Zeta Global Holdings Corp - Ordinary Shares - Class A Stock

Chris, Winnie, thank you for joining us. Good morning, everyone. Let's get started. I'm Ron Josey. I lead coverage here of the Internet sector at Citi. And look, always excited to have Zeta, the Zeta team with us today. Chris and Winnie represents Zeta. Zeta is, I think everybody, for those who don't know Zeta. Zeta is sort of at the crossroads of a lot of different enterprise and advertising areas, so to speak, right? So when it comes to data cloud, when it comes to MarTech, when it comes to marketing tech, when it comes to advertising tech, when it comes to AI infrastructure, Zeta is there.

So anyway, thank you for coming, both of you today. And what I wanted to do is maybe we'll kick off really quickly, just Chris, tell us about what Zeta is in your words and introduce yourself. And then Winnie, I'm going to come to you next because Winnie, by the way, is the Head of the Data Cloud, I believe, at Zeta. So a treat to have you with us today. We're going to get the nitty gritty on what is the data cloud. Chris, you know everything about the business. So...

And we give her an IR badge for far too much of her time that she wants to do, she's amazing. Yes. So I'll start off with where we pioneered our footprint. But as you said, that footprint now and really the application of what we can do, we're being pulled in directions that we didn't anticipate, but now kind of going their full throttle. But initially, our primary buyer was the Chief Marketing Officer. And if you put yourself in the Chief Marketing Officer shoes, and these are with the biggest brands in the world in every industry vertical, they are trying to figure out who they want to reach, how they're going to reach them. And once they do, how do I get that individual to engage with my brand?

And they have done that through the use cases of I want to keep my existing customers for longer. I want to further grow my wallet with them, and then I want to acquire new ones. If you are a marketer or a CMO, you have to apply dozens and dozens of technologies against those 3 use cases. Whereas with Zeta, as you mentioned earlier, not only do you get as part of the out-of-the-box solution of using Zeta, access to our proprietary data cloud that you can't get anywhere else outside of the walled gardens, but you also have the ability to consolidate many, many, many different point solutions and technologies that can consolidate all those use cases on a single platform, market through every single digital channel without exclusions, even through social. And as I'm sure we're going to get into today, not just solve those marketing use cases, but we're now being pulled into the office of the CFO, into the Chief Strategy Officer, into customer success and service and starting to see the application of our data with use cases well beyond just marketing.

That's great. I mean I want to stick with this for a few seconds because, Chris, you mentioned the primary buyer was the CMO. We ended this conversation on now pulled in the office of the CFO, pulled in the office of the Chief Strategy Officer and several others. This has been a massive year for Zeta, maybe 12, 18 months in the making, probably more, just given the partnerships with Palantir, of course, with OpenAI, with Snowflake. So when we emerge or call it, expand from the office of the CMO into the office of others within the organization, how do we do that?

No better example than Palantir, right? So if you think about what Palantir does better than anyone else in the world is they create a perfect digital twin of every single piece of information that sits within an enterprise, whether it's their customers, whether it's their vendors, whether it's their own internal policies and procedures, their own people-based data, they create a perfect digital twin that is hyper machine readable that then allows an enterprise to make far more effective decisions and also makes them hyper efficient.

Now bring in Zeta. Zeta has the ability to do the exact same thing for you, but everything is happening outside your 4 walls. So when you think about the applications of Zeta and Palantir together, and we've talked about having several pilots underway and working on many, many more beyond that. These are the biggest brands that are -- that love Palantir and are now proactively going to Zeta and saying, what more can we do as an extension of what we're doing with Palantir. Those could be certainly marketing use cases, and that's where we're beginning. but I think we could be additive beyond marketing. Maybe you can give some examples of that.

Absolutely. And we actually just had a question about Ford in our previous meeting.

Ford, okay.

And so I was giving some examples of how we can create more intelligence for them outside of their 4 walls. So one, from the consumer perspective, from a marketing use case is we would know when their lease date is ending. We know what competitor dealerships those people are going to. But we even know more things about the overall market. So imagine for certain locations, maybe there's a predominant Hispanic community there and perhaps those individuals prefer Spanish language. So we can give that information over to Ford to say, in these locations, you need to make sure to staff up on Spanish language speakers to be able to speak to those individuals.

So simple things like that, but it can be also more complex where, where are the next best locations to open up because we know where is their demand in terms of new automobiles, the type of automobiles, all the information around it, life events that are happening, population changes where are people moving to and out of to help inform them along with their competitor information, where there are a lot of competitors in those same areas that we would then be able to give them data points that would help them make decisions around where do they open new locations, where might they want to divest from in certain locations.

Yes. So we're doing that in a dealership franchise model. We're doing that in quick-serve restaurants. We're doing that with retail chains. We're doing that with hotels, this kind of location-based intelligence now.

And how -- so that's a great example and great to hear the other verticals as well because the question we often get is how do you operationalize the entire Zeta that's coming with you? And so Winnie, I want to get into your background in a second, but let's continue with this idea of Ford. A lot of what you mentioned were demographic information, things that might be available otherwise. Talk to us about what Zeta brings within the broader platform that the dealers and/or Ford Motor might really benefit from specifically.

Yes, absolutely. Primarily our real-time signals. So decisions are made quickly, although automotive probably taken over a more extended period of time. But we know what people are reading online. What do they care about? Is it the features of that automobile? Is it the security? We know their financial and household profile. Are they going to need financing opportunities from you? Are they growing their family? And that's the reason why they're choosing to up level into a larger automobile. So we really create the context around the consumer and the market that helps give better understanding and glean intelligence around who those users are by enriching those profiles with more than like 5,000 attributes and signals per individual within our data cloud.

So let's dive a little deeper on that. That's very helpful. Winnie, I think you've been with Zeta for 17 years.

Yes, you got your math right.

And Chris, how old is Zeta?

Barely over 17 years old.

Barely over 17 -- going on --. So one of the pioneers of Zeta, which is impressive, and congratulations on that. Let's dive and you're now in charge of the data cloud within Zeta, which frankly holds that 535 million consumer profile, if you will. And I know it's not just a profile. What I wanted to hear more is that specific living, breathing data cloud and the application? And how do you keep the data fresh? Where do you get the data from? And maybe most importantly, why is this data proprietary to you? What does Zeta bring that...

Yes, absolutely. So we think about data in 3 core buckets: identities, which are a representation of the individual identifiers, so those identifiers associated with you. Your personal e-mail connected to your business e-mail to your digital ID. And then signals, both the data at rest, but more importantly, the data in motion, things that are frequently changing about you. So across those 3 vectors, we have a multitude of sources. We have our own demand side platform, where -- DSP essentially, where the bid auction is happening across our programmatic channels. We own our own SSP, the own supply side, that's the publisher network, while we own our own, we also plug into all the major SSPs as well.

We have our own message transfer agent that enables us to send e-mails across our IPs and our domains where we send more than 6 billion acquisition e-mail campaigns per month, and that's only on the acquisition side. That's not even inclusive of the CRM side. We own Disqus. Disqus is on more than 6 million publisher websites. So imagine there's a publisher site that has a discussion board. Most likely, Disqus is powering that. We also own LiveIntent. Think about premium newsletters like New York Times or Washington Post. You get an advertisement within that newsletter. We are powering the bid option that happens -- that goes in there as well.

We have some of our owned and operated properties, and then we also round out our identity graph with some third-party partnerships where we are really trying to create more intelligence for our customers. And to your point about how do you keep it fresh and what are you adding to it? This year, we've added a couple of different sources. So one, we've added into the information voter history data. This political cycle that's happening right now. We have which elections you're voting in, what party are you registered with donation signals, where now we're able to glean intelligence. I just did a report for Georgia looking at persuadables. What issues are starting to increase over time that the constituents care deeply about. Those persuadables really care about education right now. They're carrying a lot more about disaster preparedness as well as racial inequality. Those are the key issues that need to be spoken to, to really influence those persuadables.

We're also really fortifying our international identity graph where we are creating more connections to those identities, bringing in more signal information as well. We're working with a big international airline brand where now we can tell them for their U.S. consumers, what are their destinations of interest, for their Canadian ones, where are the differences. For their French consumers, how do we better market to these different areas. And we've also really fortified our business-to-person capabilities where we've now brought in technographic data. So now we know which companies has your competitor technologies that we can help you better conquest against.

And all of these signals as well as the connections of those identifiers are continuously broken and created based on what we see. So if we don't see that connection for a while, we'll break that connection because we see that, that identifier associated with this other identifier is no longer prevalent for that individual where oftentimes, we also see new connections happening that we can add those signals in as well.

And one last one for you, Winnie, and Chris, I'm coming back to you in a second. But over the last, call it, I was going to say 3 years, let's say, 18 months, talk to us about where you have all these signals, how these signals have really evolved to sort of build this marketing platform where you can bring it all together? And did things change significantly over these last couple of months now that you partner and have all these partnerships to unearth the data to help drive business results.

Yes, absolutely. I'll say that one of our major changes was around 2024, actually. That was like one of our main pivotal moments where AI had really advanced where it could interpret images and data a lot better before it was a little bit hallucinatory. So once that change happened, we actually pivoted our road map. We paused in developing net new and our focus was integrating AI into all of our applications. We have dozens of them. So we actually created a challenge for ourselves because it was almost too much data for our customers. So by incorporating AI, then we were able to then make it simpler for individuals to understand what are the insights and opportunities for me. But to your point, the last couple of months ago, we introduced Athena to be able to talk to our data cloud applications. So you don't have to look at each application individually and get the insights around them. Athena will help you navigate to the right applications that will help answer the business questions that you have, be able to interpret the data, combine those different data sets and give you actionable insights and recommendations all within minutes.

That's super insightful just because everything is changing here and to hear how it's changing on the ground is key. Maybe, Chris, we're going to bring it back up a notch to a certain extent. And we hear some pretty incredible numbers. Multi-use case adoption rose 90%. I think in 2025, NRR reached 120%. So once you're on the platform, you typically stay on the platform, right? More customers are allocating greater spend to the platform. One of the questions we often get is just about the visibility of the model. And we kicked off the session talking about how Zeta sits within so many different clouds, whether it be the ops of the CMO or the Marketing Cloud or acquisition. Now we're going to more. But just talk to us about the visibility that you see in the model and as a CFO, what allows you to sleep at night.

Yes, it's highly visible. So if we first kind of decompose the revenue model itself, call it, 60% of the revenue model is what customers pay Zeta to license the data cloud, perform analysis, create campaigns, build audiences. And that's the recurring part of the revenue. It's obviously highly visible, highly predictable. The other 40% is when the meter begins to run, the consumption-based model for when our customers take those audiences and campaigns they created and then put them through activation strategies. Activation strategies, meaning Zeta's AI knows which individuals in that audience are most responsive to which curriculum of activation, whether it's CTV, mobile, display, video, social.

We have now 5 years' worth of pattern data with customers. As you mentioned, the last 3 years of our net revenue retention has been 112, 114 and the last year, 120. And we're through the first half of the year above our model of 110 to 115 above the high end of that. So that's good, and we like that. But what makes it predictable is we set for the most part, our customers through pilots and then they become scaled and they become super scaled and beyond. And we have data in our supplemental earnings release materials, but I'll kind of walk you through it. Within the first 12 months, an average pilot is called $100,000. In their first 12 months, our super scaled -- our customers in general gets around $700,000 in spend. In that next 1 to 3 years, that $700,000 on average goes to $1.1 million. In years 3 through 5, it goes to 2.1 million. And in years 5 plus on the platform, it goes to almost $4 million, $3.9 million.

And there's years and years and years of this data, this pattern where once we start someone on the platform, we can reliably see how their spend is going to grow. And that spend happens, as you mentioned. It's up to this point, mostly happened through adding channels. We talked about 5 or more channels being up 50% year-to-year. But now in the last, call it, 1.5 years, once the One Zeta strategy was coming into play, which is the selling of multiple use cases, retain, grow or acquire any combination, which was up 90% year-over-year multi-use case, we're starting to see bigger deals happen. So one of the data points we shared in the last earnings call was deals that were won in the quarter. The average size of those deals was up 40% year-over-year. It's one thing to look at kind of what are the deals in the pipeline. Sellers can be aspirational at times in terms of what they loan to the pipe. But at the end of the day, what closes is what matters, and those deals were up on average 40%.

So we should expect this cohort data, the 5 years just gets better and better as more years go on.

Yes. I mean you talked about customers who will leave the platform. The reality is the longer you're with us, the more you spend, the more you use the platform.

Right. And why do you think that is the case? I know we're going to get into Athena, and we're going to get all the different tools or whatever. But what exactly is changing in Zeta's approach to market that's allowing that ramp from $100 to $700 to $4 million or what have you over 5 years?

We start by lowering total cost of ownership. So the initial value proposition is test us for $50,000 to $150,000 on something you're spending millions and millions on. Let us show you how we can lower your spend by starting your marketing strategy much lower in the funnel. We can tell you deterministically who is in market at this moment, not just who you'd be wasting spend on for someone that maybe isn't credit approved or not even in market. So we'll make you hyper efficient and will allow you to reduce technology spend.

But what allows us to get bigger over time is provable value, right? So within our platform, we create a visibility mechanism to where they can measure the efficacy of their marketing spend, not just of Zeta, but handled against every other vendor they're spending with. And it's one of the best ways we go in as we say, let's put pixels down, challenge your other vendors to do the same, prove out the spend. We allow them to measure that efficacy, not just on Zeta scoring the homework, but on their own methodology. So this last touch attribution, multitouch attribution. If they want to take the data and move it to a third party like a Verizon would do and have that vendor really test out and prove the ROI, it's ultimately the return on spend that we're helping them generate that allows us the permission to go back and ask for more and to do more with them.

Measurement is one of the hardest things to prove out within marketing. And so talk to us about how you're able to prove out that measurement so you can build the ramp.

Yes, absolutely. So we basically first ask to pixel the website to be able to then track all of the traffic that's coming in. Not only traffic that we're driving in to Chris' point, be able to measure everyone else as well. By placing that pixel on the website, we can see all that referral across those different types of providers, the combination of the channels that we see people coming in under. And then we also ask for all the conversion information. So for e-commerce, would have a conversion pixel on that page, and we'd also ask for all their offline conversions to all go into the same omnichannel attribution dashboard.

From there, we can then allow them to compare and contrast different types of attribution models because there's different preferences from one client to another. We allow them to then be able to see each of the channels broken out, the combination of those channels. I was even looking and analyzing one of the reports that we have for brands where we were looking at the combination of the channels. And weirdly, we saw that if you had a combination of display CTV and online video for this brand in particular, your performance was not as good as if you just had display and CTV or display and online video. And that indicated to us, you didn't need multiple brand awareness strategies around online video and CTV. You just needed one of them because then you're creating inefficiencies when you started to introduce more.

But that's very unique to that brand. That might not be true for another brand. So to be able to see that holistic picture across their entire ecosystem and how we can connect them, I saw another brand, a huge amount invested into direct mail. Well, imagine the people that aren't engaging with you in direct mail. Let's find them in another channel where they're actually engaged, try to reengage them based on all the signals that we see here.

So we've spoken a lot about we just got through visibility and measurement, the data cloud and all the attributes. Let's move forward a little bit in talking about something that happened this year, which was the launch of Athena. And I think that's the intelligence layer, if you will. So maybe, Chris, I'll start with you specifically on Athena, we were testing it for some time. We went live in March. Just maybe bigger picture, tell us what it is. And then I won't say early, but relatively early results thus far. And Winnie, I'm going to come to you and ask you how Athena changed sort of the go-to-market opportunity when working with clients. So Chris, first to you on what it is.

We've been -- we've provided our customers with a sixth-generation fighter jet, where most are equipped to fly a Cessna. And in marketing parlay, that is an incredibly powerful platform where there is a vast amount of data. That feedback to us was it's almost overwhelming. Athena was initially created to be this facilitation tool. So how do I get our customers to use more of what's in the corners of the platform, right, really kind of take advantage of finding efficiencies and creating value. And in those early days of adoption, we're roughly, call it, 40% of our superscaled customers are now monthly active users, not just access to it.

All customers now have access to the platform across enterprises and agencies. But who is actually using it every day and what patterns are emerging and what can we glean from that in terms of our revenue funnel. And what we're seeing is that of those 40% superscaled customers that are in the platform daily performing work, over 80% of how they're interacting is conversational, whereas prior to Athena, it was all keystrokes, it was all discovered on your own. Sometimes there was our hand helping them along the way. But for the most part, it was their hands on keyboard, really doing their discovery and their own analysis and planning and building of workflows. They're now engaging conversationally, which is driving more automations.

And in our world of how we monetize, if you think about back to our revenue model, what customers pay us that subscription fee for is the creation of audiences and the analysis inside the platform. We're seeing thousands of more audiences and campaigns being created because they're now engaging in a conversational way rather than having to do it manually.

Downstream from that should be the usage component of our revenue. So once you create an audience, put it through workflow, the next logical step in working with Zeta is now I want to go reach those individuals. Now I want to go acquire them. So the early optimistic signs are that it's performing as it's intended. It's making the platform easier to use. They're doing more output as a result of that, which should down the line lead to even further activation, more consumption-based revenue.

Got it.

Yes. And I'll add on, and I'll give a real example of a meeting I recently had and how Athena played a huge part in it, but also address to some of the points earlier around like selling to the CTO, the CFO, the other heads of the organization as well. We were with an agency where the head of their Board was on the call, their CTO, their CFO. Funny enough, mostly not marketing people were on that call, and they wanted to understand our platform and the value that we can bring to them. And one of the first things I talk about is we can white label our platform, give you the technology that you typically wouldn't have as an independent agency to be able to provide services like a CDP capability to be able to onboard your first party and to glean all the intelligence and actually monetize the platform.

Once I said monetize the platform, their CFOs eyes lit up and they're like, wait, let me play that back to you. We would be able to charge for it. They could have a subscription fee for it. It could be if you spend X amount, we'll be able to then monetize in this way. So very simple things that our platform was able to help support across those different types of decision-makers. But then the next thing I wanted to show them is they're in tons of pitches. They send us a lot of different RFPs, very quick turns in helping respond to those RFPs. I showed them how quickly and easy it is to do with Athena. So they are pitching a news media brand. And so we happen to have that data in our universe.

They specifically were interested in a product line around cooking. I looked at our intenders. So in our intenders, people who are reading about recipes, who likes to cook and who likes to entertain. And for that persona, we have all the information around them. What types of entertainment do they like, what kind of cuisines do they like? What types of food delivery services are they typically using? And I basically asked Athena to help me create a persona based on all those attributes and signals. It took all that massive amount of information, things that it would probably take you days to actually analyze and put together. It gave me this very long output. Thank you, Athena, but then it's also hard to digest just in a chat window, right?

Then I had to create a PowerPoint slide for me, clear, concise, executive ready that we could then share with them that they could then share with the end client that they're pitching. So not only creating that first persona, but even subpersonas based on the signals that we're seeing and even giving them recommendations across their different product lines, which ones should you feature when based on the trends that we're seeing. All of this is Athena did within minutes. This call was only maybe like 45 minutes long. I could show them all these things that we could do that you could then show that end client that you're pitching within minutes.

So it removes a lot of the friction, makes things more smooth and you're seeing significant adoption. And we're only a quarter in, a quarter-ish.

Yes, call it a couple of quarters, yes.

And then how does Palantir fit into this equation? So we're now on foundry. We're seeing more wins, I think. Chris, just talk to us about the Palantir integration that we've been seeing? And how do you -- how does that bridge us with the benefits and the use cases or the capabilities of it?

Yes. It's multifaceted. So I think first off, integrating Zeta's data or I should say, having Zeta's Data Cloud move from AWS onto Foundry makes the implementation process with Palantir's customers that much more seamless, but it also makes our process, our speeds to which we can deliver analysis much faster. Palantir is interesting from other partnerships that you mentioned earlier in that -- so of the ones that were kind of newly minted this year, there's Snowflake, there's OpenAI. Palantir is unique in that we also, in the course of delivering this 7-year contract, created incentives around joint selling that escalate over the years. It could be a number of wins that we have, but then the ACV tied to wins.

So when we were able to put in the press release that we would envision this being at least $100 million business on a per year basis, it was actually tied to goals and incentives. So even to little old Zeta, it's obviously pretty meaningful, but meaningful even to Palantir at those levels. The go-to-market engagement is one-to-one. So it's not just Zeta sellers going in. It is Palantir sellers with our senior sellers in the company. And David is leading a lot of these right now himself, which is awesome. He's our best salesperson. And I think it's got a lot of promise. It's early days, too, though.

So walk me through the $100 million again because that's a number that was thrown out of the press release. We've talked about it quite a bit. The goals and incentive side of it all. Talk to us about what -- break that down a little bit more.

Yes. So it's a rev share model. Their take, if you will, comes off the top, so it's not a margin impact.

Their meaning Palantir.

Correct. It is a higher-margin product for Zeta. It tends to be more leveraging our data analysis, which is very limited cost to us. I think it could lead to activation down the road. But right now, it's more on the analysis side and the business intelligence side. But it is -- we didn't disclose the tiers that have to happen over the years, but it's an escalating amount of client wins per year, and it's an escalating total value of contracts per year.

Since we're talking escalating wins and contracts and talk to us about, I think we saw 2 pilots, and we have several others in the hopper.

Yes. We're working on -- we've ring-fenced 20 specific opportunities with their largest customers with the largest marketing spend. So these are all brands spending over $1 billion a year. They are emerging to be bigger than our normal pilots. I mentioned earlier, 50,000 to 150,000 -- these are kind of add a 0 to them, which is great.

Okay. Every quarter, we'll hear more about this.

Yes. I think in every earnings call, I would be surprised if there wasn't a mention for a while.

Okay. And Winnie, with the integration of Foundry, how does it amplify the cloud, the Marketing Cloud? How does it amplify Athena's capabilities? And ultimately, that's 2 questions. Chris, we're going to come back to you. With Palantir plus Athena, maybe I should talk about Snowflake and OpenAI, but with Palantir and Athena, how does it allow you to get into other areas in the organization? So if you remember the question, Winnie, talk to about how this amplifies the data cloud.

Yes. So oftentimes are Athena right now as it is with all of our data cloud applications can interpret and make recommendations based on what it sees in our data and what it can find across the open web. Now you can imagine how much powerful it gets with the contexts of the organization applied to it, knowing what they care about, their business rules, their compliance, their rules, the regulations, what are their KPIs and their goals, having that context will make the recommendations even more powerful and insightful for those organizations.

Great. And Chris, how do we expand to other parts of the organization?

Yes. meaning outside the office of...

Outside the CMO, CFO, chief strategy officer, yes.

I think it's -- we put in our supplemental deck, we wanted to show what further use cases we're being pulled into. And these were the applications I mentioned earlier, location-based analysis, customer service analysis, what new lines of business should I be moving into. All of those are natural extensions that we could also be doing with Palantir's capabilities.

How easy is that conversation? So in other words, the CMO calls up to his or her counterpart in the tech organization. Give us a little bit more there because it's not as -- I mean, I know everyone is on the same executive committee, and we're talking about the opportunities, but it's a little bit more than that.

Specific to Palantir, and if Dave were to hear what he would tell you is that these sales engagements are going faster than they normally would. I think there's immense credibility. There is a passion, positive passion for how customers love Palantir and what they're doing for them that is breaking down some of the speed bumps that normally exist when you're pitching a big ticket item. So deals are moving faster. Deals are moving wider into the organization as a result of this partnership.

We were seeing this, by the way, even within our own sales cycles. I think a lot of times, we'll get the question is, given the macro environment, are you seeing extended deal cycles? And our deal cycle has been consistent because most of our deals close in 2 to 5 months because they're pilots. They're not subject to RFPs that could be 5 to 15 months. So that's not out of the ordinary for us, but even moving faster than what we would typically see.

Got it. We have about 3 minutes left here. I want to open up for anyone in the audience, any questions. And we can think about the questions I keep going as well. So okay. We'll keep going. So I want to get to margins specifically. But before I do, we've talked about your other partnerships that were announced, the Snowflake and OpenAI. How do those fit into Palantir's Foundry and Athena's launch?

Yes. So for OpenAI, the voice that you hear of Athena is powered by OpenAI. We are also accessing some of their newer models that we can test out earlier. And then we're also actually one of our clients on the advertiser side. So we're actually testing out their advertising within their ChatGPT model. On the Snowflake side, a clean room environment enables us to get customer data faster into our platform since they're already on Snowflake, we are as well that we can then bring in that data very quickly and glean intelligence, faster path to intelligence there.

We also enable clients to run their own BI models in that clean room environment because they'll typically have their own analytics team that they want to be able to run. Anything that they think is going to be most powerful for them. And then we also provide in their marketplace a lightweight version of our customer pulse that basically gleans intelligence on brands that first-party data as well. So really trying to create more opportunities for us across their network of clients.

So as we wrap this up and we think about the core platform with the data, the new go-to-market approach, the ability for these not RFPs, but pilots to be at a larger level. Chris, talk to us just about the margin side. Because when I see this and the ability to upsell and I hear larger numbers, it would suggest margins could start to expand. So how do you think about that?

Yes. Look, I think it's all additive for sure. So if you kind of start with the top javelin throw, which was by 2030, we want to get to at least 30% adjusted EBITDA margins, which is, call it, pretty darn close to 20% free cash flow margins. Everything we've talked through today is part of that road map to get to those levels. We have consistently grown free cash flow margins even faster than EBITDA margins and EBITDA margins have been expanding every quarter and every year while we've been growing 20%. So we haven't had kind of that -- make that trade-off for growth versus profitability.

I think we can -- over the longer term, we can get anywhere from 100 to 300 basis points of gross margin improvement. I think we can continue to be very efficient in G&A. I think we can continue to be very efficient in sales and marketing and as well as R&D. We talked about last quarter, almost 90% of new code that was generated was done in a fully automated way. So each one of those OpEx areas as it relates to expense to revenue, like you saw last quarter, should continue to see strong efficiency gains that get us from, call it, this year, a little north of 22% adjusted EBITDA margins on our path to 30%.

Okay. Look forward to continued updates on that market because there's all these new changes to the market to the business. So I think we're at time. Thank you, Chris and Winnie, for joining us today.

Appreciate it.

Look forward to next time.

Awesome. Thank you.

All right. I think we are ready to get started. Thank you, David, Chris, for coming. I really appreciate you guys taking time out of your busy schedules to be here.

It's our pleasure to be here. Thank you.

Yes. And with that, we can get right into it. So marketing was one of the first applications of the Zeta platform. You've increasingly talked about the company more as this intelligent AI infrastructure asset. How do you think about what Zeta can become over the next 5 years beyond the core Marketing Cloud?

That's a great question. Trying to keep my voice here. I've been talking -- you've had me talking since 8 a.m. this morning. When you think about the evolution of our business, you have to go back to some of its foundation. When we originally started the business, we started the business as a business intelligence company. We actually just focused on marketing because it was a $1 trillion TAM. And it's been a very, very positive thing for us.

As we think about the evolution of our business, it really revolves around intelligence. We think there's going to be 2 types of software companies. There are going to be software companies that create intelligence and have proprietary assets, and then there are going to be workflow management tools that they're going to have problems in the future because you've got to be able to take data, create it, synthesize it and then make it actionable. And that's really what I think we do best at Zeta.

When you look at our core marketing business, today, we return between 600% and 700% return on every dollar spent through our platform in marketing. Now what we're seeing with the adoption of Athena, and we gave some updates for Q2, but we're seeing faster adoption than we expected, and we're seeing better return on investment than we expected. And when you think about Athena, I think Athena is really the future of our business. And not to get too esoteric, but if you go back, humans in sort of our current form have existed for about 400,000 years. For almost all of those years, we've communicated via voice, right?

And then you look at some of this enterprise software, most individuals use between 3% and 5% of the capabilities of their enterprise software. I always sort of joke with this type of an audience, what percentage of your Bloomberg terminal you actually really use, right? So Athena supersedes the creation of the keyboard of the 1950s, which I think has created meaningful friction between humans and technological adoption. I think we're now at 83% of people utilizing Athena are utilizing it from a voice perspective versus a text perspective. And we're seeing, as I said, a substantially higher uptick in sort of adoption and a meaningful increase in return on marketing spend for enterprises that have adopted Athena.

At Zeta Live, we're going to make 3 of the biggest announcements we've ever made. They will be the evolution of Athena to answer your original question, where we start helping enterprises to run their entire businesses, where it's not just customer acquisition, retention, monetization or business intelligence. It's the actual seamless integration and superior operating of their core businesses by using Athena and our Data Cloud.

I think one of the focuses of you has been getting more enterprise -- big large enterprise wins. And I think one good example of that is Gap. Gap selected Zeta as part of their broader marketing stack transformation. So what's kind of changing in the market that's causing some of these large enterprises to reconsider their incumbent marketing platforms now?

Well, first of all, they made us their system of record. So I'm not sure we're part of the ecosystem. I think we're the core of it. But what I would say is 2 things have really changed. One, AI has now become so prevalent, but most enterprises don't know how to adopt it in a way that creates return on investment. That's what Zeta does for a living. Because we're able to create such a high return on investment, enterprises that in the past might not have worked with us are now saying yes.

The other issue is we're in, I think, the largest sort of replacement cycle of existing marketing and technology stacks I've ever seen. And I thought it was peaking a year or 2 ago, and it's actually now accelerating. And it's accelerating at a time where I would tell you 5 years ago, nobody had any idea who we were. I started every meeting with Zeta who, meaning who are you and why are you here? We then moved a couple of years ago to sort of why Zeta? They know who we are. They know why we're in the room, why are we going to pick you? We've now crossed over to Zeta now. It's really become enterprises feel they need us to be a part of their tech stack or they're not going to keep up with their competitors.

Now we'd like to get the must-have Zeta. That's sort of the holy grail for us. I'm not sure we'll ever get there. But that happening, simultaneous to the opening of testing with AI for return on investment, simultaneously to the evolution of our brand, simultaneously to this massive upgrade cycle for technology and marketing clouds, I think, has been the driver, as you know, better than most. Last quarter, we grew the business 44% top line, 56% EBITDA growth, 73% free cash flow growth. And we all know that numbers are outputs. The input is, is the business working. And our business is really working right now. I actually -- I think it's working better than at any point we've ever operated the business.

Yes. And then data has been part of Zeta's differentiation for a long time. As customers increasingly bring their own data in through platforms such as Snowflake and Palantir, how do you think about the role of Zeta's data moat, I would call it, evolving?

Shockingly, 2 of our biggest announcements this year, right? In the first 6 months of this year, we announced that we are amongst OpenAI's first enterprise partners that Athena was going generally available, both for the enterprise and the agency. We announced our strategic partnership with Snowflake and probably the largest deal we've ever announced was the Palantir partnership. Both those companies are incredible companies. We've, in fact, re-architected all of our Data Cloud on top of Foundry, and we've adopted Palantir's ontology to help onboard data and streamline it between us and enterprises.

But in no cases, does our data go into their platform. They don't see it. It's the same thing with Snowflake, unless we're using a clean room sort of a no-copy, which has been a massive win for us to be able to go in no-copy and go joint to market because it's accelerated the time to test. It's lowered the barriers to testing, and it's now got clients saying yes, much, much faster. Our entire business is based on enterprises giving us their first-party data, but merging it with our data. and that's where the magic happens. So the fact that they're able to move their data faster through Snowflake, Palantir is a massive win for us because they're partners of ours. We don't expose our data even to the enterprise client.

So when we combine the data, we remove the personally identifiable information and we replace it with the Zeta ID number. That does a number of things. A, it allows us to import the 5,000 to 7,000 incremental data elements that our clients might not have on their end users, demographic, psychographic, credit card transactions, credit scores, all the way through. And not only does it allow our algorithms, which we've built internally -- we've built all of our own inference-based models, it allows the algorithm to get smarter, smarter, smarter.

But if the enterprise were to ever fire us, they would lose 100% of that knowledge. It never transfers to them. I think that's one of the reasons we had a net retention rate of 120% last year and why we continue to have a much higher net retention rate than I think we projected to at 110% to 115%.

Yes. I have to spend a little more time on AI. So Zeta offers prebuilt agents across campaign creation, audience building, analytics, quality assurance. Which agents are gaining the most meaningful production usage rather than just experimental engagement? And then how do you think about the value that those individual agents provide versus what Athena provides as, kind of, that aggregate?

Well, I was going to say #1 by far is Athena. Like we saw 40% of our super-scaled clients, right, adopt -- our monthly active users Athena which -- and they're doing things that we've never seen before in the platform from a scale perspective. It's -- they're building audiences. They're looking at new technologies. They're looking at new activation capabilities at -- I mean, rates we've never seen.

So I think Athena is really going to be sort of the most powerful one. As it relates to other agents, they're all subagents that then sort of work together. So audience building agents then lead to activation capability agents that then lead to agents that help you lower your churn. All of them sort of begin to compound and work together. The other thing that's been really cool from my vantage point -- I'm a nerd. So the fact that I think it's really cool might not mean others think it's really cool. But the uptake of our build your own agent has been astronomical.

So we're on our third iteration of our AI agent studio, where our clients can now use Athena to voice enable build their own agents, to build and sort of handle tasks that they otherwise would have had to do manually. You go back to the average enterprise user uses 3% to 5% of most enterprise software. The more agents they're building, the more they adopt Athena, the more they're able to use the entire capabilities of the platform and the data ecosystem, the better the return on marketing spend they're seeing, the more they spend with us, the higher our net retention rate. So it's sort of -- it's a nice funnel.

And then to go into that metric that you were talking about, 40% of super-scaled customers that became monthly active users of Athena within like 4 months of launch. So how do you think about what's reduced the friction of adoption relative to maybe some of the other enterprise AI products that we've seen that are taking a bit longer?

I think David mentioned it's the voice enablement. And of those 40% active users, we -- what we know is that they're growing faster than the rest of the set of super-scaled customers. But -- and David made kind of a slight mention to it, but just to give you a sense for how our revenue funnel works, customers pay Zeta on a license and a subscription basis to effectively use the Data Cloud to create automations around audiences and campaigns.

Of those 40% monthly active users, 83% of the engagement is voice. And what we're finding is creating thousands more audiences and campaigns, which if you now think about downstream from that, roughly 50% to 60% of our revenue is that subscription to the Data Cloud and the platform. The other 40% is the monetization through consumption. That consumption happens once you create an audience and put a campaign workflow in place, then the software of Zeta helps you understand deterministically which individuals in that audience are most responsive to CTV, display video, to mobile, to e-mail.

Once that curriculum of workflow begins to operate, that drives the meter of usage with them. So it's an early positive indicator in terms of the intent of the Athena platform is to create a facilitation tool, more easy to use and access more of the platform and it appears to be working.

Yes, we're already seeing it flow through. But it's still early days, and I think we'll see it continue to flow through. I mean, listen, the business is really working right now across the board. When we set out to grow the business, we really look at it normally like how do we get to -- half our revenue growth is from existing customers and half our revenue growth is from new customers. And it's certainly been much higher than we expected on both fronts. We're adding more customers than we expected to that are spending substantially more money than we originally expected them to do that, which is a good combination when you're running a business.

And why do you think it is specifically about voice that is kind of changing things for your customers? Like what are they not doing on the traditional interface that they're not doing with the voice interface?

Yes. So I think the big thing there is we didn't just create a voice interface to navigate the platform. We actually did that a few years ago. We built this platform called ZOE. I think this is standard for...

Zeta Opportunity Engine.

You're so good at the acronyms. We have so many acronyms -- I name them and then I forget. We have more acronyms now than I can keep track of. But ZOE was effectively a voice enablement to better just navigate the platform. Customers who adopted ZOE spent 250% more than customers that did not adopt ZOE, and she was rudimentary at best.

The big difference with Athena is, Athena has been built so you just have to tell her the outcomes you want. You can literally tell Athena, I'd like to create 2 million incremental customers this quarter, and I'd like to lower my cost to create customers by 15% while doing it. She will then navigate the entire platform, the entire Data Cloud and, in real time, show you what you need to do, to do that. You can then say, great, Athena, let's start with the test of $500,000. Could you e-mail me the reporting every hour on the hour because I'm going to be out of the office because I want to see the return on spend here, activate now, and she will activate for you. She will then send you the e-mails every hour on the hour.

I don't want to get too ahead of what we're announcing. But ultimately, we believe that you should be able to access Athena through any UI, not just through ours. So you'll be able to go into whatever platform you're using. I'm going to get in trouble for this later with the tech guys. But the reality is you're going to be able to talk to Athena and say, great, the test is working, let's go live. And that's where I think it's going to be very, very interesting.

And I know you guys love to talk about ROI. So how do you think about the incremental revenue opportunity you're getting from this higher usage that people are getting with Athena versus the cost that you have to absorb to support higher AI usage and kind of what it implies for Zeta's longer-term margins?

Well, I think it's interesting to note that last quarter, just -- it's not your exact question, but as you know, I'll answer the question I generally want to versus the one you're asking me. But I'll start with last quarter, 89.6% of all new code generated by Zeta was generated on an automated basis. And yet, to answer your question, less than 1% of our revenue total as a company was spent on token utilization. Well under 1%.

So when you think about it, we've built workflow management tools starting with Spade, rolling to [ Loom ] to allow for workflow management to the best foundation models at the lowest token utilization while doing automated QA before it goes to an architect who then pushes a button and makes it generally available. Our ability to have built the vast majority of our AI internally. All of our inference models are homegrown and home built. So as we scale as a customer, our AI utilization expense has not gone up at all. In fact, I would say it's gone down as a percentage.

Yes, more or less. Over the last few quarters, I think it will continue to go down. We were able recently to cut a very interesting deal with one of the foundation models that will allow us to scale very, very large at a fixed cost.

So you talked a little bit about this earlier, but Athena is kind of evolving from just answering questions to really executing actions and recommending things for your customers. Where are you seeing customers being more comfortable delegating decisions to Athena versus where are they still kind of a little hesitant and want more human in the loop?

I think that is a great question, which is -- not to say your other questions weren't great. But the reality is that we are seeing a lot of exploration with Athena. I mean, clients are really comfortable exploring in their data sets, in our data sets, in the merged data sets, where we're not seeing everything yet is activate now. We're seeing them spend more time making sure that they're double, and triple checking the numbers. The accuracy levels have been at the five nines, to be clear. And we're even sending in some of our sort of FTEs to help clients really understand the accuracy level of the outputs from the explanation -- exploration, I should say.

I think -- I equate it to when we all started using the Internet. You were probably a little young for this example, but I say that as a compliment. You lost 90% of transactions on the credit card entry page. People were terrified to put their credit card into a transaction. I mean, I lived through this. I'm sure a lot of us had for many, many years. Today, you just -- who cares, right? You put your credit card in, even if it's a site you've never been on, you just put it in and you know it's going to be fine. And if it's not, your bank will cover it and so on and so forth, although most of us go back to Amazon and just push the buy now.

We're seeing a similar thing with Athena. She's getting them to that last page. They're looking at it. They're doing the activation, but they're double, and triple, quadruple checking it before they're going into the activation mode. I think that will continue to scale.

As consumer discovery is kind of shifting more towards AI-generated answers, and we have this conversation internally a lot about agentic commerce and the implications of that. What role do you think Zeta can play in helping brands remain visible and kind of influencing their own demand?

So let me start by saying I don't think the HTTP layer of the Internet is going away. I do think commoditized products, and I shuddered to say commoditized and then travel because we have so many clients in that vertical. But you know what an airline ticket is, right? You don't really need to go to the website to see the 4 or 5 different offers. You can go agent to agent in travel and that type of thing. I don't think my wife is ever going to go to her agent and say, pick me out a black dress for Saturday night.

I also think that as AI begins to -- or continues to proliferate, managing the HTTP layer is going to become less and less expensive because you'll be able to auto generate your code for that HTTP layer. So I don't think it's going away. I think there'll be more agent-to-agent commerce. I think that you'll see smaller publishers have challenges in a post-Gemini world. We're already seeing that in a meaningful way, and yet our Data Cloud is growing exponentially. Why? We've launched what I think will be one of the biggest GEO platforms in the world. Today, we integrate from a GEO perspective into Claude, into ChatGPT and into Gemini. So across the whole platform.

And many of our clients are using it. I think GEO is going to be sort of the next SEO. That doesn't excite me that much. I mean, it's real money, but it doesn't excite me. What excites me is the data extraction we get because we're now getting more data out of the GEO extraction than we are in, I would guess, the bottom 2 million of the 5.2 million publishers who use us on a daily basis. And I think that's going to continue. I think that enterprises are trying to figure out new ways to market in a post-Gemini world.

And I think our business has been a big beneficiary of that as we've expanded and grown exponentially in connected TV, in messaging, even e-mail continues to grow at a very rapid pace. So I think -- once again, I think the agent-to-agent commerce is good for Zeta. I'll remind you, we have a big e-commerce business. And I think that our data extraction will continue to actually grow, but it will grow in different places than it's grown over the last 10 years.

That makes sense. The Palantir relationship -- already helped Zeta engage some very large enterprises, and you've talked about that. But I wanted to get more into the tech stack side of things. What does that combined product kind of enable Zeta to do that kind of couldn't do before or provide independently?

The Palantir partnership is changing the game for Zeta. From a technological perspective, we took our Data Cloud and completely rearchitected it on top of Foundry. That did 2 or 3 major things: A, it allowed us to onboard customers much faster. B, it's making our Data Cloud much faster because we've adopted their ontology. And C, I think the go-to-market, the 7-year go-to-market arrangement is going to change the game for us from not just direct to their clients.

I'm already seeing the halo effect in other components of our pipeline, where we've -- I mean, today, I think 24%, 25% of the Fortune 500 use us, 51% of the Fortune 100 use us in every single onboarding of a new enterprise. Bar none. You have a data privacy workflow, you have a data security workflow and you have a data use case workflow when you have as much data as we have. I have never been through a more arduous workflow than the creation of the contract and then the merging of the platforms with Palantir. They looked at things that I've never even considered. And I do think people forget that when you have 500 -- what do we say publicly?

