Viant Technology Inc. (DSP) Earnings Call Transcript & Summary

All right. Hello, everybody. Thank you for joining us. My name is Matt Cost from the Morgan Stanley U.S. Internet team. I'm very happy to be joined by Chris Vanderhook, Co-Founder and COO of Viant. Thanks for being here.

Yes. Thanks for having me.

Just before we begin, for important disclosures, please see the Morgan Stanley research disclosure website at morganstanley.com/researchdisclosures. If you have any questions, please reach out to your MS sales representative. And with that, maybe for those in the audience who may be newer to Viant, maybe let's start with a quick overview of the company, where it sits in the ad landscape. And you've been with the company, I think, since it was founded 25 years ago or a little more even. Talk about how it's -- how the landscape has changed?

Well, yes, so my brother and I co-founded the company together. Tim is usually here, but he likes the vacation. But we -- so yes, we started in 1999, actually, the first iteration of our company. Back then, we were -- this is early, early Internet. Everybody was 468 by 60 pixel banners, pop-ups. I was wild. There was no third-party ad serving. Google didn't even exist. It was the Wild West of the Internet. And kind of the -- I'll fast forward this story. But we actually sold the business to Time Inc. in 2015. That was quite a trip, and we thought that was going to be the end of it. And then we really saw the opportunity on the DSP side. And while we were there, we acquired a company called, Adelphic out of Boston. And they had great tech. They were self-service DSP. We saw that landscape consolidating big time, and we knew that there would only be a handful of companies. Long story short, we ended up. They get acquired by Meredith, hostile takeover bid of one magazine company over another. It was insane. We bought the company back in 2019. And we took it back. We used our own capital and bought the company back. We took the business public in 2021. But what we were at that point was just a DSP. And today, the landscape is there's a handful of DSPs that exist in the world that are self-service enterprise-grade DSPs. That's Google, that's The Trade Desk, it's us, Yahoo! and a new entrant into the space in Amazon.

Great. So maybe let's talk about addressability. I think it's one of three major strategic areas of focus. But let's start there because I think it's where you have some pretty unique assets from a data and signal perspective. Talk to us about Household ID, about IRIS_ID and the other unique data assets that you're leveraging as part of your go-to-market? And how are those differentiated from those competitors that you just mentioned?

Yes. So if you look at digital advertising, if I go back to when we first started, we used to go to advertisers and say, why are you buying in linear television or traditional media. Digital is data-driven. You can measure everything. You can target ads, everything. And really what we were doing was laying the groundwork, which later became known as addressability. So the #1 thing that we think the largest platforms will have to have is addressability. Marketers want to be able to address or target their audience they're looking for men 18 to 34. Women, 25 to 49. Whatever it is, you need to have -- you want to deliver highly addressable advertising, so there's no waste. And I know that people who are like, well, doesn't everybody do that in digital? No. they don't. So our Household ID, basically this is household level addressability. We think that is a more privacy-friendly approach that we've gone down, and we also have a big focus on CTV, which I'll get into. And those are household level devices. And so Household ID has 80% scale of every bid request that we see. We get 15 million bid requests a second. 80% of the time, we resolve that back to the household. We know which household that is. That's very powerful. In CTV, it's actually 90%. We can resolve back to who the household is instantly. This is very powerful for marketers because they're not wasting their dollars on households or people that aren't in their target. If you compare that to our head-on competitor, which is The Trade Desk, they have UID2. Their idea is public in the bidstream. We see it. They have about 20% addressability. The next closest one is LiveRamp, and they have about 30 to 35 depending on the day. That's their level of addressability. We've been at this since 2013. What this process is, is what's called identity resolution. I have some data in a database, a name and address, an e-mail or whatever it is. And can I resolve that back to who this household is right now? And it's somewhat complex, but just the headline is that we resolve 80% of the time, that's our scale, and the competitors are in the 20% to 30%. So we're instantly much more addressable. The second thing is to building a really good ad product that's going to deliver great returns or sales or Outcomes for advertisers. I want to know what content you consume. And I don't take my cues from the Trade Desk or Yahoo!, we really study -- I think who does a very good job of -- in their ad platform, especially in using AI, which we'll get into, is Meta. Meta does a very good job. They have something called Advantage+. And Meta knows two things. They know who you are when you're logged into the app, addressable. And they know what content you're consuming, because it's on their platform. And if you've ever used like reels, it's somehow, you're consuming whatever niche piece of content watching and they're really good. They have an algorithm that feeds you more of that. And then they match an advertiser that closely associates to that content. That's what drives the power of Meta ads business. It's completely lost on people. And lost on even our competitors in the space, like no one is really interested in that. So there's a company that we -- Tim and I actually invested, there's a VC firm, that invest in like ad tech companies called Aperum. We're an investor in that, and they invested in a company called IRIS. I met the two founders of that. I thought it was really interesting. They work with content owners in CTV. They go to them and they say, hey, send me Paramount, you have 20,000 titles in your VOD library. Send us every video we will create -- they stamp it with what's called an IRIS_ID. Behind that IRIS_ID, what they do is they take that video, they run computer vision. And they say, what is this content about? It's season 1, episode 2 of Yellowstone. It's a Western. It has Kevin Costner in it. In this scene, there's fishing. The next scene, there's a bar fight, blah, blah, blah. What we're doing is pulling out all the content intelligence about that video file. So that a marketer, when they go to target in CTV, they can show up relevant. My customer Cabela's, when there's a fishing scene, they want to then show an ad about their fishing rods and all that. If I have a customer that sells women products, I'm probably not going to show it here. They probably don't want to buy that. This is -- offers incredible content intelligence that helps us in CTV. And that was the thesis behind the acquisition. Problem was they were a small company. They only had 7% penetration in CTV. In our last quarter, we reported we were in the high 30 percentiles of penetration. We have earnings next week. I'll give the update on where we are there, but we're signing more and more content owners every day. We have every OEM. Publicly, we've announced we have every major television manufacturer who's adopted that. And we have a bunch of large content owners that we'll make announcements of that as well.

