So, here we are, live at Possible Miami, 6,000 attendees, David, and they're swarming South Beach, and the thing you hear, other than talking about the Zeta-branded coconuts that everyone's on top of, the buzz phrase on everyone's tongue is AI. AI, AI, AI. So, what are companies and the market getting wrong? What's just hype from your perspective?
Well, I mean, listen, you can go back to multiple technology cycles that have gone through a similar hype, right? I mean, in my lifetime alone, it started with dot com, then it was cloud or mobile, then cloud or mobile, I forget which came first at this point. But if you think about the dot com era, I think there's a lot of similarities, and a lot of people don't like hearing this, but I think there's a lot of similarities. You had a new technology coming in that everybody said was going to put everybody out of business. Yeah. Walmart's going to be gone. You know, you're going to have JP Morgan Chase is going to be gone. FedEx will be gone. Everything will magically get from where it's got to go to where it's got to go. And in the end, companies that adopted this new technology have become the most powerful companies in the world. Companies that turned their backs on it are gone. Most of the original dot coms didn't make it, right? But you have companies like Amazon, Meta, Apple, AT&T that have really thrived in an internet era. You saw the same thing in cloud. You saw the same thing in mobile. In AI, you are going to have two types of companies. You're going to have companies that adopt artificial intelligence and win, and you're going to have companies that do not adopt artificial intelligence and for the most part, you know, fail or go away. When you look at what I think you're subtly referring to is I lovingly call the SaaS-pocalypse, and we've got a lot of that. That's certainly been a fun few weeks for us. What I would tell you is that you're going to have, once again, winners and losers. Companies that are software-based, that have built workflow management tools without proprietary data, without intelligence, are going to have a major pivot ahead of them to win or survive. Companies that have massive amounts of proprietary data and create intelligence and really high-quality outcomes, like Zeta, are going to adopt these technologies and thrive in the next era.
And I'm not talking thriving for a quarter or two. If you look at our data assets at Zeta, I think this is a moat around our company for the next 10 to 20 years. And listen, we're going to have to adapt to new technologies in 10 years, because you're going to have quantum coming in. And when quantum and AI get together, it's going to hyper everything. And we're going to start seeing technologies that we can't even contemplate. And, you know, we pivoted to artificial intelligence in 2017. When we went public, this is one of my favorite stories. When we went public in 2021, we did this huge banner on the side of the New York Stock Exchange, right? Cuz you know, you get to do that. And our banner said, 'Data plus AI equals intent.' And I joke you not, multiple people came up to me that day and said, 'Who's Al, and why is he in charge of your data strategy?' It's a true story.
Well, let's talk about the expectations for AI, David, because in the analogy that you just referred to, those who adopted and got on the internet and the names that you threw out there aren't just big, they're mega, mega, mega players, right? And then you got those who just failed and they died. Yes. But that middle group, right? What do you think about AI? Is this about growth to being mega, mega, mega or death and then there isn't a middle?
No, no. That's so that's a great, great question, man. The answer is you're always going to have middle companies. Right? For, you know, you've got Walmart, which obviously thrived. You've also got Target. You've also got Kohl's. You've also got a bunch of other places cuz you don't really, I think, have Sears anymore. And I do think there's JCPenney. But my point is that companies have survived and they make it through and they've adopted and all of those companies today have an e-commerce platform. All of them, right? But, you know, I believe Zeta is going to be one of the mega megas. Like we really believe we're building the next sort of generational changing marketing company and technology company in what we do and we believe that we have people, we have the data and we have the technology to do that, to help our customers. Because at the end of the day, the biggest value we drive is today for every dollar a customer spends through our platform, we return an average of 600%. Our goal is to get that to 1,000% with Athena and I think we'll be there. And we have some customers already there based on some of the early findings of the Athena launch. But, you know, as you look at those types of return on investment, clients move more dollars to you. That's just how business works. If you succeed and you drive incredible return on investment, clients do more with you. If you do not, they do less with you, right? And I think you're going to see a lot of companies end up in that middle. And you know where I think the middle is going to be? It's interesting. There are companies like Zeta and others like Palantir, obviously OpenAI and Anthropic and others that have made data AI foundational to their businesses. It is literally at our core architecture is AI driven. You're then going to have a lot of companies that keep their existing architecture. And they have algorithms that sit outside of it. I have to put this down to do this, but you might have a platform that they have to step out of the platform to do a data dip into a repository which then goes to an algorithm which then creates the intelligence that then tells the platform what to do. That latency destroys return on investment where we have to make a decision for our clients in a millisecond or two. You're going to have a lot of those middle layer companies that do well. But it's going to be a step out of the architecture. Now, the APIs that they're using to go in and out might get faster and they might be able to look at additional data points in faster periods of time at some point, but they certainly can't at this point. And I think that's where the middle tier of companies are going to be. And I'll go out on a limb here and I will tell you the vast majority of the companies I believe end up in the middle are the companies we today consider the incumbents.