525 million (sic) [ 535 million ]

So 525 million (sic) [ 535 million ], we have a lot more than that. But say publicly, 525 million (sic) [ 535 million ] people who have opted into our Data Cloud, multiply that by 5,000 to 7,000 incremental data elements per person, multiply that by the trillions of signals that we're extracting from the 5.2 million publishers and the GEO clients. And you get into sort of some of the levels of the data that we have in our Data Cloud.

I think at our peak period, we process -- we do 7,500 computations per millisecond, multiplied by every moment of the day. So from a Palantir perspective, we actually are managing that outside of our pipeline. I'm personally managing that. Elias Davis, who's one of their top guys and one of the best humans that I've ever known, who helps to run their commercial business reports directly to Alex, and I have been attached at the hip. I've said publicly that we've gone into 3 meetings together. We've had 3 yeses.

I have never seen a group of clients who love their technology provider more than enterprises love Palantir. It is amazing. And by the way, they go in and say, this is our solution to help lower your cost for marketing and CRM by 50%. We think you should test this very quickly and then scale with it. And they're like, okay. So it's been great. But as Chris has said -- Chris, why don't you talk about the pipeline numbers for Q1, Q2? Is that okay?

Yes. So we've kind of setting context in the first quarter, our sales pipeline was up 40% year-over-year and accelerated to 60% year-over-year. That is independent of the Palantir share opportunities that are being managed separately. So it's to kind of keep good frame of view into the kind of the core outside of the partnership ecosystem.

But what's interesting about the pipeline, you can have aspirational goals if you're a seller and put what you want in the sales pipeline. We're not lost on that. But if you look at the deals that have been closed and the pattern of deals being closed, David made a mention of that, but I think it's a good evidence point of the replacement cycle from an IT perspective is the deals that were closed in the second quarter were 40% bigger than a year ago.

And if I look at the sales pipeline, the value of those deals, it's greater than 25% in terms of average deal size. So I think to us, that says that our sellers are doing an even better job of showing our prospects and our customers how much more they could be doing with Zeta. Other points of evidence for us of that is they're using more than one use case. That statistic as of last quarter was up over 90% year-over-year, and customers spend roughly 3 to 5x more as they build on use cases. And then customers that are using 5 or more channels was up over 50% year-to-year. So it all nicely hangs together when I look at kind of what does the sales pipeline project into the outer periods, which is why I think we have such good visibility into the business and confidence in our guide.

And I would say that -- I publicly said that I think Palantir can be $100 million a year in business. I think as I sit here today, that seems very low compared to what we're seeing.

And then on those pipeline numbers, obviously, your quota-carrying headcount is growing quite a bit slower than that right now. So can you speak about a little bit of the individual pieces of like individual rep productivity that have come in better than expected and some of those drivers?

By design. So over the last 4 years, we've -- from 2021 through 2025, we were adding roughly 22% more quota carriers per year. When we released our most recent long-term model, we said we did not believe we needed to add that many quota carriers to generate the similar levels of growth that we had been, that we felt like we could effectively hire half as many. And last quarter was a good example of that. We grew quota carriers by 11% year-over-year. So what makes that possible?

But we grew the business 44%, just to put it out there.

Yes. So sales productivity is really obviously quite strong. What makes that possible is we have 2 fast-growing parts of our business, those that go directly to enterprise and those that are now, call it, in its fourth year of maturity, selling directly into agencies. And that's roughly approaching 25% of our revenue, not quite there yet. Our model for agencies is a one-to-many. So to scale like we've been scaling, we do not have to add and throw resources at growth. We can do it in a very efficient way because we have high quality there. And as we get new agencies and new independent agencies, we add headcount, but we can grow within an account very efficiently.

On the enterprise side, it is more of a one-to-one, right? Because we divide the sales team between hunters and farmers. Those are going after new logos and existing customers, which is why half of our quota carriers are deployed equally, which is why half of our growth has come from new customers and existing customers in the last 4 years or so.

And I would add to that, the quality of salespeople we've been able to attract over the last 5 years is unlike any people we ever thought we would get to work with us. What -- salespeople are motivated by success, right? And what we would find is we would beat Salesforce, Adobe, Oracle in an RFP. And literally within 60 days, the head salesperson would call us and say, we want to be with you. So the ability to bring in substantially more senior people and really build behind them has allowed us to do that. We also have a core group of people who close a lot of deals, and we tend to be out there a lot.

And something that I think David and the sales team and our President, in particular, has done well over many years is you look at our revenue where it's diversified. We serve 15 different industry verticals, and there's no concentration in any 1, 2 or 3. But we, years ago, went away from this generalist salesperson. This idea that you can walk into Goldman Sachs on Monday and United Airlines on Wednesday and even halfway sound like you understand their business model, let alone how they use data to go market and go acquire customers. And that led us down a path that we are going to go hire industry-level experts, and that's also selling.

And by the way, we just brought in one of the top health care guys in the world to really focus on health care because we think that's a vertical we can really, really grow in.

You've spoken a lot about the outperformance in revenue that you guys have been able to achieve over the past couple of quarters. How do you think about the split of that outperformance across customer additions, ARPU expansion, [indiscernible] good cross-sell, new partnerships? And how do you think about those drivers also going into next year? And where do you feel the most confident?

For investors, we give you the model. So you don't have to dissect it every quarter, and we don't have to be kind of weird in how we talk in code. We give you our long-term model, but underpinned by that, we tell you what are those metrics that we must be on, not just every quarter, but over a multiyear period to be getting there on time or faster or if we're behind, it will be very clear. We're ahead in all metrics, and that is how fast should we be adding customers? Our model is to add between 4% and 8% more customers year-over-year every 90 days, we've been averaging 15%.

We want to grow their spend with us between 12% and 16%. We've been averaging at the high end of that. Net revenue, we want to drive between 110% and 115%. We've averaged over the last 4 years, 115% plus -- last year it was 120%. So when I look at the building blocks of what allows us to guide like we do with the stated level of conservatism that we say, we say our guidance is going to have embedded in it 2 to 5 points of conservatism. That gives us headroom on all those metrics to effectively be at the low end to get to our number. And if there's upside, it tends to mean we're at the high end.

Yes. I mean, our net retention rate has been running much hotter than we expected, which is always very nice, right? When your existing clients are spending an average of 20% more in a given year and your stated goal is to grow organically by 20%, that gives you a good head start on that. So that's why we've been growing.

I think last quarter, we were at the Rule of 68 as it related to growth plus operating margin at 67%, 68%. And we've been, I would say, pleasantly surprised by that. I'm not surprised we're adding more customers, and I'm not surprised we're adding -- growing the ARPU. It's just the net retention rate has been a really powerful metric. And even in the last 2 quarters, we've been above the -- we don't break it out, but you can do the math. We've been above 115%.

Makes sense. I'll try and squeeze one more in. We started this conversation talking a little bit about the proprietary data advantage that you guys have. As you kind of expand the use cases of your platform and as AI has kind of changed a little bit, maybe what are relevant advertising channels these days, how do you see the data advantage becoming more valuable to customers? And how have you, kind of, adjusted the business?

Yes. I think that's a great question. What I would say is our data moat is the biggest moat we have in our business. You've got 20 years of building or buying platforms that create trillions of fresh data points per month. We're not relying on an old antiquated database. We're literally creating new data every second of every day.

And in a world where effectively the whole open Internet is being ingested by foundational models every moment of every day. We made a really conscious decision a few years ago to never sell our data or expose it to a third party. And I'll tell you, during the few years that we were re-architecting the business and we went from making real money to nothing for a few years, I was pretty honored, and I know Chris was too, and Steve Gerber. A few weeks ago, a Harvard case study published on Zeta and the decision we made and they centered it around the Board meeting where we pivoted the whole company.

And there were years where the private equity guys were like sell the data, sell the data. We're like, no. And I would tell you the fact that we've never exposed our data to any large language model ever and will not is mission-critical to our long-term moat and our ability to do what really matters. It's not about the data. It's not about the AI. It's about creating actionable intelligence.

And if you're not creating actionable intelligence, you're going to fall by the wayside as a technology company going forward. And that's why I think our pivot into becoming an intelligent AI infrastructure company, launching the business intelligence use cases and some of the big launches we have coming soon are going to be so game changing, not just to Zeta, but I think to the businesses that choose to license our technology.

Yes. Well, thank you so much. Please join me in thanking [indiscernible] thank you so much.

Thank you. Great job.

Good day, and welcome to the Zeta Global Second Quarter 2026 Earnings Call. Today's conference is being recorded. At this time, I'd like to turn the conference over to Trey Campbell. Please go ahead, sir.

Thank you, operator. Hello, everyone, and thank you for joining us for Zeta's Second Quarter 2026 Conference Call. Today's presentation and earnings release are available on Data's Investor Relations website at investors.zetaglobal.com where you will also find links to our SEC filings, along with other information about Zeta. Joining me on the call today are David Steinberg, Zeta's Co-Founder, Chairman and Chief Executive Officer; and Chris Greiner, Zeta's Chief Financial Officer.

Before we begin, I'd like to remind everyone that statements made on this call as well as in the presentation and earnings release contain forward-looking statements regarding our financial outlook, business plans and objectives and other future events and developments, including statements about the market potential of our products, potential competition, revenues of our products and our goals and strategies. These statements are subject to risks and uncertainties that may cause actual results to differ materially from those projected. These risks and uncertainties include those described in the company's earnings release and other filings with the SEC and speak only as of today's date.

In addition, our discussion today will include references to certain supplemental non-GAAP financial measures, which should be considered in addition to and not as a substitute for our GAAP results. We use these non-GAAP measures in managing our business and believe they provide useful information for our investors. Reconciliations of the non-GAAP measures to the corresponding GAAP measures, where appropriate, can be found in the earnings presentation available on our website as well as our earnings release and other filings with the SEC. With that, I'll now turn the call over to David.

Thank you, Trey, and welcome to the team. We are very excited to have you. Good afternoon, everyone, and thank you for joining us today. I will start with the headline. We delivered our 20th consecutive beat and raise quarter, delivering the rule of 64 as a company and the rule of 49 excluding M&A. And in the second quarter, our year-over-year revenue growth accelerated to 44%, up from 35% in the second quarter of last year. This is the result of Zeta increasingly becoming the system of intelligence for our customers. The second quarter performance once again shows this strategy is working.

Second quarter revenue was $443 million, representing year-over-year growth of 44%. That's up 28% year-over-year, excluding M&A revenue. Adjusted EBITDA was $92 million, up 56% year-over-year with margin expanding 170 basis points year-over-year to 20.7%. And based upon this strength, we are once again in the midpoint of our 2026 revenue guidance by $33 million consecutive [ beat and race ] quarters reflects more than just strong execution. It is evidence of growing and durable demand for our platform that turns data into intelligence into decisions and decisions into measurable outcomes.

Investors have historically viewed data as a market technology company. Marketing is where our platform was first applied and where we've built our leadership position, but that description no longer fully encapsulates who we are today. Zeta has evolved into an intelligent AI infrastructure platform. Marketing is our first application, not our limit. Our platform combines proprietary data AI, workflow automation and activation into a single operating system that helps enterprises make better real-time decisions and take action. The foundation is our data cloud, built on proprietary data covering more than 535 million individuals globally, trillions of signals and 20 years of model tuning.

And to be clear, this data is owned, not rented because nothing you rent can be a moat around your business. Athena serves as the intelligence layer, allowing customers to interact with that foundation using natural language and embed AI directly into their everyday workflow. It truly allows customers to focus on outcomes versus navigating the platform. Our activation platform then turns those insights into actions across channels while continuously learning and getting smarter via the results. This has contributed to the strength of our net revenue retention. And with data, business intelligence, AKA, ZDI, we are also extending beyond helping enterprises to acquire, grow and retain customers to a fourth use case, enabling organizations to transform business and customer data into intelligence, insights and real-time actions.

To be clear, this is not static dashboards or old school business intelligence used to explain what happened. VBI helps predict what happens next and act on it in real time. In major sports and entertainment company uses ZDI today to understand its fans share of entertainment spending and engagement across live events and streaming. This helps quantify the value of current distribution partnerships identify the most attractive future streaming relationships and strengthen its negotiations at renewal. A leading energy drink brand uses ZDI to quantify the incremental value it drives for retail partners, including customer growth spending and long-term value. This helps demonstrate the brand's impact at retailers and identify opportunities to deepen distribution and partnership investments.

Taken together, these capabilities position us as the intelligent AI infrastructure layer that sits at the center of enterprise decision-making. This transformation has been accelerated by 4 strategic catalysts. First, is our OpenAI partnership. Our collaboration with OpenAI validates the AI strategy we have been building for years, enhances Athena's capabilities and creates new commercial opportunities. As foundation models continue to improve and become more widely available, we believe the durable advantage will come from what's built around that proprietary data governance, workflows and decisioning systems that turn intelligence into outcomes. Those decisioning systems are our own inference models, smaller, purpose-built models that we have been training on the Data Cloud since 2017.

OpenAI powers Athena's voice, the decisions run on our models and no large language models ever touch the data in our data cloud. The second is Athena. Launched earlier this year, Athena is fundamentally changing how customers interact with the data platform. Athena is true voice enablement and fully conversational instead of navigating dashboards or conducting manual analysis, customers can simply ask for outcomes. The platform knows their business decides, acts and learns from every single result. That dramatically lowers the barrier to adoption and expand use across teams, channels and use cases.

We have seen that users who interact with Athena through voice exemplify this rapid adoption. Over the last 60 days, Athena's voice users are interacting with the platform at a 500% higher level than non-Athena users, we are also realizing the benefits of AI within data. In the second quarter, 90% of our new code generated was automated helping our teams to innovate faster and continuously enhance the platform for our customers. The third is Snowflake. We deepened our existing partnership with Snowflake and we now have over 100 shared customers, reinforcing Zeta's role in the enterprise data ecosystem, where customers increasingly want to connect data environments with intelligence and activation.

Zeta is uniquely positioned not only to help customers store and analyze data but to act on it. And the fourth is Palantir. Our partnership with Palantir significantly expands Zeta's enterprise opportunity. Palantir provides the oncology, governance, an enterprise AI infrastructure that large organizations require. Data contributes proprietary customer intelligence, proprietary data, identity, decisioning and activation. And we have reached 2 important milestones in this partnership already. Our Data Cloud was fully integrated with Foundry as of July 31. And we have already received multiple agreements for our initial combined sales with several other meaningful opportunities in FLYHT.

Together, these four catalysts are accelerating Zeta's transformation and strengthening our position as the intelligent AI infrastructure layer for the enterprise. We believe the defining characteristic of an infrastructure platform is not just innovative technology. It is a platform that becomes more valuable, the more customers use it. every decision that powers makes the next one better. And because those decisions run on our own inference models, not rent and tokens, we make millions of them a day at infrastructure economics.

That is the flywheel driving our results. and we're seeing it play out across our existing customer base. The leading telecommunications provider recently expanded its relationship with Saba into real-time personalization. Adding grow and retain use cases to its existing footprint. It is a powerful example of our land-and-expand model, start with one use case, prove value and expand across the customer life cycle. Athena is beginning to accelerate that dynamic to give investors greater visibility into its impact.

Chris will provide more detail on our initial framework for measuring key metrics like revenue contribution adoption and usage later in this call. We'll continue to expand on our reporting around these metrics in the back half of this year. GAAP is a powerful example of our platform evolution. As part of its broader AI-led transformation GAAP selected data to help architect its next-generation marketing stack with Athena, at the center of how customer data, decisions and execution come together. Our role extends well beyond powering individual campaigns.

We are helping GAAP to remove silos unified decision-making across its iconic brands and build a more intelligent marketing engine that learns, adapts and acts in real time under a multiyear agreement as GAAP system of record. That is what it means for Zeta to become the transformation agent for the enterprise. This momentum was also evident at Cannes Lions, which served as an important proof point for the growing market interest in Zeta and Athena. We launched Athena for agencies, hosted more than 115 executive meetings and delivered 37 live Athena demonstrations resulting in a record sales pipeline coming out of the event.

The conversations we began in can will continue at Zeta live. Where we will provide an even deeper look at the future of Zeta and intelligent AI infrastructure. Zeta Live will be in New York City on October 8 and I'm excited to announce we already have 2 incredible headline speakers, Olympic Gold Medalist entrepreneur and founder, Lindsey Van, an entrepreneur and global superstar, Kevin Hart.

Zeta Live has always been our opportunity to show customers, partners and investors where we are going next. This year, we expect to introduce the next generation of Athena continuing the extension of Athena as a super intelligent agent, answering questions and powering workflows to a system of intelligence that knows, decides, acts and learns. The opportunity in front of us is no longer just about monetizing marketing. It's about helping enterprises turn fragmented data into intelligence into decisions and decisions into measurable growth and cost savings in their businesses. The past year has been an inflection point for Zeta, bringing together capabilities and investments we have been building for years. What we are still in the very early stages of this opportunity.

AI is changing how enterprises operate, how software is consumed and what businesses expect from their technology platforms. Posada, that creates an opportunity to expand our role from helping customers execute marketing programs to become the intelligence layer that enables them to move faster, make smarter decisions and drive better outcomes across the enterprise. We continue to be the disruptor in this new ecosystem. As always, I want to thank our customers, partners and shareholders for their continued support and to Team Zeta, thank you for your hard work, your commitment and belief in what we are building together. Now let me turn it over to Chris to discuss our results in greater detail. Chris?

Thank you, David. I'll echo by welcoming Trey to the team. As we announced last week, Tray is perfectly suited to help lead Veta's evolution into an intelligent AI infrastructure company. and I'm thrilled to see that fall step up to lead FP&A for Zeta. In my conversations with investors, one thing is clear: a new framework for investing in companies is emerging. Investors are prioritizing their time with companies gaining share and delivering durable, predictable growth. They're increasingly screening for companies generating free cash flow and positive GAAP earnings and they're ultimately backing companies with defensible AI moats proven by results.

Data embodies each of these characteristics, and the second quarter's results make this evident. Q2 was our 21st straight quarter of greater than 20% revenue growth, excluding M&A and political candidate revenue. Our increased second half revenue guidance continues that trend. Q2 was also our highest-ever free cash flow, paired with positive GAAP net income, showcasing the quality of our earnings and driving the largest full year guidance raise to free cash flow and GAAP EPS in our history, and we're listening to shareholders.

We're introducing an initial framework to measure adoption and monetization of Zeta's AI, one underpinned by revenue. because moats are ultimately proven by how long customers stay and how much more they spend over time. I'll cover all of this in detail, along with updates on our pipeline, sales productivity and drivers behind our increased Q3 and full year guidance. Let's start first at the top line, where Q2 revenue came in at $443 million, up 44% year-over-year or 28% excluding M&A. That beat our guidance by $23 million or 5%, driven by faster growth in both superscale customer count and ARPU each exceeding the growth rates in our 2028 model.

Superscale customers grew to 197, up 17% year-over-year, more than double our 4% to 8% 2028-mile growth rate driven by demand for Athena. Customer gains were especially strong in consumer and retail, telecom and health care. Superscale quarterly ARPU expanded to $1.8 million, also up 17% year-to-year and above our 12% to 16% long-term 2028 model. And we're seeing some interesting usage dynamics unfolding ARPU. First, them engagement is increasingly voice first. 83% of customer interactions are now spoken. Reinforcing our thesis that natural language will become the primary interface for marketing and business intelligence use cases and has the propensity to drive higher utilization on the platform.

Second, the One Zeta sales initiative is gaining speed. Customers using more than one use cases are up 90% year-over-year. Customers using 5 or more channels are up more than 50% year-over-year. Cross-sell and upsell deals won in the quarter were up 43%. And we saw double-digit revenue growth across e-mail, CTB and social as well as double-digit revenue growth across all 3 marketing use cases, retain, grow and acquire. And third, demand for the platform was broad-based across industries. Eight of our top 10 grew more than 20% year-over-year on a trailing 12-month basis with consumer and retail, financial services, automotive and health care all accelerating from last quarter.

Also notable in the quarter was the significant expansion of the sales pipeline and strength in seller productivity. At the end of Q2, the total sales pipeline was up more than 60% year-over-year. and up over $100 million compared to just 90 days ago. On a per seller basis, pipeline creation is up more than 100% year-over-year as 1 data and Marigold cross-selling take hold. This is driving higher average contract values on deals won in the quarter, up more than 40% compared to last year. And overall deal sizes in the pipeline increased more than 25% year-over-year. driven by higher attachment rates across channels and use cases, all while quota-carrying head count increased by just one versus last quarter to 198, up 11% year-over-year.

That gap between pipeline growth and head count growth speaks to the strength of sales productivity. And importantly, this growth and revenue upside came with impressive operating leverage. Specifically, we generated $92 million of adjusted EBITDA, up 56% year-to-year at a margin of 20.7%. And an increase of 170 basis points versus last year and $5 million better than the midpoint of our guidance. Marigold restructuring actions and integration savings drove total data expense to revenue ratio efficiencies across R&D, G&A and sales and marketing, improving 30180 and 250 basis points year-over-year, respectively.

That cost of revenue came in at 41% as expected. This was 10 basis points better sequentially and 300 basis points higher year-over-year, reflecting strong social channel adoption by agencies. Cash grew even faster than adjusted EBITDA in Q2. We Net cash provided by operating activities was $69 million, up 65% year-over-year, with free cash flow of $58 million, up 73% year-over-year, a margin of 13.1% and free cash flow conversion of 63%. We also generated positive GAAP net income in the second quarter of $8.2 million compared to a net loss of $12.8 million in the same quarter last year. resulting in GAAP earnings per share of $0.03.

In the second quarter, we prioritized using cash to repurchase shares, deploying $29.9 million to buy back 1.6 million shares. And year-to-date, as of July 30, we've spent $74.6 million on share repurchases with approximately $89.4 million remaining on our authorization. Solution in the quarter was just 0.1%, and we remain on track to hit our normal course net dilution target of 3% to 4% for 2026. Finally, we closed a new $1 billion credit facility, including a $250 million Term Loan A and a $750 million revolving credit facility that remains undrawn, giving us the capital flexibility for M&A, share repurchases and disciplined investment.

To that end, we've been investing in AI for nearly a decade. So today's introduction of metrics showing how AI adoption drives deeper platform usage, longer customer relationships and higher net revenue retention is not new. It's just more visible than ever. super-scaled customer adoption of AI is ramping nicely. Since the theme is launched for enterprise customers 130 days ago, more than 40% of our super-scaled customers are already monthly active users. And together, they've generated thousands of campaigns using Athena and customers who comprehensively adopt our AI using it for audience creation, activation and other means contribute a disproportionate share of revenue It shows up in several ways.

Across our total customer base, including pilots, proof-of-concepts and scaled customers, the 20% of customers who have comprehensively adopted our AI tools account for roughly 70% of revenue. Among superscale customers, the 50% who have comprehensively adopted our AI tools drive 75% of superscale customer revenue. And these AI super users grew 4x faster and the 80% of customers still early in their AI adoption journey. But it's not only AI adoption that's ramping. It's also leading to longer customer relationships. Super-scaled customer relationships now averaged 56 months, up from 48 months a couple of years ago, and that's based on data going back to 2018.

In addition to longer customer relationships, we're also seeing them spend more, generating higher net revenue retention among AI adopters. Customers who are most comprehensively adopted our AI tools have a year-to-date net revenue retention that is 400 basis points above overall data. and more than 20 percentage points above customers still ramping in their adoption of our AI, showing that adoption is translating into stronger, more durable expansion. With tailwinds from AI adoption, higher sales productivity and strong operating leverage, we're once again raising our top and bottom line guidance for the third quarter and full year.

In doing so, we'll maintain our typical conservatism, building in a 2% to 5% cushion that assumes minimal go-get revenue from partnerships and uses 2028 model growth rates for customer and ARPU growth. To be clear, that new customer additions and ARPU growth exceed our 2020 model growth rates, driven by rapid Athena AI adoption or newly announced partnerships that would put us towards the high end of that 2% to 5% range. For the full year 2026, we're increasing the midpoint of our revenue guidance by $33 million to $1.818 billion representing growth of 39% or 25% year-over-year, excluding M&A and political candidate revenue. For the third quarter, we now expect revenue of $471 million at the midpoint, up $10 million from our prior guidance. representing growth of 40% or 23% excluding M&A and political candidate revenue.

You'll note, we're continuing to maintain our original second half guidance for political candidate revenue in Q3 and Q4 of $7 million and $8 million, respectively. For adjusted EBITDA, we're increasing the midpoint of our 2026 guidance to $405 million. up $8 million from our prior guidance, representing a year-over-year increase of 45% and a margin of 22.3%, up 90 basis points year-to-year. For the third quarter, we now expect adjusted EBITDA of $115 million at the midpoint, up $3 million versus our previous guidance.

We are also increasing the midpoint of our 2026 free cash flow guidance to $255 million, $20 million higher than our previous full year guidance, representing year-over-year growth of 55% and a margin of 14% and a conversion of 63% and tracking well towards our 2028 conversion target of 65%. Finally, we're raising our full year GAAP EPS guidance to a midpoint of $0.10, well above our prior range of $0.02 to $0.04. It's worth noting this guidance increase excludes the impact of a potential onetime tax benefit from the release of a valuation allowance that has a reasonable probability of occurring later this year which could represent additional material onetime upside.

With that, I'll close where I began. Demand for Zeta is robust and durable and we're expanding our platform organically and through partnerships to accelerate share gains. That durable demand and record pipeline improves our visibility, giving us the confidence to raise guidance across the board. That visibility supports disciplined investment and our balance sheet is well positioned to support our growth priorities and focused execution drives greater profitability and cash generation. which has us pacing ahead of our long-term data 2028 model.

With that, I'll hand the call back to the operator, so David and I can take your questions. Operator?

[Operator Instructions]. And our first question comes from Jason Kreyer with Craig-Hallum.

Congratulations on another great quarter. Chris, you had talked about one and how that's accelerating the adoption of multiple use cases. I think you said 90% growth in multi-use case customers in the quarter. What do you think that looks like over the next several quarters? How does Athena continue to optimize customers and maximize both channels and use cases?

Great question. And thanks, Jason. The short answer is I think we've got continued runway because of One Zeta and the Marigold cross-selling continue driving both multiuse case adoption, which as you said was up 90% year-to-year, but also 5 or more channel usage was up 50% and cross-sell and upsell deals within the quarter itself, was up 43%. If you look at the trend for ARPU growth over the last couple of quarters, it's been nicely above our 12% to 16% long-term model, and I see that continuing.

And we're seeing Athena really just the tip of the spear it's just getting started. It's one of the single most powerful tools we've ever had for cross-selling against use cases and channels, Jason.

Maybe a follow-up for you, David. On the DBI, can you just talk about how you go to market with existing customers. Like what does that cross-sell conversation look like? And what are the key functions in BI that are really speaking to customers.

It's interesting because we've had customers that have been using the ZBI for a few months now before we announced it because we really wanted to get use cases in and tried it. So the use cases I gave in my prepared remarks are real-world use cases that they are paying us to develop and create with them. What we're seeing is that customers no longer want static business intelligence. They're not looking to have large numbers of people come in, put data into a user interface, spit out a static report. What they're looking for is how do they make real-time decisions, how do they better negotiate on contracts that they're influx on like broadcasting for sports networks. How do they figure out how to better invest marketing dollars and develop retail partnerships like with energy drink companies and so on and so forth. So we're actually getting pulled into the ZBI use cases more than we've been selling them in, and we're now sort of doing a combination of both together.

Jason and investors on the call, we've included slides in the earnings supplemental to David's point, to bring those existing customer BI use cases to life. It's actually Slides 32 through 38 and David mentioned there's dozens of customers spanning real estate intelligence, customer experience, market sizing, loyalty growth, business expansion and business measurement, just to name a few.

We'll go next to Matt Swanson with RBC Capital Markets.

Great. David, we talked a lot about that transition from data who is AY to data now. Can you just talk a little bit about when you get names like OpenAI and Palantir and Snowflake, probably about as good of a trio of partners in this day and age you could find how that helps you in your kind of go-to-market process of skip the evangelizing but also what it can do for initial deal sizes?

Yes. Thanks, Matt. I would tell you that it has been game-changing for us. I mean the two engagements that we signed with Palantir happened in real time. the Palantir guys walked us into two of the largest advertisers in the world. And literally, we walked out with some of the largest test case and use cases we've ever had as a company based on the fact that they chose us to be their marketing platform partner. And if you listen to Alex's remarks yesterday, they are over-indexing on partnerships.

One of the interesting things that I don't think people understand is that the Palantir relationship will be one of the biggest we've ever done because there isn't a boardroom today that doesn't know who they are and isn't interested in use cases that they can do with them with us as their marketing partner, it's become an incredible entry point. And if you look at the number of customers they published just yesterday, that is the number of marketing opportunities we effectively have in partnership with them.

Add into that, the OpenAI relationship, which has been one of the biggest we've ever signed and the Snowflake evolution, we are seeing more at bats than ever. if you saw the 17% growth in super scaled customer count, that's not even the beginning of what we're going to be doing with these new partnerships in the evolution from Zeta Who to Zeta Now, these are game changing even when we're not in the room with the Palantir OpenAI and Snowflake people. When we're out there, we're coming into meetings now and everybody is -- the first thing I get is, wow, how did you get that deal done? And I have to say to them which one, and we all laugh and sort of move on. But I think this is sort of as they said in the movie, Casablanca, the beginning of a beautiful friendship.

That's great. And then I think we touched on it a little bit when you were talking about the BI and the different uses for data. But when you talked about how your platform is just currently being utilized for marketing, but it's not necessarily a marketing platform. How far do you think some of those adjacencies can go in terms of use cases? And have you seen within any of your larger customers, any use cases that really made you think about product development?

Yes. Well, every day, we're getting asked questions by our largest clients, can you help us fix this problem or answer this question in a way that we've never thought of and it's causing just an incredible amount of excitement. I think it's important to note that if you look at the first quarter of this year, we were super proud to announce that using our internal workflow management tool spade, we were able to literally automate 75% of all new code generated. If you look at the update, the actual number for Q2 was 89.6%.

I think we rounded it to 90 in my prepared remarks. But this is changing the game for how we take interaction with clients and turn it into product development almost in real time. And if you look at this very large sports league that is using us, we were looking and we were in the room with them and their CMO and they're like, listen, how do we better negotiate our streaming rights with the two partnerships that are coming up for bid and how can we equate viewership and mine share for their they own multiple products, multiple leagues that are owned by one holding corporation.

How do we help them to better show it's not just who's watching it in that moment, but it's the halo effect that comes for the broadcaster by having them on, and we were able to put together a solution in hours that came back and they believe will result in millions of dollars of incremental revenue to them on the renegotiation of those streaming contracts. So it then becomes a flywheel because then we get part of marketing for those streaming rights. So it really becomes a major flywheel in and around the company, Matt.

And we'll go to our next question from DJ Hynes with Canaccord.

David, maybe we can just follow up on the last point. I mean Zvi and Business Intelligence is a massive category, right? I mean you talk about getting pulled into this and there's a lot of different directions you could take it. How do you think about formally productizing some of the use cases? How do you sequence which ones make sense first? Just high-level thoughts on kind of the multiyear playbook would be interesting as you think about building out that fourth use case.

Yes. And obviously, as usual, great question, DJ. What we're seeing is that our data cloud is able to really help enterprises to make better decisions around certain things very, very quickly. We're seeing big examples inside of retail. What products should they be moving into retail? How do they use co-op dollars and marketing dollars to get better shelf space and how to create deeper and more meaningful relationships with the end customer. So it starts with the business intelligence around a retail use case, you then end up identifying meaningful opportunities for them to move inventory around geographically, which retailers they should be doing more with maybe which retailers they should invest less with and then we end up being able to market to the end user to drive them into the retailer on behalf of the product and vice versa.

So yes, we are seeing logical productization much like we've talked about in the past, where, DJ, you've heard us say, yes, our platform can help almost any vertical but the more industry expertise our salespeople have across the 15 verticals that we operate in, the better they can sell, the higher the sales productivity per rep and so on and so forth, we're seeing multiple use cases by vertical that we're building into the ZPI and every new request from every customer becomes productized immediately, and that goes back to the ability to do 90% of your new code on an automatically generated basis.

So it's really building a very interesting flywheel as we're growing the ZPI not just from a revenue perspective because by definition, it's our fastest growing use case right now because it was starting from a smaller number. But the reality is it's becoming products every moment of every day that we're adding them in from a client question perspective.

Yes. Very clear. And Chris, maybe a more tactical question for you. Look, obviously, we're heading into U.S. midterm cycle. I'm just curious what you're seeing in terms of political advertising demand, how much visibility you have there today and kind of how that feeds into the guidance.

From a guidance perspective, DJ, we've held to that consistent political candidate guide of $7 million in the third quarter and $8 million in the fourth quarter purposely because we wanted what to come through in the numbers. was that the strength of the business unfolds and the actuals, the raise is all tied to the strength of the core business as well. The visibility is good in terms of -- we expect it to be very robust. It comes into the pipeline pretty late because these can be programs that spin up quickly, and we execute even more quickly.

We get paid ahead, which is nice. So it's good for free cash flow. But we're optimistic that the number we have in there is conservative. And we've got -- for the full year guide is maintain our normal level of conservatism. So despite this being the largest raise of the year of $33 million revenue, it still maintains our normal 2% to 5% cushion.

I think it's also important to note, we raised revenue by $33 million. We raised free cash flow by $20 million. So you're seeing a disproportionate percentage of incremental revenue into this business at this point drop to the bottom line. You saw a 170 basis point increase in operating margin. And you saw a meaningful increase in free cash flow at 73% growth.

And we'll move next to Ron Josey with Citi.

This is Jake Hallock on for Ron. Congrats on the great quarter, and -- so my first is on the open AI partnership. So curious how deeply is OpenAI integrated into Athena and Zeta's broader platform? Could you help us better understand the capabilities data brings to open AI's ads ops? And is there a direct revenue opportunity associated with either side of that partnership?

Yes. So that's three big questions, Jay. But let's start with OpenAI powers the voice component of Athena. No large language models ever see the data in our data cloud or our clients' data. We keep all of that totally and completely safe. And now the data cloud is built on foundry with Palantirs ontology. So you're looking at a really interesting use case. What does OpenAI bring to Athena it personalizes Athena.

So now Athena gets to know the user at the enterprise better and can begin to understand what questions they have before they even ask them and help them better navigate to outcomes instead of having to navigate a platform in the entire way. As it relates to the integration to their ad platform, we are actively serving ads, and it is a meaningful revenue opportunity. It is also scaling very, very quickly. I think it's also, as Chris pointed out, part of the conservatism we've put in going into the year where we really haven't included pretty much anything from either Palantir or open AI as it relates to our forward guidance. And next week, we have a follow-up meeting with Open AI. We've got 20 people in a room, literally brainstorming what else can we do together. I think they consider us one of their most important enterprise partnerships, and we certainly look at them in the same way.

That's so helpful. And I know it's a 3-parter, but can I just sneak in none more here. You did like I appreciate the update on the 90% of new code generated through AI and Athena operating largely on in-house inference. Could you just touch on how those capabilities and that AI code is changing your product development speed?

It's just -- I mean, accelerating it at a pace that I never thought humanly possible, Jay. When you look at the fact that Yes, we increased operating margin by 170 basis points. We added engineers in the quarter. We did not eliminate engineers in the quarter. So the fact that we're even able to add horsepower to that while simultaneously having 90% of all new code being auto generated is pushing sales cycles -- I'm sorry, pushing development cycles, I should say, to what would have been years at 1 point to what is now months and new products inside of the ZPI as we productize it to hours versus what could have been months in the past.

I think when we get to Seta Live on October 8, we're going to have a massive unveiling that I'm incredibly excited about that's a product offering that will continue to evolve Zeta as a company in a very meaningful way, and it's something that would have taken years that we're now going to be able to have done in months.

And our next question comes from Elizabeth Porter with Morgan Stanley.

Great. So I just want to do another follow-up on ZBI and how that's changing some of the market opportunity. We used to talk about data having 1% of the marketing wallet, but an opportunity to be closer to 10%. And just as you're going kind of beyond marketing to the broader business intelligence, how should we think about the wallet share that's up for grabs and how that changes the ZB? And just as a follow-up, how should we think about the sales cycles as you expand from more of a CMO-focused product to just broader touch points across the organization?

Elizabeth, let me start by saying welcome back. We've missed you. Look, from a ZDI perspective, and I want to be clear, we're still looking to get to 7% to 10% of our clients' marketing wallet share. That's not changing. We believe that we are on our path to building a $10 billion business with a 30% operating margin with the vast majority of that dropping to free cash flow on the marketing component of our business alone. When you look at the ZDI and the opportunity that it's opening, what we're seeing and what really happened Elizabeth is we started getting pulled in that direction by our clients at first that then accelerated massively with the Palantir partnership, whereby putting our data cloud on top of foundry and the ability to onboard seamlessly our clients' data in ways that we couldn't before, and using their ontology in a way that we couldn't have done without Palantir as a partner, that's opened up this entire new use case from a productization perspective.

I think we're going to look at two different sales cycles. I think we're going to have existing clients that are going to adapt or adopt the ZDI very quickly. Then I think we're going to have new clients that we're going to go out and get where you're going to start with the CIO or the CTO and that could take a little longer. As you know, our normal product cycle is anywhere from 90 to 180 days. I think this will be very -- maybe start out on the longer end of that for new customers. but we'll start on the shorter end of that for existing customers. And quite frankly, we've seen much, much faster than that from an adoption perspective, from existing customers who are adding GBI as their third or fourth use case. Does that make sense?

That does. And I wanted to do a quick follow-up on Athena for agencies. It sounds like you had a lot of high interest in France this summer. So when we think about the base of agencies, is there a certain proportion that a good client base for Athena Agencies? Is there any sort of specific subsegment? Or do you think that most agencies could adopt Athena over time? And what are you learning about the onboarding cycle that could potentially drive some faster penetration?

So to answer your question, I was shocked because I thought it would really start at the midsized agencies from an entry point perspective. We're actually seeing the larger agencies adopt it faster and at scale very, very quickly. It's it really changes the game for them from a navigation perspective and the ability to showcase for their customers our data cloud and the ability to drive substantially higher return on marketing spend. As you know, Elizabeth, we always talk about sort of we deliver according to Forrester, 600% to 700% return on marketing spend. We're seeing clients who are adapting -- adopting Athena at substantially higher rates than that in partnership with a very large agency that adopted Athena and a very large airline, we're seeing that airline today at a 1,400 return on marketing spend post the agency bringing Athena in and us putting together the project together.

So I actually thought it would be small to midsize. It's starting very, very large, and it's scaling quite frankly, a little faster than we expected.

And we'll move next to Arjun Bhatia with William Blair.

And I'll have my congrats on a great quarter. David, I can start with you just on talent here. It sounds like, obviously, you did the tech migration and your onboard on foundry, but I also sounded like you're closing some deals already on the cross-sell. So I would be curious to hear just what that pipeline looks like into Palantir's commercial customers? And just as you look across different verticals, are there any inside that commercial base that get you more excited than at was from an opportunity perspective?

Yes. I'll be honest. I publicly said that I plan on doing the first 20 deals in partnership with Elia Davis, who's become a very good friend of mine, who's one of the top guys at Palantir reports directly into Alex's office. And we've been going out. And I would tell you that if I put the numbers of the deals we're working to into our pipeline, it would artificially skew our pipeline up too much at this point. it is that big an opportunity.

So I'm sort of working on it on an internal pipeline that I'm running in my own office at this point. I'm embarrassed to say I'm doing it inside of Claude. But the reality is that this is maybe the biggest opportunity we've ever had in front of us. Certainly, the biggest opportunity we've ever had from a partnership perspective, and as you know, we have 15 different verticals that we operate against. Not one of them is a massive concentration. So when I look at their client base, I started by saying, why don't we just start with the 20 customers who spend $1 billion a year on marketing to consumers.

And the list was so much longer than '20, we had to pare it back to start going out there and getting into it. The two deals we closed were at 100% hit rate. Met with two, closed two. We've got some other very large ones in flight. And we see this as a meaningful and very large opportunity that, as Chris said, is not baked into the numbers just yet.

All right. That's great to hear. And then Chris, maybe one for you or I guess maybe for David also, but the -- I would love to hear sort of your updated thoughts on capital deployment with the new credit facility. And I think you're pointing out M&A is obviously a potential use case, which you've been sort of consistent on, but I'm curious if there's any sort of change in the type of acquisition these or assets you're looking at in the market?

Chris pointed to me, so I'll answer it. What I would tell you is we are going to continue to focus on our 5 pillars of M&A that is not going to change, although you can look at the fact that for us, at this point, we have to continue to evolve the type of deals we've done or we'll do because some of the smaller deals we've done over the years just wouldn't move the needle for us at this point. I will tell you, Arjun, you've heard me say this many times over years, I believe transformative M&A transforms both companies for the worse. So we are not looking to go do one like huge transformative deal.

We will continue to do small to midsized deals where we're adding great human capital, great data sources, incredible products that our clients want to buy and clients that want to buy our products and so on and so forth. This does give us optionality around the buyback. It could cause us to accelerate it in the short run, depending on how the stock trades. But the reality is it puts us in a position that we have a lot of flexibility as a company I mean you can see the cash position is also very, very strong at the close of the quarter, and we're projecting I'm pretty sure this will be a record free cash flow quarter for us, right?

Yes. So we -- last quarter, you can do the math. We bought more than 50%, I think closer to 75-plus percent of our free cash flow and share repurchasing last quarter. We're going to continue to buy the stock back at these prices because we think we continue to be the best investment we can make with our free cash flow is purchasing back our existing shares.

And we'll move next to Callie Valenti with Goldman Sachs.

