David, congratulations to you and the team on 20 consecutive beat-and-raise quarters. One of the biggest takeaways from Q2 was a meaningful increase in Zeta's 2026 GAAP EPS guidance, where you went from 2 to 4 cents — only a 33% raise, only haha — and went to 9 to 11 cents now for 2026 on GAAP EPS. What is driving that improvement in profitability? What are you most proud of this quarter? And also, organic growth moderated slightly, 1% down. How should investors view that deceleration, and what factors impacted the numbers? And how do you see organic growth progressing going forward?
Let me start with your last question first. Organic growth didn't really go down. I mean, it was a few basis points below the quarter before, but our average organic growth rate over the last, you know, four to six quarters has been 23 to 26%. It hit 30 once, but the reality is it's been in that range, and I think it's going to continue to be in that range. Right, we've given guidance to mid-20s organic growth, and we've beaten by 2 to 5% in each quarter. I'm not worried about that. As a manager, you're going to see it accelerate in some quarters, and other quarters it's going to abate slightly. But the reality is, what really matters to me is where are we on the rule of 40. Last quarter, yes, organic growth was 29%, but we were at the rule of 48. This quarter, yes, we were 28% organic, but we were at the rule of 49. We added 170 basis points to operating margin. So, you know, you're seeing a business that, when you're growing at these rates, it's very difficult to sort of regulate it to 20, 30, 50 basis points here or there as it relates to growth. I'm very, very proud of the quarter we put up. You know, 44% top-line growth, 56% EBITDA growth, and 73% free cash flow growth. I'll also point out, Ronnie, in the raise that we put up — let's call it a 32, 33 million raise, which is one of the biggest raises in our company's history — we also raised free cash flow by 20 million dollars. Other than us and Palantir, I can't think of another AI-native company that is growing free cash flow substantially faster than it's growing revenue right now, because everybody's taking all of their profits and then some and plowing it back in and are taking themselves to negative free cash flow. So I am not worried about the difference of 28, 29, whether it's, you know, in that range over the next few quarters. What we're really targeting is how do we continue to grow this business greater than 20% on an organic basis for many, many years to come. And I think the partnerships we've put in place, the deals we're doing, and the way we're executing really show that that's where we're going to be. Our goal is not were we at 28 or 29% last quarter. It's how am I building a 10 billion dollar a year business with a 30% operating margin and the vast majority of that flowing to free cash flow. That's one that I really am passionate about because I'm just so proud of the quarter we put up and how much the team is executing. This sort of discussion about 50 basis points of growth, you know, sequentially is not something that I'm worried about.
Crushed it, hahaha. Mr. Zeta, what's your question? So David, let's talk about Palantir. You mentioned Palantir already, but on the call you said that Zeta officially has now two joint go-to-market clients with Palantir. There's two agreements on the table. You're two for two, 100% hit rate. Congrats on that. That's awesome.
And we have three more meetings next week.
You did mention that, and I did want to note that. On this topic of clients, realistically for 2026 and forward-thinking, maybe 2027, 2028, as this deal starts to mature, what should Zeta retail audience and Zeta, you know, just shareholders in general expect for the floor? Like, how many clients do you anticipate internally, or just, you know, if you kind of wanted to elaborate a little bit more each year?
Listen, I've said publicly that I think that this is 100 million dollars a year in business. I would tell you, post my first two meetings, that's looking super low. I just, I think it's going to be more than that. You know, I said, let's look at of the 800 clients, which ones spend greater than a billion a year in total marketing, business intelligence, so on and so forth. And I said, let's try to get to 20 targets. It was so many clients we had to whittle it to 20, like we had to bring it down to 20. And I'm spending a disproportionate percentage of my time on this, and I think it's the best use of my time because I think once I figure out — first, I want to sell a bunch of them, I want to get them up and running, and I want to show them how well we do it. Then you've got testimonials, then you've got reference clients in that group. The other great news that nobody's really talking about is we have a ton of client overlap. So there's a ton of our clients that already work with them that I think will be able to massively accelerate, which will be another big driver to net retention rate. So I'm, you know, if you can't tell, very bullish on Palantir. Obviously, their results were insane. You know, it's funny because one of my biggest worries going into our print was like, how does anybody top that, right? Like, I mean, the crazy thing is, you know, Palantir grew at 94%, they trade at approximately 60 times revenue. We grew at 44%, you know, as of yesterday we traded at 3.3 times revenue. So it's a little disproportionate. Even if you look at like CrowdStrike, you know, they're growing slower than us, their net retention rate is lower than ours, our cash yield to market cap is 100% above theirs, and they're trading at 41 times revenue, and once again we're trading at 3.3 times. So I think that we continue to get a bit of a rap, and whether it's marketing or whether it's, you know, erroneous information that was put out on us a couple years ago, I'm not sure yet. But I think we're finally starting to get through this stuff. And, you know, I think as we continue to execute as a company, which is what we are going to do, you know, we're going to continue to see meaningful wins in the marketplace. And, you know, back to what Ronnie asked, I'm not myopically focusing on one quarter. I'm trying to build one of the very few software platforms that gets to 10 billion. Even Palantir's not at 10 billion yet. They're going to be there very soon, but they're not there yet. I mean, you can count on one or two hands the number of software companies that have gotten to 10 billion. And if we do that, our stock is, you know, many, many times what it's trading at today. And, you know, as far and away the largest shareholder, I'm focused on that.