Great. Let's shift over to ViantAI. I think that since really you went public, you've been pretty consistently talking about autonomous advertising. And over the course of the year, past couple of years, you've been launching these AI products with the goal of kind of a fully autonomous ad platform, if I have that right. So talk about the various components of ViantAI, the launch cadence and then where we stand with that.

Yes. So we saw this migration. The industry went from managed service, where people manage absolutely everything and then DSPs came out, and then it went to self-service. And what that meant was agencies wanted to control every decision. They wanted to pull all the levers. They wanted all the transparency and self-service has been amazing for the industry for 10 years. But now these systems, these DSPs, they're like Bloomberg terminals. If somebody new started, when I look at these DTC, e-com companies, they want data-driven advertising, but they don't have like reams of staff and traders, they don't. They want automated systems, they just deliver themselves. That's what they want. And so again, we looked at the future was going to be autonomous. These platforms are overly complex, and it's too much for a human, trader or a room full of human traders to make optimal decisions. And so we came out with ViantAI and we started that about 2 years ago, all with the goal of moving towards full autonomy. The first thing we launched was AI bidding. And 70% of our customers are the traders, they belong at an ad agency for us. And the first thing that they do is they set up a campaign it's complex. Then what they do is bid on every publisher, every audience segment. They're trying to bid the right price at the right time. And that actually is, I don't know, there's no -- that's an exercise of utility for them because this is a 24/7, 365, 15 million bid requests per second, a human can't possibly change price at that speed. So we launched AI bidding 2 years ago. And it's just an AI model that predicts a bid the lowest possible price for that ad request, what's the price that we should bid. We have 85% customer adoption on that product. And what we're saving customers 40%, our model does versus human bidding. And so that's been a tremendous product that we have out there. The next thing we launched was AI planning. Agencies have tons of people that work in planning. And a traditional media plan takes at best, 6 to 8 weeks, tons of market research. It's very expensive. And they're designing a media plan for their marketer. We do it in 60 seconds. You give us the URL of the advertiser or product that you're going to advertise your budget, your flight dates and what's your goal? Your goal might be to raise awareness, your goal might be to generate sales. Whatever the goal is, in 60 seconds, it builds out a full media plan. It's incredible. And we use all of our unique data signals behind that, all the bidstream data, our Household ID, IRIS_ID that I talked about, and we build plans in 60 seconds. From there, the trader builds a plan and it clicks a button to build it in the DSP. We save -- a typical ad campaign may take anywhere between 2 and 7 days to set up in a DSP. It's very time-consuming. So we eliminate all that time. We have about 30% customer adoption of our AI planning. The third thing we did was we launched AI measurement analysis. Just think if you're in a DSP, there's so many reports, what are trader is doing after setting up a campaign and then bidding, they're trying to get insights as to what's driving performance, and they want to double down on that. But that's very complex for them to get at a lot of that data. And so what we want to do is completely automate it. You just basically chat with it, you make great prompts and it gives you all the insights. You could say, hey, for my CTV campaign, what are the top-performing publishers? What are the top-performing IRIS_ID and segments, which shows are driving the best performance? What should I do to improve performance? And it makes recommendations to you. We have great customer adoption on that. We have tons -- that's a phenomenal product. And then the last thing we did was AI decisioning. This is our move towards full autonomy. We want to create a decisioning architecture, decisioning system that will make human-like decisions, but it will do it at the speed that a human can't. We launched -- so we committed that we would launched that in 2025. We did that in the fourth quarter. We launched a product called Outcomes, which is the first fully autonomous ad product that we launched. And what makes this possible is an innovation that we built and we -- it's called a Lattice Brain. Most autonomous systems that are out there today, whether it be self-driving cars, or guided missile systems or unmanned drones, they all have some type of Lattice architecture that makes decisions based on a set of inputs in real time, but they make human-like decisions. I like to use the self-driving car example because everyone gets that with like a Tesla. But it's the same thing in these advertising platforms. We want to make human-like decisions, but at a speed that humans can't. And what we're trying to do is compress what we call decision latency. Where in these current systems, we make recommendations or our platform make recommendations to traders every day on how to improve performance. But then they think about it. And then they sometimes got to take that recommendation to their boss and say, well, hey, we want to make this change. Boss might call a client, and the end advertiser and say, hey, we need approval to do this or that. And then they think about it. And then a week later, they might make the decision. Most of the time, they don't make the decision. And then ad performance, campaign performance leaks all over the place. And we want to compress that, and that's what our Outcomes product is about. It's our first fully autonomous product. It makes -- we ran a series of about 20 pilots in the fourth quarter, and the performance results were incredible.