So new AI products have generally taken a while to ramp in usage for many software companies. Are there a couple of key things you would point out that you think have made your customers adopt new AI tools faster than what we're seeing in the rest of the market?

First of all, Callie, congratulations. I'm sorry, congratulations. We're super happy to have you covering us. return on investment. I think one of the things you see in our business strategy is our ability to show a return on investment is second to none. Most AI products where you're making a large investment into whether it's infrastructure, software or technology take years for enterprises to show return on investment. Our -- for use cases show massive return on investment.

As we've talked about, our marketing use cases, show a 6% to 700% return on marketing spend effectively day 1 and if you look at the ZBI, we can't quantify it just yet, but we're giving real-time business decisioning that drives massive incremental profits into our clients in real time. So I think that's been one of the big differentiators for us versus companies that are out there selling very expensive technology that will take years to pay off.

Yes, that makes a lot of sense. And then just one more for me. We're hearing more companies talk about headless architectures as agent kind of potentially disrupts the UI or traditional -- how do you think about this in the context of data? Like what makes sense for you wouldn't make sense for you?

I mean, listen, we've been no code for many years around here. So when you look at what they're talking about as it relates to head list -- that's not something that I think we end up dealing with on either side. I don't think it becomes a competitive force for us, and I don't think it's something that we would move into quickly. From our front, what we're really focused on is what percentage of our new code can we generate and make generally available to our clients. And we've gone from what was a to 90% in the first quarter of this year to the second quarter of this year, that puts us in a massive competitive advantage.

As our competitors are trying to catch up to where we were a year, 2 years ago, we're already moving to where they won't be able to get for many, many years. And I think our no-code architecture has allowed us to do that.

And we'll take our next question from Terry Tillman with Truist Securities.

This is Lu Cordis on for Terry. To start considering the big strategic win with GAAP, how is RFP activity? And what are you seeing in terms of large martech replacement cycles, potentially aiding revenue and business in the second half for helping enhance visibility into 2027.

Thanks for the question. Just real quick. RFP activity is very strong. You heard about the pipeline stats that we shared that's built into greater than 60% year-over-year pipeline growth. I think it also speaks to how the average contract value in the pipeline is up because more of it is RFPs. I'll turn it quickly to David to talk about broadly what's driving that environment.

Yes. I mean we are right in the middle of what looks to be a marketing cloud replacement cycle. And Gap was a perfect example. Gap had been with 1 vendor sales force for quite some time. There were 3 other vendors that we displaced in addition to Salesforce to become the system of record inside of GAAP. And when you look at these very large organizations, they don't want to use 4, 5, 6 different vendors, including a software provider or a professional services firm, activation platforms, data, CDP, all separate these very large enterprises want next-generation technology.

And today, we are the only marketing cloud that has data and AI as native, foundational to the application layer. So as other entities have to step out of their platform through API integration to use AI and then to get to the data sources and then back to the AI and then back to the platform, that latency destroys return on marketing spend. Our foundational platform can answer in a millisecond what other platforms can answer at all to be quite frank. So we're actually seeing RFP velocity go up and we're seeing closing go up as we saw in this quarter. And we have a record pipeline right now.

It's by far the largest we've ever had. So I think we're very well positioned for where the market is and where it's going.

Great to hear. And if I could sneak one more in. Given the increased adoption with Athena and AI-enabled workflows, can you help us think about the puts and takes on gross margin performance in the back half of 2026?

The gross margin performance in the back half of the year is largely dependent upon mix. So we talked about how the second quarter came in where we expected, given where the direct mix was in the quarter of where that efficiency plays in, by the way, not just in R&D, but also across G&A and in sales and marketing as well. As we saw in the second quarter, and we expect to continue to see very strong unit economics on our expense to revenue ratios that showed up in very strong adjusted EBITDA margin. We're continuing to see efficiencies in CapEx, which then obviously flowed through the higher free cash flow. And then the good work on dilution and on stock-based compensation generated a very positive GAAP net income.

And we'll go next to Jack Nichols with KeyBanc Capital Markets.

David, maybe to start with you. How are you thinking about the opportunity for Athena's heavy users today post foundry infrastructure pivot? And what does that mean for these heavy users on the new platform? And what's the biggest risk in the coming months during the transition to lift the move to foundry?

Well, so to be clear, the transition is done. It's seamless to our clients. We rearchitected the data cloud on top of it and adopted their anthology. So it is a massive benefit to answer your question. So when you look at Athena at the top of our tech stack is now going to be Athena powered by open AI. She will then interact with this Zeta platform in addition to the data cloud. Every time the Zeta platform and the data cloud access data will now be inside of Palantir's architecture, which moves extremely faster than the architecture we were working on before. So we'll be able to answer more questions smarter and faster.

We'll be able to onboard new client data faster and with better orientation, which will allow for higher levels of intelligence faster. As you know, the longer a client has worked with us traditionally, the smarter the platform has gotten the faster the return on investment. With Athena and foundry we're seeing that happen faster. Instead of taking years of managing questions, it can now take days to get to the same level of return on marketing spend and the same intelligence that used to take us years.

And that, I think, is going to drive much faster adoption. And once again, heavy users had a 400 basis point higher net retention rate than our other clients for Athena today. I think you're going to see that number continue to evolve.

That makes some sense. And maybe for Chris. What kind of usage trends of a then are baked into the super scaled ARPU growth to achieve the organic guide?

What we have line of sight to. So -- and that, by the way, same is said for the partnership agreements. We talked about the multiple layers of conservatism. What we built in the guidance is what we signed already. So we're not leaning into anything on a go-get perspective on either AI adoption or newly signed partner agreements that are still yet to be closed in the pipeline.

And up next is Matt Bullock with Bank of America.

I was maybe -- I was hoping you could maybe put a finer point on what's working well in terms of driving improvements in sales rep productivity. And then can you help us think about the outlook for quota-carrying rep head count for the remainder of the year to address the pipeline you talked about?

Yes. So first off, 1 data continues to be gaining speed. So we launched next, call it, 15-ish plus months ago. And we're starting to really see the benefit of our sellers attach more channels and more meaningfully more use cases into their deals. We talked about pipeline creation per rep being up around 10%. And -- that's a big source that also we're call it 6 months now from acquiring Marigold, all the hard work on the integration has been done both within the G&A sales and marketing R&D structure, now we're starting to see the benefits of the cross-sell and the upsell activity, namely really around their loyalty products and selling Zeta's acquire and grow use cases in. So I feel like we're at a place where -- we've got the right tenure of reps, meaning the right balance between those are in their first 12 months, 12 to 24 months and greater than 24 months where the type of productivity we're seeing now can really continue to go throughout the year.

I also think our hyper-focusing by vertical by salesperson has really unlocked a massive opportunity. That was sort of an aha moment for us, Matt, where we start to see when we bring in people who have industry expertise in a vertical and they sell their productivity goes through the roof.

And we'll go next to Richard Baldry with ROTH Capital Partners.

You talk about in the generative engine optimization side. Do you think that's helping you win client wallet share? Or is those dollars really sort of moving from one channel to another?

Because we never did search engine optimization, it's 100% upside to us, Rich. So the other thing we're finding is that new clients are really excited about that. And we're one of the very few companies that has an API integration into Claude chat PT and Gemini. So we're able to help across all of those platforms in real time through one GEO user interface. And then we're able to serve the marketing, obviously, in the GEMINI and into open AI as a subset of that. So you're able to really move the flywheel in a way that I don't think others are able to do. I think new customers see that as a shiny new thing and I think it's very exciting. And we're seeing existing customers adopt it as a part of their marketing strategy.

Actually, maybe when you think about how fast moving those generative engine corporations are that you're partnering with and Palantir itself is a different use case, how fast that company is moving you've really grown without adding a lot of head count in recent years. Do you think there's a point in which to try to keep up to those opportunities, you need to add a bit more to the head count to support the fast growth in those areas.

We're just not seeing that right now, Rich. We're seeing productivity in the company go up at a rate that I didn't think possible not only we seeing, as I said, 89.6%, call it, 90% of our new code generated. We did a really interesting deal this quarter where we signed one of the first sort of ubiquitous enterprise agreements with OpenAI, where all of our global employees now have access to all of OpenAI's products.

And not only are we doing that, we're doing it in a way where the used cost for everybody is baked in. So we've got use cost for everybody. If power users go over a certain amount of token usage, the platform actually refers to their manager to approve additional token utilization I think as we're building productivity tools like that, we're seeing sales productivity explode -- at the same time, we're seeing total AI cost for the company internally well under 1% of revenue. And we think that's something that's going to continue. So I spend a lot of time on this, as you know, I mean I think we'll add head count, but I think we'll add head count at a slower pace than we grow revenue, and we'll continue to grow EBITDA and free cash flow at a much faster rate than we're growing revenue, if that makes sense.

We've actually had a head count just thrown revenue much faster. And even though that AI-based usage is up across the company, unit costs for that AI are down almost 40%.

And that's because of the relationships we've been able to negotiate.

And we'll move next to Scott Berg with Needham.

I'll skip the [indiscernible], move to something a little different, I guess.

You don't want to talk about Palantir?

I do. We'll speak on it plenty I'm sure, David. But my question is on your expansion opportunity, especially within your superscale customers, the slide in your deck, in your presentation deck in the quarter, I thought it was kind of interesting that -- you've been kind of range bound on the ARPU for your superscale customers over the last 7 quarters. It's kind of bounced around from $1.6 million to $1.8 million just kind of back and forth.

We know you guys are doing a good job of expanding with some of your customers. I'm trying to help under -- just trying to better understand that dynamic, I guess, some of your new customers may be coming in with a slightly lower ARPU on that superscale level to just kind of balance that out? Or maybe the metrics not reflecting some maybe numbers or some expansion deals that were maybe expecting in the back half of the year?

It's more the latter, and I'll go quick just so we can get other questions in, but there's a Slide 8 in our earnings event. What's masking that, Scott, is that as new pilots and proof of concepts become superscale, so they cross that $1 million threshold they are still a distance from where our more mature, those superscale customers that have been platform, call it 2, 3, 4 years that have an ARPU that is 4, 5, 6x greater than those that are in the kind of early part of the journey.

So what you'll see on the slide is the average size in terms of ARPU of a customer who has been on the platform less than 12 months is $700,000 as opposed to those that have been on the platform for 4 or more years, that's now approaching 4 million per. So we're actually seeing those superscale customers get bigger, become a bigger and bigger part of our overall revenue. and have a greater and greater share of our net revenue retention.

And we'll go next to Clark Wright with DA Davidson.

was great to see better-than-expected organic growth results this quarter. How much of this growth is coming from continued success with agencies versus your direct enterprise sales motion?

I think it's really well spread out, Clark. We continue to see the agency business at approximately 20% of revenue, and we continue to see direct-to- enterprise at approximately 80% of revenue. So haven't seen that skew meaningfully, but we're seeing meaningful organic growth across both those components.

Awesome. Helpful. And then there was a sequential step down in direct platform revenue from 75% to 72%. What caused that this quarter? And is that something we should expect going forward to be at that level?

Just about to add to David's answer. What drove that was new sales and expansions with agencies. In fact, just within this quarter, since we closed and within the timing of this earnings call, we closed and expand with another very large [indiscernible] and as you probably know, Clark, those newer agency signings tend to begin with social and social channel grew very rapidly in the quarter again and drove the integrated platform revenue mix to be higher and then as a byproduct of that the direct platform mix to be at 72%. But it was where we expected based upon how we saw the pipeline, but it's driven by a very strong agency adoption of social as the initial channel.

And we'll go next to Naved Khan with B. Riley Securities.

This is [indiscernible] calling in for Naved. To start, it's great to see you've got a really strong uptake in tenants making it generally available, but some kind of like costs haven't scaled at the same pace as usage. As you said does continue to scale, how should we maybe think about the way that cost scales or token costs passed directly on to customers? Or does usage become cost of revenue consideration?

No. So we -- I would tell you that we've done an agreement with OpenAI, where we have a tremendous amount of visibility into our cost. And in this particular product, it's not focused on token utilization. So it's focused on sort of a license for the product that is fully embedded into Athena. I feel very comfortable that we will be able to continue to keep our total AI cost well under 1% of our revenue, while simultaneously growing revenues at obviously substantially faster paces than that even.

Understood. That's really helpful to hear. And then separately, so 8 of your top 10 industries grew over 20%. Can you maybe speak to how those areas are pacing so far in June and July?

I don't want to get -- first of all, the performance within our quarter in 2Q was pretty linear. And I would expect third quarter to be the same way. I don't want to get into projecting third quarter. But what I can tell you is the momentum that they have. We talked about within 8 of the 10. And by the way, 1 of those 2 that didn't grow over 20% was advocacy, and that obviously has a lot of tailwinds going into the second half of the year. So I'd expect that to be 1 of those greater than 20 next quarter. But there were several of our industries that actually accelerated in their growth from a trailing 12-month basis ending the first quarter where we ended in June. Those were consumer retail, financial services, automotive and health care, all industries with, obviously, a lot of marketing spend behind them.

Obviously, we wouldn't have raised the quarter by $10 million if we didn't think we had a lot of visibility into the quarter, Ethan. So as Chris said, we don't want to comment on 1 month, but nothing that happened in July would lead us to believe that we shouldn't have raised the quarter in the year as much as we did or we would not have done that if that makes sense.

Thank you so much. And that concludes our Q&A session today. I'll turn the conference back to David Steinberg for closing remarks.

I just wanted to close on how incredibly proud I am of the Zeta team to be able to continue to execute 20 for 20 quarters as a public company, 20 quarters beating and raising to continue to execute over that period of time with that level of excellence to continue to see accelerated sales growth to be able to do partnerships with 3 of the world's most important companies within just a few months if you had told me a few years ago that we would be in a position to announce partnerships like OpenAI, Palantir and Snowflake, I would have I would have been blown away by that alone.

That shows what's happened to data as a brand because none of them would have trusted us if they didn't trust our brand and they didn't trust our business. So thank you again to all of our Zeta people to all of our clients and especially to our partners, we appreciate everything that you guys are doing for us and with us as a company. Have a nice day, everybody.

And that concludes today's call. Thank you for your participation. You may now disconnect.

Hi, everyone. This is Ron Josey talking. I can't be on camera today for some technical reasons. But with us today, I've always liked to have with us David Steinberg, the Co-Founder and CEO of Zeta Global; and CFO, Chris Greiner.

David, Chris have joined us, and we've got some very interesting conversation to hear as Zeta evolves more to an infrastructure AI platform. I believe everyone should have the disclosures on in front of them, if you don't already for me at least, please feel free to e-mail me directly or let me know. But David, Chris, thank you for joining. Always appreciate talking with you all.

Ron, thank you so much for having us. This is obviously very, very big news for us, and we were so glad that you had the availability to help us.

So absolutely. And again, I wish I were on camera, but let me start off. I believe Zeta is in entering the quiet period, I can't really comment on 2Q trends. So no financial update today. And the goal of today is really just how Zeta is evolving into an infrastructure layer. So with that, David, obviously, we've had a lot of good track records of earnings, and we'll table that for a few weeks. But when we talk about what's going on at Zeta and when we all met really about 1.5 weeks ago in Cannes, it was about...

It was very hot there, Ron.

It was very hot. Indeed. Well, when we met 1.5 weeks ago, I believe Zeta just announced -- you just announced your partnership with Palantir. What I wanted to do is the first question is, my goodness, we've seen Athena launch in March. We've seen partnerships with OpenAI, Snowflake, now Palantir and the pending integration of Foundry, a lot is going on, on Zeta. So maybe level set us in terms of how Zeta is transforming itself into being that AI infrastructure platform. And what exactly does that mean? And then as we go through here, I have a ton of questions about the partnerships, about where Zeta is going, but maybe bigger picture, just talk to us about the evolution of Zeta and how really 2026 has changed everything.

Yes, Madison, let's go to the what's changed slide, which I think is very helpful. If you think about just the last few months, we've announced a multi-tier -- multiple partnerships with OpenAI. We took Athena generally available, both for enterprise and then agency. We announced a major growth to our Snowflake partnership and then the biggest thing we've done to date, which is Palantir. I said OpenAI was the biggest thing to date when we did that, but Palantir now usurps that.

The reality is none of this was an accident. We've been building towards being an infrastructure/business intelligence platform for quite some time. As I've shared with you in the past, Ron, I've always thought of us as a business intelligence platform that started focusing on marketing because the total addressable market was so high. But with the combination of the launch of Athena and then Palantir, and I think there's been some confusion, which is one of the things we wanted to say, we should have Palantir fully integrated for Phase 1 and have all of our existing clients on the Data Cloud moved over to the Foundry platform, I would say within 30 days of today, maximum maybe 40, 45.

So we've been working on this for quite some time behind the scenes. And a lot of people were thinking this was going to be late this year or early next year. It will be done, not only will we have Palantir fully integrated in the next 30 to 45 days, we'll have at least one joint launch enterprise client from their side this quarter is what we believe and very possibly more. But when we do move the Data Cloud over to Foundry, which will happen in the next -- fully integrated in the next 30 to 40 days, all of our clients will then have access to the Foundry platform. Once we have that as our bottom layer infrastructure, moving from our 3 main use cases, which is customer acquisition, customer retention and customer monetization to what will now be our fourth official use case, which we're calling the Zeta Business Intelligence or engine.

So ZBI, a new one for me to remember. I should remember it, I sort of came up with it. But the ZBI is going to change the way business intelligence works. So today, according to IBIS, business intelligence is a $35 billion addressable market in the United States alone on a yearly basis.

Today, the way business intelligence works is the software ingests old data from the past, I call it rearview mirror data. It then puts it into a format that a bunch of data scientists can ingest and look at and try to make decisioning going forward. By layering the Zeta Marketing Cloud and the trillions of signals that we're ingesting, we're now going to be able to forecast business intelligence changes with substantially greater accuracy in the future.

So we think we're going to completely change the game in the business intelligence space by using our Data Cloud plus our clients' data in the same consumer data platform that we're currently utilizing, but literally inside of Foundry, which will allow us to gain access to all of our clients' data in real time and allow for real-time decisioning and real-time business intelligence. So when you think about the 4 big things that have changed over the last year that have led to this announcement and this evolution, it starts with OpenAI, then you have Athena, then you had Snowflake. And then, of course, you have Palantir, which will have fully up and integrated in the next 30 to 40 days.

So David, that's a lot. First of all, it's pretty exciting to see. So let me try to break this down a little bit more. When we were together in Cannes maybe a week ago, 1.5 weeks ago or so, we talked about the transition happening. The 30 to 45 days, that seems a lot faster than what we and many others were thinking. So you've been working on this for a while. I just want to make sure I understood that correctly. So Phase 1, all clients moved over to 30 to 45 days. Did I hear that right?

Yes, all our existing clients that are using the Data Cloud will move over to Foundry in the next 30 to 40 days. It will be seamless to them. They won't even know what's happening. What they'll see is 2 things right away. One, intelligence will speed up. And two, we think this will drive to a substantially higher return on marketing spend right away. We're already seeing with clients that have adopted Athena a substantially higher return on marketing spend, and we think that the addition of Foundry is going to accelerate that.

We have one client that is now over 1,400% in return on marketing spend. This is real world, real examples. Chris is going to get into it in a bit, but we're going to be building an AI scorecard. We're not prepared to roll it out today. But we're looking -- Madison, would you go to the compatible slide?

Now the compatible slide, the slide of the other companies that are AI infrastructure companies that trade publicly, correct. So if you look at this slide, you can look at the cohort of companies that we believe we should now be compared to. And you can look at the metrics that correlate to us and these companies. And in all cases, we're either 1, 2 or 3 as it relates to Rule of 40, we're at 48, whereas it relates to net retention rate, we're at 120%. And the one that's really exciting is free cash flow yield of 5%, with, I believe, the average being 2%.

So you can look across the board, our company is already achieving the rules that are necessary to be an AI infrastructure platform. But when you think about the real benefits of this and you look at Palantir plus OpenAI plus Snowflake plus Athena, we think the first big move is going to be a meaningful increment in return on marketing spend while we're simultaneously rolling the ZBI out to all of our existing enterprise clients. It opens up a new sales track into the CTO, CIO or CEO in addition to the CMO. It opens up an incremental $35 billion revenue opportunity. But I would say even more importantly than that, it's going to embed us even more deeply into our enterprises, Ron.

And that's actually a key point that I was going to ask a little bit later on, but you talked, David, about expanding the TAM, the $35 billion, now being able to sell the CIO and CTO. So you have really 3 different owners potentially within the organization as you now for those 4 different solutions, acquisition, retention, monetization and the ZBI.

So talk to us about the go-to-market approach in terms of how Zeta is positioned today now that in about a month, 1.5 months, we'll be on Foundry, things will go a lot faster. You talked about latency. And now we're also talking to CIOs and CTOs. I'd love to hear the approach here because it is a newer part of the organization. How do you get the introductions? Is a CMO doing this on a nice -- on a warm introduction? Talk to us more about that process.

Well, first of all, we -- most of the big deals we close, we have CEO relationship, and we end up going through the CTO or CIO's office. So we already know all of these people because when the CMO brings in a new component to the tech stack, the CMO brings us to the CIO or the CTO. And often, I know the CEO and/or we're able to build a relationship with them and come in that way.

What I would tell you is that the sort of sales motion on this should be very similar in that we've already been doing this for a number of our clients. We have one great use case recently where a very large hotel client of ours wanted to figure out where to put their next 500 hotels. So we pulled all of their most loyal customers. We looked at every one of their credit card transactions in geographies that they did not have a hotel against their 4 main competitors, how many room nights were they consuming at their main competitors. We then drew a circumference or 5-mile concentric circle around that area, and that was where they decided to put their next 500 hotels.

We then -- as they've opened a couple of those, they bought a couple, quite frankly, in those areas. They're still building the ones that they're building. They used us to launch the marketing to do conquesting to move their loyal clients away from their competitors in the geographies that they open locations.

The other big thing we're looking at, Ron, is inventory management and what organizations are going to manufacture next versus what should be discontinued, how should inventory be shifted on a geographic basis based on where people are. All these different things are natural workflows in the ZBI coming out of the existing relationship in the consumer data platform.

And so that's actually a good segue into the new ZBI because when you talked, David, that example with hotels, that is now going into business strategy. That is now directly touching revenue as opposed to through the marketing side. And so one of the things that I always find fascinating at Zeta is just your current Data Cloud, I think it's 535 million consented individuals across trillions of signals results to real people. And so one of the things you said when we were just starting to talk here is just the Data Cloud in the Foundry, but then also working with client information to put everything together.

So talk to us a little bit more just around the opportunity with Zeta's Data Cloud into the Foundry, but also leveraging the data that the clients have because that proprietary, that 1P and 3P data is really a significant differentiator from others, I would imagine.

Yes. I think our Data Cloud is our biggest differentiator and our biggest moat as a company. I mean if you look at the other 3 or 4 companies in the world that have a comparable or larger data set, they're all worth over $1 trillion. So when you look at this data, by putting it on top of Foundry, it's going to speed everything up and allow throughput of data outside of just marketing data. So now we'll be able to get all of their business data put into Foundry and into the Data Cloud as a subset into the consumer data platform.

So what I simply like to talk about is how Foundry is a game-changing foundation for our new skyscraper that we're building. It's sort of the foundation that will allow us to go higher and allow us to expand out, whereas in the past, we've been very sort of -- I don't want to say pigeonholed, but we've been very thought of as a marketing platform for obvious reasons. The vast majority of our revenue came out of marketing. And marketing is going to continue to be an incredibly important application for Zeta for many, many years to come and probably forever, right?

But at the end of the day, the ability to get even stickier, I shudder to say stickier because I want to reiterate, last year, our net retention rate was already 120%. But as we think about how do we get stickier inside an organization and how do we drive incremental revenue and margin inside a customer, we've always looked at it as how do we solve our clients' biggest problems. And by adding Foundry on top of -- the Data Cloud on top of Foundry with Athena on top of the Data Cloud, you'll now be able to, in a very conversational way, ask Athena how to solve your biggest business problems. How do I turn my inventory more quickly? How do I shift inventory for specific products or services to markets that they need to go to? What new products should I be developing? What new products should I not be developing? Where can I go get the world's best employees to help work for my company as I try to evolve my organization? All of these are questions that the ZBI will answer with Athena.

And that again opens up to more organizations and use cases. I want to ask about the data and Athena, but before we do, let's make sure we're on the same page on Athena for those on the webcast. We launched in March. I think we had a pretty large client win here recently. But maybe, David, just step back for a second and tell us how Athena's capabilities enable the ZBI or the Zeta business intelligence layer so that you can obviously build that skyscraper on top of Foundry. So I would love to hear more about what Athena does, what you've seen thus far, and then we can get into examples.

Yes. So I would start by saying that Athena is meant to break down the friction between the human and the platform. So as I sort of talk about, I don't want to get too theoretical here, but for 400-and-some-odd thousand years, humans have been communicating via voice. It wasn't until the 1950s that we created the keyboard. Today, the average enterprise uses between 5% and 10% of the enterprise software they license just because they don't understand the feature sets, they don't have to use them. Most of our clients know how to fly a really good Cessna, and we've given them a stealth fighter.

So by going to Athena, you can ask Athena in a purely conversational way, how to solve your problems. Athena, I'd like to create 2 million incremental customers this quarter, and I'd like to do it while lowering my marketing cost by an average of 7%. How would you recommend I do that? And instead of the individual at that enterprise having to navigate 20 or 30 screens to do that, Athena does it for you, changes the screen to just give you your answer. It's going to do the same thing with business intelligence. You'll be able to ask Athena, how do I solve my biggest business problems and Athena will change the UI in front of you to give you the answers.

And to me, sort of this intelligence AI infrastructure, which is really what we've built is the differentiator in Zeta. And I think a lot of that's gotten lost over the conversation about marketing or not marketing. To me, marketing is a very large and necessary component of very large enterprises. We've grown our business now for many years in a row. We're on a 4-year compounded organic growth rate of greater than 30% over the last 4 years. And there have been ups and downs in that period, but yet we've still grown through them because we continue to take massive market share from our competitors. Back to the large win, if you jump to the slide that has the -- we start with what's changed, which are the 4 things that have changed, which is OpenAI, Athena, Snowflake and Palantir. Now let's talk about what proves it. That's -- now let's go to the prove it slide. Thank you.

So if you think -- we started with OpenAI as a technology partner, then they awarded us a meaningful component of their marketing serving business. We started with Palantir. We moved very, very quickly. We have a 7-year joint go-to-market with Palantir. And when Elias, who's one of the best partners I've ever worked with, Elias Davis was on Alex Karp's personal team, and we really worked with Alex and his guys, Zach and Luke, to really get that deal over the finish line. He was -- when he was interviewed and he was asked about the term of the agreement, he said the contract is 7 years, but we expect to work with Zeta forever. We feel the exact same way.

And then, of course, to win a client like the Gap, which was a massive Salesforce shop, they use Salesforce for pretty much everything. We displaced 4 vendors in there. Salesforce was one of them and then Braze was another, and then there were a couple of others that were in there. And they made us the system of record. I would tell you, it's an interesting story because over the last 8 years, we've probably bid the Gap RFP at least 2 prior times and came in second both times. We came into this one, quite frankly, with a bit of trepidation because these RFPs for enterprises of this size and scale are very complex. There are a lot of work. They take a lot of people's focus for a number of months.

But they said, listen, we're open to meaningful change, and you really should bid it. So we came in. We were, I don't know, 3/4 of the way through the RFP process, and we demoed Athena for them, and they were like blown away. And I am 100% confident that Athena was the reason we went very quickly from this very long RFP process to them saying, okay, send us pricing, which is a very, very good buy sign. We then ended up closing with them. We're happy to partner with Sapient, who we work with regularly. And to become the system of record at Gap.

But it all came down to the Data Cloud plus Athena was really what changed the game for us with the Gap. And once again, what changed? Three new major partnerships and then a new game-changing product launch. What proves it, companies like OpenAI, Palantir and Gap trusting us as either their major partners or their vendors of choice. These are deals that even 4 or 5 years ago, we probably could not have closed as a company because we were not in a position to do that. And I will tell you, I'm looking at the road map, and I'm looking -- we have -- we've built an entire forward deploy engineering team now that is sitting inside the Gap.

We have Palantir people helping us there. We have OpenAI people helping us there. All these forward deployed engineers are working together. We've got a room of forward deployed engineers with OpenAI. We've got a room of them with Palantir. We've got rooms of them with Gap. And I'm very, very excited about sort of this evolution of the business and have gotten a real chance to spend a lot of time with the Palantir people in particular. And it's from the absolute top to everybody I've met with, they are just absolute rock stars. It's been an incredible partnership for us as we embark on this with them.

And as a system of record, this is and -- forgive me with this question, but is this a system of record for acquisition, customer retention, monetization and ZBI? Or is this something you can envision?

So today, it's a contract that's very focused on CRM and very focused on monetization. But we're already working on major projects on the other things. And I expect to expand out from what we're working on today. But once again, to be named the system of record was a big thing for us for a company of this size and scale. And we always want to think of ourselves as the challenger because I think that keeps you a little more hungry, Ron.

But the truth is we're really becoming the leader in this category. If you look at the Forrester report, you look at the wins, you look at 50-plus percent growth in Q1, 29-plus percent organic growth in the first quarter of 2026. You're really seeing us take meaningful market share from our competitors. And as we expand into business intelligence, we expect to be able to do that there as well.

Sure. Maybe -- I've got maybe one other question for you on Palantir. We can go probably another hours on the Palantir partnership. I want to get Chris on this, too. So David, first to you, you mentioned the cross-sell and go-to-market. I wanted to hear a little bit more about when you've got Athena plus Palantir, I wanted to talk to us a little bit more about the broader cross-sell and joint go-to-market approach for both companies, Zeta and Palantir and Palantir and the customer base there.

So once Foundry is live with the Data Cloud, which, as I said, should be in the next 30, worst case, 40 days, any existing Palantir enterprise customer will be able to just turn this on. There's a few exceptions for very highly regulated industries that we've sort of made a decision together. We're going to sort of not start with, and we're not starting with government for a whole host of reasons. We're starting -- but I would say the vast, vast majority of their enterprise clients are -- we're already opening discussions. We've already made meaningful headway with a number of clients. And I would expect to have at least one of their clients turned on this quarter, in addition to having the full integration done probably by the end of July or very early August.

It's just once again that Palantir is an organization that really puts their money where their mouths are. And one of the things Alex is really focused on and the things that he and I have chatted about is how he wants to take the best technology possible and help his clients solve their biggest problems and help them save the most money in their organizations. They didn't really have a marketing solution until they decided to partner with us. So I think that from a pipeline perspective, you're going to see a meaningful uplift in the pipeline as you see Palantir come online.

The truth is the first few deals I'm doing myself with very senior guys over there and women over there because I want to build the sales motion personally. But these deals are so large in nature that doing the first few myself, so I can really build the sales motion and then start to bring others into it, I think will accelerate things, and I'm very excited about where we are.

That's great to hear with that insight and color. And then, Chris, I think we've talked about $100 million ARR at some point. Help us a little bit more on the road map. This is part of the Palantir partnership, of course. How do you think about that $100 million? How should we think about that $100 million now that we actually have a great foresight or at least a vision for 30 to 45 days going live here?

Yes. I think the best way to think about it is kind of really zooming out and you think about our long-term model and the visibility that we have into it. So the short answer is this adds to our confidence that we can get there early. When we rolled out our first long-term model, Zeta 2025, which was get to $1 billion in revenue, we got there a year early. And we built the current model with the same foundations of conservatism and reliance on our own organic selling and new channels that we introduce in new use cases, as David just talked about with Zeta Business Intelligence.

So I think this is just one of many levers. We've given ourselves ample room in our model for this to scale and grow and perfect the go-to-market model. But we think of it really as just more upside and the ability to get there faster.

And -- again, Ron, I just want to say we've already fully budgeted for all the expense in this contract for this year and the out years. We've moved money from other vendors to Palantir, and it's going to be -- this should really be additive to operating margins in the years to come.

And that's the -- I think you and I, we've spoken about this as being cost neutral in '26 and potentially op savings in '27. Is that a fair comment?

I think it will help us continue to drive down our percentage of operating expenses in sales and marketing to revenue, which we've done now, I think, 4 or 5 years in a row. And I think that's a trend that will continue. Chris, do you want to add to that?

No, I think you're right, David, you said it well.

Perfect. So we'll come back to Palantir in a second. I want to make sure we get to all these partnerships because this is what has changed the game a little bit here in '26. Talk to us about the OpenAI partnership. I think it spans both product, I think voice capability for Athena, advertising and others. And so I would love to hear more about the partnership with OpenAI and how the team over at Zeta is leveraging GPT to build Athena or to come to go to market faster.

So it's interesting. We just did a third deal with OpenAI this past week where we started using OpenAI for all of our employees instead of Copilot, which was a big move for us. We've been a Microsoft shop for quite some time, which, listen, even furthers the partnership. But what I would say is it started because we were looking for the world's best voice engine. We had years ago bought -- built a product called ZOE, Zeta operating engine, I think, that was our first evolution into voice. And it was like being on an old walkie-talkie, Ron, you'd say, ZOE, how would I do this? Stop. And you'd sort of go in and out. Clients who use ZOE spent -- correct me if I'm wrong, between 300% and 350% more on the platform.

Yes, between 200% and 350%, yes.

So they're spending meaningfully more in utilization fees on the platform. And we said voice is really the future. How do we solve that? And for us, there's no better partner in voice activation than OpenAI. We do use a bunch of the other models. We partner with Claude. We partner with Gemini and OpenAI, and we use Cursor and so on and so forth.

But the reality is that we felt like OpenAI had the best voice interface. I do want to be clear also to clear something up. We never use any large language model to see any data in our Data Cloud. Our entire Data Cloud operates on our own proprietary models that we've been building since 2017. But I would tell you the vast majority of what we do from an AI perspective, and we think the future of business-based AI is inference-based artificial intelligence. And large language models are obviously very, very important.

But you don't need to ingest the entire Internet to figure out if somebody is going to buy -- get a new credit card, buy a car, churn off a wireless platform, right? We need 30, 40, 50 really good, really fast pieces of information that we're able to do with individual agents that are looking at all of this trillions of times. But when we wanted to build a voice UI, we didn't have the capacity to do it in the time frame in which we wanted to roll out Athena.

So OpenAI started as our partner. We use their voice engine. It gets smarter about the individual using it, whether they're at an enterprise or an agent. It begins to understand their inferences and get to know them so it can better navigate the Zeta marketing platform. Whenever Athena does a data pull or is creating information or changing the UI, it's on our models and our platform. So that's where it started.

And that went really, really well. In fact, I think we were one of the first enterprise applications to go live with OpenAI. And they were shocked when we went generally available on time. They were like, wait, you actually did this in the exact date. We did it 8 days early. But you did this before you actually committed to doing it, that really got their attention, and we had already been talking a bit about helping them on the advertising side. I think one of the things that also is not understood is how powerful our GEO platform has become. -- so many of our clients want GEO. And GEO in its heart is a ZBI product. It's an intelligence layer product for business intelligence, where it's helping enterprises to be better recognized inside of the large language models.

And our platform integrates into OpenAI, Claude, Gemini, really every large language model. We're also getting meaningful data out of the API that makes our platform much, much smarter and starts the flywheel there.

So that was sort of Phase 2 was getting into the advertising space with them. And then this past week, we just made ChatGPT available to all of our global employees and we're sunsetting Copilot as a part of that. So we felt like it was the right thing to do. We've got our people in the room with them all day. We wanted to make sure we were working together to support each other's businesses, and we think it's going to be a very powerful lift for our employees.

I mean, I think voice as a new interface is a very interesting concept. And I want to make sure I heard the impact from those clients that use ZOE. If you could just repeat maybe the benefits on the platform. And then I wanted to maybe get a picture -- a question, David, bigger picture. Do you think voice is ultimately the best complement to the UI going forward? Is it possible that people use voice more than keyboards over time? Two questions. One is ZOE, and then the voice.

Let me start with your second question first. I think that, once again, enterprises and the individuals who work there just want their problem solved. Whether they use a keyboard or they speak to the platform, they're just trying to get to a solution, whether it's around customer acquisition, retention, monetization or business intelligence, okay?

The easiest way to do that is to speak to it. But one of the interesting things I don't know if everybody knows about Athena is you can just as easily go into the Athena chat box and type your question. You don't have to use the voice. The thing about Athena is getting to the answer and utilizing all of the available data sets in the Data Cloud, all of the intelligence in the platform to drive the best results in any of the 4 use cases we focus on, right?

So some clients like to type into Athena the exact same things that other clients like to say. It's just what I think they're used to. It also, I think, depends if you're sitting with other people around you, it could be inconvenient to have to speak to it, whereas if you're alone in an office, it's much easier to speak to it. So we want to make sure that it was convenient.

But to answer your question, I think solving complex problems and allowing enterprise individuals to use up to 100% of the feature set and the data sets inside of our enterprise software platform is the real key to in our 3 marketing use cases, getting to the highest return on marketing spend. And as I said, we've already got clients north of 1,400% return on marketing spend who have adopted Athena. And then as it relates to the ability to put it into a chat box, you can do that or you can speak to it. Well, I'm sorry, what was your other question, Ron, I apologize.

Just maybe Chris can repeat the customers, the clients that use ZOE spend, was it 300% more or 3x more? I missed -- I want to make everybody hears that on the webcast.

Yes, it was between 200% and 350% more for those customers in the first generation of Athena when they leveraged automation.

200% to 350% more. And so now we're in next iteration. That's perfect. And so maybe I want to take a step back here. I do want to get into longer-term goals. But maybe, David, if I could take a step back, and this is a year where we've had the partnerships with OpenAI with Palantir with Snowflake, the launch of Athena, newer business models launching. If we take a step back and...

And by the way, Ron, the year is only half over.

Love it. So the question is more, David, when you think about the moat that Zeta has, what would you say is the biggest competitive advantage that Zeta has as a company?

Yes, we have a lot of moats around our company. But I think the one that is the hardest to replicate is the Data Cloud. This is not something that could be built overnight. You couldn't throw money at building this type of scale. You look at it, I do believe that in the years to come, there are only going to be 3 or 4 real winners in the large language model space.

I'd like to remind people that at one time, there were over 1,000 companies in the United States manufacturing cars or automotive parts. that's sort of consolidated down to 5 or 6 depending on how you look at it. And I think you're going to see a very similar thing in large language models. I think the biggest differentiator is going to be the data that you feed into your own models. And the fact that we have 535 million individuals who have consented to be in our Data Cloud that we have an average of 5,000 to 7,000 data elements per person, the fact that we're now sitting as a first-party component of the technology stack for over 5 million publishers sitting on trillions of pages of content that synthesize back what individuals are doing without needing a third-party cookie, meaning everything we do is first-party and deterministic.

It's just a game changer as it relates to the ability to take our clients' data, our data, merge it into one ecosystem and allow our proprietary models to do the work. These are the same models that are going to do the work around business intelligence. The information we ingest about credit card transactions, clients' credit score, real-world location, all of these things are mission-critical to the ability to make high-quality business decisions.

The other thing that's really interesting is if you think about sort of once again, the evolution of the business, the purchase of the loyalty program that we did late last year in the purchase of Marigold, we're now feeding SKU-level data into the models for the enterprise clients that use us for loyalty. So we're seeing massive game changers there. And back to loyalty, loyalty is part of the ZBI, right? You've got a platform that is ingesting intelligence. Yes, it's trying to get the consumer to buy more. But more than anything, you're ingesting every single transaction down to the SKU level.

What should you be doing from an inventory management perspective? What products should you be creating and rolling out? What products are not selling that you should be killing? All of these are decisions that we made purposely over the last year as we've tried to pivot the business or evolve the business, I should say, from what has been pure marketing into what we believe is going to be marketing and business intelligence.

That's great. That's super helpful because we are emerging. And so one of the things that we get, the questions we often get is what is an AI infrastructure, intelligent infrastructure company. What does this mean? I think if I put what you were saying back, it seems it's speed to market, it's ingesting data, it's analyzing the data and doing more than just one function. But I'd love to hear your thoughts.

Yes. The way I think about it, let's go to this slide again on the compatibles, Madison. But the way I think about it is when you're selling somebody a software product, you're a line item in the P&L that the CMO has to justify every year to the CFO. If you're infrastructure, you are mission-critical and necessary to the operation of that business. And I think that's the single biggest difference.

And back to why we think net retention rate is such an important metric inside of what differentiates an infrastructure platform from one that isn't. And by the way, I'm not talking about semis and we're certainly not building hosting centers. But the reality is that when you look at these other organizations, they are the ones that we think of as infrastructure. They are the ones that are embedded in their clients. And the big difference is not only are you mission-critical to the operation of your clients' businesses, but you're a revenue-generating source not seen as an expense.

And Zeta has been that for quite some time. When you're returning 600-plus percent return on marketing spend, your clients look at you as a revenue generation platform, not the other way around. But now when you look at the adoption of Athena and our goal of getting all of our clients to an average of over 1,000% return on marketing spend, that accelerates even further. And when you look at adding this business intelligence layer, it's going to add that even further. So when we refer to infrastructure, we're referring to organizations that are fully embedded and are a component of the client's tech stack that every year, the line item is not even looked at. The CFO just says, yes, that's infrastructure approval.

That's great. I think that's a really key differentiator seeing not as an expense, but as a revenue-generating source. That's the ideal. I think you used the term service of record or that's the...

System of record, yes.

System of record. Chris, one of the questions we often get is Zeta's monetization model, and I believe it's utilization focused versus subscription. Just talk to us about the visibility of these 2 models. Why is Zeta's utilization and why that's the right approach going forward?

Yes. So first, we'll continue to have a mix of each. But what David is focused on throughout this call is the outcomes that we generate for our customers. And that's why utilization is an important component of the revenue model. If you think about utilization, it's really the point of impact where customers experience the byproduct of the data and the AI working in tandem. And the fact of the matter is our flywheel spins faster when our return on marketing spend is superior to the competition.