So OpenAI is beginning to roll out advertising inside ChatGPT. They're one of the first ones to start doing this, and we know that you are partnered with OpenAI and Gemini. Gemini does it too. Ronnie, oh awesome. This is going to help brands navigate obviously this new age of generative engine optimization, kind of starting to replace search engine optimization. Two separate things, interestingly enough.
And I'll jump in just to quantify. GEO is getting our clients into the results of the chat. The advertising will sit next to it. So you would think of GEO as search engine optimization, and you would think the ad serving as search engine marketing, just to give you the two analogies that you would really think about. And by the way, I would tell you 99 out of 100 people wouldn't understand the difference, but I want us as a group to be able to do so.
So how has the partnership evolved so far? Is Zeta helping OpenAI deploy and scale advertising already, and what could this opportunity ultimately mean as a singular pillar for Zeta?
Yeah, so let me start by saying we are actively participating in a meaningful way in both GEO and ad serving into the ChatGPT environment. But when you separate them, which I'm going to do for just one second and then I'll get to your OpenAI question, we're doing GEO today into Anthropic, into OpenAI, and into Gemini, and we have API access into all three. So we're able to help our clients get their brands into the results that are cranked out by the foundational models that we just discussed today. That's a really important part of how does a brand stay top of mind share in an environment where it's getting harder and harder to get traffic delivered off of Google, just because Gemini is answering most of the questions on platform. Two, we've entered into an agreement where we also are connected into OpenAI's API — say that three times fast — to serve ads on behalf of our clients, which, unlike anybody else out there, I'm not sure anybody else is doing this, we're able to target to a deterministic individual inside of ChatGPT. So we're not just looking at the results; we're able to look at the person who's actually doing the search. I don't know another company where we connect today. The ZID ID connects to the Med ID, it connects to the Google ID, it connects to the OpenAI ID, it connects across the 5.2 million publishers that we're native to their platform, which is over a trillion pages of content on the internet, and it connects to a device graph that can rearchitect to a deterministic individual inside of connected television. We can hit an individual across any digital methodology and have a deterministic attribution capability. OpenAI is going to be a very important component of how ads are served going forward. I think ultimately you're going to see Anthropic come in as well, and Gemini is already doing it. And we partner with Google so we can serve ads into there, but they're primarily, you know, doing Google ads as you would imagine. But I think OpenAI is going to scale very rapidly. What I would tell you is that I think that, you know, companies are going to have to justify results when they go from being hyper-scaling private companies to hyper-scaling public companies. And I think for OpenAI, advertising is going to be a big part of it. And, you know, I remember when Netflix said, 'We will never do advertising ever,' okay? You know, and you hear this from different platforms now. You've got two streaming platforms — you know, I know the names, but I'm not sure it's public yet — there are two streaming platforms that we work with today that are building a free component to what they do, all ad-supported. So I think you're going to see foundational models across the board begin to move ads into them, and I think that'll be a big opportunity for Zeta down the road.
You know, David, it's kind of funny that you mentioned that because I wrote recently about OpenAI, and if the requirement of them to go from IPO to IPO is that they need a sustainable revenue engine, and I said that's very beneficial to Zeta Global because Zeta can help provide them with this programmatic and deterministic placement, which is, you know, it's a win-win, right? It's a win for OpenAI, it's a win for Zeta.
And by the way, I want to say, just to reiterate so everybody knows, no large language model ever sees our data, ever. We use only our own inference models to touch our own data. And it's one of the reasons I think we really have, you know, what is, you know, really when you think of sort of architecture, you know, you think of sort of enterprise-grade infrastructure. You know, infrastructure companies can raise guidance by 32, 33 million and say 20 million of that's going to drop to free cash flow. And that's where we are as a company today. It wasn't just a rebrand to sort of intelligent AI infrastructure; it's what we've been for a long time. We just were doing it sort of exclusively in marketing, and now we've branched out from that. So when you look at the sustainability of our relationships with our marketing partners, we are a real force in that. And I think it's something that will benefit OpenAI, will benefit Zeta. Today, not only do we have a deep and meaningful relationship with OpenAI around powering Athena, which is our most important go-forward, you know, software and intelligent agent, not only are we serving ads, we've recently entered into an enterprise-wide agreement to put OpenAI in every one of our employees' hands in a fixed-cost scenario, which I don't know if they're doing with that many others. So when I look at our token utilization and I look at our cost on how we automate — because remember, last quarter, Q1 '26, we were ecstatic to get to 75% of new code auto-generated. In Q2, it was 89.6% auto code generation. We did that with token utilization, all AI expense combined being well under 1% of revenue because of Spade and because of other internal workflow tools we've done. Even putting every OpenAI into the hands of every one of our clients will still be way below 1% of revenue in token utilization and AI cost.