Okay, we've got to continue that because you have gone out on a limb. What is it about new product development, right, that you're doing that the incumbents aren't doing? And how are you charging for it? Cuz I do want to come back to pricing.
Yeah, so I mean, a lot of this is the innovator's dilemma. This has been going on for hundreds of years, right? If you're super successful and you're generating billions of dollars of revenue on your platform, you are not likely to throw that platform out to create an entire new architecture. And that's what you have to do with AI. It's not you can't just add it. People think you can, but you really can't. What we did differently was a decision we made in 2017. Where we brought in Nish Gor, who's going to speak today, and Chris Tanan, I'm sorry, Chris Monberg, and a few other players who are in our company today. Steve came in at that time and we put together two assets. One was called Discuss, which was the world's largest commenting platform and one of the largest data repositories of opted-in first-party data. And we bought Boomtrain, which was one of the top natural language processing platforms in Silicon Valley. And we literally took the eight platforms we had bought to that date. We were very profitable at the time, comparatively speaking, not like today, but and we had to go to our private equity partners and be like, 'Well, guys, we're going to go from making, call it, 50 million a year to nothing for the next 3 years so we can actually run these platforms for cash flow while we build an entirely new architecture that's going to put this thing you've never heard of called artificial intelligence as native to the application layer.' And you know, that was probably the only board meeting I have ever run, or certainly in the last 18 years, that was not a unanimous vote. Fortunately, it didn't really matter because I controlled the company, so we did it anyway. But, that being said, you know, we didn't even launch our new platform till late 2020. And really, 2021 is when the Zeta marketing platform came into its true life. We tried and failed three or four times. It just wouldn't work. And then finally, interestingly enough, the tech caught up to us and our vision. And by 2021, it was fully embedded. And it was great. And that's the big difference.
And by 2021, you're going public, right?
Yeah, June of 2021. We were growing already and the platform went live in 2020. I mean, a lot of people were using it. I didn't think it was where it should be, but it was. The good news was it was better than most of our competitors already. And now we've continued to evolve it. And by the way, if you look at what we're doing internally, back to this whole sort of SaaS apocalypse narrative, at the end of December, Google put out a press release that they were at 50% automated generated new code. We were at 57.5% already at that point. We were ahead of Google. We're now above 75% of our code is auto generated. We built this incredible internal tool called Spade, which don't ask me what the acronym stands for, but that actually has automated our ability to use any algorithm. So, we could use Spade to use Claude to check our security protocols. We could use Spade to auto generate code using OpenAI. And then we have a program called Zippy. Great name. Yeah, I don't know where I come up with these names, but we have Zippy, which does all of the Q&A before it pushes live. And we're literally seeing pods specific to Athena, pods at a 4 to 500% average productivity increment over just 12 months ago. So, while others are talking about this like destroying SaaS-based companies, companies like us that are using it, we're moving 3 or 4 years ahead today. We're going to push code for Athena in the next 3 months that would have taken us 3 years.
We're going to come back to Athena and I really spend some time there, but before we do, I do want to just come back to this point you made, David, about the degrees of freedom you had as a private company versus now. Public companies typically don't get patience from their investors to fail three or four times before something goes right.
Fortunately, we haven't done that as a public company, so that was good.
You were also, when I think about it, at a moment where your growth was so strong that delivering 600% return on investment to your customers made sense as a public company as the growth play. Now, as an investor myself, as someone who covers the markets, I hear that kind of return and I ask, 'Well, Zeta, why are you not taking price to grow share?'
Yeah, listen. So, it's a good question. In this the last 3 years, we've grown this business on a compounded growth rate of greater than 30%. We've grown EBITDA by greater than 50% on a compounded growth basis and we've grown free cash flow on greater than a 75% free cash flow basis while simultaneously delivering a higher return on investment to our clients by 100% than they can get from anybody else. It's a good mathematical formula. We've added 500 basis points to operating margin in the last 3 years and we believe we'll add another 500 basis points to operating margin over the next number of years. So, the way to think about it is today our existing customers, forget about like global town, our existing customers spend 100 to 110 billion dollars a year on marketing. And last year, I had a 130 basis points of wallet share. I want to get to 1,000 basis points of wallet share. How do I build a 10 to 15 billion dollar year business with a 30% operating margin and the vast majority of that dropping to free cash flow? If I start looking at nickel and diming my clients to make an extra few points of margin, that's going to make that harder. So, the counterpoint is we believe that the right number to get our clients to get to that wallet share is a 1,000% return. We believe that you get your clients to a 1,000% return on marketing spend, it's game over. Everybody's going to move everything they can to us. And then we can make a little more money on top of that, right? So, if we could get to 1,200, I don't think our clients are going to begrudge us making a few extra basis points, but I really believe that we've done the right thing. Listen, we've been public now for 18 reported quarters thus far. We've beat and raised 18 quarters in a row.