So when we see those returns, which we can see inside the platform, our customers can see the returns they're seeing in real time, it makes utilization a very important and predictable component. It's about 40% of our total revenue base, whereas recurring software is about 60%. Combined, that revenue is predictable because so much of our revenue, over 90% is driven by customers that have been with us over a year and almost 3/4 driven by customers that have been with us over 3 years.

And I think the -- what's important to me, too, is just how reliably both of those components, recurring and utilization have been growing. In fact, over a multiyear period, recurring revenue has grown 36% and utilization revenue has grown 25%. So not just predictable, but both growing at a very high and rapid rate.

Well, the funny thing is that -- and I'm not saying it's exactly the same thing, but you could look at our utilization fees as a Zeta token. And that might be something. And we're looking at the scorecards and what we're building. But when you think about it, we're charging them to utilize the platform. We're not charging a per seat fee which has allowed us to scale. It's the same thing that the large language models are doing with tokens, right?

You've got some companies that have decided to move to pure utilization and some companies that are pure SaaS. We've always been sort of 50-50, 60-40, and it's been very healthy for us. And listen, we've reported 19 quarters as a public company. We've beat our guidance and raised our guidance 19x. At some point, people are going to start giving us credit for having high-quality visibility into our numbers.

Right. No, that's -- and to that, I mean, I think there's a growth target or at least say this 20% organic growth and free cash flow margin expansion, I think, are some of the key North Stars of the business. Just talk to us a little bit more about why those are the right numbers. I think I'm sure I'm going to get -- getting questions, why is that the right number? Why can't we grow faster? Maybe David?

We grew 29% in the first quarter organically. We're not modeling to that for the next 5 years. What we're saying is we believe we can be a 20-plus percent organic grower at scale for many years to come. We're not just looking at this year or next year. We're looking at many years. So we see that as a minimum number. And quite frankly, I think we've surpassed it. As I said, we're on our fourth year of an average of 30% plus organic compounded growth, higher than that for EBITDA and higher for that for free cash flow.

But listen, the market is not really sort of paying us to project higher at this point. And we'll continue to leave some conservatism in our model. Chris already insinuated that, and I think we've shown that. And we'll always try to beat and raise our numbers. Our goal will be next year to be talking about 23 consecutive quarters, the year after that, '27 and so on and so forth. Chris?

Yes. David, you said it well. I mean we'll often get questions too, Ron, on what are the drivers that give you that level of confidence. And as David said, it's been 4 straight years over 30%. It's been 7. This will be the seventh straight year of growing revenue over 20% organically and expanding free cash flow margins.

So when you think about the model itself, the top line model is really driven by 2 parts. First is our ability to see and predict and grow new superscaled customers. The key component there is we are not beholden to RFPs, which can be extremely large sales cycles. Things go bump in the middle of the night, someone leaves on vacation, you don't get a deal signed. We signed $50,000 to $150,000 proof of concepts that have a faster return on marketing spend. And a superior return on marketing spend that allows us to then get our foot in the door and then grow reliably, which then in turn leads to the second part of our revenue model, which is how customers grow their spend with us. David talked about the addition of new channels, the addition of new use cases. We've consistently beat on both what our model growth rate is for adding scaled customers.

Our model is to grow between 4% and 8% year-over-year new customers. We've grown double that on the high end. We've averaged over 15% and to grow their spend with us year-over-year between 12% and 16%, we've averaged over a multiyear period at the high end of that 16% range with last quarter being 21%. On the bottom line, we know what we deliver from a channel perspective, the margin profile. And we've begun to move work where it can be delivered in centers of excellence, not just by labor forces that have better price points, but more importantly, labor models that are relying on automation. And that's what's been allowing us to grow free cash flow and now GAAP earnings even faster than revenue.

And so I have to ask, Chris, because I think you mentioned we have longer-term targets, and we've achieved those early. I think we have 2028 targets out there on revenue. And that was before the Palantir announcement and before the potential ARR that's coming down the pike. And so help us understand a little bit more how you feel. I know you said meeting earlier, but how you feel the target will hold?

Yes, the best thing about it, Ron, is when David and I get to sit with investors and analysts like yourselves, we don't have to go off feel. We provide you with metrics, right? We tell you just what I stepped through, how fast we should be adding scale customers, how fast their spend should be growing, what the net revenue retention rate of the business should be. Quota carriers, the mix of our business on every one of those components, we publish what the model is to get to the long-term target, and we publish how we're performing against it every 90 days. And we are in a position because of the execution that David talked about to be ahead on every single one of those metrics. And I'll leave it with that bodes really well for getting to the 2028 targets in a timely, if not faster manner.

And so maybe as a follow-up to that, when we think about NRR, and we talked about the comparable slide that you brought up earlier, what do you think are the distinctive characteristics that drive that NRR going forward? Obviously, we have these new products. I'd be curious about the upsell capabilities and things along those lines.

I think a lot of -- yes. I'll just start, Chris, and then I'll throw it to you. I think a lot of it to date has been this massive return on marketing spend, where clients are looking at for every dollar they spend in our platform according to an independent Forrester survey, we're returning 600% return on marketing spend. And as I've said, we're seeing upwards north of 1,400% for clients who have -- some clients who have adopted Athena. So I think we're going to continue to see that happen as you add new products like business intelligence, it's going to continue to drive a higher net retention rate.

Chris?

And the fastest-growing parts of our business, Ron, are exactly what David described. Customers that are using 4 or more omnichannels with Zeta are growing 40% year-over-year. Customers on more than one use case are growing over 50%.

That's super helpful. And then I have 2 more questions, and then we'll let you go. We've got 4 minutes left here, but the questions keep coming to mind. David, I have to ask with the 14x for some clients who have adopted Athena versus the, call it, 600x -- 6x overall, what do you think is the difference? Obviously, Athena is the answer. But what did Athena do to enable that 14x?as a faster time to change approach if things weren't working? Was it just out of the gate, things were just better? Help us understand that because now 14x is something that we'll be thinking about quite a bit.

And I want to be clear. So if Chris were here, he'd be kicking me, where the good news is we're in different states, so he can't kick me here. But what I would tell you is that not all clients are at 1,400%. We've got a number of [ clients ] who are at 1,400% who have adopted Athena. I think the biggest thing is most of our clients don't know how to use all of the features or access all of the data sets. So instead of going into a platform that you have to navigate everything, you can now literally say, Athena, how can I create meaningful greater customer growth at a comparable cost? What marketing would you cut and what would you add?

By the way, Athena can ingest all of their marketing, whether it's linear, whether it's their relationships with other platforms, and it's able to, in a very neutral way, show them how we can drive substantially higher return on marketing spend than they are currently getting from other vendors. I would tell you that new use cases and new data sets have been the single biggest Athena wins thus far as it relates to meaningfully stepping up the return on the marketing spend.

That's great. And maybe as one of these last ones, and that gives us a lot to think about the 14x and the new use cases and data and also how voice enables users to find more of the features. What I wanted to maybe wrap us up on is we have Zeta Live coming here soon in early October, if I'm not mistaken. And so I think, David, you sort of teed up to like we're only halfway through the year, and we had some pretty big announcements. So talk to about what we might expect for Zeta Live here and going forward?

Well, listen, I don't want to spill any tea and get my team super agitated with me. But we've been pretty open about the fact that our goal is to turn Athena into a full operating system for our clients' businesses. I'm not sure we'll get there by Zeta Live, but we're certainly trying.

And what I would tell you is that I demoed Athena and her capabilities to a number of Fortune 500 CMOs or decision-makers and 4 of them during the course of meetings stopped the demo and asked if they could license this platform to run their entire company. And I said, that's the goal, right? And if you think about the first quarter of 2026, we publicly announced that we auto generated greater than 75% of all of our new code. We've created a platform called Spade, a base level programmer can go into Spade and say what type of code they want.

Spade then looks at which large language model is best to develop that piece of code at the lowest token utilization. So by way of example, for cyber, it might be Claude; for complex programming, it might be GPT; for a publishing program, it might be Gemini. We use Cursor inside of Spade. It then designs the code. It then sends it to [ Zippi ]. I swear I don't know who makes these names up, but it was probably Chris Monberg.

But [ Zippi ] then QAs all of the code, sends it to an architect. So you've actually bypassed the engineering layer. The architect that gives it a review pushes one button and it goes generally available. It took us 9 months from when I came up with the idea for Athena to launching her generally available. 3 months later, we launched an entirely differentiated platform inside of Athena for agencies. That's how fast things are accelerating.

So as you think about Zeta Live, I'm not even sure everything we're going to roll out. But we've got some incredible talent. We've already announced we've rented the David Geffen theater inside of Lincoln Center. So we're going to have more room than we had last year. We're going to be doing an insane dinner with some unbelievable entertainment. I don't have permission to disclose it last year. The Chainsmokers on the deck of an aircraft carrier will be tough to beat, although I think we might have it. But what I would tell you is that, the single most important thing is that the 2,000 people that will be in the room where it happens, will be the decision-makers that make the decision on greater than $100 billion a year on marketing spend. That's who attends Zeta Live. So super...

Remind us the date?

What?

Remind us the date again? October...

Sorry, Ron. Yes. October 8 in New York City, if you'd like to attend, please go to our website and sign up. We do have some curation layer, but the reality is that I think it's going to be an unbelievable year. We've announced Kevin Hart. We've announced Lindsey Vonn. I mean, talk about tenacity in Lindsey, talk about a guy who understands entertainment and business in Kevin Hart. As I've said publicly, I think beating Tom Brady and Serena Williams last year is going to be tough, but I think both of them are great.

And by the way, we have 2 or 3 more game-changing keynotes who have said yes verbally that we're trying to get through the process with so that we can announce them. But I will tell you that last year and the year before, the product road map presentation that Chris Monberg and Neej Gore give, we have people sitting on the floor. You couldn't get into the room. So we're moving it to a bigger room this year so that hopefully, we can get more people in. But I'm incredibly excited about what's going on at Zeta. It's hard to believe when you look at the level of partnerships, the level of growth, the level of productivity, the new products rolled out that were only in the first week of July. It's -- and I'll say what I always say, Ron, we are just getting started here at Zeta.

Well, that's a perfect way to end it. Thank you for taking an hour of both of your time to do this. And as we enter the quiet period, we're in quiet period, I guess we'll be talking in the next few weeks on earnings, which is exciting. And with that, that's probably a great way to wrap it up. I don't know, David or Chris, if you have anything else to say? Or with that, maybe we'll just call it a night.

Chris, do you want to add anything?

Thank you, Ron.

Ron, thank you very much. We really appreciate you putting this together. When you asked us about it, we thought it was brilliant and thought it went very well.

Wonderful. Thank you for your time. I certainly learned a lot. So I look forward to the next update.

Thanks, Ron.

All right. Thank you. Bye-bye.

Are we ready to go?

Yes.

Yes, we're live. Okay. So David, Chris, thanks you for joining us. I'm Ron Josey, covering the Internet sector here at Citi. And look, I had a ton of questions prepared for this, and there's late-breaking news for -- really for company. So David, why don't you take the stage and tell us what was announced?

This morning at 7:00 a.m., we announced we're fully integrating our platform with Palantir, and we're signing -- we signed a 7-year joint go-to-market strategy where we will be marketing to the vast majority of their customers. Putting our data cloud on top of Foundry is going to speed it up dramatically as it relates to data ingestation from our existing and new clients, and it's going to allow our algorithms to train even faster. We've built a joint forward deploy engineering team where we have one, they have one. And as we go to market, this is really one of the biggest pivots we've ever made as we move the company from what is a marketing technology company to what is really an AI infrastructure company.

If you think of our business, for those of you who don't know, we are a very large marketing -- I shouldn't say very large to this group. But as it relates to marketing platforms, we have a marketing platform, and we have a data cloud. Today, 51% of the Fortune 100 and 24% of the Fortune 500 use our platform to better manage customer acquisition, customer retention and customer monetization. We do that through what's called the Zeta marketing platform and through the Zeta Data Cloud.

In 2017, we made the decision to totally throw out our old platforms, when we ran them for cash flow, and we've created an entirely new platform that in 2021, we made Live which is the only marketing cloud that has data and artificial intelligence as native to the application layer. It's literally foundational. And when people tell me they hear this in every meeting they go to, I suggest they go back 5-plus years ago when we went public. We put a big sign on the side of the New York Stock Exchange that data plus AI equals intent. Everybody was asking me "who's Al? And why is he in charge of your data strategy?" But we've been doing this for a very long time.

Today, one of the biggest and most powerful statements about our platform is for every dollar that's spent on our marketing platform, we return an average of 600% to 700% return on marketing spent. That's an independent study that was done by Forrester. Our goal is to get that to 1,000%, an order of magnitude return for every dollar that's spent. And everything we do as a company is to try to get there.

A few months ago, we launched Athena, which is our voice-enabled super agent. If you think about enterprise software and not just us, any enterprise software company. I always make the joke about the Bloomberg terminal. What percentage of your Bloomberg terminals features and functions do you use on a regular basis? The vast majority of enterprise software people use between 5% and 10% of capabilities.

Athena you're no longer navigating a complex platform. You're going to Athena and simply saying I would like to create 2 million incremental customers this quarter, and I'd like to do it by lowering my average expense by about 7%. Athena will then access every feature, every function in every data set to come up with the best possible outcome. It will then change the user interface to show you the results. You can then say to Athena, amazing. That's a little big for me. I don't want to start with that. I'd like to do a test. Can you activate the test? And every hour sending an e-mail so I can keep track of the return on marketing spend per hour.

When you look at taking the Zeta Data Cloud and Athena and putting her on top of Foundry, which is Palantir's platform, it's going to massively speed her up. That speed every time we can remove incremental latency, we can return substantially higher return on marketing spend. And also the credibility that comes from partnering with Palantir and going to market, we think, is going to be a game changer for us as a company.

David, talk to us about how this came about, working with Palantir. We've been talking about Athena now for some quarters. We just launched not too long ago, and now we have a pretty big partnership here. So talk to us about the genesis of this and sort of where -- how you were thinking and when this came about.

Well, we'll try to get this going here, so I don't die. For those of you on the webcast, it is hot. So we are doing this as a web webcast from a Reg FD perspective because we literally just announced the Palantir relationship.

So listen, we've been building our business in a way that we've always really looked up to Palantir. We've always thought they were the type of company we would like to aspire to be.

So I was originally introduced to the CEO of Palantir through a very close mutual friend. We started talking and working together. And then I got to know Elias, who works directly with Alex. And the synergies just became so clear. Like if you look at what Alex said in the press release we put out, and I'll paraphrase here, Palantir plus Zeta should change the entire marketing landscape.

And when you think about it, Palantir's entire strategy, I mean it's -- I shouldn't sum it up to their entire strategy, but one of their biggest selling points to their client is we can help you meaningfully lower your cost and be more efficient with your business. Zeta is the best platform in the world for managing and lowering marketing and CRM expense. We're usually able to lower our clients' expenses anywhere from 50% to 75% on marketing or CRM expense.

I don't know if you saw yesterday, Ron, but Gap announced they were throwing 4 vendors out and replacing all 4 with Zeta as the system of record for Gap. That's the type of deals that we're winning at this point and you put Foundry underneath it. It's game changing. One of the interesting things about the Gap deal was we had bid the Gap RFP three times and come in second three times. And I was like, do we really need to waste our time with this again? And I ended up getting introduced to their CEO. And I said, "Listen, we're probably not going to bid it this year. You guys seem great". But they've been with Salesforce. I mean, they were a Salesforce shop forever. And he said, "bid it, just bid it".

We came in, we bid the RFP. Halfway through the RFP, Athena became generally available. So we went in and we did a full demo in the middle of the RFP process of Athena, the voice agent. And the next day, they asked us for pricing. So you start to look at Athena and where she is going in the marketplace, in partnership with Palantir and where they are in the marketplace.

This will be our fourth year of greater than 30% compounded organic top line growth and greater than 50% compounded organic EBITDA growth. And I think that Athena in partnership with Palantir is going to make us a 20-plus percent organic grower for many, many years to come.

That's a great segue. I think there were some dollar amounts maybe in the press release. Chris, talk to us about how you see the partnership. Obviously, David, just put some numbers and growth rates around there, but talk to us more about the dollar amounts, but then I want to get into execution too. So let's talk dollar amounts and then executions.

Yes. So I'll ladder down from our Zeta 2028 plan, which is to get to at least $2.3 billion by 2028. I'll remind investors that we got to our first long-term model to get to at least $1 billion a year early. Integral to the current long-term model we're operating in is partnership expansion. We've announced a number of partnerships over the last several months, this being the biggest inner history. But those partnerships being a vehicle to not just get to that longer model, but to get to there faster.

David had mentioned in the press release, we see this becoming at least $100 million business over the years to come annually, and there's joint selling incentives and partnership to drive that type of outcome.

So let's talk about the process to get there then and dollars will be dollars, but talk to us about -- we just had the announcement, Athena is 3 months old?

Yes. It's crazy.

So it seems like quite a catalyst and then we just talked about Gap. But David, talk us through the implementation process from here on out. What can we look forward?

So one of the big questions that -- I mean, I've been asked in the last hour. So it's not like I could say we've been doing this for a long time at this point was, is this going to mean to a meaningful increment in expense this year? The answer is no. In fact, we were able to move budget from other providers, to Palantir to have it being neutral this year, and we think this will help us lower operating expenses further starting next year and out.

Based on -- we've been obviously building this for quite some time. We're announcing it today. But I think we will have this live by the end of the third quarter of this year. And we're targeting even sooner than that. And we've already really, at a very high level, started identifying customers we want to jointly go to. And I think the joint go-to-market strategy is going to be a real game changer for us.

The other thing that's really important to note is when we turn it on. Our existing as of the end of Q1, 189 superscaled customers will all benefit that day. The speed in gestation, the model the model is speeding up, Athena's capabilities speeding up will all be available to every client.

That's great. That's fascinating. As we think about that scaling up I wanted to ask a little bit more just, I think you said, David, when we first started this webcast, speed will drive greater usage of Athena. And we just talked about OpEx may be going down with the scale.

Speed drives greater return on marketing spend, which drives greater spend on our platform.

So talk to us about the speed aspect of this.

So right now, we are probably the fastest you can get out there in marketing because we're the only one that AI is native to our application layer. Everybody else has to step out of their platform through an API to an algorithm to do a query, then the algorithm steps through another API to a data repository to draw data back to the algorithm to create intelligence and then it informs the platform what to do. That latency destroys return on marketing spend.

By adding Foundry underneath and replacing other infrastructure we've used in the past, we're going to even meaningfully speed up our ability to process data. And that will allow our algorithms to train faster. We're already seeing -- for those of you who don't know, Athena went live in March, and then on Thursday, we announced Athena for agencies. So now it's available to 100% of our total client base.

We're already seeing clients who are using Athena spending more on our platform and showing meaningfully higher than a 600% return on marketing spend. We think that Foundry's addition to that is going to supercharge that even further.

To that 10x goal...

The goal is to get 10x. I'm not saying we're there or I'm not saying we're not there.

It's a path to get there.

Yes. What I would say is we have some clients who are there, and we're trying to get that there.

So we can take questions from the audience. If there's any questions, I've got one in the back. I'll just repeat it, if you want to say.

Really exciting announcement this morning. May be a little bit more on the specifics of the Palantir deal. I assume you are now becoming a customer. Can you talk about how much you're going to be spending with Palantir? Can you say it's built on Foundry. Is there some sort of like OEM rev share [ premium ] for Palantir? Obviously, a really sophisticated data intelligence platform [indiscernible] to...

We're aware. And you're a client of ours.

What exactly [indiscernible]...

Well, so first of all, I cannot...

If you could repeat the question for the [indiscernible].

How much are we spending with Palantir? And is there a rev share? Or how does the joint go-to-market work? Let me start by saying, we're not going to disclose how much it is, what I will tell you is it is neutral to Zeta's guidance from an expense perspective for this calendar year, and we expect it to lower our expenses going into next year on. So this won't be like, Oh, you'll get surprised by expense creep or picking up; B, we are both heavily financially incentivized to jointly sell this product together, to be clear.

Palantir's strategy, and it's hard for me to say this with a Elias sitting here, but Palantir's strategy is always to help clients with cost and efficiency. Today, and I think one of the things Elias really liked about it, I know one of the things Alex really liked about it is you've got a $1 trillion TAM that's effectively opening up to the Palantir-Zeta partnership that we can go in together as partners and say to their enterprise clients, we can help you meaningfully lower one of your largest cost as an enterprise, while simultaneously increasing quality.

So the answer to your first question is it's neutral to lower. The answer to your second question is, yes.

Maybe to that, but let's take a step back and talk about Athena because it all sort of comes together here but we're 3.5 months in. Maybe David, just what surprised you the most about this platform? And you talked about the client win yesterday that was announced. What has changed, I guess, right? #2 to #1.

It's funny. So honestly, I -- and we're a big company at this point. I don't invent most things. I happen to invent Athena. It was something that I came up with last April. We were able to go from the day we decided to build this to generally available with Athena within 9 months. Within 90 days of that, we rolled out an entirely new platform for Athena for agencies. If you look at Zeta in the first quarter of 2026, we announced that by using our internal workflow management tool, SPADE, we generated 75% of all of our new code on an automated basis.

The goal with Athena is to become a full operating system and to truly become the infrastructure that artificial intelligence for our clients. The Palantir deal is mission-critical to that.

Now to actually answer the question you want me to answer instead of the question I want to answer. The thing that surprised me the most is the reaction from people because I've been so intimately involved in the creation of what we call the birth of Athena, you see the little steps every day, it gets better. And all of a sudden, somebody sees it for the first time.

Winnie and I did a demo for somebody yesterday, CMO of a very large company. And he was literally like this is the greatest thing I've ever seen. He's like, this blows me away. And I was like, great, you should move all your budget to us. And of course, he started laughing. But we'll start with the test and we'll move up from there. But human reaction to Athena is the thing that surprised me. What didn't surprise me is people who are adopting it are spending far more on the platform and enterprise clients that are adopting it are showing amongst the best return on marketing spend we've ever seen. But that didn't surprise me.

And we've left 10 minutes in this fireside for Winnie to do a demo for your team.

Are we at that level now?

I think we have until 5 after, but just know that we've got that ready to be queue now.

Winnie is going to do a demo of Athena so you guys can get a little taste of what she looks like.

So we've got maybe 5 more minutes of Q&A.

That gives us 12.

12 minutes of Q&A. Okay, great. So maybe last one on Athena. I believe, what was interesting is that I think every beta customer became a multi-use customer when you were doing it. And so I -- what we're trying to understand is how adoption goes forward and monetization of Athena? And I might have this wrong, are we monetizing Athena now? Or where are we on?

We we're monetizing Athena every moment of every day. I mean the way Athena works is instead of charging a modular fee for her. We are letting clients use her as a tool inside of the platform.

Let me give you some interesting statistics. A number of years ago, we rolled out our first ever voice interface. It was called ZOE. And clients who adopted ZOE spent between 300% and 350% more on our platform than clients that didn't adopt ZOE.

Now I'm not suggesting all of our clients to adopt Athena, we're going to grow by 350%. But what I am suggesting is clients who utilize her are not just spending far more money in utilization fees, but they're using multiple use cases. So for those of you who don't know our business that well, we have three main use cases, and it's pretty simple, customer acquisition, customer retention, customer monetization.

For the end of -- for companies?

For enterprises and agencies. Yes. Although 80% of our revenue is direct to enterprise, about 20% is agency. Those are round figures. Bottom line, what we found was clients who use multiple use cases, and we were pretty siloed for a while because we were selling that way. We've now broken through the silos. And when customers use multiple use cases instead of one, the data trains the algorithms substantially more effectively. So the data that informs customer acquisition further informs retention, which further informs monetization, which further informs acquisition.

We're seeing a massive uplift in return on marketing spend for clients who adopt multiple use cases. We call it, as you know, Ron, One Zeta. And then last year, we hired a gentleman named Ed See. Ed was the Head of McKinsey's global CMO practice. So we stole that away from McKinsey to become the head of -- we call him Chief One Zeta Officer. I think his technical title as Chief Growth Officer.

But with Athena, Athena does the cross-selling automatically. So if you're using a use case around customer acquisition, you might say, "Athena, I'd like to drive 2 million incremental customers this quarter. I'd like to do it in an average of 7% savings on a cost to create a customer". Athena will walk you through that whole process at the end of the activation. So you'll say, "please activate that".

Athena will then say, by the way, do you know that I've identified 700,000 of your existing customers who are in process of churning off your platform to your competitors' products? Would you like to save them? And it automatically creates a loop into the One Zeta. So that's why we're seeing with Athena, a meaningful uptick in One Zeta clients. What was the growth in One Zeta clients in the first quarter?

Well, we didn't -- growth in multi-use case, which would be One Zeta was 50% year-over-year.

So 50%.

And to that 50%, maybe we can talk a little bit about the sales pipeline, which I think was up 40% year-over-year in the last quarter. So as we sort of migrate from big news today to Athena's live, and we're seeing results and wins and everything to implementing sort of the go-to-market and everything else in the company, pipeline was up 40%.

So talk to us just about how you view the pipeline going forward. We'd love to have some insights on conversion rates, if that's possible. Anything that can help us better understand that pipeline number and metric?

Yes. Our -- as you said, our pipeline is up, and that's our model, frankly. We have quarters where it's up 30%, 40%, 50%, 60%. This was indicative of what we've been seeing. We'll often get the question from investors, how is the market reacting to a tougher macro or choppier macro times. When we look at our sales pipeline and specifically the industries that are closest to consumer discretionary wallet, they are our fastest growing industries, growing even north of 40%. Athena...

And I think that's because we're a cost savings, right? So when things get choppy, clients actually grow with us. I'm sorry, I [indiscernible].

Yes. No, and they become far more data-driven. What we've also found while the longer-term benefit of Athena is higher ARPU or higher usage, which then translate to revenue. What we're seeing at the top of the funnel, Ron is better conversion. So it is an accelerant to the sales process. As David mentioned, with the Gap, but with many other enterprises, putting Athena as a demo at the very first step of the sales process, which is probably a good jumping off point for Winnie, is a conversion factor that is allowing us to accelerate time to close.

If we've got -- so great, we've got 10 minutes left. 11 minutes left.

You could probably go 5 minutes longer because we started late.

Maybe we can ask if there's any audience questions before we get to Winnie. If not, we'll go to demo. We have one in the back.

Have you guys disclosed or talked about like how we should think about the impacts to gross margins like the economics of the relationship? Should this show up even in your numbers like -- or is it something that will only show up once customers start using it? Like any kind of [indiscernible]?

Yes. I think -- I mean, as we're going to get it live next quarter. It's not currently in our projections at all for this year. So it could be potential upside, maybe, maybe not. Next year, I think it's going to be very additive to top line revenue and the revenue should come in at a very comparable gross margin to our other gross margins. Maybe slightly higher for a few reasons, but it will not contract.

It will be -- it will aid at a lower OpEx, as David mentioned. We'll get synergies off of it, which then translates to better free cash flow and adjusted EBITDA margins year-over-year.

All right. Well, with that, thank you for the questions. Maybe Winnie, you take it away?

Great. Thanks, guys. We're going to switch over to our demo and showcase. I'll be navigating on my laptop at the same time. So I'm Winnie Shen, EVP of our Data Cloud. We've been doing a ton of these demos here at Cannes, as Das mentioned, some really great resonance with our clients.

So our Zeta Marketing Platform is where our entire ecosystem lives in terms of Athena as an operating system, where our data is, where all of our activation capabilities are all of our intelligence, all housed within this one platform. So we're going to transition over to our platform.

And before we get started, what I'm going to walk you guys through is, first, how do we glean intelligence from all the massive amount of data that we have. We have the ability to reach about 245 million people in the U.S., 586 million globally. We enrich those profiles with thousands of attributes and signals. We then transform that massive amount of data into intelligence. And from there, once we've unlocked opportunities, we then want to determine how do we activate off of it, how do we optimize it, how do we find the right creates and then measure all of it.

So I'm going to actually take you through that entire flow. Athena will pick up all the noise in the background. So I'll try to mute us and I'll pause and try to explain some of what's happening in the background as I ask Athena questions. I'm going to do this from the lens of travel since we've all traveled here today, we thought that would be a good use case. I'm going to do it from the lens of United Airlines just so that it makes it more real. None of the data I'm showing you today is any customer data. We're just using it as an example to help showcase some of the capabilities. So with that, let me get Athena started.

So for those of you who don't know, while we're setting it up, today, our proprietary data cloud has 550 million people globally opted into it, 245 million Americans, which -- with an average of 5,000 to 7,000 data elements per person, which includes demographic, psychographic, transactional, credit score, behaviors across 5.2 million publishers that have embedded our technology as a first-party component of their tech stack.

So we're feeding in trillions of marketing signals on a daily basis to power everything she's about to show you. All of this data is 100% proprietary and has never been shared with a large language model ever nor will it. We do not sell our data to anybody at any price at any time. We simply use our data to train our own proprietary models. And we've been doing that since we founded the data cloud and rolled out our AI many, many, many years ago. Sorry, Winnie.

No. Perfect set up.

Hi, Athena, can you take me to the application that looks at always-on intelligence, and I want to look at comparisons in online travel destination research comparing United Airlines intenders versus visitors versus transactors.

[AI Agent - Athena]

Among United Airlines intenders, online travel research leans heavily into broad destination inspiration, such as United States and North America travel, regional travel, family and seasonal travel plus big dream categories, cruises, national parks, theme parks and Europe-Caribbean trips. United Airlines visitors show more research around specific U.S. leisure hubs and short call getaways, e.g. Florida, California, Texas, Las Vegas, Orlando, New York, San Diego, Palm Springs, indicating strong interest in domestic vacations and repeatable fly to Sun destinations.

United Airlines transactors are more skewed towards practical trip enablement and concrete planning with higher relative focus on hotel reservations and mid-scale lifestyle hotels, car rentals and ridesharing, vacation packages and deals and travel services content like TravelAdvisor and TripAdvisor. Intenders over-index more than visitors and transactors on long-haul international inspiration, e.g. Australia, Africa, France, Italy, Spain, Greece, Portugal, Brazil, Caribbean, Indian Ocean and on aspirational experiences such as private jet travel and luxury hotels.

So this first part, she basically has compared 3 different reports all in one conversation. That meeting Das referred to, the reaction was, that was just a day's worth of work that we just compacted into a few minutes. Athena just tend to go a little bit. But basically, she's looked at people that are reading about traveling to United versus people that are visiting their actual terminals versus people that are spending with them.

She's identified the differences amongst them and identified an opportunity with the international travel. You'll notice that Spain -- you guys probably can't see. I will just tell you about it. So from the people in the back of the room, you'll see that Spain travel actually indexes high for intenders. We don't see that with the current people that are actually with United. So I want to pull that forward and see where does that ranking happen in our generative engine optimization. So I'm going to push it forward.

Okay, Athena, that's a great opportunity on the international side. Can we go GEO and see how does United rank on questions related to the best international travel brands?

[AI Agent - Athena]

United is the #1 brand in the entire network. On the generic international long-haul overseas prop specifically, it is mentioned in 100% of them. Its average share of voice is regularly the highest in the answer pool. It owns the broad best U.S. airline for international travel territory, beating American and Delta on most network reach and reliability prompts. For example, United ranked #1 for share of voice for one airline for all international trips, best long-haul international and wide international route networks.

Well, how does United compare when it comes to Spain travel?

[AI Agent - Athena]

On the 20 Spain, Madrid, Barcelona prompts, United strength evaporates. United is still mentioned in nearly 100% of responses, so it's not invisible, but it rates #1 on 0 of them. It's its third or fourth place in almost every case. United's average SOV is well behind the leaders. Iberia dominates Spain. The Spanish flight carrier leads around 15 of the 20 prompts hitting 61% to 70% SOV on the strongest. And American is usually ahead of United too.

American wins the most nonstop to Spain prompts, thanks to its Iberia oneworld joint venture. United's best Spain showings are only #2, Barcelona and Honeymoon to Barcelona. Overall, United is the category leader for international flags generally, but on Spain corridor, it's a clear laggard, consistently third or force behind Iberia and American.

So what we're seeing our Generative Engine Optimization is that we can hypothesize the type of questions people are asking in LLMs because we see such a massive amount across the open web.

So for those of you guys who don't know what GEO is, it's actually the ability to read the output from the large language models. So this is data coming right out of direct integrations we have with OpenAI, Anthropic and Gemini.

So one of the things our clients this year are really interested in is how are they doing in the large language models. And that's what this platform is showing you. It could just as easily show you this through any other methodology and we have already announced that we're able to serve ads into OpenAI.

So I actually think we're the only platform in the world now that matches to the MetaID, we match to the OpenAI ID, we match to the Palantir ID, the open web, connected TV, messaging, so on and so forth. But GEO is the hottest thing our clients are talking about right now and where they're investing, our GEO business is one of our fastest-growing businesses inside the company.

Exactly. It's been very powerful for us as a unique opportunity with our capabilities and innovation. And from here, I want to then understand, okay, with current travelers, are they actually interested in traveling to Spain? Or is this white space for me with acquiring net new customers? So I'm going to continue the conversation.

All right, Athena. So I want to check in leading indicators for my United visitors if they are actually researching about Spain or no.

[AI Agent - Athena]

Within this United visitor audience, interest in Spain travel itself is not surfaced as a top direct leading indicator signal. So Spain-specific intent is not present versus other destinations. Destination-wise, Europe hubs like France Travel and Italy travel are showing more pronounced steady growth as leading indicators, suggesting European near-term trip planning is prevalent yet country-specific.

Broader vacation content is also on a gradual upswing, indicating these customers are actively researching reviews, routes and tips for upcoming trips, including Europe, but not clustering specifically on Spain yet.

So what you're seeing here is people that are actually visiting, the level of interest in Spain is not there at all. They're not researching anything about Spain. Now if I flip it, Athena called out France, for example, as one of the destinations. So then if I switch over to France and look at those signals, you do actually see intent. So it's not that they don't have international travelers. They ranked high in that, in GEO. They -- as leading indicators for France and Italy. They just are not seeing that in Spain.

So this is a great opportunity for me to extend. Now I'm going to take you guys through how do we actually create a media plan around this opportunity here. Any questions so far as we move into the actual execution of it all? Yes, please.

[indiscernible] Can you say how much of this was like someone [indiscernible]. Was this moving something that's a little bit more technical for a broader audience [indiscernible] starts asking questions [indiscernible].

Yes. We like to call it democratizing intelligence. So you don't need a Winnie. I don't want to get myself out of a job to ask that. You don't need me to help navigate...

We're keeping you here forever.

Okay. Perfect. Put that on paper.

So to be a little more clear, we are seeing a 75% reduction in man hours for clients that are utilizing Athena. So what used to take 100 hours now is taking 25 and it's allowing different levels of the organization to access data that they could have never gotten to before. They would have had to go to a third party in their organization to put a report together to bring them 1/10 of this information would take hours to create.

How does the [indiscernible]. If you think about that particular [indiscernible], there's a portion of time required by Athena to [indiscernible].

This does all of that automatically.

And then there is [indiscernible], so which one seems more [indiscernible]?

So that's a great question. The most reduction we're seeing today is in segmentation. It's in the creation of the right audience for what they're trying to get to because that's where you had SQL and that's where you had a lot of big data pools. As it relates to the execution of the marketing, which she's about to show you, that's probably not as big a reduction because it wasn't as time consuming to do that.

And the preparing of the data [indiscernible] segmentation.

Correct. The data cleansing is all in the segmentation. Everything you're going to see and she's been doing thus far is totally cleansed data down to a deterministic individual. We're not looking at third-party cookies. This is literally looking at 245 million active Americans as they go through 5.2 million publisher sites where our JavaScript and our first-party tracking pixel sits on every page they publish. So you're talking trillions of pages. Let's move into the execution because I want to make sure we get to that.

Great. So Athena, I have $1 million budget to actually use with my partner hotel brand. Can you tell me how I should actually be spending on the different channels and media budgets?

[AI Agent - Athena]

Pulling up Performance Advisor, here's the recommended channel mix and where to rebalance spend.

Great. Can you actually project out for me for the $1 million and the destination opportunity that you identified, create a media plan for me to drive those vacation bookings, please.

[AI Agent - Athena]

This plan is designed to move travelers from inspiration booking. We allocate 65% of the budget or $650,000 to CTV because Zeta Data Cloud signals indicate this audience has a strong propensity for premium streaming content. That investment is expected to generate high video completion rates of over 99%, ensuring United Airlines brand message is fully consumed in driving awareness and consideration and scale. We then dedicate 10% of budget or $100,000 to OLV to extend incremental reach beyond CTV and reinforce key messaging through additional video touch points.

Finally, 25% of budget or $250,000 is allocated to display, where retargeting audiences exposed to video messaging helps increase engagement and drive stronger click-through and conversion rates. Together, this creates a connected full funnel strategy where CTV builds demand, OLV reinforces it and display converts high-intent travelers into bookings.

So we basically took the KPI of interest to get those bookings to happen and then look at if it is a new client without any data, we would look at our data cloud and what are the recommendations based on what's going to drive conversions. If it's already a client that's running, we would leverage all the data that they have in the past campaigns to then make recommendations on we want to create awareness first with CTV and online video and then pair that with display to drive to conversion.

All of this is totally dynamic, and it is recommended based on the deterministic individuals that are being targeted and how we know is the best way to communicate with that exact person. It's not just saying, "let's try these different things". It is all 100% data-driven.

Exactly. And from here, we know what is going to be the right type of campaign that we're going to run. Now I want to understand the right types of creates to run.

Okay, Athena. Can you help me actually pick out the best assets that are suited for this particular plan?

[AI Agent - Athena]

Based on their creative scores, this is the top-performing asset for the plan, a video built for high-impact CTV storytelling that scores strongest on engagement and brand impact, making it the best fit to lead inspiration and drive booking intent.

[Presentation]

So that was selected based on what we saw in signals as what's going to be most resonant to this audience.

So depending on the enterprise, we can either auto generate content in real time, creative or we can have as large a repository as that client would like us to have, mostly because we work with very large enterprises, they have an agency. And their agency is doing their creative. We never want to disrupt the relationship with the agency. So we'll take their creative. It goes into a very large repository of all of their creative and then the system automatically chooses the creative it thinks will resonate most to this audience.

It's really interesting, like if you look at one of our clients, which is a large seller of pet products through retail stores. God forbid, you send a cat image to a dog person and vice versa. I use that example because it really makes sense, right? If somebody is a dog person, you want the hero image to be a dog. If somebody is a cat person, you want the hero image to be a cat. That is like 101 class-wise of what we're doing here. This is 1,000x to data being driven into the decisioning for which creative gets picked. Winnie?

I'm also going to check, Athena, can you show me what I should actually advertise for display as well?

[AI Agent - Athena]

Here's the recommended display created for your plan. A 300x250 banner built to reinforce your brand and drive booking consideration as audiences moved from CTV-led inspiration toward conversion. The overall creative score is strong at 81, including brand guideline adherence and reusability across channels.

Okay. Can you go ahead and build out the campaign and prepare the campaign for launch?

[AI Agent - Athena]

I've built a campaign using the forecasted media plan and recommended creative. The campaign is set to run for 30 days with a $1 million budget, $650,000 allocated to CTV, $100,000 to OLV and $250,000 to display. This gives the campaign a balanced mix of premium video reach, incremental video exposure and lower funnel booking support. I've applied the high-intend consumers audience, assigned the recommended CTV, OLV and display assets, connected the vacation booking event and set CPA and VCR as the campaign goals. Campaign setup looks good and is ready to submit for activation.

Perfect. Now that the Zeta team has approved the actual campaign, let's go ahead and start monitoring the performance. Can you go ahead and create a dashboard to help me monitor the campaign performance?

[AI Agent - Athena]

Sure. Here is the dashboard to monitor your omnichannel campaign performance.

Okay. There's a lot of metrics here. Can you help me interpret how my campaign is actually performing?

You'll note this campaign, which is real...

[AI Agent - Athena]

Sure thing.

Had a 14x return on market [ expense ].

[AI Agent - Athena]

Looking into simulated dashboard over the last year, you spent about $1.1 million, generated roughly 12,413 bookings and brought in about $14 million in revenue. That works out to a cost per booking of around $88.84, a return on ad spend of about $12.77 and an average order value around $1,134. Overall, those numbers suggest strong efficiency and revenue impact.

Perfect. What about the best channels that I'm operating in?

[AI Agent - Athena]

Based on the data, display is your top channel with about 7,965 bookings, followed by CTV at roughly 4,357 bookings, OLV trails with just about 91 bookings. Display and CTV are clearly performing best for you.

Perfect. And what about the audience?

[AI Agent - Athena]

The strongest audience segments are summer travelers with about 2,984 bookings, travel enthusiasts close behind at roughly 2850 and look-alike audiences with around 2,372. Travel intent and retargeting audiences each delivered about 1,830 bookings and families with kids contributed around 553. These are the groups driving the bulk of conversions.

Perfect. Thank you, Athena. Great. So hopefully, you guys were able to see how do we take that massive amount of data, find opportunities for our clients to be able to transform that into a media activation plan, execute it, find the right -- well find the right creates, then execute it and be able to measure the impact of every dollar spent.

Today, I think you guys saw the power of this and how fast it is. When it's put on top of Foundry, it's going to be substantially faster. It's going to be substantially smarter.

So the reason the Palantir deal is so important to us is, A, the capabilities they bring to us from a found -- easy for me to say, a foundational level; B, the joint go-to-market, which is going to be very, very powerful; and three, our real move from being a marketing platform into an AI infrastructure platform.

