketing automation with us and now moving into the media space. And that gives us, again, a tremendous amount of revenue leverage and stickiness amongst those enterprises.

Matt Pfau, Head of Investor Relations

Yeah, and Brian, we called that on the earnings call. We saw from a customer growth perspective, the fastest growth was customers using four or more channels, as well as customers using two or more use cases. So I think broader adoption of the platform is certainly driving that. The other thing I'd highlight is that the number we reported does include a headwind from LiveIntent, so we didn't quantify exactly what that was. We did say in the fourth quarter that LiveIntent drove a five-point headwind to ARPU growth, so there was certainly a headwind from LiveIntent in the second quarter, as their customers are, on average, smaller than Zeta's customers.

Brian Schwartz, Analyst — Oppenheimer

In terms of the product, just thinking about when you land and bundling products, is there a common product or a common use case that you land with? Is there an interplay with the different channels and products or like certain products like messaging, data management, do they always land together? Does customer success in certain use cases or modules pull in other modules or use cases?

Nij Gord, Other

So I think generally what you'd find is there are two types of lands. One would be RFP-driven, and that usually involves marketing automation in your vernacular, messaging and data management. And then there's another kind of land that involves media, right? So media means I need to acquire customers most generally. So these run side by side. But what we find is that when we land with marketing automation or CDP, which would be RFP driven lands, they always get exposed to the media offering at the time of the land or sometime during the pitch. And that becomes of interest to them very, very quickly. They say, well, if we're using this to manage our existing customers and you have all this intelligence, why couldn't we be using this to acquire new customers that look like our best ones today? So very common to expand that way. On the media side, if we land with a pilot around media, what they often say is, wow, you have all this great intelligence. What happens if I give you my first party data file? Could you build a lookalike against it? Could you help me segment that data file into new categories? So then the media land starts turning into what looks more like a data management land. So they flow bi-directionally, but generally an RFP-based land would be around CDP or ESP or marketing automation. And then the other kind of land would be a media-based land, which can also be driven by an RFP, but the cycles around that would be much faster.

Brian Schwartz, Analyst — Oppenheimer

The last question for the session is just thinking about M&A. You know, it looks like you've had a really successful M&A here with Live in 10 and kind of expanding the platform. How do you think about the future strategy with M&A? Are you more interested in doing like a Live in 10 where you're expanding the breadth of the platform? Or would you like to go deeper into your capabilities, you know, within your different MarTech product sets?

Nij Gord, Other

So one of our core philosophies at Zeta is that anything we do needs to be able to be integrated into our stack, right? So we look for capabilities that either are going to be fully complementary to the current offering or they make the current offering better. But we're not looking for things that sit on the side or need to be loosely coupled. We think in this new world, technology needs to be integrated so that AI can operate most effectively on top of it. Nothing disruptive right now from our vantage point, but certainly opportunistic. So there's going to be a lot of AI fallout from companies that have good tech that couldn't make businesses work, let's say, in the next 24 months. that's going to be ripe for the picking, right? So there's going to be stuff there we're going to like, we're going to want to integrate to reinforce things we're already doing. And there may be some complementary, again, channels, use cases that fit well into the MarTech architecture that we'd also look at over that period of time. Anything, Matt, to add?

Matt Pfau, Head of Investor Relations

Yeah, the other thing that's important for us, Brian, is the ability to cross-sell an acquired customer base. If there is a customer base we're getting, and you can see this in live intent, where we're able to sell them Zeta's products and we're able to take Live Intense products and sell it to our own customer base. So that's another key criteria that we look for.

Brian Schwartz, Analyst — Oppenheimer

Nisha, I did get one question from the field that if you're willing to take one more, it's about large language models within the platform. And the question I'll just read it is, do you use your own large language models or do you partner with some of the other large language model suppliers in the market, or is it a combination of both?

Nij Gord, Other

It's both. So when, you know, as of about eight years ago, we started building our first large language model at Zeta, and that model is used to essentially categorize interest intent on the internet. So when you're browsing and you're part of our consented graph, and then we can understand what you're reading, what you're shopping for, and then give you scoring against that, that's internally built. Now, we've also obviously complemented that with all of the production LLMs and new foundation models that have come out from ChatGPT and Perplexity and Claude and Snowflake and others and Facebook and others. So those are available to you in the platform when you build agents. And what you can do is you can decide which model you want to use. Again, the interesting thing in our platform is those models can speak to Zeta data. They can speak to the Zeta LLM and scoring we've created for a user. They can also speak to customer data, and then eventually they get put into eugenic workflows where they're taking large percentages of someone's work that might be the busy work or the really time-consuming work, and then making it into something that the eugenic system can actually deliver against.

Brian Schwartz, Analyst — Oppenheimer

answer to your question very simply is both great um well we're out of time i want to thank nij i want to thank matt from zeta global for spending time and sharing with us a very exciting story going on uh this year at zeta global thanks brian