So let's stick with Outcomes there for a second. I mean it sounds a little bit like Performance Max or Advantage+ from Google and from Meta. I guess when you're talking to advertisers about this product, where are you differentiating yourselves? And what is that process like for pitching Outcomes?

Yes. So -- okay, about $400 billion are spent in the United States in digital. About 70% of that is performance-based advertising, is where those dollars are going. And those are -- that's predominantly in three companies: Google, Meta and Amazon, right, sponsored listing ad on Amazon, that's performance-based. We want to go after that market because it's largely untapped by anyone on the open Internet. So it is true that some of the more automated ad products are Advantage+ for Meta and Google's PMax or Demand Gen that they have. And it is aimed to compete with those companies. They operate those automated systems, but they do them in their own walled gardens against their own content that they own. And we want to bring a solution like that for the open Internet. And I think that, that is just an incredible opportunity, and we think that we have advantages over them. And really, those advantages, we know that we have to deliver performance that are on par or better with them. I think a lot of the performance that you will get -- that a marketer will get in Meta and in Google operates off of the thinking that whoever showed the last ad gets credit for the sale. I tell our customers all the time. I promise you that Google, Meta and myself, we know who's about to buy your product. But just because I snuck an ad right in front of them before they purchase doesn't mean that your total business is going to grow. And so we're aimed at -- we're a buy-side only player. We don't have any content that we own. I only represent the marketer, and my whole goal is to drive business performance for them, but not business performance in a dashboard. I want business performance that you report to Wall Street. And this is an insight that a lot of companies had they've been in digital for years. They've been seeing customer acquisition costs go down and down and down, get more and more efficient. But results they report to Wall Street don't tell the same story. There's lots of customers who have this problem. And we want to drive incremental growth in their business, not the same growth or the same sales they were going to get anyways. And Meta and Google pray on less sophisticated advertisers who don't know the difference. And we are out, when we deliver our Outcomes product, it's about delivering new growth to the brand, not the same sales. And we think that the vehicle that does that most effectively is actually in the CTV channel. And so I think that those are some of the ways that we're different than them.

So it sounds like you're kind of in the business, at least with this product of eliminating, I don't want to say, wasteful spending, but certainly helping people focus on the spending that they can actually tie back to dollars on the P&L. So it's a more focused way of doing it. I guess what's the financial opportunity for Outcomes?

The Outcomes, I think it's absolutely massive. Just start with the 70% that we -- I believe we currently don't touch today. There's probably about 10,000 advertisers that buy on the open Internet. There's probably 1,000 or less advertisers that buy in television. Meta has 10 million customers. So that right there, like I really like that opportunity. To me, that's sort of like freedom because if I produce a return on ad spend for you of whatever your goal is, when I do that, you give me more money. In the television business, if you raise unaided awareness by 2 points, you don't get more money. Like the brand has a fixed budget amount they're going to spend for the year. So we like being able to go after this performance opportunity because it heats -- that money is heat seeking. They don't care who they spend it with. The other insight, too, on Meta, these e-com and direct-to-consumer -- of the $10 million, I believe most of their growth is driven by a few hundred thousand of these e-com and direct-to-consumer companies. And they are -- most of these companies spend all their money with Meta. They're only on Meta. They run their whole business there. But a lot of them tap out on performance, and you can see it if there's a whole, they call it a DTC on X, like these groups on X. And they put a bunch of content out and you see them regularly. In the fourth quarter, they get absolutely squeezed on Meta's pricing, performance dips, they have nowhere else to go. We think that it's an unbelievable opportunity to go to them, deliver an autonomous product, yes, in Outcomes. But we can bring them into CTV, and we see it across all customers. The incrementality or the net new sales that you get in a channel like CTV, it absolutely trounces Google Search, which has 0 incrementality. I have not had one customer where I've seen any notable point gains of incremental customers from Google-branded search, which is 40% of all search buying. And in Meta, you typically see around 20% incrementality in Meta. But in CTV, on average, we see somewhere around 150% to 200% growth on average.