I want to go back to what I said earlier about Athena, 9 months from idea to generally available, 90 days to launch her as fully and up agency capable product. We are turning her ultimately into a full operating system where this data will inform all of our clients' decisioning.

Let me give you a perfect example. This is a real example we've already done. A very large client, who's a very large hotel chain, was trying to figure out where to put their next 500 hotels. We researched, we took every one of their existing clients. We looked at every credit card transaction that they did in every market that they were not. We segmented it to their competitors from a landscape perspective, how many room nights were they consuming from their competitors in a 5-mile circumference around that competitor in markets they did not have hotels. And they made the decision that those markets where they had the most hotel nights for their most loyal customers were where they should put their next 500 hotels.

As they've opened those hotels, they're using us to launch the marketing and do conquesting against the competitors and move those consumers to them. That is the convergence of business intelligence and marketing in that case. But when you think of Zeta, we really think of ourselves more as a business intelligence/infrastructure company that chose marketing because it has a $1 trillion TAM. So it's something that we've been focused on, and we've been working through. You had a question back there, and then I'll go to you.

Yes, about the reference architecture. So it sounds like Foundry's [ going to ] get you guys efficiencies, [ obviously ] better answers, et cetera. But what was -- what's the [indiscernible] without giving too much away [indiscernible] architecture [indiscernible].

Yes, a lot of AWS. And by the way, we love AWS, we love Snowflake, we love a lot of our existing vendors, but we have moved some of the spend we were doing with others or spend we expected to increase to Palantir to replace those vendors in our tech stack.

Did you have a question?

Yes. I just want to -- can you help me just to understand the impact of the partnership with OpenAI and [indiscernible] levels [indiscernible].

The question is, how does Palantir affect OpenAI. It's not going to affect it in any way other than today, we use OpenAI to be the voice engine for Athena. The reason that's important is every time an individual enterprise or agency individual uses it, it gets to know them better. So it really begins to understand inference in addition to just what you're saying, okay?

So when we put Athena on top of Foundry, they'll stay independent, but the data that's fed to Athena will move much, much faster. And there's no problem there. As it relates to the advertising ramp of the open AI relationship, it's ramping as we expected it to. We're trying -- everybody is right now trying to figure out how to do it.

This is very emblematic of a number of partnerships we've started over the years. It actually reminds me years ago when we helped to build the API that helped a little company called Meta get into the advertising I think it's one of the reasons we're the only platform that I know of that matches the our ID to the Meta ID. And that, of course, is an amazing company and been an amazing partner. But when they first started, they were trying to figure it out.

So we're seeing it ramp nicely. We have a lot of demand for it. Enterprises want to be in it. I think a lot of it is sort of the new and cool factor. So we're super excited. I think we're the only platform that is partnering with OpenAI both ways. Let me -- go ahead.

Just on the comments earlier about reducing the Palantir [indiscernible] cost savings. What is the...

In out years, we're not suggesting this year, it's going to lower cost.

Is the like system and other technologies [indiscernible] consolidation, is that for application SaaS factors or is it more like, we talked about this data infrastructure [indiscernible].

Yes. We don't consume a lot of SaaS-based technology because we are a SaaS platform, right? And most of our architecture is our own. The only time we've used third parties in the past has been around cloud compute, primarily through either AWS or Snowflake, which, by the way, we've openly talked about our partnership with Snowflake. And I think it's pretty well known we're an AWS shop.

Although yesterday, when Gap announced that they were making us the system of record. We're doing that in partnership with Google Cloud. We're totally sort of, I don't know if Switzerland is still considered totally neutral, but we're very much like Switzerland in that we're trying to stay very, very neutral.

The reason Foundry made sense from a migration perspective was the quality of their architecture. And quite frankly, a lot of their architecture maps really well to our infrastructure today. And we've already done -- I mean, listen, we wouldn't have announced this today if we hadn't been building this for quite som time, right? So it's where we've been in process.

[indiscernible] some of the holding companies we met, they say they're still cautious about building tools [indiscernible].

Great.

[indiscernible] LLM data that just...

We don't feed our data into any large language model ever nor I believe any of our enterprise clients would allow us to do so. And so we totally agree with that. We're only using OpenAI as the voice interface for Athena. She never sees the underlying data. She's seeing an output that's coming from our data cloud.

By way of example, everything you saw here, you could have navigated through our platform using a keyboard a few months -- I mean you still can, if you could figure out how to do it. So none of the output or non of the underlying data goes to the large language model.

So this -- could there be a risk like kind of iOS model or moments where you have to have big [indiscernible].

So when iOS moved to a no cookie system, our growth that year was only 62%. So we don't rely on third-party systems to track people. So we do not fear that type of a thing. When people say to me, what are you most worried about, it's a data breach. For a company like us, where 51% of the Fortune 100 are trusting you with all of their data, if you have a breach, trust could be lost, and that's where you have a major problem.

So what do we do? Our network operating center operates on 4 continents, 24/7, 365. Our Head of Data Security, Dr. Jeff Nimeroff, who has his PhD from the University of Pennsylvania, used to help to run all data security for eBay before joining us, has designed a siloed data system with 28 different silos. Each of one has a different encryption key to connect the PII to the different data points and we invest very heavily in that.

The other thing that Foundry gives us is another added layer of data security. It's one of the best data security platforms in the world, whereas a number of our other competitors -- I'm sorry, our other partners didn't have that layer. Did you have another question?

Last one is going to be I think, early on when we saw Tiktok ads, we thought it was hard to kind of predict how these [indiscernible].

Yes. And we partner with TikTok, we partner with Snap. We partner with Meta. We partner with LinkedIn.

[indiscernible] tells he premise of the question but really I wanted to know based on what you're seeing where the LLM is with some of the [ churns ], some of the economics performance experience, how would you compare that to what's out there in the market?

Well, it's hard to compare it to TikTok because TikTok is so viral, right? You don't get that type of a viral nature in the natural outputs of the existing large language models. That being said, we are seeing very high click-through rates, but they're only showing -- we're only serving the ads in a very small percentage of what is a large inventory pool. So it's not going to be a small amount, but we're seeing higher click-through rates. It's still early on the return on marketing spend to be honest. One last question. Do you want Ron to sum us up or?

Just in terms of the return on ad spend [indiscernible].

We say marketing, but yes.

In terms of the third-party verification of just numbers...

It comes from Forrester. This is a Forrester independent study that they published showing that the average return on marketing spend was an average of 2.4x to 2.7x, and we were at 6x to 6.5x. We were more than 100% higher than #2 in their study.

[indiscernible].

I don't know. 2 minutes? I have 2 minutes. You can get the report from Forrester. It's available through them. I -- personally, I don't recall if they put a narrative around it, but...

What they did was they actually took a number of customers from different industries, looked at different channels that they were using and the different use cases of acquire, grow and retain. And the outcomes in that study by customer range from anywhere as low as 4.5x, I want to say, to as high as 9x.

By the way, when Forrester named us the #1 marketing automation company, again, third year in a row. I'll point out in the year that they just named us number one, we did not spend meaningful dollars with them as a client. So that's something I always like to point out. We're big fans of Forrester, but I like to be clear about that. They said, when putting us the highest and furthest to the right in their study, we take the most complex marketing problems and make them simple. And I think that's ultimately what clients want.

The CMOs are always under pressure. They're usually the lowest tenured employee in any CXO suite. So our job is to make them the hero of their own environment and ecosystem and keep things as simple as possible while driving the highest possible return on marketing spend. By the way, what she showed you was not like a demo. This was like a real interaction with the platform. The outputs are from a real campaign that ran and you'll note that it was 14x return on marketing spend. She did not cherry-pick that. We have a lot of clients who were well above that order of magnitude return, the average is 600% to 700%.

That's a great place to end it. David, Chris, Winnie, thank you for the time and congrats on the deal. Enjoy the heat. Yes.

All right. Hello, everybody. My name is Brett Miller. I'm the Co-Head of Software Investment Banking at JPMorgan. Thanks for being here. Today, it's my pleasure to lead a discussion with David Steinberg and Chris Greiner of Zeta Global. Zeta is an AI-powered omnichannel data-driven cloud platform that provides enterprises with consumer intelligence and marketing automation software, helping brands personalize marketing across channels, including e-mail, social, web, chat, connected TV and video. David, the company's CEO and Chairman, co-founded Zeta in 2011 and took it public in 2021. Chris, the CFO, joined Zeta in 2020 after serving as the CFO of LivePerson and Inovalon. David and Chris, thanks for being here today.

Thank you for having us.

First off, big picture, what is Zeta trying to become for the enterprise? Is it a marketing operating system, a data cloud and AI decision engine or some combo?

Well, let's start with yes. I mean we're -- our real vision is to become the operating system for our clients' entire marketing ecosystem. And if you look at Athena, our new sort of AI-driven super agent, she really positions us to do that really well. And if you look at the way we operate today, we're ingesting 100% of very large enterprises first-party data. We're matching it with our data. In the United States, we match it greater than 92%. And that's where the algorithms start training inside of the consumer data platform.

So once we do that, all of the business intelligence and all of the data for our clients exist in that ecosystem. And I think one of the things people don't really understand is we think of ourselves as an AI-driven business intelligence company that started focusing on marketing because it's a $1 trillion TAM, right? It was just a very good place to start. It's not where we expect to finish.

I spent a lot of my time today on the debate about the SaaS apocalypse and whether SaaS is going away. What is your view on this debate on the disintermediation by AI?

Well, I think it's going to be a lot like the dot-com, except I don't think we're going to follow that type of bubble because most of the enterprises that are making these investments have $1 trillion balance sheets. They're not start-ups. But what I would say is companies that adopted the Internet are the ones that have flourished from a technological perspective. And if you think about it, I remember -- and yes, I'm old. I remember when the dot-com boom was coming and they said it was going to kill Walmart. It was going to kill JPMorgan Chase. It was going to kill Federal Express.

In all those cases, those organizations adopted those technologies and became juggernauts. Now I do think I mean, first of all, right now, all funds are flowing to AI hardware, right? So that's one challenge that software companies are dealing with as public companies. The other really interesting thing is today, nobody is willing to take a risk on who's going to be the winners, who's going to be the losers.

And everybody is sort of throwing the baby out with the bathwater. If you think about companies that are going to win, there are going to be companies that have moats around their business. If you look at Zeta, we have 555 million global individuals who have opted in to be in our data cloud. The only other companies that have data at our scale are all trillion-dollar companies. You're talking about Meta, Google, Amazon. And you can go through that list. We use that data. And of course, we're ingesting the data today, 51% of the Fortune 100 largest companies in the United States use our platform, and they're trusting us with their first-party data. We're using that data to create intelligence. And according to an independent Forrester survey study, for every dollar a client spends on the Zeta Marketing Platform, today, we returned 600% return on market spend.

Now if you look at the adoption of Athena, which rolled out in the first quarter, we did say in the first quarter, we saw our Agentic workflows go up by 700%, and Athena was a very big driver of that. Clients that are using Athena are seeing an even greater return on investment than that 600%. So I think over the next quarter or 2, investors are going to start to figure out there are going to be some winners. There are going to be some losers. I think companies that are workflow management tools don't create intelligence, don't own data and don't have meaningful defendable assets are going to have to figure out how to pivot their businesses to survive. And I think you're going to have companies like Zeta that are creating intelligence, creating massive return on investment.

And once again, I don't see Fortune 500 companies turning over their first-party data to large language models. I just think that's going to be tough. The other thing, of course, is we're already partnered with OpenAI. We've already announced an enterprise partnership with them. We already work with Anthropic. We work with Google. We're already in that entire ecosystem.

You just teed up about 20 questions in that response. So thank you for that. So at CES, you did announce, you just mentioned it, the major partnership with OpenAI, which powers Athena. Tell us about that relationship, how it's evolved, where it's going?

Well, I'm super excited to mention today that we have now executed an agreement to partner with OpenAI and help them run their advertising. We're going to bring our enterprise clients into the OpenAI ecosystem, and we'll begin to serve ads there on behalf of our enterprise clients.

Anything more about where that's going and like what you can be doing generally with it?

Yes. I mean, listen, we think OpenAI is going to be one of the winners in the next generation. And the fact that now we've executed partnerships with them around Athena and we've executed partnerships with them around serving ads into the ecosystem. I think it's going to be game changing to our organization. If you think about it today, we are the only company I know of that is fully integrated into Meta, into Google, into the open web, into messaging, into connected TV and now connected into OpenAI, both through serving ads and through our GEO platform.

Incredible. We've mentioned Athena a bunch. Why don't we just up level for a second and give just a quick description of Athena. And then also like what is the workflow that it's changing for your customers?

So Athena is -- well, we'll step back a little bit. If you think about it, humanity is always communicated through voice, right? I mean I shouldn't say always, but for, call it, 400,000 years. In the 1950s, we created this thing called the keyboard, sits between most humans and technology. And I think most of us find it very, very frustrating. Athena was created to be a voice-enabled, fully conversational agent who helps our clients navigate our entire platform, not just by trying to figure out how to use different features and functions, but by literally asking for outcomes from the platform and Athena can change the screen in front of the client. So a client might say, I'd like to create 2 million incremental customers this quarter, and I'd like to do it in an environment where I can lower my marketing cost by an average of 7% per customer acquired. And Athena walked through how a client can get to those outcomes.

So the way I equate it, and I think many people in this room probably use Bloomberg terminals, right? What percentage of your Bloomberg terminals capabilities do you use on a daily basis? 1%. That's -- I've heard 1%, 3%, 5%. It's the same thing for most software, right? So we've built a fighter jet that our average client knows how to fly assess. And Athena becomes the copilot for how to get to those most valuable components of the platform that they don't know how to use on their own. And what we're seeing, once again, Athena rolled out in Q1. She was beta for much of it. She rolled out generally available in March. Even with that limited period, we saw a 700% uptick in Agentic workflow. And we saw Athena as most of it.

Now I mean, we did -- our business grew 50% top line in the first quarter, 42% EBITDA growth, 42% free cash flow growth. This will be on a 4-year average basis, this will be our fourth year of greater than 30% organic top line growth and greater than 50% organic EBITDA growth. So what we're doing is working because our clients are consuming more of it. They're growing with us. And as we show a greater and greater return on investment, we believe we'll continue to see that growth.

Yes. We'll come back to the metrics and let Chris brag about those in a minute. On the 700% increase in Agentic interactions, like in the 1% of the Bloomberg uses your analogy, what are you seeing about how customers want to use AI through that early signal?

We're seeing a lot of clients move right to how do I get to this outcome versus trying to navigate the platform to get to the outcome. And then Athena is able to control the entire experience and she actually changes the screen to show audience types, activation capabilities and all of the different return on investment in real time as you experience it. So that's a big one. Another big one, interestingly enough, has been analytics.

We're seeing a lot of clients ask for analytical outcomes that we haven't seen them ask for before. And for those of you who don't know our business well, we have 3 main use cases. it sounds complex, but it's pretty simple. Helping enterprises to create customers, maintain their existing customers and further monetize their existing customers. And we do this by activating across multiple channels. A channel might be connected TV, a channel might be online video. It might be online contextual. It might be now plugging into OpenAI.

It might be plugging and targeting into Meta. All of those things are sort of use cases versus channels. If you've looked at our business since our IPO, we've been public for 19 quarters, I will point out 19 consecutive quarters, we've beaten our guidance and raised guidance. But the way we've grown our business is by adding channels. And we've gone from an average of about one channel to an average of over 3 channels in that same period. What we found is when enterprises use multiple use cases so they do customer acquisition and customer retention and customer monetization. They spend 250% more on our platform. But the reason for that is because the flywheel for intelligence starts with one of them. So the data that informs acquisition further informs the data that's necessary for retention, which further informs the data for monetization back to acquisition. So the return on investment is substantially greater than our average of 600%. The single best tool to ultimately get our clients from an average of 1.4, I don't know the exact number, but sort of, call it, just about over 1 use case to 3 use cases is Athena. Because after you've asked Athena for outcomes, around acquisition, she might say to you, by the way, did you know that this 1.2 million of your existing customers began the process of churning off your platform and discovery for your competitors' products? Would you like to save them? And that's how it's already starting.

So you're not explicitly monetizing Athena yet. So talk about that decision of when does this become a line item in your P&L versus when is it just something that drives retention and ARPU?

Well, it's driving ARPU growth. I mean, we had a 21% ARPU growth in the first quarter, which -- and last year, our net retention rate was 120% as a company. We guide to 110% to 115%. We were actually well above that in the first quarter. But when you look at what Athena is doing, she is driving a massive uptick in our utilization fees. So if you think of our business, about 60% of our revenue is subscription and 40% is utilization.

If we were a large language model, we would simply be calling our utilization tokens instead of utilization because it's the exact same model. It's -- as we run messaging or we place an ad or we run a connected TV ad, we charge a fee for that. We've batted it around. There will be an opportunity at some point to build, I think, analytics products and extensions around Athena.

But today, if you look at it, our 189 global superscaled customers will spend approximately $110 billion on marketing this year. At the middle of our range, we have approximately 170 basis points of wallet share. Now we have some clients that we have 700 to 800 basis points of wallet share, and I don't see any reason we can't get all of our clients there at some point on average.

So if Athena can help us go from 170 basis points to 200, 300, 400 basis points, the return on investment to us is so much massively greater as you go from what I think last year, we were about $1.35 billion.

And then this year, we're projecting $1.75 billion. $1.78 billion. You're looking at sort of the business is not just growing at 40% or 50%, it's also dropping a disproportionate percentage of that to the bottom line. So we think Athena is better served today in driving that massive uptick in utilization.

Any plans for the future and how you'll monetize AI differently than you're doing today?

Yes. I mean there's a lot of different ways. And back to the original premise, as you think about Zeta, we think about ourselves as a business intelligence company that happens to focus on marketing. As we roll out business intelligence products, which is in our long-term road map, we'll begin to charge for those products, and that will be commensurate with other AI products.

And in terms of managing token costs, one of the sessions as of this morning, one of the companies said that they basically had no idea how this is going to evolve and they're learning in real time. Like how do you think about how.

We are not that. We are fully and completely aware as to, how we utilize tokens and how we program. What I would tell you is we built an internal platform called SPAD, and SPAD is an internal proprietary workflow management tool. Last quarter, we publicly announced 75% of our internal new code was generated automatically, 75%.

We're ahead of Google, which is even shocking to me. It works because of SPAD. So a programmer goes to SPAD and tells SPAD the type of code it wants to create. SPAD then looks at all of the available large language models, which include ChatGPT, Claude, Gemini, all the way through Cursor, and it makes the decision on which platform is right to create that piece of code and is utilizing the least tokens in its generation.

So for security, it might use Claude, for programming, it might use GPT, for publishing, it might use Gemini. Once it completes it, it sends it to another program called Zappi. I don't know where we come up with the names for these things, but Zappi then QAs the entire thing, sends it over to an architect who reviews it, pushes a button and it goes generally available.

So literally, we're doing stuff. We've talked about this, and I'll sort of mention today. We're going to be rolling out the next generation of Athena, which will be Athena for agencies. The agencies make up about 20%, call it, 25%, less than 25% of our revenue. And the product itself would have taken us 2 years to develop 2 years ago. We're going to have the entire product done in less than 4 months because the pace at which we're able to accelerate code generation and the way we've been able to control cost while doing it.

So I mean, by way of example, we've grown the business again on a compounded 4-year growth rate of 30% greater organically. We've grown headcount just about 10%. And I think we can even slow headcount growth even further, and we expect to be a 20-plus percent organic growth rate company for many years to come.

Speaking of the quarter. So great quarter, another great quarter. Rule of 67, growing 50% year-over-year and over 20% of that is organic.

29% was organic...

Which is over 20%...

Which is well over 20%...

It's well over 20%.

I mean, it's over 20%. It could be 20.02%.

It's approximately 30%. That sounds better. Chris, if you were to pick a couple of drivers of the quarter and then also talk about where the growth -- that organic growth is coming from.

Yes. I think what stood out to us as we even looked at it internally, certainly as we spoke to it externally was just how many parts of the growth ecosystem of Zeta were working in tandem. We talked about the 29 points of organic growth and the balance that we had, 14 came from existing customers, 15 from new.

If you look at the industries that rely upon Zeta's platform, we're in 15 different industry verticals. Our top 10 generate about 90%, 9 out of our top 10 verticals grew over 20%. When you think about the KPIs that power our growth, the customer additions and the customer spend expansion, we had 19% growth in super-scaled customer count. That's more than 3x greater than what we require from our model.

And then as David mentioned earlier, from an ARPU perspective, 21% growth, our model is 12% to 16%. David linked it back earlier to channel expansion and use case expansion. Two data points that we shared with investors in the quarter was those that are using more than one use case. And if we were sitting here a few years ago, that would have been less than 10% of our portfolio.

Today, that's double that. We're approaching 25%. And that metric, number of scaled customers using more than one use case grew 50% year-over-year. And those using 4 or more channels, could be any channels in tandem, grew over 40% year-over-year. So you had many different parts of the portfolio all working and executing really well that led to a much bigger beat than what we would typically have of, call it, 2 to 5 points on the top line.

The other thing really is Athena. We're seeing Athena out there. And we're taking meaningful market share from competitors. And one of the challenges we have as a public company, right, is are you a CRM platform? Are you an activation platform, or you an ad tech company? Are you a marketing technology company? Are you a data platform? The answer is yes. Because our goal is to displace all point solutions in the marketing space. And if you look at our prepared remarks from our most recent call, we announced what I think is the most marquee client win we've ever had.

It's a retailer with greater than 3,000 global locations and manufacture and sell their own apparel across multiple brands. This is a company based in San Francisco. It had been a total Salesforce shop, and we were able to rip and replace them out.

We also replaced 3 other vendors in that because Salesforce, can't provide everything a modern marketer needs outside of a very sort of narrow CRM and CDP capability, whereas we're able to bring CDP, data cloud, best-in-breed data activation to CRM, customer acquisition and customer retention.

We were told we were in the top 3 going into the end of the sort of ninth inning of the process. And we asked if we could come in and do an early demo of Athena. We did the demo of Athena and they called us the next day for pricing. So it's really -- I think Athena is going to be a major, major game changer for our company.

Sounds like it. On the bottom line, you did grow your EBITDA by 42% year-over-year and your free cash flow by 48% year-over-year, which is pretty remarkable. What are the biggest drivers of that leverage?

Yes. What we've done a really good job of, and David mentioned earlier, is automating more and more of what we have. We have a kind of a philosophy in the company of where is value created. And we're creating the world's best products, and we put resources and investment in engineering and the tools they use and then selling it as fast and efficiently as you can in the market. We put resources in quota carriers.

Everywhere else, we cut, we trim, we automate. So that's allowed us to invest and to have the best product in the market. Sell at what is a very productive rate in the market and expand adjusted EBITDA margins and free cash flow margins faster than the rate at which we're growing the top line. The rest of the year, we'll get back to very strong adjusted EBITDA and free cash flow margins as well.

It's also important to note that we built a lot of this stuff ourselves. So as everybody is now talking of the evolution of AI, right, it's starting with large language models, and it's moving to what's called inference models, right? And I don't want to bore everybody in the room with this.

But we've been building inference models for 5 years. The idea that one algorithm can focus on one task with the data set that is necessary to complete that task is really what inference is. We don't need to ingest the entire Internet to figure out what we're doing because we have so much proprietary data we can focus on.

We're not sort of trying to boil the ocean here. Now of course, we love our partner, OpenAI. We love what they're doing, and they're trying to do something on a very, very big basis for that.

Yes. So the quarters are often driven by shifts in the mix-shift. So last quarter, you talked about agency wins and social being a big driver. Why don't we spend just a minute on mix shift and how -- what that did for the quarter and how you think that's going to evolve?

Yes. We disaggregate revenue in a number of different ways, but one of which is what we call and report as direct revenue contribution and integrated platform revenue contribution. Direct is nothing more than when customers use Zeta data and Zeta channels to do their activation. And that regularly contributes 70% to 75% of our revenues. I'd call it a gross margin of 70% to 75% on average.

The integrated platform revenue is when customers rely on Zeta's data, but the ability to go market inside the walled gardens. Meta is a great example. Dave talked about others. That's about, call it, 25% of our revenue has a lower gross margin profile. A lot of the inventory gets passed through to the customer.

What we found empirically is that when agencies, which have been, call it, in the last 2 years, a new buyer of Zeta and they've grown and scaled to approaching 25% of revenue, they first test Zeta data and they first test our platform with social. But what we've seen is as they grow their spend, as they give Zeta more and more brands within their ecosystem, they do that by expanding on Zeta's channel.

So these 2 large wins, like patterns in the past, started in social, and it's our expectation that as they grow, they'll migrate to direct, which will then take some pressure off of the gross margin side. But it's really important from an investor's perspective, even though social has a lower gross margin profile, it's still accretive to adjusted EBITDA and free cash flow.

And it's important to note that we're really the only platform of our type that plugs into these walled gardens. So agencies start with us because there's nobody else who can actually do this. And then as they do it with us, they see the power of our data, and we're able to expand out into other use cases that are much higher margin.

So you've touched on this once before, maybe twice before, but you talk about, in your words, super-scaled customers, which is my favorite metric that any company displays.

We're going to move to super duper scale.

Even better. So on the super-scaled customers, you're at 189 of them, I believe, and you're scaling that ARPU pretty significantly in that cohort of customers. So one, how are you continuing to keep the momentum gaining those customers? And what is driving that ARPU expansion?

Yes, there's a -- for those that know our story well and for those that are new that want to pick up on it really quick, there's a slide in our investor deck that shows that the longer our customers are on our platform, the bigger they become. There's 5 years' worth of this slide that the vast majority of the time, we are not beholden to an RFP because we are proud to start with a $50,000 pilot or $150,000 proof of concept.

And what we found is that those very small pilots in the first 12 months almost get to $1 million in spend, which would then flip them over to that super-scaled category. But beyond that, our fastest-growing set of customers are those that have been with us 5 or more years.

So to David's point, it's taken some of those longer tenured customers to find out where are all the corners in the platform that I'm not using today that I should be using. Athena unlocks that to everybody. So it has the potential to take those that are earlier in their life cycle and have them climb to that much higher level of spend faster.

By way of example, I was at a conference just a couple of weeks ago, we were with a very large new partner. And I had 2 or 3 salespeople in the room, and they were all really nervous about getting this deal done. And the CMO of this very large company asked me how large I thought the test should be to really get to a statistical sample.

And I said, "Listen, I wouldn't start with any more than $10 million." If we do more than $10 million, it's really going to get us past the point of diminishing marginal return. But if we start with that, it will give us really good statistical data. And they were like, "Okay."

So we're not just taking existing clients and growing them, which has been our traditional business. To Chris' point, he's totally right. But the credibility that's now coming with our current size, scale and return on marketing spend is leading clients out of the box to start at bigger numbers than we've seen before.

And on that dynamic, your net revenue retention rate is remarkably consistent and considerably high, 110% to 115%...

Last year, it was 120%.

Right. It's the consistency of it, I think, that I would call out at a time when a lot of other companies had seen degradation in that metric. And then in some cases, a recent recovery. How have you been able to maintain such consistently high net revenue retention?

I think that goes back to what Chris was saying, right? When you start with a client -- and by the way, the client that spend -- we're asking for $10 million has a $4 billion annual budget, right? So it's still statistically similar. A lot of the companies we start with, we're starting with a small pilot new scale. That actually leads to a higher net retention rate, right, because they're growing and growing.

And I think that -- listen, we -- one of the things people don't understand is we build every one of our clients a consumer data platform. We ingest all of their data. We match it with all of our data. We match it greater than 92%. We then remove the personally identifiable information from a regulatory and keeping the consumers' data private. The algorithm then gets incredibly smart on how your customers came to be your customers.

What did they read, what did they search, what transactions did they do? We then go into our data cloud. Let's say, just in the U.S., we have 245 million opted-in people. And we'll say, "Okay, great." These 50 million people are already your customers, don't waste money marketing to them. Think about how often you get an ad for a product you already have.

Then what we do is say, "Okay, these 2 million people are in market for your product, but only 1 million of them would be credit approved by your criteria." So now you've gone from 245 million potential customers down to 1 million, you hyper target them. And as they buy, it feeds their data back into the CDP through the CRM integration and the algorithm gets smarter.

If a client were to fire us, they lose 100% of the learning. All of that data goes away. So what happens is the longer a client is with us, the smarter the algorithm gets, the better we can either lower their churn or help them create customers at a lower cost -- the higher the rate they move budget to us, the higher our net retention rate.

When you think about sort of AI disintermediation, I couldn't see an environment where a Fortune 500 company is going to feed all of their first-party data to a large language model. And then, of course, all of our data, all 555 million people multiplied by 5,000 to 7,000 data elements, multiplied by all of their web behaviors across trillions of active pages on the Internet has never been fed into a large language model ever, nor will it.

That's our proprietary data for training our models in partnership with our clients' data. So that's why our net retention rate is so high, and it's one of the biggest moats around our business.

On the guidance, I'm going to fold a few questions together here. So one, 19 consecutive beat and raise quarters, as David mentioned. And then after this last quarter, you did once again raise your guidance around revenue, EBITDA and free cash flow. What made you comfortable doing that? And how have you managed to achieve that level of, again, just remarkable consistency?

We are very transparent, and we give ourselves 2 to 5 points of growth cushion in our model. So it gives us doors A, B, C, D, E and F every quarter. So we know what those doors are in our model. We tend to put our guidance at the lower end of what our KPIs say we need to do.

This year, we added layers of conservatism for Athena. We wanted Athena to be pure upside. We wanted Marigold and the synergies that we drive to be upside. And then we wanted political, which we break out anyway to be upside. So the fact that we beat by $26 million in the first quarter, we felt comfortable rolling through the year.

We obviously felt comfortable rolling the next, or I should say, adding to the 2Q guidance by $4 million. Holding 3Q and 4Q, or holding out quarters consistent, is consistent with what we do as our model anyway. So I know we've gotten questions on going from 29% organic growth to, call it, low 20s in the 2Q, 3Q, 4Q.

That's totally consistent with how we've guided in the past and wasn't meant to infer anything related to a slowdown. It's just we are keeping our process and our practices consistent as our guidance.

And unlike other years, because of the Marigold acquisition, we actually gave our guidance for this year early. So if you take that raise for this year plus the first quarter raise for this year, we've already raised this year by $55 million on the top line.

So -- and about consistency, what a lot of people don't understand is, yes, 60% of our revenue is subscription. We've got a lot of visibility into that. But even in our utilization fees, we have 1-year contracts on most of that.

So we're not including it in subscription because it's only 1 year, but it gives us unbelievable visibility into our business. And quite frankly, even in what appears to be a tough economic environment, we continue to see our growth at the rates that we expect it to be at.

David, I'm going to ask you a broad question about the competitive environment. So you compete against legacy marketing clouds. You compete against potentially horizontal AI platforms. What do you want to highlight on what's happening?

What are the moats around Zeta? And then also this week, there was obviously the LiveRamp acquisition. Like how does that change that competitive dynamic? That's a whole lot in 1 minute.

Yes. Well, so let me start by saying we love LiveRamp. We've been partners with them for years. We love Publicis. We've been partners with them for years. I see this as a major win for Zeta, both with our existing client and our partner there. And I think a lot of the other agencies are going to be looking for new solutions, which Zeta can provide for them.

So I think we're in a very unique position to win both ways. And once again, we -- I will say I thought this was a brilliant acquisition by Publicis. This was a really good decision on their part. And we spoke to them at length yesterday. We're seeing them multiple times this week, and I think that's going to work out very well for us, once again, as I expressed.

We haven't seen any horizontal AI competitors at this point. Just to be very clear, we don't even see them moving into what we're doing. We're having lots of conversations with the horizontal AI platforms to integrate them into what we're doing and integrate them into our stack as we've already announced with OpenAI today on the ads side and had already announced working with them on the Athena side.

As it relates to the marketing clouds, listen, these are great companies. I sort of joke we compete with a few small companies, right, Salesforce, Oracle, Adobe, Trade Desk. The truth of the matter is most of their technology is built for where the market was, not for where the market is going. We're in the biggest replacement cycle I've ever seen for marketing clouds, and we are winning at a disproportionately higher rate, and we expect that trend to continue.

Incredibly exciting time at Zeta. Chris, David, thanks for being here.

Thank you, Brett.

Greetings, and welcome to the Zeta Q1 '26 Earnings Call. [Operator Instructions] As a reminder, this conference is being recorded. I would now like to turn the conference over to your host, Matt Pfau. You may begin.

Thank you, operator. Hello, everyone, and thank you for joining us for Zeta's First Quarter 2026 Conference Call. Today's presentation and earnings release are available on Zeta's Investor Relations website at investors.zetaglobal.com, where you will also find links to our SEC filings, along with other information about Zeta.

Joining me on the call today are David Steinberg, Zeta's Co-Founder, Chairman and Chief Executive Officer; and Chris Greiner, Zeta's Chief Financial Officer.

Before we begin, I'd like to remind everyone that statements made on this call as well as in the presentation and earnings release contain forward-looking statements regarding our financial outlook, business plans and objectives and other future events and developments, including statements about the market potential of our products, potential competition, revenues of our products and our goals and strategies. These statements are subject to risks and uncertainties that may cause actual results to differ materially from those projected. These risks and uncertainties include those described in the company's earnings release and other filings with the SEC and speak only as of today's date.

In addition, our discussion today will include references to certain supplemental non-GAAP financial measures, which should be considered in addition to, and not as a substitute for, our GAAP results. We use these non-GAAP measures in managing our business and believe they provide useful information for our investors. Reconciliations of the non-GAAP measures to the corresponding GAAP measures, where appropriate, can be found in the earnings presentation available on our website as well as our earnings release and our other filings with the SEC.

With that, I will now turn the call over to David.

Thank you, Matt. Good afternoon, everyone, and thank you for joining us today. We delivered our 19th consecutive beat and raise quarter. This consistency is not driven by a single product cycle or a short-term tailwind. It is the result of a structural shift in the market. AI is no longer a feature. It is driving a replacement cycle where enterprises are demanding fewer systems, measurable results and applied intelligence that works today. We are winning in this environment because of the system we have built, proprietary data that improves with every interaction, intelligence that compounds with every decision and a platform with AI at its core that allows customers to consolidate vendors into a single unified operating system. This differentiated approach has been recognized by Forrester, where Zeta was once again named a leader and also reflected in our customer advocacy with an NPS score in line with market leaders, up 23% from our assessment in the prior year. Both come on the heels of Forrester study showing Zeta returns an average of 600% on marketing spend for its customers.

Athena by Zeta is an accelerant. It is the user interface that brings AI directly into marketing workflows and removes the barriers to enterprise-wide adoption and impact. Signs of this were evident in Q1 with beta customers plus strong early adoption of Athena contributing to the revenue beat. Our first quarter performance, once again, shows we are the disruptor in the AI-driven marketing ecosystem.

First quarter revenue of $396 million, representing year-over-year growth of 50% and up 29% year-over-year ex Marigold, our fourth straight quarter of revenue growth acceleration, excluding acquisitions and political candidate revenue. And adjusted EBITDA was $66 million, up 42% year-over-year. 19 consecutive beat and raise quarters combined with a 4-year revenue CAGR of 30% reflect more than just consistency. They are evidence of sustained demand in a market consolidating around platforms that can deliver measurable outcomes at scale. And that visibility is reflected in our outlook. After raising the midpoint of our range for 2026 revenue guidance last quarter by $25 million, we are again raising it by $30 million, representing growth of 37%. These market share gains are evidence of a shift in the competitive landscape as AI moves from feature to a new way of doing business.

Athena is designed to accelerate our share gains by bringing intelligence directly into workflows, turning answers into actions and ultimately changing how marketing is planned, executed and optimized. Athena is currently available to all of our enterprise customers, and its impact is already evident in sales pursuits and results. The number of Athena demos to potential new clients increased dramatically throughout the quarter. The promise of Athena is influencing decision-makers and helping Zeta win deals as customers want to invest in applied AI, not road map AI.

One new customer we closed in the quarter commented, "Leapfrogging to the future requires thinking differently today and committing to execution. Interacting with Athena made it clear that Zeta has already made this leap, bringing its vision to life and positioning us to accelerate into a fully agentic marketing future."

Athena was a driver in one of the largest deals we have ever closed. The customer is a leading global apparel retailer operating across multiple brands, each with unique customers and over 3,000 locations worldwide. Zeta's platform was purpose-built to handle the complexity required by the largest enterprise companies, and this customer was able to consolidate down from 4 vendors to 1, Zeta. As the legacy Marketing Cloud replacement cycle begins to accelerate, this particular client was a marquee win.

We are also seeing rapid adoption among existing customers. Early feedback and usage shows that customers view Athena not as incremental functionality, but as transformational technology. As adoption increases, Athena learns from more data, outcomes improve and usage deepens, driving ARPU expansion and ultimately reinforcing the same flywheel that has powered our growth. That flywheel is powered by more than just Zeta's AI models. It's driven by the data and infrastructure behind them.

Zeta SuperGraph, our proprietary identity and intelligence graph, unifies data across the enterprise and enables a complete deterministic view of the consumer that we believe is difficult to replicate at scale. This is translating directly into wins where access to our data is a key driver for customer decisions. For example, our SuperGraph was instrumental in a win with a leading online retailer of pet products in the United States that serves millions of active customers with a highly personalized e-commerce experience, a broad assortment of over 100,000 products and a rapidly expanding ecosystem that includes autoship subscriptions, pharmacy services and pet health offerings. In addition, our proprietary data and the intelligence it generates was a key component in the expansion of a Fortune 100 telco client, expected to drive an 18x increase in spend with Zeta in 2026 versus 2025. As Athena brings that intelligence to our customers in real time, the impact of this data advantage only grows.

This foundation of data plus AI continues to power One Zeta. We are consistently seeing that the land, expand, extend model takes hold as customers begin with a single-use case and scale across the platform over time. That expansion is driven by the modern CMO mandate, do more with fewer partners, improve ROI and simplify execution across the organization. The result is larger commitments, deeper adoption and a growing role for Zeta as the marketing operating system and core infrastructure.

That momentum is showing up in the data. Super-scaled customer ARPU was up 21% year-over-year, well ahead of our target range. Net retention rate remained above our target range of 110% to 115%. And the number of super-scaled customers using more than one use case was up over 50% year-over-year at scale. It also creates a reinforcing cycle. Consolidation drives adoption, adoption drives results and results drive further expansion. This is the One Zeta model, and it continues to be a powerful driver of durable growth.

What stands out for me this quarter is the strength we are building across every part of the business. At the center of this is Athena, which is already beginning to change how our customers operate and how we compete. Together, our data, our platform and our leadership in AI are positioning Zeta not just to participate in this shift but to define it.

As always, I want to sincerely thank our customers, our partners and our shareholders for your continued support of our vision. And to team Zeta, thank you for everything you do. It was an honor to be recognized as a Great Places to Work for the third year in a row. This is a reflection of your hard work and collaboration.

Now let me turn it over to Chris to discuss our results in greater detail. Chris?

Thank you, David, and good afternoon, everyone. Our results, once again, demonstrated the durability, predictability and profitability of Zeta's growth. Revenue growth, excluding acquisitions and political candidate revenue accelerated for the fourth consecutive quarter to 29% in Q1, further cementing the durability of our growth and market share gains. Broad-based strength across the business is resulting in improved visibility, leading us to, once again, raise our 2026 outlook, underscoring the predictability of our growth. Even in doing so, we're maintaining our typical conservatism. And we also saw free cash flow conversion improved to 63%, generating $42 million in free cash flow, demonstrating the increasing profitability of our growth. These results surpassed even our internal stretch goals, coming in $26 million or 7 points above the midpoint of our revenue guidance for the quarter.

As I analyzed the strength of our quarter, what stood out was how balanced the upside contribution was. It was not 1 or 2 isolated benefits. Instead, in baseball parlance, it was a lot of singles and doubles, which in my opinion, is healthier. Here are some examples. In terms of revenue growth, excluding Marigold's contribution, approximately 14 points of growth came from existing customers and 15 points from new customers. From an industry lens, 9 out of our top 10 industries grew faster than 20%, with more discretionary industries continuing to be at the upper end, demonstrating why in tougher macro times, data-driven, lower marketing funnel, high ROI attributable marketing is paramount. And finally, as it relates to how customers use our platform, e-mail, connected TV, mobile and social all grew double digits, all while each use case, acquire, grow and retain also grew double digits.

Now let me dive deeper into our KPIs, income statement and balance sheet. Total super-scaled customer count grew to 189, up 19% year-over-year and an addition of 5 customers sequentially. This exceeds our Zeta 2028 model of 4% to 8% super-scaled customer count growth. Super-scaled customer additions were especially strong in advertising, marketing, travel and hospitality. Super-scaled customer ARPU was $1.7 million, up 21% year-over-year. This also exceeded our Zeta 2028 model of 12% to 16% ARPU growth. Strong ARPU growth in the quarter was driven by an increase in the number of customers using multiple use cases, which was up over 50% year-over-year, as well as customers using more than 3 channels, which increased 40% year-over-year. Both are great examples of the One Zeta sales motion working and how Athena can unlock more of the platform's capabilities for our customers to use.

The forward-looking sales pipeline is also robust, going into a season when Athena will be front and center at multiple industry conferences. In fact, Athena demos were a crucial differentiator versus incumbents and RFP competitors in each of our marquee enterprise and agency wins in Q1. And we expect Athena to play an even bigger role in adding to the sales pipeline, which is already up 40% year-over-year with a subset of discretionary industries up even more, those like retail, advertising, travel, restaurants, furniture and resorts to name a few. This outsized sales pipeline growth in discretionary industries is consistent with what we've seen in previous periods of macro volatility and is another proof point that in times of uncertainty, customers consolidate onto fewer platforms that can drive measurable ROI with AI-driven efficiency.

Now moving on to revenue mix. Direct revenue in the first quarter was 75%, above the 73% last year and in line with our target of 70% to 75%. Our GAAP cost of revenue in the quarter was 41%, a 190 basis point increase year-over-year and 50 basis points sequentially. The increase in cost of revenue was driven by new agency wins, driving a higher initial mix of social as a channel. This is consistent with the pattern of business we've seen and spoken to previously when new agencies platform on to Zeta. This is because we offer a substantially more efficient and effective solution for social and has become the first of many channels adopted by new agencies as they migrate. As new agencies scale over time, not only does their aggregate spend increase, but they do so by adding Zeta-owned channels like e-mail, display, video, mobile, CTV and others.

It also bears repeating, while social has a higher cost of revenue, it is still accretive to both adjusted EBITDA and free cash flow margins. Further, social drives high customer stickiness as well. In the first quarter, adjusted EBITDA was $66.1 million at a margin of 16.7%, 100 basis points lower year-over-year and $5 million better than the midpoint of our guidance. Marigold integration is progressing rapidly and tracking ahead of our expectations. We took aggressive steps in the quarter to execute operating synergies, which should begin to benefit our adjusted EBITDA margin in Q2 and into the back half of the year. At the same time, Marigold's revenue came in better than we anticipated, and we're seeing encouraging traction from the One Zeta approach of cross-selling Marigold's loyalty product along with Zeta's grow-and-acquire use cases to the combined customer base.

Another area we spoke about last quarter was becoming GAAP net income and EPS positive for the full year of 2026, specifically generating between $0.02 and $0.04 of GAAP earnings per share. Our first quarter results have us pacing towards the high end of that range. In Q1, our GAAP net loss was $13.2 million, an improvement from a net loss of $21.6 million in the first quarter of last year. GAAP loss per share was $0.06, coming in ahead of our expectations for the quarter with forecasted costs related to the integration of Marigold being the primary driver and not seen as recurring over the rest of the year. First quarter net cash provided by operating activities was $49.7 million, up 43% year-over-year, with free cash flow of $41.7 million, up 48% year-over-year and representing a margin of 10.5%. This represents a free cash flow conversion of 63%, a 270 basis point improvement from the first quarter of 2025. This also includes a roughly 13-point working capital headwind driven by longer agency payment cycles standard for their industry.

During the first quarter, we repurchased 1.5 million shares for $25.7 million and have approximately $138 million remaining on our share repurchase authorization. We expect to remain active buyers of our stock, especially at these price levels, subject to market conditions and other priorities. And we continue to make significant progress in reducing dilution and stock-based compensation expense. Excluding Marigold, our dilution in the first quarter was 0.1%, and we remain on track to achieve our normal course net dilution target of 3% to 4% in 2026.

Relatedly, with most of management's previously issued equity now fully vested post-IPO, Zeta's Board of Directors and Compensation Committee, in consultation with an independent compensation consultant, approved a new long-term equity incentive plan for management. This performance-based plan secures continuity of Zeta's named executive officers and management for 6 years and incentivizes management to achieve its long-term revenue and adjusted EBITDA margin objectives while adhering to its principles of lowering dilution, reducing stock-based compensation as a percentage of revenue and achieving GAAP positive earnings. Furthermore, named executive officers who received these incentives will not be rewarded any further equity for the next 6 years.

Now on to our increased guidance. For the full year 2026, we're increasing the midpoint of our revenue guidance by $30 million to $1.785 billion, representing a 37% growth rate or 22% year-over-year growth when excluding Marigold and political candidate revenue. None of our guidance increase is related to political candidate revenue, which we continue to assume will be $15 million in 2026 with $7 million in the third quarter and $8 million in the fourth quarter. Additionally, we continue to take a conservative view of Marigold, contributing $47.5 million per quarter to 2026 revenue for the remainder of the year. Our revenue guidance also includes minimal contribution from Athena. And as shared earlier, we have taken into account our typical conservatism of 2% to 5% in setting our outlook. For the second quarter, we now expect revenue of $420 million at the midpoint, $4 million higher than our previous guidance and representing year-over-year growth of 36% or 21% when excluding political candidate and Marigold revenue. For adjusted EBITDA, we're increasing the midpoint of our 2026 guidance to $397 million, up $6 million from our prior guidance and representing a year-over-year increase of 43% at a margin of 22.3%, an improvement of 90 basis points over 2025.

For the second quarter of 2026, we now expect adjusted EBITDA of $86.6 million at the midpoint, up from our previous expectation of $84.9 million and representing growth of 47% and a margin of 20.6%, up 155 basis points year-to-year. We are also increasing our 2026 free cash flow guidance to $235 million at the midpoint, up from $231 million, representing year-over-year growth of 43% and a conversion of 59% of adjusted EBITDA, which likely has upside.

And here's the broader point. A 19-quarter beat-and-raise track record is obviously something we're proud of and continues to demonstrate our consistency and strong execution. We also recognize the times we're in, specifically the need to underwrite investments in companies with strong free cash flow generation, durable revenue growth and share gains and demonstratable moats. Q1 was an excellent jumping off point for these emerging investor frameworks. Not only did free cash flow set a record in the first quarter, but we are also tracking to the high end of our 2026 GAAP EPS range of $0.02 to $0.04 and long-term 2028 targets. As it relates to durable growth, this was the fourth quarter in a row we accelerated revenue growth, excluding acquisitions and political candidate revenue. And in terms of exhibiting our moats, our marquee wins with enterprises and agencies this quarter came at the expense of legacy marketing clouds and legacy DSPs, where Zeta's proprietary data and Athena operating system were capabilities our competition could not match.

With that, I'll hand the call over to the operator for David and me to take your questions. Operator?

[Operator Instructions] And the first question comes from the line of DJ Hynes with Canaccord Genuity.

Congrats on a fantastic quarter. Nice start to the year. David, I want to ask you a competitive question. So obviously, there's been more public backlash against the Trade Desk and the agency ecosystem. But I think for those living in the marketing and ad tech world, like that's been going on for a while now, right? So 2 related questions here. Number one, like how much do you think Zeta has benefited from that dynamic? And then second, if the Trade Desk figures out how to get pricing right or at least make it more transparent, does this rebalance competitive dynamics at all? Or is the horse already out of the barn there?

Well, let's -- I mean, first of all, DJ, thank you. We appreciate it. Could not be happier with this quarter. And I think it really speaks to kicking off the year right and Athena really was a massive driver here. I want to separate the conversation about the agency and other technological platforms like the Trade Desk that are out there and struggling a bit because the agencies continue to thrive and they're not really having any issues from our vantage point.

And I just got back from 3 days at the POSSIBLE conference, where I did 54 meetings in 3 days, hosted 4 dinners and 3 cocktail parties, which is why I'm losing my voice going into this. I think that -- and I don't want to speak to any particular platform, but I think the horse is out of the barn. I think that organizations that have built workflow management tools that do not have proprietary data, they do not have proprietary native artificial intelligence are going to really struggle in this next evolution of where sort of marketing is going as it relates to intelligence. Because if you're not creating intelligence in today's world, you're not winning. And I think that we are a direct reason that a number of our competitors are either growing slower or shrinking as we take meaningful market share.

Chris, in his prepared remarks, was very clear about the fact that we had a number of meaningful agency wins in the quarter that will continue to run out through the rest of this year and into future years that are starting with social. We're starting to see those move over to programmatic and connected TV as well. So I think if you separate the agencies, which are doing well and thriving from the technological platforms that have based their business on workflow management, I think they are going to struggle, and we are going to continue to beat them handedly in the marketplace.

Yes. Perfect and helpful color. Chris, I want to follow up with you. So David gave a bunch of great anecdotal data points around Athena and the early success there. The product is not explicitly monetized, right? So what are the signs that we all, as investors, should be paying attention to from a financial perspective that will signal to us that Athena is moving the needle for Zeta?

Great question, DJ. I'm glad you asked. There's a couple of leading indicator data points that I think you can already begin to look at. So as part of the press release, one of the data points that was called out was a 7x increase, and this is just in the first week of Athena's general availability. We saw a 7x increase in the type of -- in the amount of agentic interactions on the platform, coupled by 60% of the AI usage on our platform being driven by Athena. How that should ultimately translate to the usage part of our revenue can be seen through ARPU expansion, some of which you already started to see. So if you look at ARPU in the quarter for super-scaled customers, it was $1.7 million. It was up 21% year-over-year. But if you look underneath that, what drove it are exactly the dynamics that Athena was engineered to be able to do, which is to make more of the platform available and visible for the customers to be able to exploit. If you look at multiple use case customers, it was up over 50% year-over-year. If you look at the customers that are using 4 or more channels, that's up over 40% year-over-year, which again are not just great examples of Athena as an unlock, but also Zeta working well.

And by the way, she's just getting started, DJ.

Yes, totally. Congrats, guys.

Thanks, DJ.

And the next question comes from the line of Gabriela Borges with Goldman Sachs.

David, I want to ask you about the people and process part of Athena, meaning the technology you've demoed, it clearly is able to do a huge amount of knowledge work. My question for you is, how do you then change the end user behavior? What does the training and enablement look like? How do you encourage more people to use it once your customers have already decided to adopt it?

First of all, Gabriela, what a great question. First, let me say that we were incredibly proud that we were able to make the product not just generally available on time, but available to 100% of our enterprise clients, which was not a small task. At the same time, to your exact point, we have a learning and development group that is literally purpose-built to train our clients and get them up and running on this new product. And they've already done. Our top 30 clients have been onboarded through that group, and we're going to be adding all the other clients as we continue to expand out.

The other thing that's really important is we're doing a weekly leaders -- I'm sorry, a weekly learning and training program to all of our clients that's virtual, recorded and they're able to then watch it at their convenience inside of the platform when they want to. But we're doing sort of a ask Athena question of the week. Every week, a new question that you could ask Athena goes out to all of our customers so that they can begin they process of using her. And I know I'm sure you've seen the demo, you know how incredibly intuitive she is to use. So what I would say is we're really focusing on it from a relationship management, learning, development, both in-person, virtually and weekly follow-ups. But the intuitive nature of her is, I think, one of the reasons you saw her so powerful. Just in the first week she was live, Athena drove 60% of all AI utilization across our platform. That is off the charts for a new product from an adoption perspective.

Very good. And Chris, the follow-up for you on inference cost. So Athena has pieces of Zeta proprietary technology. And then I believe you have some third-party technology in there, too. How do you think about managing those inference costs or optimizing those inference costs? And then same question for your R&D team, your engineering team. We're at the stage where we've been through more token usage is generally better, but also can sort of outrun budgets very quickly. How do you think about managing that internally?

So Gabriela, I'm going to take that one. Sorry, Chris. First and foremost, as the world moves from large language models to inference-based AI, our platform is purpose-built for that as the operating system and infrastructure for our clients. The vast majority of our queries, Gabriela, are done on our own platforms, on our own data. So we are not buying tokens as we roll this out to our customers. It's fully embedded, which is one of the reasons I think you're seeing us project substantially higher growth to profits and cash flow than we are even to revenue. We have put ourselves as sort of the perfect spot as the market moves to inference-based AI.

As it relates to our internal consumption, we have built a platform called SPADE. Don't ask me what it stands for. It is an anagram. But the reality is that SPADE is a tool that we custom built inside of Zeta and is being utilized by a very large percentage of our engineering team, where effectively, an engineer would go into SPADE, they would create the construct for the code they are trying to develop. SPADE would then automatically choose the most efficient and best large language model to do the coding itself. So if it's security-based, we might choose Claude. If it's general coding based, SPADE might choose ChatGPT. If it's complex publishing, SPADE might choose Gemini. Now of course, Cursor is sort of at the center of this as we think about where that's expanding.

The code is then auto generated by the LLM, which is the only time we utilize tokens. Everything else is sitting on our own platform, our own cloud. The LLM creates the code. It then goes to a program called Zippi because obviously, we have great nomenclature capabilities. And Zippi automatically QAs the code on our platform once again. Once it's done, it sends it to one of our senior architects. They review the code one more time, and they can make it generally available.

To put it in perspective, Gabriela, at the end of the first quarter, Zeta was already driving 75% automated new code creation. I believe that puts us even above Google as it relates to that. And the pods working on Athena today, I know for a fact are up from a productivity perspective between 400% and 600% year-over-year from an output and productivity perspective, all the while, the vast, vast majority of the compute and of the tokenization is on our own platform. So as you look at our growth, we will not experience some of the constriction of margin or additional CapEx. We've already allowed for everything in the projections that we've got, and we're very, very comfortable with where we are externally and from an engineering perspective.

And the next question comes from the line of Arjun Bhatia with William Blair.

Congrats guys on a very strong quarter here. David, I have 2 questions, maybe I'll just do them one at a time. The first on awareness. It seems like the customers that are using it are getting great value out of it. It's early, but for this to have a material impact, for the company as a whole, you have a fairly large revenue base. Like how do you roll this out to all your large customers? Where is it right now in terms of customers having awareness and knowing what Athena can do? And how do you sort of plan to progress that?

Yes. So great question, Arjun. First of all, from an awareness perspective, I would say that our marketing team today is doing the greatest job it's ever done in the history of our company. I just came back from the POSSIBLE conference where you couldn't walk 5 feet without seeing the brand Athena and without seeing Buy Zeta. And it was really exciting. We did an Athena Suite. We did a Zeta Cafe Powered by Athena. And we're starting to see that we're moving to that next evolution of our brand where it's sort of moved to it's getting the must-have data in our industry. And I didn't think I would say that this early.

As it relates to internal awareness, we have built an internal learning and development team, which is doing nothing but training and onboarding our clients. One of the things we're going to be rolling out in the next few months, which I'm super excited about, is an Athena certification. We're going to certify the individuals who work for our clients on Athena utilization. They'll get a full certification that they can put into their resume, and we're very excited about how that's going to be rolling out. So we're also doing sort of a hint of the week, tip of the week, question of the week. It's going out to all of our clients. I would tell you, in all of the years I've run this company, which is a long time now, I have never seen a faster uptake of a technological product that we've rolled out, and it's really been exciting, Arjun.

Awesome. That's great to hear. And then maybe switching gears from Athena for a second. Marigold, that also looks like it was off to a strong start. I think you beat your sort of Q1 target on that front. But where are we on the cross-sell there? And what's the early traction you're seeing on, I guess, the 2-sided cross-sell, both into your base and into Marigold's base?

Arjun, I'll take that, and David will wrap it up also. So a couple of places where you can see where it's evident that the cross-selling is working. So we talked about the number of multi-use cases. It's nicely contributing to the growth that we've seen across the base of super-scaled customers. But I think more broadly, if you look at the areas that we talked about being purposely conservative around Marigold, it was around the potential for their SMB and mid-market customers that were on the enterprise platform we anticipated churn. We're not seeing as much as we thought, which is good. There were products that -- and geographies that we thought we would have less growth on and would also see churn. That hasn't happened yet. And then just more broad normal churn at the enterprise level, and it stayed healthy. And by the way, a lot of that is being driven by Zeta's interactions with those customers and partnering with Marigold's people.

So it's been really interesting, Arjun. We've seen a meaningful uptick from existing Marigold clients with us integrating the data cloud into the platform. So the first thing we did and we had it done within 90 days was a full data cloud integration into their platforms, which allowed clients to begin to access data sets that they've never had access to before. So we've seen meaningful growth there.

As it relates to cross-selling, we're really -- we're making progress, but not a lot of that is in the numbers yet. These products are complicated, and they're very big. I think you'll see more of that as the year progresses. But I think -- I mean, to say we're very excited about how well we're performing with the asset would be an understatement. And a lot of that today is a result of the data cloud integration.

Now whether you want to consider that a cross-sell because we're bundling the Data Cloud in to drive additional utilization or not, that's up to you. But to us, as we're rolling out loyalty to all of our global clients, and we're starting to take sell-through and roll it out to the Live Intent clients and all of the different things we're doing, that's in the early stages, and I think will drive meaningful growth in the future.

And the next question comes from the line of Jack Nichols with KeyBanc Capital Markets.

Maybe pivoting back to Athena. I was wondering if you could walk us through the early adoption trends among the enterprise customer base, specifically around how they're deepening engagement with the platform and then existing use cases today? And then I've got a quick follow-up.

Well, first of all, welcome, Jack. It's great to have you on coverage. We really appreciate you. Second, we have been really blown away by the early adoption of Athena. We made it generally available to 100% of our enterprise clients, and we saw a 7x increase in agentic interactions from our clients in the first week of Athena alone. So we think of that as pretty good. 7x is always something we aspire to. But our long-term goal is for Athena to be the operating system of our clients' businesses, and we're just getting started on that. But early adoption has been very, very exciting.

That makes sense. And then pivoting quickly to Marigold and thinking about the recurring revenue mix, as those customers adopt the Zeta platform, should we expect that mix to trend down or up over time or kind of remain in line with the 2025 60% expectation disclosure?

Jack, it's Chris. I'll take this. And as David said, welcome. It should go up is the short answer. And I think a really interesting proof point that you'll see in the Q tomorrow is just how substantially RPOs went up quarter-to-quarter. They went up $66 million just from fourth quarter to the first quarter. Obviously, part of that is Marigold, which then helps with visibility. But I think an interesting thing for the audience here to understand is another large piece of that was not only these marquee wins that we talked about with the apparel retailer and the e-commerce pet retailer, but it was also agencies beginning to now also sign long-term committed contracts.

That is an exciting proof point for us. It adds to the recurring revenue, which then obviously adds to visibility, which both of those came into our confidence to be able to raise the guidance that we did on the top line by $30 million while continuing to keep to our 2% to 5% conservatism.

And the next question comes from the line of Clark Wright with D.A. Davidson.

Awesome. I wanted to maybe quickly touch on the consolidation story. You noted on one of the marquee wins this quarter that you consolidated 4. And I recognize over the course of the last few quarters, you mentioned consolidation being a key piece. Can you talk about the use cases that beta continues to solve for and how you see that expanding over time?

Yes. Thank you, Clark. Listen, when John and I founded this company, I don't know, 18, 19 years ago at this point, our vision was to put everything a marketer needed into one user interface with one reporting infrastructure. And I would tell you that because of Athena, I think we are finally there. And our ability to consolidate anywhere from 8 to 12 different vendors into one user interface and one reporting infrastructure has never been stronger. In the case of this global company because it's a retailer and a manufacturer of their clothing, we displaced what I think many people think to be certainly the longest serving of the marketing clouds. They made, I think, their acquisition first in the space as they built their marketing cloud. And in fact, this particular client used that company for everything. They consider themselves a you know what shop, so to speak. So decoupling their marketing cloud from everything else they were doing, I think, was a very difficult decision.

We also displaced another competitor of ours who tends to be more focused on mobile. They tend to be a little easier to displace because they're so singularly focused on mobile. And neither of those companies brought any data or any activation capabilities to task.

When you're working with one of the large marketing clouds and you displace them, you're almost always also displacing a professional service provider, who they have to then spend millions of dollars on to customize their platform versus our platform is pretty much ready to go from a cloud perspective. So that would be a really good example of a -- and we see this as one of the most important wins in our company's history, and it goes back to not just our ability to consolidate other vendors, but to do everything that each one of those point solution does better than they do while simultaneously putting everything into one place.

Got it. That's helpful. And then if I could just add one more. Over the long term, you talked about increasing wallet share with customers. Do you think AI accelerates the rate of share capture, increases the total wallet share or both?

I think both. I mean, remember, the single greatest way, Clark, to get market share is drive meaningful return on investment to your clients. The Forrester study that came out that said we have a 600% return on marketing spend, we're seeing early adopters of Athena at a materially higher return on investment than even that. The higher we drive return on investment, the more wallet share we're naturally going to get. And as you know, our existing global super-scaled customers will spend well over $100 billion to $110 billion on marketing this year. And at the middle of our range, we'll have, call it, 150 to 170 basis points of wallet share. I believe we can get that to 700% to 1,000% of their wallet share in the years to come. The key will be driving better return on investment. Artificial intelligence, specifically Athena, plus our data as a moat into our business is going to drive return on marketing spend up meaningfully, which we think will then drive wallet share.

Our next question comes from Jason Kreyer with Craig-Hallum.

Great job. So I wanted to stick with the point on wallet share because you announced some major wins and you've announced some major wins over recent quarters. But I'm curious, when you look at the aggregate data representing somewhere less than 2% of wallet share, how big of deals are you winning today? And how big a deal do you think you can win over time just in terms of the wallet share of those customers?

It's interesting, Jason. I would say the last few wins we've had have been at a comparable wallet share to our current wallet share, but the clients are spending 4 or 5x as much per year on marketing and CRM. So they represent some of the largest deals we've ever done right out of the gate. Does that make sense just mathematically?

At the same time, what we're starting to see is some of our clients who have been on the platform for 2, 3, 4, 5 years are getting to that 7% to 10% of wallet share and higher. And we're using that as a road map for how do we take new clients there. So the wins are much bigger than they've ever been, but I'm not sure they're much bigger wallet share only because the companies are so big that we're winning. Now that will give us meaningful upside as they're on the platform, and Chris does a much better job than I do, talking about how ARPU grows the longer a client is with us, and these clients are starting at probably the highest ARPU we've ever seen clients starting.

That also drives, Jason, with our sales pipeline. So we talked about its growth. But if you look at deal sizes and particularly the annual contract value of deals are up pretty substantially year-over-year.

Perfect. Maybe one quick follow-up, David. You've been doing AI for a long time, but it seems like the release of Athena has certainly put you in a different conversation within the AI industry. I'm curious, how has that translated to conversations with customers? And like do you feel like Zeta is becoming more of an AI thought leader in the marketing ecosystem and that's driving that engagement?

It's interesting, Jason. In some ways, being a native AI company has been complex for us over the last few years because everybody is rolling out shiny new products, most of which are not real, but most of the people are rolling them out. And we've always been seen as sort of like AI is under the engine.

Athena is the hood ornament to what we're doing as a company. She is now us announcing ourselves with authority that we are not just an AI company, we are the leader and the disruptor in the AI space. And with the launch of Athena as a marquee product, it has changed the game for the way people are seeing us.

And I will tell you, the 2 client wins we talked about in the prepared remarks, there is 0 chance we would have been in the room if we had not launched Athena or started talking about her at Zeta Live. And there's -- I don't think a chance we would have won these accounts without Athena showing that we are the leader in artificial intelligence as it relates to marketing.

From an internal perspective, we're also one of the best users of AI. I mean back to what I was saying around the SPADE internal platform we've built. If you had told me a year ago, we'd be auto generating 75% of our own code while simultaneously driving the type of quality products we're driving, I would have said that's just not possible. SPADE has made that possible. And it's really been very interesting how we've done that in an environment where we're still using a very de minimis percentage of tokens versus what many of our competitors are doing, which is going to allow us to continue expanding our operating margin as we've done over the years.

Our next question is from Matt Swanson with RBC.

I have my congratulations for the quarter. I think the metric that really jumped out to me was the increase in multi-use case. And I know that's something we had kind of talked about with Athena and its ability to kind of create this organic expansion motion. Given that, that 50% increase was for the full quarter, like is Athena a real part of that? Is there other parts of your go-to-market driving that? If you could just kind of touch a little more there.

The great news is Athena is just getting started. So we had a great trajectory going into our launch. Now I will tell you, every client that was on the beta became multi-use case. So it was -- but that was not a lot of clients, right? So as she rolled out to generally available, we saw an uptick there. But I think that's continued upside to growth in multi-use case. And the One Zeta team continues to just do an exceptional job. I'll remind you, Matt, we really started on the One Zeta mission just 18 months ago. So you've got a massive tailwind coming out of the work we've been doing there. And then I think Athena is going to supercharge that.

And Matt, I think the reason why you picked up on it, but for others, empirically, what we know is that when customers use more than one use case, their ARPU is 3 to 5x greater. So I think you're right on that being an exciting data point.

Yes. No, I appreciate that. And we'll make sure to take note that 100% of Athena users will become multiuse case. That's what I heard.

I wouldn't go quite there. I mean we -- obviously, that's the goal, Matt, but we certainly didn't say that just yet.

Yes. The other one I want to talk about is the independent agencies. I know you called out advertising as a key vertical for you guys. I think your willingness to kind of share the credit with agencies and allow them to white label some of your technology has been part of the reason you've been so successful there. I guess with Athena, how much more can that help you in those deal environments as a lot of these independent agencies are trying to compete with the big holdcos and so on?

One of the key wins we had, Matt, in the quarter was with a large independent. If you look at business done a year ago with them was 0. Business done within this quarter was 8 figures with Athena being, again, something that was visible to them as something they could also exploit for their benefit. The same was true with a very large new agency that began piloting Zeta in 2025. The spend was material, call it, a little less than $2 million, but that new agreement that was signed is more than 10x that size. So both the independent as well as the large agency continues to have a lot of runway. In fact, amongst the 5 large holdcos, the number of brands we're working with year-over-year grew by 50%.

And I just want to say, Matt, we're actually big fans of the agencies. They provide incredible services to their clients. And we've had clients approach us to go direct. And we always try to bring the agency back into it because we think it's a very healthy relationship when it's the 3 of us.

And listen, we're good if the agencies make their money because they're providing meaningful services. But as it relates to our business, I'll remind you, none of the agencies really focus on the retain, which, as of last quarter, is about 60% of our business. So we have real greenfield opportunity there as it relates to the activation, which sort of create customers, monetize customers. We're very, very happy to partner and give the credit to the agencies because they've built incredible businesses, and we're very excited now to be working with pretty much all of the large holdcos. I think now it's all. And well, certainly the biggest ones. And then to be partnering with a select number of independent agencies. There's a lot of them out there, but we want to work with only the best.

Our next question comes from Naved Khan with B. Riley.

This is Ethan Widell calling in for Naved. To start, can we talk about the ideal customer profile for Athena. I'd imagine there are 2 kind of distinct value propositions there, a, where Athena can drive efficiency gains for your larger enterprises that already have sophisticated marketing teams and whatnot; b, more small and mid-market players where Athena creates access to capabilities that these customers don't necessarily have in-house. So like which of these is management really seeing more of early traction-wise? And what's kind of the ideal customer size that you're leaning into with your early sales motion?

Well, it's interesting you put it that way. I mean, today, to be honest, Ethan, we don't focus on midsize. We're really just focused on very large enterprise, although Athena opens up the midsized market to us at some point because you're very intuitive to understand that the cost of layering Athena out to midsized companies is so de minimis to us that would allow us to move into those -- that vertical -- or I'm sorry, into that sort of category without having to meaningfully hire people to do it. But today, we focus solely on very large enterprise.

So I would tell you the 2 things very large enterprises have really focused on is, a, and you're totally right, efficiency. What they're finding is it takes 70% less labor to manage the Zeta marketing platform with Athena than it did with hands-on keyboard. So you're effectively able to take 70% of your marketing workforce and retask them into other functions where they can be more valuable to your organization. We're also seeing that because -- and this is something I talk about a lot, Ethan, but when you buy software, whether it's us or it's Bloomberg or somebody else, you're buying a stealth fighter, right? We're all spending to build a stealth fighter of a platform. And most of our clients know how to fly assessment with -- I mean think about a Bloomberg terminal, the vast majority of their customers only use 5% to 10% of the capabilities. As you look at our platform, being able to fly that Cessna, we're still delivering a 600% return on marketing spend. As clients are able to use Athena as their copilot, they can get right into the cockpit of that stealth fighter, and they can then fly the entire platform, which is driving meaningfully higher return on marketing spend than even that 600%.

Got it. That makes a lot of sense. And then coming out of first quarter, I think you mentioned 9 out of 10 top industries grew more than 20%. I know you spoke to some customer consolidation being a benefit there. But are there any verticals that you see showing any signs of softening, particularly anything sensitive to the macro and geopolitical risk going on right now on the discretion?

Yes. Short answer, no, Ethan. The 9 out of the 10 were effectively say 9 out of 10 that ended last year. The 1 out of 10 that wasn't growing over 20% is 4% of revenue. So it really gives you a sense for the vast, vast majority of revenues on all of the verticals we support are performing in a very healthy way.

Our next question is from Terry Tillman with Truist.

I'll just keep it to one question because I know we're running over time. And maybe I'm getting too far ahead of myself, but I like hearing about 40% increase -- 40% plus increase in the sales pipeline. And I think you said your discretionary markets where you have a lot of activity is even higher. Is it too early to start to say because of the emphasis on agentic AI in general that's in the market, plus you have Athena that's now credibly in the market and in production, could it start to tip the scales and move in some of this funnel activity faster and you actually close new deals quicker? Or is it just too early or I'm just way too optimistic?

I don't think you're too optimistic. But I do think it's -- from an expectation setting perspective and frankly, from a data-driven perspective, and this is a multi-quarter statement I'm about to make. Our deal cycles in good times and in less good times have stayed consistent. What we're seeing is more opportunities in RFPs, many more at-bats than we were given a year ago and certainly 2 years ago. Those by nature take longer. But again, our strategy many times is to work around those processes through pilots and proof of concepts. Those deals are getting bigger, as David said. So yesterday's $100,000 pilot is today is $1 million. But I wouldn't say right now, it's an accelerant, but it's in addition to the pipeline.

And I would concur. But I do want to be clear, Terry, we're getting at bats that we would have never gotten a few years ago. So it's sort of -- it's working really, really well.

Thank you. This concludes the Q&A session and our call. You may disconnect your lines at this time, and have a wonderful day.

Awesome. All right. Thanks, everyone, for joining us at day 1 of the Morgan Stanley TMT Conference. My name is Katie Kuser. I'm on the software research team here at Morgan Stanley. Super excited to be joined by the Zeta Global team, David Steinberg, CEO; Chris Greiner, CFO. Hey, guys, great to see you. Thank you.

Great to see you, Katie. Thank you for having us.

Thanks for being here.

Yes. It's great to be here.

Awesome. A quick disclosure statement before we get started. For important disclosures, please see the Morgan Stanley research disclosure website, morganstanley.com/researchdisclosures. Awesome. So with that, thanks, guys, for being here. Great to see you.

Great to see you.

Maybe for investors that are newer to the Zeta story, just provide a quick overview of the business. relatively complex. So maybe talk to who the end customer is. What about the Zeta portfolio lets you kind of uniquely straddle both marketing and advertising? And maybe just why the business model is differentiated relative to some of the legacy martech players that we all know. we only have 34 minutes.

So let me start by saying Zeta has built a software and proprietary data cloud platform that helps very large enterprises more cost efficiently manage their customer acquisition, customer retention, customer monetization. 51% of the Fortune 100 and 24% of the Fortune 500 use the Zeta Marketing platform today. That is not an aspirational number. That is the number that we work with. And -- what we traditionally do is by combining everything a marketer needs into one user interface with one reporting infrastructure, we're able to deliver a substantially superior return on investment. A recent Forrester survey or report, I should say, showed that for every dollar spent on the Zeta marketing platform, we return 600% on their ad spend as it relates to returns. So I think that's 100% to 200% better than our next closest competitor.

Awesome. And maybe just to hit on kind of the AI debate at the top, right? This is a question that's on all of investors' minds, whether AI is representing a threat to the industry, a transformational opportunity for legacy kind of vendors here. So maybe Zeta has been building AI into its platform for a number of years. You're doing a lot on the innovation front with Athena, which we'll get into. But maybe just high level, like where do you see Zeta positioned on the AI threat versus opportunity spectrum? And I think kind of the key question is what gives you confidence in the durability of the moat going forward, just given how fast the kind of tech?

I mean, obviously, there's this new, what I would call, silly narrative that large language models are going to disintermediate all enterprise software. And I guess I joke at sort of the silly Wall Street narrative du jour. There tends to be one every year or 2. As it relates to Zeta, we have 3 massive moats in our business. First and foremost, we have 552 million active people who are opted in our data cloud today with an average of 5,000 to 7,000 data elements per person our first-party tracking pixel sits on 1 trillion pages of content and ingest trillions and trillions of proprietary marketing signals that synthesize to a deterministic individual that we have never and would never share with any large language model.

Second, in order for our clients to operate with us, which include some of the largest companies in the world, as I've said, 51% of the Fortune 100, we take 100% of their first-party data. We ingest it into a consumer data platform. We then merge our data with it to enrich it. I don't see any of the Fortune 500 companies delivering their first-party data to large language models anytime in years to come.

And then three, we have a 600% return on investment today. Our clients don't look at us as an expense to be disintermediated. They look at us as a revenue source. We are a revenue center for their businesses, and we are driving meaningful growth into their businesses.

And then fourth and final answer is working with a Fortune 500 customer, Katie, very difficult from an onboarding and management perspective, you have to get through their data security groups, data privacy, procurement, legal, accounting, CIO, CTO, CMO. That is a very difficult skill set to build that we have really become experts at over the years. And it took us a very, very long time to get there. So data, return on investment and the ability to operate inside of very large enterprises is a massive moat to our business.

I'll finish by saying we do believe that the large language models are going to massively benefit our business. So -- as I like to say, I started my first company at 21 years old in 1991. I was the youngest person in the room for the first 20 years I ran businesses. I am now universally the oldest guy in any room I walk into. There are some exceptions. But the reality is that teasing a friend.

The reality is that when you think about it, I lived through the mobile era that was going to destroy technology. I lived through the dot-com area where the Internet was going to destroy Walmart and JPMorgan Chase. I lived through the beginning of cloud computing, which everybody the difference between the winners and the losers from my vantage point, has always been the ability to create intelligence Companies that create intelligence and outcomes that adopt these new technologies.

And as you said, we adopted artificial intelligence in 2017. We started working with it back then. When we went public in June of 2021, our banner on the side of the New York Stock Exchange said data plus AI equals intent. Everybody was asking me, who's Al and why is he in charge of our data strategy. That's how sort of far back we go in AI. But we've announced a major partnership with OpenAI, where we're using them to be the voice foundation and the conversational super agent inside of athena, but it's all our models. It's all our data that's training it. We use Anthropic inside of our engineering team. In fact, our engineering team today is 125% more productive than it was just 12 months ago. Now it delivers 150% increment, but 25% of it, you have to Q&A out, right? So -- but the net 125 is very, very strong. We use Microsoft tools. We use Git. We're sort of using everybody across the board to drive productivity. And I think that this is going to be a renaissance for Zeta.

Yes. Yes. Great context. I guess, Chris, I want to kind of bring you in and think about how this translates to the model. Q4 results, very strong last week, growing 28% on a normalized organic basis, guiding to 26% kind of sustaining that 20% growth again. I wanted to ask in the context of how you kind of laid out a framework at your Analyst Day late last year, durability, predictability and profitability of your growth. I thought that was a very thoughtful framework.

It's kind of the quality that investors are looking for. So maybe just talk through how you kind of exemplified that in '25 and how you're thinking about it going forward?

Yes. We used to have a metric of how many straight years we were growing over 20%. We've now added the 3 straight years of growing over 30% on the top line. It's actually 28% even organically on a compound basis over a multiyear period. So I think the ability to show that our market is growing, call it, roughly 10%, we're growing 2 to 3x that consistently. So that puts us in a durable take share position. On the predictability of our business, I mean, it's one thing to beat and raise because you could do that through setting low expectations and or low, low expectations and doing kind of okay. We set growth every year at least 20%. And as you said, we surpassed it by giving ourselves that typical, which we disclosed a typical 2 to 5 points of buffer. So we try to demonstrate through our track record, how predictable our business is. But we also give the metrics that allow us to be predictive in our models to investors. So over 90% of our revenue comes from customers that have been with us over a year. We show investors how those cohorts behave over almost now 5-year period so that you can do the same level of modeling that we do internally.

And then on the profitability side, it was record adjusted EBITDA, which is great. But as a company, we've more and more anchored to free cash flow. It was a record free cash flow margins and a turning point in the company getting to positive GAAP EPS, not just for the quarter but then being able to project it forward for '26 is getting to at least $0.02 to $0.04 of positive EPS. And I think that's just kind of us running from here.

At some point, people will start to give us credit for visibility, right? We've been public for 18 quarters, we've beat our guidance and raised guidance 18 consecutive quarters. So we're obviously -- we have at least some level of visibility into the business.

Yes. Definitely, we'll dig into some of those points that you brought up, Chris. But 1 of the other ways we kind of look at the share gain is by wallet share within your customers, often talk about kind of addressing 1% of the available marketing, advertising, wallet today with potential to kind of drive that up. What are the key strategic initiatives we kind of talked to the impressive ROAs, but anything else that you can kind of highlight on the strategic initiative side to drive that out.

Yes. So to be clear, our 603 global enterprise clients spent $100 million in marketing last year, and we had 1.3% of that wallet share, put us to about $1.3 billion in revenue last year. We're projecting $1.755 billion this year, up from the 1.3% last year, which would -- and our clients are growing about 10% a year. So you would say that grows to about 1.6% to 1.7% of wallet share. Our long-term goal is to get to 10% of wallet share or build a $10 billion a year business with a 30% operating margin and 75-plus percent of that flow into free cash flow. We're not going to do that over the next few weeks, but we do think we're going to do that over the next number of years.

What we're seeing is that the higher we drive the return on investment, the greater the percentage of budget that our clients move to us. And I think the next step function in return on investment is going to be the launch of Athena. Athena, which is our conversational super intelligent agent, say that 3x fast, is really a conversational voice platform that allows you to manage the Zeta marketing platform. So if you think about it, we've built an F-22 fighter jet at Zeta but our average client knows how to fly Ascena. And they're flying that testament they're returning 600%, but they're not using anywhere near the capabilities of the fighter jet. Athena is going to be our clients' copilot in managing the fighter jet. So you'll be able to say to Athena, Athena, I'd like to create 2 million incremental customers this quarter. I'd like to simultaneously lower my cost to create a customer by an average of 7%. What new data sets am I not accessing that I should access to do that or what new channels can we experiment to get there.

Not only will Athena speak back to the client, it will change the entire user interface to show the answers. By the way, there's great demos of this on our website if you want to see it.

The clients that are already using it in beta are seeing a substantially higher return on ad spend than the 600% we're showing, and the best quote I was given that it is totally and completely revolutionized their workflow. So the ability to speak in a conversational way with the platform have it answer you, have that interface then change the platform, be able to say, yes, Athena, I like those new data sets. I like that channel, let's start the activation, but I don't want to start with $10 million, I'd like to start with $500,000. And I have to be out of the office for the rest of the day. Could you e-mail me hourly with the return on investment, so I can track it remotely, and we can make sure it's staying on track before I expand this. That's all available in version 1 of Athena, which will be generally available by the end of this quarter.

Awesome. Maybe taking a step back and thinking about the data asset that feeds a lot of this roast goodness. Very robust proprietary database. Maybe talk us through what having this database does for you? How is it compiled? And then when we think about the breadth of the data set, definitely very wide. But when we think about the depth of it, right? How do you guys think about the need to kind of go out and acquire more data? Do you feel like you have enough to keep delivering that insight?

Well, so first of all, our data cloud which has 552 million globally, people opted in, 242 million in the United States. We touched 2 billion people a quarter. We just don't track them because they're not opted in. We do get there -- we do see some of their data from a modeling perspective. We have 20 different platforms that plug in to publishers that we partner with the publishers. We give them software that makes them more profitable, allows them better interaction with their end users, allows their end users to share their data out and drive large volumes of traffic to the publishers, and in exchange, they put our first-party tracking pixel on every page they host. They host our Java script on every page and then a host of other things that we do.

So what I would tell you is today, we have pretty much every credit card transaction, we have all of their psychographic, all of their demographic. All of their web behaviors, searches are all synthesizing in for those 552 million people. As it relates to, is it enough? Listen, some data isn't valuable until 7 years later. Some data is worthless in 10 seconds. So the more data we can ingest, the more powerful the models can be and because we never share our data with any large language model ever, it's proprietary to us and it's a major moat to our business.

What I would say as well is, whenever we look at an acquisition, when we think about buying companies, we bought 18 companies in 18 years. I'd say it's highly probable there will be in 19. We look at data sets and we look at what's there. And when you think about the most recent acquisition of Marigold, where we picked up some unbelievable enterprise assets, you picked up Geodigital, you picked up Selligent, you picked up sale through. But I'd say the most valuable is the loyalty program, where many Fortune 500 customers are currently using us for loyalty. That is now at giving a SKU-level transaction data to train our models on. And my entire goal is to continue to drive a greater return on investment. My goal is to get to an order of magnitude, return on investment for our clients, 1,000%, and I believe we can do that over the next couple of years as it relates to our data sets, the evolution of our internal AI, partnering with third-party AI partners, and I think Athena will be the tip of the spear.

And maybe talk a little bit about the 2 platforms that you have to activate this data, direct integrated. Maybe talk a little bit about channel expansion.

So I want to be clear. We have 1 platform. So you asked the right question, but we have 1 platform. We have 1 user interface, 1 reporting infrastructure. The SEC mandated that we reported in 2 separate segments. One we call on our platform. One is our integrated platform. Our integrated platform is mostly our partnership with Meta and others where we match the Zeta ID to the Meta ID. I don't know really anybody else in our industry who does that. And listen, Meta takes a disproportionate percentage of the revenue as they should, which makes us even more valuable to them. As it relates to how we sell it, I would tell you that when we started with our agency holdco clients, they didn't have a solution for seamlessly managing their social media marketing. Most of the competitors in the programmatic space, they're just focused on the open web, and they're all trying to figure out -- they're all trying to figure out connected TV, whereas we came at it with a fully integrated solution that included social, mobile, display, online video, messaging and connected TV. When a Client activates through us using our data and our setup tools to a third-party walled garden that we partner with, we count that as integrated. Whereas on platform, it's going directly through us to the publisher.

Right. And I think the kind of next set up for why Athena can be so interesting is when we start to think about the cohort progression right that Chris were talking about early. So maybe when you are thinking about the use case expansion with your customers, what are you seeing there? You talked to an 80% kind of increase in incremental use case expansion on Q4. So maybe what's driving that strength in the use case adoption?

Yes, it's interesting. Outcomes, better outcomes, but we just conducted our most recent NPS survey. And what the survey comes out almost indisputably is the more our customers use what's available on the platform, whether that's more channels on or use cases or more of our data, the happier they are. Our ability, and I think it was seen through our net revenue retention rate of 120%, which was coming off a prior year record 114% is a good indication reflection of that usage that happiness that loyalty of the customer. So what's been driving that, to your point, for most part, up until 2024 was channel expansion. So we're getting an average of, call it, 3 channels per skilled customer. This year, 2025, I should say, was the year in which we saw an inflection point in use case adoption. That's important to our financial model because when customers use more than 1 of our use cases, they spend roughly 5x more. And that could be going from retention to grow or retention to acquire or any different direction.

We are now at a point where in the most recent quarter, 80% -- so 80% growth in number of use cases, greater than 1, which reflects almost 1/4 of our total customer base, and that's up from, call it, 13% of our customer base using more than 1 use case a year ago. I think for us, the 1 Zeta sales motion, which we're organized, all sellers can sell any channel, if you will, on the truck. But we do have them organized by use cases, that's how our customers, for the most part, organized. They have existing customer set of technologies and people and new customer acquisition set of technologies and people focused on that. We have a thin layer of -- think about it as fighter pilots that sit above all of those use cases that are very strategically going at existing customers and then pro-market to where we see we can add immediate value by getting them to use more than 1 use case.

And if you think about it, when a client uses us for customer acquisition and then chooses to add retention, the data from acquisition informs retention. Then the data from retention informs monetization, which then both of those inform acquisition and the return on investment, not just return on ad spend, the return on investment goes up exponentially when you use us for all 3, which is why they spend 500% more. We actually discovered this almost by accident. When we looked inside of the NPS score a year ago, and we said, wow, we better get behind this. And we brought in a gentleman named Ed C, who ran the Chief Marketing Officer practice for McKinsey, very difficult to get a practice leader out of McKinsey, but we brought him in his technical title as Chief Growth Officer, but I call him Chief One Zeta Officer.

And he's doing 2 things. One, they're going in and they're doing this. This 80% lift in customers who use us for multiple use cases with not an accident. But what's more important than that, Katy, is, I believe, Athena is going to be the tip of the spear for multiuse case.

Let's go back to what we were talking about earlier. So now we've completed how we're going to add 2 million customers at a 7% discount, Yada yada, athena can then literally pivot and say, by the way, did you know 1 million of your existing customers are currently doing research with your competitors' products and thinking about churning. Would you like to save them? And that's how you go from 1 use case to automatically because we're tracking all the data.

Awesome. Yes, that's helpful. Maybe for a little bit of context around the actual technology of athena, right? How would you characterize the evolution from ZOE, which was previously the kind of innovation engine here and then into what's kind of fundamentally different about Athena? And then started to get at what's really interesting from a use case perspective. Maybe, Chris, if you could kind of just talk to monetization, if you kind of look at historical precedents, there was a pretty material uplift for low usage. And so kind of how are you thinking about it?

Why don't I start with the technology and how it works, and I'll turn it over to you, Greg, give the percentages on Zoe return. What I would say first and foremost is when you look at the fighter jet that we bought, it is so complex. Like it's not just us, it's tech companies, we get so excited about building our platform, and we put every feature in it and every function at it. We think it's amazing and then clients get in front of them, like, what the hell do I do with this? And we're a product of that, right? So we found that clients are using a small percentage of what our platform can do because that's what they know how to do. And -- the concept of sending out a learning and development group to train these people how to use the entire platform. When this is what they do an hour a day. They're not doing it 100 hours a day, right? Not possible.

But somebody will tease me later for that. The bottom line is that by adding the voice enablement that can also control the platform -- it's going to open up all of the data sets and use cases and channels that they don't currently use. Today, all of our AI is internally built -- you've got large language models out there. You've got small language models out there. I sort of jokingly say we build midsized language model, where we're -- we're looking at trillions of data points, but we don't need to ingest the entire Internet to decide if this individual is going to get a credit card and will credit approved for our clients' credit card, right?

So when you think about Athena, she is native to the application layer for the Zeta marketing platform, if she wasn't she couldn't control it. Like you couldn't have an API in and out and have that work. It has to be native. But we're using OpenAI's voice enablement platform, which is the best in the world to make her truly conversational -- the other thing that's going to happen in V2 is you'll be able to access her as a client inside of your chat app inside of your phone. So it's going to be native going both ways. And I think that's going to be an incredible uplift to return on investment to get from that 600% to 1,000%.

Chris, the intent from the very first version, if you will, of athena, which was a Z opportunity engine or so. It was to make life easier for our customers to use more of the platform. There was so much data and intelligence to be extracted, but it was sometimes hard to navigate. So though we made that discovery process faster and easier, athena now kind of hypercharges that. And we wanted to understand -- our intent was to make life easier, which met you use more channels, more use cases, -- was that actually happening on the platform. So then we took a sample of customers in the dozens that had Zoe's capabilities.

And then we looked at it over a month, 6 months, 9 months, 12 months, beyond a year of what was their change in ARPU over that period of time, those that had Zoe but weren't using it and those that had Zolan were using. We found that it was a 2 to 3x greater ARPU on those earliest adopters of Zoi then those were not using it. So as we begin to kind of have the fun with numbers on what athena could generate, it's -- you just kind of start at a base level 2, and you can really model up because athena has really made it multiplier easier to navigate the platform.

And to be clear, at this point, we've included a de minimis amount of Athena increment into the projections for this year because it goes generally available by definition, we're saying the end of the quarter, but it's now by the end of the month, right? We're in March. And we're very, very much on track for that. I feel highly confident.

And the reality is that we'll have to see. What we're doing is we're going to launch it, we're actually birthing athena sort of like the birth of venous we're going up the birth of athena -- we'll see if we can pull off about a Chilean event. But the reality is that when she comes out, we've got a full learning and development package that we're going to be rolling out with it to all of our clients and we'll start with our top 30 customers, and we'll physically go visit them with the learning and development group to start getting everybody up to speed.

So listen, if they adopt it quickly and they see higher return on investment, I think it's highly probable that they will spend more than we're budgeting.

Wanted to zoom out a little bit to some of the bigger picture kind of industry buying patterns that you're seeing from customers. The RFP sets that you guys throw out continue to be very, very interesting...

Only up 100%.

Maybe like anything particular in the competitive dynamic that's driving that? Any differences in kind of win rates that you're seeing over the past few quarters? And then not just kind of more RFP wins, but larger ones. So kind of what's kind of driving that?

You know what's been interesting is, for many years, we got RFPs. But you could tell, they were literally written by the incumbent, right? Like the incumbent sitting in the room, writing the NFE, so they cannot lose it, right? Because this might be a Salesforce shopper, it might be an Adobe shop or it might be an Oracle shop, and they're not going to not be those shops. And by writing the RP, those large enterprises probably have a rule, they have to go out every 3 years or every 5 years or so on and so forth. That's changed.

For the first time, I would say the vast majority of the RFPs we're receiving, the companies are willing to bifurcate Marketing Cloud, from the core products of the larger software companies. And I think a lot of that is because their architectures at this point go back, in most cases, 15 years. You're still talking about exact target response oseneolane here. And what happens is you can't integrate AI into the platform. You can talk about Agent Force from now until you're blue in the face, but it doesn't make any difference if it doesn't really work, right? And Salesforce is marketing cloud shrunk last quarter. That was their results. we're not interpreting them. That's what they said. Adobe and Oracle don't break it out, but I don't know what they're doing.

What I would tell you is the architecture that we have built has data and AI is native to the application layer. When our competitors put AI on their platform, you have to step out of the platform to an AI algorithm, do a query. The algorithm has to do a data dip into a data repository back to the algorithm to create intelligence and then it forms the Marketing Cloud what to do. That latency destroys return on investment. Because AI and data, we launched a whole new platform in 2021 that literally, we can process thousands of data points in milliseconds, and decide, should this person see this ad, get this message or get this connected TV ad, that is creating this 600% return on investment.

So I want to be very clear guys like Salesforce, Oracle Adobe are some of the greatest companies in the world, their core products are best of breed. I don't even think they have meaningful competitors yet in their core products. But their marketing clouds are now the side hustle to their side hustle, and they're just not getting the investment. They would have to re-architect the entire thing, which they could, at some point, choose to do, but then none of them own any first-party data. They're just using their clients' data to train the algorithms. We are very, very focused on the sort of 1 plus 1 equals 4 as it relates to our clients' data plus our data -- and so not only are you seeing 100% increment in RFP volume, which we talked about. I think we're seeing RFPs we're going to win at a greater percentage than we've ever won before.

Yes. And I think 1 of the most compelling data points from the print was how balanced from a vertical perspective it was 9 out of the top 10 kind of growing over 20%. I guess are there any key areas from a vertical perspective that you're thinking could be delivering outside share gains? Anywhere specifically you're thinking about for '26?

It's continued to strike us positively that the verticals that are closest to consumer discretionary, which are the brands that have to make the most -- kind of get the most ROI for their dollars invested were our best-performing verticals again. They grew over very strong comps in the year prior. One vertical that I'd call out because we've been intentional on incremental investment, not just in people but healthcare is 1 that you absolutely cannot have a general son. He must have a domain expertise in health care. By the way, across health plans, across health systems, et cetera. But we've also been intentional in finding partnerships around data there. So that was a vertical that did not grow 20% in 2024, that did grow over 20% in 2025 that we're optimistic about.

And I would just say, 1 of the things we decided when we sort of pivoted this business in 2017 into what it is today, was we wanted to operate across every vertical, right? We effectively operate against 15 different verticals that I'm not sure what percentage of GDP, those 15 make up, but I know the top 3 make up half. So it's got to be a disproportionate percentage. What I would say is we're very well balanced. We usually, there's a year where a couple of the verticals are flat. A couple of the verticals are growing a little bit. A couple of the verticals are skyrocketing growth. The last couple of years, we've seen really solid growth from over 10 of our verticals in pretty much every quarter. And -- it's been interesting, even some of the verticals that their industries have been challenged. They're moving more budget to us than ever before because they need the higher return on investment. And then the guys who are skyrocketing are doing the same thing because they want to supercharge the growth. So it's been great.

I would tell you, Katy, I believe, based on the business we've built, the assets we control, we have the people, the technology and the data to continue to be a 20-plus percent organic growth company for many, many years to come. Yes. And you're not just growing well, but you're expanding margins as well.

So maybe with the last few minutes, just talk about kind of the key leverage drivers and how you're thinking about balancing AI investment was still expanding I'll start and what's you turn.

But let me just start by saying a lot of the investments we made in AI, we made from 2017 to 2021. So everybody is now trying to catch up at 3,000% more cost per item. So we were able to get in before the price is really skyrocketed. A lot of people after our call said, well, if you're really delivering 600% return on investment, why are you not raising price? And to me, what a lot of people don't understand is the 40% approximate of our business, that's utilization fees. We effectively buy the marketing for our clients, and it transitions through our platform. So we decide what the margin is there. Do we want to manage it to growth or do we want to manage it to profitability? I believe that over the next few years, we will surpass 1,000% return on investment to our clients. And I believe we'll be able to keep all of the margin that exists above that, which will start to drive meaningful increment to gross margin.

Chris, I didn't leave you much time, but I'm sorry.

I'll go super fast. We've created a model to where we don't have to sacrifice growth or profitability or vice versa. We give as much investment we can to the innovation team. We try to be smart in how many people and who we hire on the marketing and sales side. The seeker sauce is everything else in the middle should be 0. And we're not there, but that allows us to really refine get efficiency in the middle over invest in the end, and that's resulted in great growth and profitability.

And by the way, we've grown the business, I think, on a 3 or this will be the fourth year of a compounded 30-plus percent growth rate greater than the 50% EBITDA. And last year, we grew free cash flow by 78%. So, yes, I feel like we're doing a good job managing to that.

Very impressive. Awesome. We'll leave it there. Thanks so much.

Thank you, Katie. Appreciate it.

Greetings, and welcome to the Zeta Q4 2025 Earnings Conference Call. [Operator Instructions] Please note this conference is being recorded.

I will now turn the conference over to your host, Matt Pfau. Thank you. You may begin.

Thank you, operator. Hello, everyone, and thank you for joining us for Zeta's Fourth Quarter 2025 Conference Call. Today's presentation and earnings release are available on Zeta's Investor Relations website at investors.zetaglobal.com, where you will also find links to our SEC filings, along with other information about Zeta.

Joining me on the call today are David Steinberg, Zeta's Co-Founder, Chairman and Chief Executive Officer; and Chris Greiner, Zeta's Chief Financial Officer.

Before we begin, I'd like to remind everyone that statements made on this call as well as in the presentation and earnings release contain forward-looking statements regarding our financial outlook, business plans and objectives and other future events and developments, including statements about the market potential of our products, potential competition, revenues of our products and our goals and strategies. These statements are subject to risks and uncertainties that may cause actual results to differ materially from those projected. These risks and uncertainties include those described in the company's earnings release and other filings with the SEC and speak only as of today's date.

In addition, our discussions today will include references to certain supplemental non-GAAP financial measures, which should be considered in addition to and not as a substitute for our GAAP results. We use these non-GAAP measures in managing our business and believe they provide useful information for our investors. Reconciliations of the non-GAAP measures to the corresponding GAAP measures, where appropriate, can be found in the earnings presentation available on our website as well as our earnings release and our other filings with the SEC.

With that, I will now turn the call over to David.

Thank you, Matt. Good afternoon, everyone, and thank you for joining us today. We delivered our 18th consecutive beat and raise quarter. And I want to be clear about why this keeps happening. It is not a single product cycle or a favorable compare. It is the compounding effect of a system, proprietary data that improves with every customer interaction, intelligence that sharpens with every decision and now an interface in Athena that lowers the barriers to enterprise-wide adoption.

The flywheel is what drives durable, predictable and profitable growth. Fourth quarter revenue was $395 million, up 28% year-over-year ex acquisition and political candidate, an acceleration from the third quarter.

Adjusted EBITDA was $95.1 million, up 35% year-over-year, and we had positive GAAP earnings. With these multiyear rates of revenue growth, we are taking market share. This is evidence Zeta is the AI disruptor in marketing.

After providing initial 2026 outlook that was already ahead of consensus, we are once again raising the midpoint of our 2026 revenue guidance by $25 million to $1.755 billion, reflecting year-over-year growth of 35%.

Our momentum is driven by AI shift from feature to infrastructure across the enterprise, and Zeta is built for this transformation. Our AI investments, which began 8 years ago, are yielding better and better results. Marketers are recognizing the strong return on investment delivered by Zeta and increasingly view us as a revenue center for their business, not a cost center.

A recent Total Economic Impact study by Forrester confirmed a 600% return on ad spend, which we believe is contributing to the wallet share gains from our competitors in an already quickly growing market segment. This is a reflection of what happens when proprietary data, intelligence and activation work as one operating system, not three separate tools. Zeta is our client's marketing operating system, and we expect these returns to grow as our AI capabilities advance and Athena fully launches. The impact is already visible in our performance.

Net revenue retention hit a record high of 120% and in 2025, up from 114% in 2024. And RFP volume more than doubled year-over-year to a new record. Customers are spending more and new prospects are showing up faster, both signals of the same thing, the Zeta operating system is working, and working at scale.

Building on this momentum, at Zeta Live last October, we introduced Athena, our super intelligent agent built specifically for enterprise marketing. And since then, the pace of interest, engagement and opportunity has continued to increase.

At CES last month, Athena had a significant presence and customer engagement was exceptional. Feedback was overwhelmingly positive, with customers consistently recognizing Athena represents a fundamentally new way of working.

Early Athena users are reporting significant time savings in segmentation, production, analysis and substantially better return on investment. This is the kind of workflow transformation that drives deeper platform adoption and greater utilization. Athena enhances the Zeta Marketing Platform that is an intelligent operating system for growth, one that can listen, reason and act on behalf of marketers in real time. We are very encouraged by this early customer feedback and remain on track to make Athena generally available by the end of the first quarter.

At CES, we also announced our partnership with OpenAI. We view large language models much like cloud infrastructure, foundational technologies that enable innovation with real differentiation coming from the tools, workflows, data and operating systems built on top. Our partnership makes OpenAI's technology foundational to Athena, but powerful models are only part of the equation. AI is only as effective as the data that fuels it.

As personalization moves to true one-to-one, identity and intent become critical. That's where Zeta SuperGraph comes in, a proprietary deterministic identity and relationship graph within our Data Cloud that unifies data across multiple sources. This SuperGraph creates a moat that widens with every improvement in AI models across the market. Built over the past decade, Zeta SuperGraph operates at scale across more than 245 million U.S. adults and 535 million globally, with more than 1 trillion signals, the vast majority of which are available only to Zeta.

As AI demands higher quality deterministic data to deliver real personalization, this asset becomes more valuable, not less. Our AI and data advantage also helped to fuel One Zeta. One Zeta is no longer just a strategy. It is a repeatable sales model. In the fourth quarter, the number of scaled customers using more than one use case was up over 80% year-over-year, and the opportunity in front of us continues to expand.

Today, we serve 51 of the Fortune 100, up from just 44, 1 year ago and over 120 of the Fortune 500. Collectively, these clients alone represent well over $100 billion in annual marketing spend, significantly expanding the long-term One Zeta opportunity.

Athena further amplifies this strategy by removing the friction across the Zeta Marketing Platform, making it easier for customers to adopt, expand and scale of multiple use cases.

And as I will discuss in a moment, the Marigold acquisition adds another important accelerant to One Zeta by expanding the data, use cases and value we can deliver to customers.

Taken together, the combination of our AI leadership, the continued maturation of One Zeta and the momentum coming out of Zeta Live and CES, these are reinforcing our powerful growth flywheel. This momentum pushed our pipeline to record levels coming out of Zeta Live, and we have already closed $39 million of business directly attributable to the event, putting us well on our way to our goal of $100 million total.

I will close with an update on Marigold. The integration is progressing well, and we continue to anticipate Marigold being accretive to free cash flow and adjusted EBITDA in year 1. We are actively engaging with Marigold's enterprise clients through a One Zeta lens, identifying opportunities to add value, expand use cases and deepen those relationships over time. We are also seeing strong interest from Zeta customers in adopting Marigold's loyalty product.

As I reflect on what Zeta accomplished in 2025, I am so incredibly proud of this team. We exceeded our initial revenue guidance by 5% and our free cash flow guidance by 27%. We expanded existing partnerships and forged new ones, increasing our leadership in the marketing ecosystem. And we developed the most important product in our company's history with Athena. Together, we expect these achievements to extend Zeta's position as the defining AI disruptor in the marketing ecosystem. As always, I would sincerely like to thank our customers, our partners, team Zeta and all of our shareholders for the ongoing support of our vision.

Now let me turn it over to Chris to discuss our results in greater detail. Chris?

Thank you, David, and good afternoon, everyone. Our results once again demonstrate the durability, predictability and profitability of Zeta's growth, which we expect to continue as seen in our increased 2026 guidance.

Revenue growth, excluding LiveIntent, Marigold and political candidate was 28.2% in the fourth quarter, up from 28.0% in Q3 and up from 27.4% in Q2 and 25.9% in Q1. And 2025 was the sixth straight year in which revenue grew greater than 20%, underscoring the durability of our growth and the sustained market share gains we are taking. Results once again exceeded guidance. And just 2 months into the year, we are raising our 2026 outlook, reflecting the visibility and predictability of our business.

For the full year 2025, we expanded adjusted EBITDA margins by over 200 basis points, achieved the highest free cash flow margin in our history, generated $199 million in net cash provided by operating activities and turned GAAP net income positive in Q4, demonstrating the profitability of our growth.

Now I'll discuss our results in more detail. In Q4, delivered revenue of $395 million, exceeding the midpoint of guidance by $14 million or 4 percentage points higher than our forecast, solidly within the 2 to 5 points of cushion we typically leave ourselves. The full year's revenue was $1.305 billion, up 30% year-over-year or 27% when excluding LiveIntent, Marigold and prior year political candidate revenue. This exceeded the midpoint of our initial 2025 guidance by $65 million or 5%.

Our platform is relied upon across virtually all industry verticals. For the first time as a public company, 9 out of our top 10 verticals in 2025 grew over 20% year-to-year, quite a testament to our broad-based leadership. Some of our fastest-growing verticals in 2025 were travel and hospitality up 105%, advertising and marketing up 70%, automotive up 60% and consumer retail up 46%. And worth noting health care, an area of new investment, grew over 20% and is showing strong momentum.

Total scale of customer count grew to 602, up 14% year-over-year and an addition of 30 customers sequentially. We ended the quarter with 184 super-scaled customer customers, up 24% year-over-year, well above our target of 4% to 8%. Super-scaled customer additions were broad-based across industry verticals, including travel and hospitality, services, technology and media and consumer and retail and were driven by continued momentum of our One Zeta initiative. Scaled customer quarterly ARPU of $625,000 increased 8% year-over-year and over 15% when normalizing for political candidate revenue in the year ago period. Super-scaled customer quarterly ARPU of $1.8 million was up 5% year-over-year and also up mid-teens when normalizing for political candidate revenue last year. Both growth rates are in line with our ARPU target of 12% to 16%.

Now I want to double click on the importance of super-scaled customers for a moment. Since our IPO, Zeta's growth has been fueled by super-scaled customers whose spend has grown to represent a larger and larger portion of total revenue. Back in 2020, customers spending at least $1 million annually, represented approximately 70% of total revenue. By 2025, that figure is now approaching 90%. And over that same period, more than 90% of total revenue growth has come from the at least $1 million super-scaled customer cohort. This dynamic is leading to a natural evolution in the reporting of KPIs. In 2026, we will exclusively focus quarterly reporting on super-scaled customer count and ARPU.

In many ways, today's $1 million-plus customer is 2020's $100,000 customer. In fact, one could imagine a new cohort of $10 million-plus customers emerging down the road. The growth and prominence of super-scaled customers is by design. It reflects how we manage the business internally, shows the efficacy of our land, expand, extend, hunter-farmer sales motion, along with the power of One Zeta working in unison.

Our customer-centric flywheel propels the progression from pilot to broad-based platform adoption, powering our very strong net revenue retention rates. First, we allow customers to start small with pilots and proof of concepts ranging from $50,000 to $150,000, which can begin with Zeta Data usage, a CDP or channel activation. Second, we provide these pilot customers with transparency into measurement and ROI. Third, our software and AI learn from past programs and form recommendations to continuously improve return on ad spend, which all leads to creating new audiences, adding more channels and branching into new use cases, each incremental to ARPU.

This is best brought to life by Slide 10 in our earnings supplemental. Customers on the platform in their first year quickly scaled to $709,000 in 2025, representing 111 scaled customers or 6% of total revenue. These customers often start with 1 channel and 1 use case.

Customers on the platform in years 1 to 3 spent an average of $1.1 million in 2025, representing 217 scaled customers or 19% of revenue. These customers begin to scale beyond 1 channel, but remain generally on 1 use case.

Customers in years 3 through 5 spent an average of $2.1 million in 2025, representing 65 scaled customers or 11% of revenue. Customers in this cohort are moving into omnichannel experiences.

And lastly, customers on the platform 5 or more years, the earliest adopters of the ZMP spent an average of $3.9 million in 2025, representing 209 customers or 64% of total revenue. This cohort makes up the largest percentage of revenue and the fastest-growing ARPU, up 39% year-to-year.

Shifting to revenue mix, direct revenue in the fourth quarter was 74%, in line with the year-ago quarter and our target of 70% to 75%. Our GAAP cost of revenue in the quarter was 40.4%, a 50-basis point increase year-over-year and 100 basis points sequentially. The increase in cost of revenue was driven by strong sequential and year-over-year growth in social and connected TV.

In the fourth quarter, we generated $95.1 million of adjusted EBITDA at a margin of 24.1%, 174 basis points higher year-over-year and $4 million better than the midpoint of our guidance. This marks the 20th straight quarter of expanding adjusted EBITDA margins year-over-year. For 2025, adjusted EBITDA was $279 million, representing a margin of 21.4% and 44% year-over-year increase.

We also generated positive GAAP net income. For the fourth quarter, our GAAP net income was $6.5 million, an improvement from a net loss of $3.6 million last quarter.

Fourth quarter net cash provided by operating activities was $64.1 million, up 47% year-over-year, with free cash flow of $55.8 million, up 76% year-over-year and representing a margin of 14%. This represents a free cash flow conversion of 59%, a significant improvement from 45% in the fourth quarter of 2024. This also includes a roughly 9-point working capital headwind driven by longer agency payment cycles common to their industry.

The improvement in both adjusted EBITDA margin and free cash flow margin conversion in the fourth quarter exhibits the strong operating leverage of our model, which we believe puts us firmly on track to achieve our Investor Day target of a 30% plus adjusted EBITDA margin and greater than 70% free cash flow conversion in 2030.

For 2025, our free cash flow was $165 million, a margin of 12.6% and up 78% year-over-year.

During the fourth quarter, we repurchased 1.9 million shares for $35 million. And for the full year 2025, we repurchased 7.9 million shares for $120 million. Since January 1, 2026, and up until mid-February, we have repurchased an additional 1.5 million shares for $25 million and have roughly $139 million remaining on our share repurchase authorization. We expect to remain active buyers of our stock, especially at these levels.

We continue to make significant progress in reducing dilution and stock-based compensation expense, just as we said we would. We ended 2025 at the low end of our guidance range with total net dilution of 4.3% or 2.2% excluding Marigold. Additionally, we improved the ratio of stock-based compensation to revenue from 19% in 2024 to 14% in 2025. And we remain on track to achieve our normal course 3% to 4% net dilution target in 2026.

Now on to guidance. We are raising first quarter and full year revenue, adjusted EBITDA and free cash flow guidance. Details can be found starting on Slide 19 in our earnings supplemental.

But before discussing the numbers, I'd like to highlight a few key aspects of our business model. First, our revenue is tied to the volume decisions made, not seats, whether those decisions are made by a human or an agent. This foundation and flexibility to adapt to the rapid pace of AI innovation provides durability to Zeta's growth. Second, strong pipeline visibility and sales productivity support our confidence in the year ahead, and we continue to guide with planned conservatism of 2% to 5%. And lastly, as we integrate Marigold, we expect to realize further operating leverage.

For the full year 2026, we are increasing the midpoint of revenue guidance by $25 million to $1.755 billion, representing a 35% growth rate or 21% year-over-year when excluding Marigold and political candidate revenue. None of our guidance raise is related to political candidate revenue, which we continue to assume will be $15 million in 2026 with $7 million in the third quarter and $8 million in the fourth quarter. Additionally, we continue to take a conservative view of Marigold, contributing at least $190 million in 2026 revenue. And lastly, our revenue guidance includes minimal contribution from Athena-driven revenue with its broad-based adoption representing incremental consumption revenue upside.

For the first quarter, we now expect revenue of $370 million at the midpoint, $8 million higher than our previous guidance and representing year-over-year growth of 40% or 22% when excluding political candidate and Marigold revenue. The linearity of revenue is it spans each quarter of the year aligns with historical averages, so there is no front or back-end loading of revenues or growth rates.

For adjusted EBITDA, we are increasing the midpoint of our 2026 guidance to $391 million, up $6 million from our prior guidance and representing year-over-year increase of 40% at a margin of 22.3% and an improvement of 92 basis points over 2025. For the first quarter of 2026, we now expect adjusted EBITDA of $61.5 million at the midpoint, up from our previous expectation of $60 million and representing growth of 32% and a margin of 16.6%.

We are also increasing the midpoint of our 2026 free cash flow guidance to $231 million, up $7 million from our previous guidance and representing year-over-year growth of 40% at a conversion of 59% of adjusted EBITDA, which likely has upside.

Additionally, we believe our fourth quarter positive net income on a GAAP basis represents an inflection point, and we expect to generate positive GAAP net income for the full year of 2026, a significant milestone for the company.

Today, we would also like to provide updated Zeta 2028 targets to account for the acquisition of Marigold. We're raising our revenue target from $2.1 billion to $2.3 billion, representing a CAGR of 23%. We are also increasing adjusted EBITDA target to $573 million, which implies a margin of 25% and a free cash flow target of $371 million, which implies an adjusted EBITDA conversion of 65%.

I'll conclude with this. In environments like today, it is more important than ever to control the controllable. For us, this boils down to hiring and retaining the industry's best talent. This is how you continue to lead. Delivering customer outcomes with unmatched industry ROIs. This is how you garner loyalty. Generate best-in-class retention rates and win market share, and doing what we say we're going to do for investors. This is how you earn trust and show durable, predictable and profitable growth.

Now let me hand the call back over to the operator for David and me to take your questions. Operator?

[Operator Instructions] And our first question will come from Terry Tillman with Truist Securities.

I had a question and a follow-up. I must ask about Athena. And again, I know that it's not generally available yet, but when you all put out a press release a while back, it did actually have a custom already quoted in there, which I thought was interesting. But you talked about a couple of agents at that point, Insights and Adviser. Can you remind us, is there going to be kind of a whole slew of agents that you will release? Or are these going to be the 2 primary wins and there is a monetization structure around that?

And the last part of this Athena long-winded question is, with the business that you've signed year-to-date, has it had an influence in some of these deals getting across the line, even if you're not monetizing? And then I had a follow-up.

So thank you, Terry. Let me start by saying that those 2 agents are first because they will drive probably the greatest benefit to our customers. So when you think about Athena and you think about large language models and how we are using them to win in the marketplace, today, the Zeta Marketing Platform is like a 747. Most of our clients know how to fly a Cessna. The beauty of Athena is the ability to fly the entire 747 just by narrating and speaking to Athena who can automatically do it.

So to answer your question, these will be the first 2 agents. We will have it generally available by the end of this quarter. We have a number of clients as early users today. The initial feedback has been, it is game changing from a workflow management perspective, and it is driving substantially higher return on investment than the existing 600% return on ad spend that the platform is generating.

As we roll out additional agent function built into Athena. She'll be one agent, but with different feature sets, if you know what I'm saying. It won't be different agents sitting under her. Everything will be driven by Athena. We're going to continue to drive out functionality based on things that we believe will drive the highest utilization rates, the highest move to One Zeta. And I think you'll see that really start very quickly out of the gate soon.

That's great to hear. And I guess, Chris, maybe a follow-up. You framed it well in terms of maintaining that 2 to 5-point cushion. But I think for the year, sometimes -- you've talked about maybe your top 5 kind of spending verticals and maybe the assumptions you're assuming for the rest of the year off of your top 5 verticals. Can you maybe share kind of another way of looking at conservatism, looking at some of your biggest verticals and what you're assuming the rest of the year like consumer retail and travel and hospitality?

Yes, Terry. I mean what we found was that those verticals that have been -- are closest in proximity to consumer discretionary, you saw through the prepared remarks, were and have continued to be throughout the year, our top performing verticals. This was a first for us in our history to have 9 out of our top 10. And that's, by the way, on a trailing 12-month basis. That's not cherry-picking in a particular quarter, grow over 20%. Our guidance does not assume that. Our guidance is probably closer to half, which is even below what we've been historically.

The other way I'd encourage you to think out the conservatism in our guide is we've reiterated that continued 2 to 5 points of cushion. We have also kept what we believe is a conservative view of political candidate revenue at $15 million in the guide. So of the raise that we put through, none of it was a change in our assumption of that $15 million. So that should be conservative. We continue to be conservative around Marigold.

And to the extent, as David talked about, we get really strong continued adoption from Athena and that adoption has proven to generate increased ARPUs and usage, we're assuming minimal contribution right now on the guide. So that should also be incremental upside. So I feel like this is a very nicely derisked outlook to start the year.

We'll go next to Zach Cummins with B. Riley Securities.

Congrats on the strong end of the year. David, I just wanted to ask about the deals that you've already closed coming out of Zeta Live, I think almost nearly $40 million in business. I mean any particular trend that you can highlight, whether it's customers in a particular vertical or a particular solution set that's driving the early momentum. And then just curious on how you're thinking about that continued progression through those pipeline opportunities in the coming quarters?

The one consistent, Zach, is they all really love The Chainsmokers. No, in all seriousness, it was across multiple verticals. We saw a very big win in telecom, which I think will continue to get bigger. But I would tell you that the $39 million is sort of a just getting started number, and it was really amplified by the Consumer Electronics Show. I've been running this company now for just over 18 years, and I have never experienced the type of experience we had at the Consumer Electronics Show this year. It was everybody knew who we were, the move from Zeta who to why Zeta to Zeta now has really begun to take shape.

And when you look at a 600% return on investment, you're really seeing us as the AI disruptor in the marketing space. As I say, we are the disruptor, not the disruptee in this space. And I think that first $40 million against what I think will be over $100 million. Now of course, I've got to make sure I get Chris comfortable with that as we look at what we're going to spend on this year's Zeta Live, which might be a bit more than last year, we have some really cool stuff planned.

But the company is really re-architecting to make the vast majority of our annual time, energy and capital investment as it relates to events, not all, but the vast majority around the Consumer Electronics Show, the possible conference, Cannes Lion, and Zeta Live. And I think that if you look at our record pipeline in addition to the ways that is at this point a bit conservative, I think you're going to see us continue to win in the marketplace.

Exactly. What's interesting about that Zeta Live pipeline, with the $40 million that's already been closed, there's still another 200 opportunities, distinct opportunities tied just to Zeta Live valued at over $130 million in to go pipelines to get that extra $60 million from. So as David said, that's looking like a very achievable number.

Understood. And my one follow-up, Chris, is just around the gross margin numbers that we saw in Q4. Any onetime impacts here in the fourth quarter. And now with Marigold in place, how should we think about the right gross margin expectation through 2026?

Yes. Look, I even zoom out a little bit more, Zach. When you look at our 2028 model, each year, we should be getting between 100 and 300 basis points of gross margin improvement. Marigold is a tailwind to that. When you look at the fourth quarter's gross margin profile, you had our integrated platform revenue growing 25% as well as our direct revenue mix growing 25% year-over-year. We had very strong channel quarters in social and very strong channel quarter in connected TV, both of those below the overall corporate average, but good indications for multichannel, omnichannel usage.

Understood. Best of luck with the rest of the quarter.

Thanks, Zach.

Our next question comes from Jason Kreyer with Craig-Hallum Capital Group.

Congrats on 18%. I'm excited to hear about 19% next quarter. Wanted to just ask the evolution on One Zeta. I think you're integrating some of that sales functionality into Athena. So what does that mean for customers? Or what does that mean for the cross-sell experience when you combine One Zeta with Athena?

Let me start by saying thank you, Jason. Appreciate it. And our goal, as always, when we start a year is to finish the year 4 additional beat-and-raise quarters by the end of the year. So we're at 18%. Our hope is to be sitting here with you guys next year talking about 22%. The number of enterprise clients that use more than one use case in the fourth quarter of 2025 was up 80%. That's 80%. And that is a massive testament to One Zeta.

You're seeing the team headed by Ed See, and of course, Steve Gerber, that are really, really getting traction here. Now that's before the next level of uplift, which will come from Athena. So I think that as you see customers moving from one use case to multiple use cases around the One Zeta strategy to remind you, we see an average of a 200% to 300% revenue uplift from those clients from the past.

Now I don't know if every client is going to do that going forward but we do continue to see a meaningfully higher spend from customers who are moving from one use case to two. I think also one of the things I was most proud of last year. And this was not a cherry-picked quarter. For 2025, our net retention rate was 120%, and I was always happy at the 110% to 115%. So when you look at 120%, that means that One Zeta is really working, and our clients are scaling very, very quickly. Chris?

Yes. David, you said it perfectly. We had -- if you look at the total percentage of scaled customers, Jason, that are now on more than one use case, we're almost at 25% compared to, call it, 13% a year ago. So just really exciting progress made this year.

Awesome. Appreciate that. One follow-up for me on Marigold. So now that you got that closed, curious what the international expansion plan looks like and how you go after international from kind of a pull versus push mentality?

Yes. I mean what we're seeing now, Jason, is international is really happening very naturally, as our clients are mostly multinational enterprises. The adding of the Marigold international assets should accelerate that, but at the same time, I want to point out, the vast majority of our revenue is in the United States. This is the largest advertising market in the world by dollars and we are taking meaningful market share.

Last year, the average growth rate of the marketing ecosystem was about 10%, we grew 30%. So as we are disrupting the marketplace, we're taking greater market share in the United States, and we're seeing a natural progression and a very nice growth rate internationally, which is a reverse from prior years where we had struggled there. So I'm hopeful for international, but I would tell you that we certainly are not projecting it to be a massive component of this business for many years to come.

We'll go next to Arjun Bhatia with William Blair.

Congrats on a strong Q4 here. Maybe I'll continue on the kind of Athena line of questioning. David, you're clearly having success already with One Zeta, right? The rise in multi-use case customers is increasing. I'm curious, just as you think about '26, is it possible that Athena can have a meaningful impact of this year? Or does it take customers some time to kind of use it and ramp up on additional use cases? Like how are you thinking about the timing of when the sort of indirect revenue impact of Athena might come through?

I think it could be' '26. We're not counting on it, and we have not put that into our projections. And certainly, you don't know. But what I can tell you is if you look at the early access clients, they are spending materially more with Athena than they were without her.

They are telling us that they're seeing a game-changing workflow environment, which is great, and they are seeing a substantially higher return on investment, which at the end of the day, when you think of our biggest moats as a company, data is number one. Number two is our ability to drive superior return on investment, call it, right now, 600% return on ad spend. And then, of course, the ability to work with very large enterprises through their data security group, their data privacy groups, their legal groups, their procurement groups, so on and so forth. I think Athena helps us continue to drive greater return on investment both from a workflow management perspective and from a return on ad spend, which will cause our clients to drive even more of their existing marketing dollars to us.

All right. Perfect. That's helpful. And then Chris, one for you. In the back half of '25, you kind of ended the year on a strong note in terms of profitability GAAP profitability, GAAP net income and in Q4, you had kind of net income positive as well. How are you thinking about how that shapes up in 2026? And what does that sort of imply for stock comp outlook next year and beyond?

Yes. Thanks, Arjun. It was an area that we had just very constructive investor feedback that was important for the company to make strides towards that goal, turning the corner this year in 2025 towards positive and then seeing ourselves kind of this being now an inflection point. I think it from a GAAP EPS perspective, call it $0.02 to $0.04 in that range. There's upside to that and obviously depends on a number of factors.

But when you think about the ingredients that went into that inflection point and turning positive, it was progress on dilution, it was progress on stock-based compensation and then more and more yield off of adjusted EBITDA dropping to free cash flow. So we're really excited about that being a turning point, and it's onward and upward from here.

And we expect to be net income positive this calendar year forward.

Our next question comes from DJ Hynes with Canaccord Genuity.

Congrats on the excellent quarter. David, I want to ask about the OpenAI partnership. I think you quoted saying, this is going to be the most instrumental partnership we've ever embarked upon. As you look out over multiple years, like can you see the potential for how that relationship may evolve over time? Like what excites you the most about that opportunity?

Well, first of all, they're an incredible -- I'm sorry, I'm getting back feedback. Can you hear me, DJ?

I can hear you, yes.

Right. So what I would say is, first of all, OpenAI is an unbelievable organization. And when we think about the large language models, we think about them much like we would think about AWS or Snowflake, where they are all at some point going to be a part of our stack. In fact, we work with most all of them, if not all, already.

The partnership with OpenAI is different because it's foundational to Athena. We're also working with them on other components of their business where we are actively talking about doing things to help them with their business while we are talking about them doing more things to help us with our business.

So I was really referring to sort of the organizational partnership when I was talking about how I think this will be one of the most important partnerships we've ever made. And as we're looking at Athena, so once again, this whole narrative that large language models are going to disintermediate enterprise software, in our opinion, is silly. We think that large language models are going to be a component of what we do. And of course, we will pay them for their products and services. At the same time, they'll drive efficiency and accelerated revenue growth into our business, which will more than make up for that.

And when you think about what we're doing here, the whole goal is to get our clients to be able to more seamlessly use our platform, drive higher return on investment and make easier workflows for them around flying a 747, which is our platform versus the Cessna they currently know how to fly. OpenAI being foundational to Athena is helping us do that, and helping us do that in a very impactful and very meaningful way. Does that make sense, DJ?

It does make sense, yes. Maybe a follow-up on the data side. I'm curious, does the ability for Disqus to collect data or intent signals change in any way with the emergence of AI answer engines? I'm just feeling like people are landing less on owned media and now just reading more summaries. Have you seen any change in the volume of comments or authenticated site visits? And does that impact your ability to collect data at all?

So it's a great question. The answer is no. We have not seen any of that. In fact, because right now, the publishers, a lot of them have a challenge, right? If you look at Google, by way of example, it used to be that 90-plus percent of all queries were directed off to a publisher, a brand or an e-commerce platform. Today, according to them, greater than 60% of all queries are being answered on their platform. That's creating a massive tailwind for Zeta.

First, it is massively driving up the cost per click, which is driving advertisers to look for new methodologies for cost-efficient management of their marketing. Second, we're seeing that publishers are more anxious for traffic than ever before. And if you look at our e-mail platform around LiveIntent and other things we're doing, we are one of the biggest drivers of traffic and then you put Sailthru in, which was one of the assets we acquired through Marigold, which is one of the larger ESPs for publishers, our Publisher Cloud is driving massive volume of traffic back to publishers at a time when they need it, and we're helping to monetize them as a part of it.

So even though they're seeing less traffic today, DJ, from Google directly, we're actually helping them to drive incremental traffic. I'll also remind you that Disqus is one of, call it, 20 different platforms we now control that are generating real-time signals on a day-by-day, moment-by-moment basis. To date, we have not seen any fall off there.

Our next question comes from Richard Baldry with ROTH Capital Partners.

It looks like quota carriers are up about 10% sequentially. I assume some of that has to do with the Marigold acquisition. Can you talk about any cross-training efforts that are needed and how we should think about more additions to the quota carriers throughout the year ahead?

Yes, Rich, we added, call it, 15-ish in that range from Marigold in terms of quota carriers sequentially. So you're right, that was a driver there. I'll let David take the next one.

Yes. What we're doing, Rich, is we're sort of teaming their salespeople up with ours. So rather than taking all of the time to train them on everything we're doing, for their clients, we've sort of segmented it to the top 30, which are, I'm pretty sure, all Fortune 500 companies. But the reality is that we're going in and we're seeing really interesting cross-sell opportunities already because of that.

Got it. And lastly maybe sort of will be gentle and call it an unusual environment for software valuations. So I'm sort of curious, your own internal preferences to allocate capital between either M&A because you've done some meaningful strategic M&A versus buybacks over sort of the near term, intermediate term?

Yes. So let me start by saying we are buying back stock very rapidly. I think we have $130 million left, give or take, on our existing buyback. When we get through that one, we will most likely announce the next one. Every time we've announced a buyback, it has been 100% greater than the prior buyback. So I think right now, the single best use of our capital is to buy back our shares.

Now that being said, we've been around for 18 years, I just did a CNBC interview earlier. And it's sort of -- we figured out we've been around for 18 years. We've done 18 acquisitions and we beat and raised 18 quarters in a row. So this is sort of the 18 cubed results as it relates to that. It is highly probable we'll do a 19th acquisition. And this is obviously a good environment to buy what I think are high-quality assets at a lower price than you would have had to pay.

We have a very, very solid balance sheet. We have a lot of cash. We are generating meaningful free cash flow at almost a 60% conversion rate of cash to EBITDA. And I think that will continue to go up as we've talked about. And we have de minimis debt. Our debt ratio was below 0. I don't want to come below 0, but it is 0 at this point. So I think we have the opportunity to do something, although there's nothing on the radar right now. Right now, we're operating our business. We've obviously given as reported guidance to a 40% growth rate in the first quarter, a 35% growth rate for the year and feeling like we're in a very good place.

Rich, we've really stepped up the repurchase program in '25. And as David said, especially at these levels, I would expect it to sustain. As a percentage of free cash flow, we did 45% repurchase to free cash flow ratio in 2024. We did 73% in 2025. So our model has always been at least 50% of the free cash flow we generate. But as David said, we're going to be aggressive in this environment.

And Rich, I don't know if you've been able to dig out of the snow but just to show the nimbleness of our organization, I woke up in Los Angeles yesterday with a 50% probability of going to New York to do our earnings, a 50% probability of the team coming to L.A. to do earnings, and we are now all in Miami doing our earnings because it was the only place we could all get to. So we tend to be on the more nimble side as it relates to this stuff.

Got it. Maybe one last one for me. The net retention number of $120 million was a pretty strong number. To keep it a little bit apart, was it more driven by volume usage or the -- more by the number of use cases sort of climbing? And a little struck that it's that strong ahead of Athena being GA because it seems like that lowers the friction to usage. So is there anything we should think about on that number and it's sort of onetime orientedness or sustainability or extensibility?

No, I mean, if you look at -- you can go into each of the Qs and get a pretty good indication throughout the year of the net revenue retention rate, you can get pretty close to it, not exact because that's an annual metric. But you'll see that it was really kind of a -- we're going in with a running start into 2026. It was a very strong end of the year.

A few dynamics to think about. First, we benefited from a lot of new customers that were added late in 2024 that very nicely scaled throughout 2025. What's also, I think, evident in the results is not just the addition of use cases and channels, Rich, but growth within brands in the agency ecosystem. If you look at brands within the holdcos on a year-over-year basis, they're up 80%. That's a big change that drives growth within just, call it, 1 scaled customer, but can have some pretty dramatic ARPU expansion outcomes.

Yes. I don't see it as an anomaly, Rich. I mean we're not going to guide to that. But the reality is that what we're seeing is we're seeing, bringing Ed See last year and building the systems around One Zeta has been game-changing for moving from one use case to multiple. And as we've mentioned before, when a client moves from 1 use case to 2 use cases, their spend goes up materially as evidenced by the 80% growth in multi-use case customers for Q4 over Q4 of the prior year. So I think that Athena will help us with that number, and I think we're going to continue to run hot there.

Moving next to Elizabeth Porter with Morgan Stanley.

Awesome. This is Katie Keyser on for Elizabeth. I just had a quick one, hoping for an update on the political and advocacy side of the business, really just in the context of what you've seen during previous elections. You guys clearly expanded the size and scale of the platform since prior midterm election cycle. So wondering what you're seeing early in '26 as it relates to advocacy.

And then maybe just for Chris, if you could comment on how that spend is expected to phase through your guidance, kind of what scenario would cause that to prove conservative. Any thoughts there would be great.

Sure, I think, Katie, and give our best to Elizabeth. From 2020 to 2022, the contribution in political candidate revenue halved. It was, call it, $20 million in 2020 -- sorry, $15 million in 2020, and it was $7.5 million in 2022. In 2024, total political candidate revenue was right around $40 million. So starting this year's guide at $15 million is probably conservative. What we've said in terms of the cadence of that spend that we expect to realize is, call it, $7 million in the third quarter and $8 million in the fourth quarter.

Advocacy is an always on industry for us now. It's obviously bigger in political candidate cycles, but we've actually made some really exciting new hires in 2025 into that area in advance of building momentum for 2026. So I expect that to continue to be an always on industry for us. It will benefit in the candidate year, but it's not driving any outsized contribution in our current guide right now, though.

Yes. And it could be upside, Katie.

Moving next to Matt Swanson with RBC.

Great. A really impressive number when you're talking about the 600% ROI from the Forrester study. Can you just talk a little bit about kind of the macro resiliency that, that gives you a business just as we're dealing with all the headlines around tariffs and everything else?

So thank you, Matt. We actually think it doesn't just insulate us from the macro environment. We think it accelerates in this macro environment. What we see is, as clients are dealing with different sort of variables in their businesses, they want to maximize their return on investment in every component of their business, with marketing being one of their largest expenses as organizations. So I think it's helped us over the last year, and I think it will help us over the next few years.

And then if I could just double-click on one aspect of Marigold, I know Loyalty was something that we talked a lot about when the acquisition was first announced. I think in the prepared remarks, you mentioned that you're seeing some early interest from customers. I was wondering if you could just expand on that a bit.

Yes. So we're -- first of all, we're very excited to take their customers and move them into the One Zeta solution. To remind you, when we bought Marigold, they were effectively 100% retention business. They didn't have anything around new customer acquisition and their monetization was very limited. So we're very excited to take their customers and move them into multiple use cases, particularly around customer acquisition. We're also seeing real interest from a large number of our existing clients to use our new Loyalty solution, and we're starting to see meaningful traction there.

We're also doing something really interesting, which is in sort of the next release, not the first release, but we'll probably be doing an Athena release a quarter over the next few years with it being generally available as we expect it to be this quarter. We're very on track for that. But we're going to be integrating Athena into the Loyalty program as well. So I'm very excited about that and its ability to seamlessly integrate them into all of our existing 603 scaled customers.

And we'll go next to Scott Berg with Needham & Company.

This is [ Lucas Metcalf ] on for Scott Berg. Just in terms of 2026 marketing budgets, what are you guys hearing from customers about overall kind of spend growth and how AI is shaping kind of their allocation decisions?

And then I guess, specifically, are you seeing AI driving incremental budget expansion, more platform consolidation? Or are you just kind of seeing a reallocation within existing spend?

Well, it's a lot to answer in one fell swoop. What I would start with is, yes, we're seeing platform consolidation. We're starting to see that our entire strategy around disintermediating point solutions into one platform is really taking hold. And Lucas, I think our 120% net retention rate last year really evidences that.

What you're also seeing is you're seeing marketing budgets from our vantage point in the United States in their entirety going up into 2026. We think we'll see marketing budgets up and to the right. And we think we will continue to take a multiple of that growth in our growth rate as a company.

Moving on to Jackson Ader with KeyBanc Capital Markets.

This is Jack Nichols on for Jackson Ader. With RFP volumes more than doubling, who are you mainly seeing in deals from a competitive standpoint? And can you talk about Zeta's right to win there? And kind of set it in a way, who are you taking budget dollars from? And then I've got a follow-up.

Yes. I mean we continue to see a whole host of point solutions depending on how the RFP is drafted, and we moved the narrative from individual point solutions to our platform solution, which is how we're winning. What I would tell you is we continue to see Salesforce, Adobe, to a less extent, Oracle and a few other smaller players in the CRM RFPs. And then in the acquisition RFPs, we're seeing companies like the Trade Desk continue to be challenged as we continue to take market share from them and others continue to take market share from them. I'm sure they'll have their viewpoint on that at some point.

But what I would tell you is that we see the different point solutions in these RFPs. And what I would tell you is we continue to win greater than 50% of the engagements and RFPs we get invited to participate in. And what you're seeing in the numbers, which obviously are growing faster than we even expected and we expected to grow fast is we're seeing far more RFPs than we've ever seen before.

So we're seeing more RFPs and we're still winning greater than half. So you're starting to see that ripple into the numbers. I'm pretty sure this will be our third consecutive 30% growth year based on our existing projections. You're talking about 3 years in a row of 30-plus percent growth and a 4- or 5-year 30% compounded growth rate, you're seeing that come in through the growth of the RFPs.

That's helpful to understand. And then maybe as a follow-up, how specifically are agencies thinking about their AI marketing strategies, maybe splitting the difference between Athena and then third-party LLMs, how are the marketing agencies kind of going about the testing and then the use case testing and deployments there?

The LLMs have no capabilities around activating marketing. I mean to be very clear. Today, the only thing the LLMs can do, it can help more efficiently code, which, by the way, our entire engineering staff is using every tool available around efficiency of coding, whether it's Anthropic or it's OpenAI or other products in Guild and we continue to use all of them, which is one of the ways we're staying ahead of our competitors as it relates to our engineering and innovation capability.

So we're not seeing any of the agencies at this point looking to move activation to an LLM because they have no activation capabilities. Even the LLMs that are driving -- rolling out their own marketing are not doing it themselves. So I think what you're going to see is you're going to see the agencies continue to consolidate platforms.

As Chris said, I think we grew our brands with the agencies we work with by 80% last year over the year before. That's not an accident, right? They're not looking to split that up with LLMs. Now I have had conversations with the heads of all the agency holdcos as of late, and everybody is looking to figure out how do they use large language models to be more efficient inside their businesses just like we're doing. And I think they're going to continue to look to do that.

But from an activation perspective, our data which is one of the biggest moats in our business, which is never fed into large language models, it stays behind our cloud, is invaluable to our enterprise and agency clients as it relates to our models being able to train on our data exclusively and the ability to better target and create better return on investment. So I'm not seeing any of that being split up at this point.

We'll go next to Koji Ikeda with Bank of America.

This is George McGreehan on for Koji Ikeda. I wanted to ask on the strength you're seeing in terms of Fortune 100 and Fortune 500, these large enterprises and the new customers coming online from those areas. What are conversations like with these large enterprise customers that presumably have a lot more room to grow with the platform over time? What are they excited about when it comes to the Zeta platform?

So thank you, George. Great question. First of all, as we announced, we are now working with 51% of the Fortune 100 and 24% of the Fortune 500. But you're right, we have a very small wallet share there with a massive headroom to grow.

Everybody that we're talking to in the Fortune 500 is focused on return on investment. How do they get the highest possible return on their marketing spend to drive efficiency into their business because for most Fortune 500 customers, especially the ones that are consumer-facing, I would say that their second or third largest global expense is marketing. And as they think about that, they're thinking about how do they invest that in a way that they can get the highest possible return.

I would tell you, even 5 years ago, Fortune 500 companies were focused not on return on spend. They were focused on the percentage of their revenue that sales and marketing represented. So if they were spending $100 -- I'm sorry, if they were generating $100 in revenue, how do they spend $20 or $22 in marketing?

As it relates to today, it's how do we spend $1 and return 500%, 700% or 1,000% against that dollar? And the attribution capabilities compiled with our proprietary data, compiled with our activation capabilities are creating a true return on investment analysis that has not been available to Fortune 500 companies in the past.

And we'll go on to our next question, Callie Valenti with Goldman Sachs.

Just one for me. I wanted to ask -- I know part of your strategy over time has been acquiring assets to add to your Data Cloud. Curious, as consumers interact more with LLMs and that ecosystem continues to mature, how do you think about the potential to capture data from these new channels through M&A or other means?

It's a great question. And yes, we, Callie, have always looked at how do we build on our Data Cloud. What I would tell you is, you hate to plant a flag and say you're done, but I would tell you that today, our Data Cloud is really the best data cloud in the world. We've got better access to information than at any point ever. We're ingesting trillions of signals. And by the way, we've rolled out our own generative optimization platform, the GEO platform where we are helping our clients to better get their information and their businesses fed into the large language models. It's something that allows us to learn a tremendous amount about how the LLMs are acting by building that GEO product.

So in many cases, we bought businesses, in many cases, we've built different platforms. In this case, we see GEO as strategic, not just helping our clients get to the next generation of marketing, which we're doing, but also the ability to learn how those models are thinking, how they're ingesting information and how do we get information back from them.

This now concludes our question-and-answer session. I would like to turn the floor back over to David Steinberg for closing comments.

I will close simply by saying I have never been prouder of running this company. We have the right people at the right time with the right technology, not just to win today, but to win for many, many years to come. We're incredibly excited about the innovations around artificial intelligence and how they are willing to fit into our platform and how we are going to be the disruptor of marketing over the next generation, not the disruptee. I hope everybody has an incredible day and an incredible week. Thank you for listening.

Ladies and gentlemen, thank you for your participation. This does conclude today's teleconference. You may disconnect your lines, and have a wonderful day.

Greetings, and welcome to the Zeta 3Q '25 Earnings Conference Call. [Operator Instructions] As a reminder, this conference is being recorded. I would now like to turn the conference over to your host, Matt Pfau, SVP of Investor Relations. You may begin.

Thank you, operator. Hello, everyone, and thank you for joining us for Zeta's Third Quarter 2025 Conference Call. Today's presentation and earnings release are available on Zeta's Investor Relations website at investors.zetaglobal.com, where you will also find links to our SEC filings, along with other information about Zeta. Joining me on the call today are David Steinberg, Zeta's Co-Founder, Chairman and Chief Executive Officer; and Chris Greiner, Zeta's Chief Financial Officer.

Before we begin, I'd like to remind everyone that statements made on this call as well as in the presentation and earnings release contain forward-looking statements regarding our financial outlook, business plans and objectives and other future events and developments, including statements about the market potential of our products, potential competition, revenues of our products and our goals and strategies. These statements are subject to risks and uncertainties that may cause actual results to differ materially from those projected.

These risks and uncertainties include those described in the company's earnings release and other filings with the SEC and speak only as of today's date. In addition, our discussion today will include references to certain supplemental non-GAAP financial measures, which should be considered in addition to and not as a substitute for our GAAP results. We use these non-GAAP measures in managing our business and believe they provide useful information for our investors. Reconciliations of the non-GAAP measures to the corresponding GAAP measures, where appropriate, can be found in the earnings presentation available on our website as well as our earnings release and our other filings with the SEC.

With that, I will now turn the call over to David.

Thank you, Matt. Good afternoon, everyone, and thank you for joining us today. For the 17th quarter in a row, we once again delivered a beat and raise quarter, driven by our leadership in AI-powered marketing. In Q3, revenue was $337 million, up 28% year-over-year ex political and LiveIntent. This is an acceleration in growth from Q2. Adjusted EBITDA was $78 million, up 46% year-over-year. And free cash flow was $47 million, up 83% year-over-year, representing a margin of 14%. This is the highest free cash flow margin we have ever achieved, and we did it while accelerating our revenue growth ex political and LiveIntent.

This demonstrates our focus on driving growth and improved profitability while investing to extend our AI leadership. Based on our year-to-date momentum and our record pipeline exiting Zeta Live, we are raising our 2025 revenue guidance by $11 million at the midpoint and providing an initial 2026 outlook well ahead of consensus. In October, we hosted our fifth annual Zeta Live, our most successful event yet with max capacity audience. Attendance was up 35%. And unfortunately, we had to turn away a large number of people.

For next year, we are currently looking for a bigger venue. We had strong executive participation, representing more than $100 billion in annual marketing spend decision-makers in attendance. At Zeta Live, we hosted 38 sessions, featured 95 speakers, delivered nearly 120 product demos and shockingly served 7,600 coffees to the 1,500 attendees. Customer feedback was overwhelmingly positive with many attendees commenting that it was one of the best events they have ever attended. Our goal is to close over $100 million in incremental business coming out of Zeta Live. At Zeta Live, we launched Athena, our next step in leading the AI revolution in marketing.

Athena is our AI conversational, superintelligent agent that becomes the intelligent operating system for our clients' businesses and ultimately for their lives. Athena marks a major breakthrough in human AI collaboration, removing the friction between the human and artificial intelligence and acting as a real-time voice-activated command center for the Zeta marketing platform. One workspace to plan, execute, analyze and optimize in plain-spoken English. AI is the new UI and Athena proves it. You speak your objectives, it builds the experience, launches the program and delivers insights on one screen, making it easier to adopt more channels, more use cases and ultimately drive greater ROI and spend on Zeta's platform.

The business impact of using Zeta is already proven. Independent analysis shows Zeta stands up 50% faster than our peers and delivers a 6:1 return on investment. And with Athena making the platform radically easier and more intuitive, we believe these numbers will only get stronger. This is how AI becomes a habit, not a headline. Adoption scales, usage deepens and the platform becomes harder to leave. We believe Zeta is uniquely positioned to develop Athena for 3 key reasons: First, data advantage. Our Zeta Data Cloud gives Athena rich real-time signals for personalization and prediction that generic AI just cannot match.

Second, native integration. Athena is embedded in the Zeta Marketing Platform, connecting answers to actions and insights to activation and measurement. Unlike our competitors who have legacy tech debt and cannot natively embed AI into their marketing clouds. Third, AI track record. AI is foundational to what we do at Zeta. We have natively built AI into our platform since 2017, and we believe this AI leadership is the reason we are winning today. At Zeta, artificial intelligence is foundational to our business, not an add-on like our competitors are building. The reception from customers, prospects and partners has been incredible.

If you did not attend the standing room-only Athena demo at Zeta Live, I encourage you to watch it in the Investor Relations section of Zeta's website. It is truly game changing. Athena is in internal beta now, and we expect it to be in client beta by the end of Q4 and fully production ready by the end of the first quarter of 2026. OneZeta is a key growth engine for us and a growth flywheel for our customers as they consolidate more data and use cases into the Zeta Marketing Platform. Customers who adopt 2 or more use cases generate greater than 3x the annual revenue of a single use case customer, have our highest NPS score and see the highest return on marketing spend.

We have accelerated this focus under Ed See, our Chief Growth Officer, and the results are meaningful. For example, a leading big box retailer consolidated acquire, grow and retain onto the ZMP at the start of our relationship. That agreement was greater than 3.5x the size of a standard deal with similarly sized enterprises. Second, one of the fastest-growing specialty retailers started with the retain use case and recently added a grow use case. As a result, the customer is expected to spend over $2 million with us in the fourth quarter compared to only $700,000 in the first quarter, a threefold increase, and we expect this customer to continue to grow into next year.

We expect the strong pipeline generated from Zeta Live, combined with the acquisition of Marigold’s enterprise software business, which includes over 100 enterprises, 20 of the top 100 advertisers in North America and more than 40 Fortune 500 companies who are all currently only on one use case, will further accelerate our OneZeta momentum into 2026 and beyond. I'll close with an update on our agreement to acquire Marigold's enterprise software business. We remain on track to complete the acquisition by the end of the year.

After spending the past several weeks getting to know the business and the team better during the closed process, I am even more enthusiastic about the opportunity this acquisition represents. Strategically, the data that we pick up, including the loyalty program, will be game-changing to training our proprietary algorithms, including our Athena, answers and generative engine optimization products and widens our data cloud moat. We hosted several of their employees and customers at Zeta Live and the 1 plus 1 equals 4 opportunity is more clear than ever.

Great numbers are the output. The inputs are great execution, products and people. The reason we are growing at an accelerated pace is simple. We have superior artificial intelligence and data to our competitors that we started developing 8 years ago, not 8 months ago. Our consistent execution over the past 17 quarters as a public company is a credit to our artificial intelligence, our people and our data. Quarter after quarter, we have raised the bar. As always, I would like to sincerely thank our customers, our partners, team Zeta and all of our shareholders for the ongoing support of our vision.

Now let me turn it over to Chris to discuss our results in greater detail. Chris?

Thank you, David, and good afternoon, everyone. The theme of our Investor Day in October was durable, predictable and profitable growth, and our third quarter results are a perfect illustration of this. Our revenue growth, excluding political and LiveIntent, accelerated to 28% in Q3 from 27% in Q2 and 26% in Q1, showing the durability of our growth. Our third quarter results once again exceeded our guidance for both revenue and adjusted EBITDA, highlighting the predictability of our growth.

And we achieved the highest free cash flow margin in our history, achieving the Rule of 40 on a free cash flow margin basis, demonstrating the profitability of our growth. With that, let's dive into the details of the quarter. In Q3, we delivered revenue of $337 million, up 26% year-over-year or 28% when excluding the contribution from LiveIntent and political candidate revenue in the year ago period. We exceeded the midpoint of our guidance by $9 million or 3 percentage points higher than our forecast, solidly within the 2 to 5 points of cushion we typically leave ourselves.

Total scaled customer count grew to 572, up 20% year-over-year and an addition of 5 customers sequentially. We ended the quarter with 180 super-scaled customers, up 25% year-over-year and an addition of 12 customers sequentially. The strong super-scaled customer additions were broad-based across industry verticals and driven by cross-sell of LiveIntent customers and our OneZeta initiative. Scaled customer quarterly ARPU of $579,000 increased 4% year-over-year or 13% when adjusting for political candidate revenue.

Super-scaled customer quarterly ARPU of $1.6 million was up 1% year-over-year or 12%, excluding prior year political candidate revenue. From an industry perspective, 7 of our top 10 verticals in the quarter grew faster than 20% year-over-year on a trailing 12-month basis, an improvement from 6 in the first and second quarters. The additional 20% plus growth industry was telecom, where we have had multiple significant wins over the past year and has been a focus of ours this year. It's also worth noting growth in consumer discretionary verticals like retail, travel and hospitality and automotive remained strong in the quarter.

And finally, all 3 of the non-20% growth industries were up year-over-year between 7% and 17%. Our direct mix in the third quarter was 75%, consistent with the second quarter and an increase from 70% in the year ago quarter, demonstrating continued success of our agency direct-to-channel adoption. Our GAAP cost of revenue in the quarter was 39.5%, a 13 basis point increase year-over-year and 160 basis points sequentially. The increase in cost of revenue was driven by strong sequential and year-over-year growth in display and video, channels where we continue to see customers investing and which remain highly effective channels for customer acquisition and growth.

In the third quarter, we generated $78.1 million of adjusted EBITDA at a margin of 23.2%, 320 basis points higher year-over-year and $7.4 million better than the midpoint of our guidance. This marks the 19th quarter of expanding adjusted EBITDA margins year-over-year. Our GAAP net loss for the third quarter was $3.6 million, an improvement from a loss of $17.4 million in the third quarter of 2024. There were $6.5 million of acquisition-related expenses in the third quarter, which absent these costs, we would have been GAAP profitable.

Third quarter net cash provided by operating activities was $57.9 million, up 68% year-over-year with free cash flow of $47.1 million, up 83% year-over-year and representing a margin of 14%. This represents a free cash flow conversion of 60%, a significant improvement from 48% in the third quarter of 2024 and 57% in the second quarter of 2025. This also includes a roughly 18-point working capital headwind driven by longer agency payment cycles. The improvement in both adjusted EBITDA margin and free cash flow conversion in the third quarter exhibit the strong operating leverage of our model and put us firmly on track to achieve our Investor Day targets of a 30% plus adjusted EBITDA margin and greater than 70% free cash flow conversion in 2030.

During the third quarter, we repurchased 1.7 million shares for $28 million and have repurchased 6 million shares for $85 million year-to-date. We also continue to make significant progress in reducing dilution and stock-based compensation expense. Just like the second quarter, in the third quarter, we had 0 net dilution and year-to-date, our dilution is 1.6% as of September 30. We remain on track to achieve both our 4% to 6% normal course dilution target in 2025 and our $190 million equity compensation expense target even when factoring in the equity we anticipate issuing for the Marigold acquisition.

Before diving into our updated fourth quarter and 2025 guidance, I want to reiterate what I said on the recent Marigold acquisition call regarding forward-looking estimates. Given there could be variability in the close date, which we continue to anticipate will happen in the fourth quarter of 2025, we have not included any Marigold-related contributions in our 2025 guidance and continue to guide analysts to not yet include Marigold in their 2025 or 2026 estimates until the transaction closes. Upon closing, we will provide guidance on Marigold's contribution for 2025 and 2026.

We want to make sure that consensus estimates do not become a mixed bag of organic and acquisition-related revenue. As we have done with prior acquisitions, we plan to clearly break out the organic versus acquired revenue for the first year post transaction upon announcement of closing. With that in mind, we're raising fourth quarter and full year revenue, adjusted EBITDA and free cash flow guidance. Details can be found starting on Slide 16 of our earnings supplemental. For the full year 2025, we are increasing the midpoint of our revenue guidance by $11 million to $1.275 billion, representing a 26% year-over-year growth when excluding political and LiveIntent, which is 5 points higher than our starting point for the year at 21%.

For the fourth quarter, we now expect revenue of $364.5 million at the midpoint, $2 million higher than our previous guidance and representing year-over-year growth of 16% or 23% when excluding political candidate and LiveIntent revenue, which is consistent with our Zeta 2028 plan to grow 20% or greater organically. For adjusted EBITDA, we're increasing the midpoint of our 2025 guidance to $273.7 million, up $9 million from our prior guidance and representing a year-over-year increase of 42% at a margin of 21.5%, an improvement of 230 basis points over 2024.

For the fourth quarter of 2025, we now expect adjusted EBITDA of $90 million at the midpoint, up from our previous expectation of $88.4 million and representing growth of 28% and a margin of 24.7%. We are also increasing the midpoint of our 2025 free cash flow guidance to $157.4 million, up $15 million from the midpoint of our previous guidance and representing year-over-year growth of 71% and a conversion of 57% of adjusted EBITDA. This conversion is up 10 points from 2024. For context, since our initial 2025 guidance back in February, we have increased revenue by $35 million and free cash flow by $28 million.

Now let me transition to looking beyond 2025 and setting our initial organic guidance for 2026. While providing out-year guidance in Q3 is not intended to become standard practice, we're doing so this year because we want to establish a clean organic baseline for 2026 before Marigold is incorporated into our guidance. At our recent Investor Day, we spent time demonstrating our track record of 5 straight years of at least 20% revenue growth and free cash flow margin expansion. The punchline is, we see another year ahead of us with each at substantially higher scale.

In terms of revenue, we're guiding 2026 to be $1.54 billion or 21% growth on 2025 guidance of $1.275 billion. Importantly, this is an organic-only view and does not include Marigold. Our guidance assumes $15 million of political candidate revenue, which we would expect to evolve and is 2x what the midterms were in 2022. We expect $354 million of adjusted EBITDA in 2026 or a 23% margin, up 150 basis points year-to-year. We see an initial view of 59% free cash flow conversion, yielding $209 million in free cash flow at a margin of 14%.

And lastly, we continue to plan for a guidance model with 2% to 5% top line buffer. From a seasonality perspective, we expect revenue in the first quarter to be $314.5 million, up 19% year-to-year and accounting for roughly the same percentage of full year revenue in the first quarter as 2025 and with an adjusted EBITDA margin of 17.8%. I'll conclude where I began. Our third quarter results, updated 2025 guidance and initial outlook for 2026 underscore the durability, predictability and profitability of our growth and reflect the confidence and momentum we have in the business.

Now let me hand the call back over to the operator for David and myself to take your questions. Operator?

[Operator Instructions] Our first question comes from the line of Arjun Bhatia with William Blair & Company.

Perfect. Congrats on another great quarter, guys, very nicely done. One question for you, David. Just at a high level, I'm curious like as you look at the next kind of pocket of spend that's out there, I imagine a lot of it is going to come from media budgets. How do you think kind of Zeta's ROI or ROAS compares to that of the walled gardens? Are you hearing more customers come to you and say, you're getting to parity? Are you already there? Because -- how do you go after that kind of large pool of budget dollars on spend?

Thank you, Arjun. I'm obviously very proud of the team for the quarter. Let me start by saying that if you look at our collective clients, you've got 572 clients that spend over $100 billion a year in marketing today. If you look at the global marketing spend, it's about 75% media and about 25% CRMS revenues. When we look at the breakdown of those 2, we continue to be substantially more profitable than anybody else in the media space. In fact, a number of our clients use our platform, which allows our data to inform the way they market into the walled gardens today, and we already partner with most all of the walled gardens.

So when you think about it, every dollar that is spent through the Zeta Marketing platform returns a 6x return in revenue to our clients. That puts us on par with pretty much any walled garden. And what we're starting to see in a post-Gemini and post-OpenAI world, marketers are looking for new methodologies to create customers and maintain the customers they have. It's been challenging as the vast majority of the questions that are now asked on the Google platform and certainly all that are asked on the OpenAI platform, ChatGPT, are answered on platform.

So what used to be sent to our clients or to other publishers is now being consumed there. So clients are very much looking for new methodologies. We continue to be at the forefront of that. And I think that was a big part of what you saw us talk about at Zeta Live, and I think it's a big part of what you're seeing in the tailwind that we're currently experiencing.

All right. Perfect. Very helpful. And then maybe one just on customer count and me nitpicking a little bit here, but it seems like in Q3, I think you added 5 scaled customers. And if I look at just the last 2 quarters, it's been hovering around 20. Is there anything sort of one-off that we need to consider there, just maybe agencies and us not having brand count visibility? And just how should we think about that metric sort of evolving over the next couple of quarters?

Yes, I'm glad you asked, Arjun. Three data points for you. First, on a year-over-year basis, that 20% growth in total scaled customer count is, a, consistent with where we've been this year and obviously north of our model of 4% to 5%. But to your point around sequentially, we had a really strong progression from scaled to super-scaled customers. We increased by 12% sequentially, which is, if you look back at our history, certainly at the higher end.

And then to your point around agencies, what's also being masked in the total-scaled customer count is we were up 23 brands quarter-to-quarter. So as we continue to have success scaling within and adding new agencies, that doesn't necessarily add to scaled customer count. But in this case, it certainly translates to higher brand expansion. So we're really happy with the scaled customer count expansion this quarter.

Our next question comes from the line of Elizabeth Porter with Morgan Stanley.

I wanted to follow up, David, on your target for the $100 million incremental business after Zeta Live. So could you provide some context around how that target compares to what you've achieved in prior years? And how you'd characterize the momentum coming out of Zeta Live relative to prior years? And then just lastly, how should we think about the conversion time line from that event into bookings and revenue? And is that already included in fiscal '26 guidance?

Thank you, Elizabeth. So first of all, last year, we closed about $57 million in business coming out of Zeta Live, which I think is one of the reasons we were able to grow 500 basis points greater through the year than we gave original guidance for thus far. This year, we're trying to get that to $100 million. I would tell you that as of today, the pipeline is already full enough to get us to that number, and we might see even more meaningful numbers than that. So as you know, we try to be consistent with our beat and raise methodology.

We don't usually give guidance for the next year at the end of the third quarter. We felt because of the Marigold deal, it was important to make it clear that we were going to grow organically 20% again into next year from a top line perspective. I do think that Zeta Live will be a part of that, but I also think it's a part of the buffer that we continue to leave ourselves where we think we can continue to do better than that.

To answer the momentum question, I would tell you my biggest challenge today, Elizabeth, is how do I top Zeta Live next year. It was just so incredibly well received. We had multiple clients just going crazy for Athena, and I just continue to see the most momentum I've ever seen coming out of any event we have ever hosted ever. It's a very exciting time for Zeta right now.

Great. And then just as a quick follow-up. On the LiveIntent piece, understanding it's a smaller portion of the business, but it looks like it was a little bit weaker in Q3 and in Q4 guidance. So wondering if there's any trends to kind of call out there on how that integration is going?

Yes. Thanks, Elizabeth. I'll take that. Look, in short, we're really happy with how LiveIntent is performing. We are accruing at 100% of their earn-out. I think if we were to do things over again, remember, we set that initial guidance skew for every quarter of 2025, about a month after we owned the asset. We've learned a lot since. I think we're also realizing a substantial amount of new product synergies that is flowing organically to the business. But overall, I couldn't be happier with how the integration has gone and the execution of the team.

And what you're not seeing in the numbers is all of the cross-sales from Zeta clients into their products and their products into ours. So we're -- I would tell you, Elizabeth, we think the LiveIntent deal was a grand slam home run, and we're looking forward to having a similar experience with Marigold.

Our next question comes from the line of Matt Swanson with RBC.

David, we've seen the compounding value from an ARPU perspective when customers take multiple products. You talked about Athena and AI being the new UI. Can you just talk a little bit about how maybe that will help with the cross-sell and for people to see the full value of your platform?

Thank you, Matt. I would tell you that I think Athena will be amongst the single biggest drivers to the OneZeta methodology we've ever had. To point out, it is, of course, fully voice activated and fully conversational. So any client by the end of the first quarter should have this fully embedded into their platform. Not only will they be able to work through the existing use case they have, Athena will already be programmed to help them with secondary use case expansion from production level launch.

By way of example, if a client is using us actively for acquisition, they'll literally be able to say to Athena, "Tell me which of my clients are most likely to churn and how do you recommend that we save them." That removal of the friction between the human and the platform is going to allow, I think, for an accelerated growth of OneZeta.

One point where that Matt shows up also in the metrics is one of the areas that as we were preparing post close that stood out to me is not only did each of our use cases grow double digits year-over-year, which has been consistent, but our 2 or more use case count grew over 100% year-over-year. So we're going into capabilities like Athena with really interesting go-to-market momentum.

Yes. And once again, with 100 global enterprises in Marigold that are all on one use case, we see that as a very unique opportunity for the specialty team we've set up under Ed See to get in there and focus on it. But I think that the integration and production capabilities of Athena are going to be really an accelerant to platform utilization.

That's really helpful. And then I'm sure the pipeline has a big -- a lot to do with this. But David, you mentioned that the '26 guide was a quarter early. So Chris, could you just kind of talk us through being one of the first people to give us an outlook into 2026? Just what's kind of building all the confidence and the visibility that you have?

Yes. We talked about at Investor Day, Matt, you'll recall that so much of our installed base not only has been with us 3-plus years, almost 60%, 5 or more years, but there's now a demonstrated history of there, on average, expanding their spend with us 15 points a year. So by starting our guide at 21% year-over-year, which, by the way, is consistent with really the last 2 years of where we started our guidance, we feel like we've left even some conservatism there. The pipeline says a lot.

And then with more and more of our revenue under the curve from really strong signings throughout the year, it just -- it felt like today was the right time. Again, wanted to be clean with our organic baseline before we added Marigold. And I wouldn't want to overlook the really strong addition, not just to revenue, but how much of that revenue for 2026 out of the gate is flowing towards increasing adjusted EBITDA and free cash flow. In fact, this year, from our initial guide of revenue, which was around 21%, so we've obviously added to that since, but that translates to $35 million of revenue and $28 million of free cash flow that's been added to the guide since the beginning of the year as well.

Our next question comes from the line of DJ Hynes with Canaccord Genuity.

Congrats on a really nice quarter here. David, Chris, I want to ask how you're thinking about sales and marketing investment and capacity build-out there. I mean, obviously, the operating leverage is great to see, but sales and marketing actually went down sequentially. And you shared some really powerful pipeline statistics around the growth there and the strength coming out of Zeta Live. So I'm wondering how you're thinking about whether it's time to lean in on growth. I mean, obviously, I guess you're going to get some sales reps with Marigold. But any thoughts there, maybe kind of plan for 2026 would be helpful.

Yes. I mean it went down a little bit by accident. We didn't mean it to. Quite frankly, we're trying to hire every great salesperson we can possibly get, DJ. The challenge is that our existing sales reps have been so incredibly productive that we've been able to continue to grow at a much faster pace than we expected to while simultaneously trying to find more salespeople. We will pick up an incredible sales team with Marigold, both in the United States and Europe and EMEA.

So it feels like we're sort of very well positioned for where we want to go, although -- once again, we did grow 28% top line, 43% EBITDA line -- I'm sorry, 46% don't want to screw us out of that last 3%, but 46% growth on EBITDA and 83% free cash flow growth, while simultaneously, DJ, making some of the largest investments we have ever made into our artificial intelligence and innovation. So feeling very good about where we are organizationally.

You'll see a pretty material tick up 3Q to 4Q with Zeta Live occurring in October versus September like last year.

Yes. Yes. Okay. Makes sense. And then, Chris, while I have you, maybe can you talk about gross margins in the quarter? I mean, direct revenue mix held pretty stable at that 75%. We saw a little bit of degradation in gross margin. Just anything we should be thinking about there as we model the numbers going forward?

Yes, mix was up nicely year-over-year. But to your point, it was -- stayed at that higher end of our range, 75% quarter-to-quarter. What we wrestle with on occasion is mix within the mix. And frankly, this quarter, we had really strong display video channel usage. That display video channel usage is lower than what we average on our total direct revenue mix profile.

But to give you a sense of just how much channel usage we're seeing, our greater than 4 cohorts, so those scaled customers using 4 or more channels was up 44% year-over-year and our greater than 5 or more channels is up over 60% year-to-year. So mix within the mix, but consistent with our model, our long-term model that each year, we want to get between 100 basis points and 300 basis points of efficiency on the cost of revenue line.

Our next question comes from the line of Clark Wright with...

Awesome. It was great to see a strong quarter of organic growth. I was wondering if you could talk a little bit more about the contributions from the agency. You talked about 23 brands being added. How should we think about the growth of the agency business and independent agencies relative to the direct enterprise relationships as we head into 2026?

Yes. We break out our business in a lot of different ways and give you a lot of metrics. One of them we're going to keep to is, keep the disaggregation of revenue down to direct and integrated. But what I will point to is as a proxy for how healthy our agencies are growing. You would see the direct mix was up over 30% on a year-over-year revenue growth basis. And then mix within the agencies, let me just look at my notes here, was up pretty substantially as well from a direct perspective. Let me just get the exact data point here for you. Matt, if you have it. let's say, direct -- agency direct mix was up from 52% direct to almost 60% this quarter.

I mean direct to enterprise -- yes, I'm sorry, Clark, I was just going to say direct to enterprise continues to be the vast majority of our business, and I think will be for many years to come. But we love our agency clients, and we continue to expand with them very nicely. So it's been a really good blend to have both.

Our next question comes from the line of Jason Kreyer.

All right. Great. So first on Athena. When you think about your 572 scale customers, are there specific groups or segments of that where you think Athena resonates better than others?

So Jason, I would start by saying I think our direct enterprise clients will probably adopt it in a more meaningful way first, which is not to say we don't think it really plays very well with the agency clients. From a vertical perspective, I don't really think it's going to be that differentiated.

I think if you have not seen the demo yet, which was standing-room only, I say standing-room only, there were people sitting on the floor who physically asked me to move so they could see the demo. We have a recording of it on our Investor Relations site. But I would tell you that when you put that type of power in the hands of a client, they're going to really adopt it very, very quickly. And we're excited about that.

And just going back to the momentum topic. At the Investor Day, you had talked about 80% faster onboarding. There's some operational streamlining you've been doing. So you just hosted Zeta Live, biggest events for building the pipeline. I'm just curious, is there a scenario where that pipeline of potentially $100 million can transition to revenue at a faster pace than we've seen in the past?

Let me start by saying our goal is to close $100 million in business, not just get the pipeline to $100 million, Jason. So we're on the same page. When you look at the fact that we've given guidance early to next year, we've now raised the fourth quarter this year, it's our 17th quarter of raising in a row. We feel like we've got the wind at our back or we wouldn't be in a position to do both of those things.

So I think that our confidence level is very, very high based on where the record pipeline is, what's coming out of Zeta Live. And yes, we are closing deals faster than we have in the past, and we are onboarding them substantially faster than we have in the past by automation and AI internally.

Our next question comes from the line of Terry Tillman with Truist...

The first question is, I was just hoping to double-click on the replacement cycle opportunity. I think there was a stat that was provided at the Analyst Day about the last couple of quarters, the RFP activity had been up quite a bit. And I don't know if that's independent of the Zeta Live and the pipeline that you built out of that.

But just maybe you could comment a little bit more around where we are in this replacement cycle opportunity and maybe Athena helps with that. And kind of related to that, and hopefully, this doesn't become a 5-part question is like maybe there's pricing and packaging and other things you could drive to kind of minimize the inertia for those folks to move to your platform? And then I have a follow-up.

Yes. So let me start by saying that the replacement cycle continues to be at full scale. And we are not just seeing more RFPs in our company's history. We are seeing the largest RFPs we have ever seen. And if you look at the legacy marketing clouds, the guys like Salesforce, Oracle, Adobe, one of them is leading the industry. The other 2 are legacy. They're great companies with legacy technology. We are the next generation of Marketing Cloud.

And we're really focused on how do we continue to win in what is an even faster accelerating replacement cycle. As it relates to Zeta Live, we came -- we went in with a record pipeline. We came out with a meaningfully larger record pipeline. It's just -- right now, we are firing on 11 of 12 cylinders. I don't say 12 of 12 because I don't want to jinx us, but we're feeling very, very good about it.

That's great to hear. And I guess I won't ask for more cylinders past 12. Maybe that's for another conversation. But in terms of the independents, I think you added 3 last quarter. I was impressed at the Analyst Day, one of the -- your folks that was on stage was an independent, and it was pretty impressive their kind of passion for the platform. Maybe you could just give us an update on the independent side in terms of agencies.