So it will be interesting to see how that develops. It sounds like a really big opportunity. I guess flipping the AI debate around to the other side, there's a lot of questions being asked right now about the durability of software business models, really across tech. We hear about it for the past couple of years at this conference, but certainly very intense this year as well. Can you talk about these set of concerns? How do you believe Viant may be insulated from this idea that AI can stand up a software solution that could replace their replicated DSP in an afternoon.

Yes. Well, first, I think this -- first, it was the SaaS apocalypse, then it was the AI apocalypse because then it was like it was going to take out all enterprise software. And then it was like, oh, wait, it's also going to take out all marketplace businesses like DoorDash. No one said Amazon. I mean, that's the biggest marketplace. And this is -- and then basically, I watched through that, our stock had absolutely hammered with everybody else. And internally, everyone is like, no, wait, we are the company that's moving towards autonomy. We are attacking the workflow of like that is what we do. We're leading that. So it is a bit of a lazy take, I think, by investors. However, just to hit it head on. I do not believe that people are going to vibe code everything. And I don't think that they're going to vibe code all software. I think you can vibe code UI. Like I have a DSP, you can vibe code in my UI and you can vibe code The Trade Desk UI. Sure. And I think that's your moat is your UI, which is for a lot of software companies. Once you get someone who knows how to use your software, you get lock in on that. We benefit from that, The Trade Desk, they benefit from that immensely. But there's so much more that I think is lost on people with this narrative. It's the last mile that everything. And what I mean by that is, when we are competing against The Trade Desk for years, my whole original thinking was we just built a -- we build a UI that's very similar to theirs, and we get to feature parity, and I should be able to win half the customers because I don't know, I can sell pretty well. And I think I'm charismatic and you'll like -- half the people will like me, half, they won't, I'll win half of them. It didn't happen. Then what I did was we introduced fine. When we go head-to-head, I'll just beat them on pricing. And in some cases, some customers we literally dropped our fees down to 0% or 1%. Someone else is doing that in the space right now, it's called Amazon. It didn't work. the level of differentiation that you have to have to win customers is immense in this industry. That's number one. And two, you have to build for the customer. You have to know what they want and you have to build for that. And that's that last mile that sure, you can vibe code something that looks like what I have, but you're not going to win any customers. You're not going to be able to customize it for their exact needs. All the integration work that you have to do with CDPs and all these data warehouse companies where all the customers house their first-party data, their CRM data. They want to use that in advertising. You're not going to vibe code that. The infrastructure that it takes, the actual technical infrastructure to handle 15 queries or ad requests per second, you can't vibe code that. So I think that, that makes it defensible for all DSPs. Now when I think about other enterprise software apps and things like that, I'm saying, well, these tools have been out there, the #1 target is Salesforce, no one like Salesforce, myself included. Anything about them. But okay, like cool. But vibe code it, please, someone just do it. The tools are here. Like why do we have to talk about the bogeyman that's going to do this. Like it's someone actually do it. If it only takes 2 days and like someone nerd a basement, like bring on the nerds, let's go. But no one is doing it. So I think that -- I don't know. I think that some enterprise apps will definitely be appended. We license tons of them internally. It's unbelievable how these businesses are created off this like little niche use case in enterprise software. And I think some of them that aren't infrastructure -- that don't represent infrastructure that aren't connected into all other data systems, if they're very skinny. I think they can be appended. Yes. And the last thing I'll say about all the AI companies, keep in mind, 60 days ago, no one knew who Dario Anodei was. No one knew who he was. He's everywhere right now. Remember what he's doing in a middle of a fund raise. I did invest in his route. But I think that the whole like AI apocalypse, like he's selling that. like to his -- like, that's what he's selling. He's raising money. Sam Waldman is doing the same thing. So a guy who's on my Board once said, be careful of the loudest guy in the room. He's always selling something.

You mentioned Amazon in there, and I want to talk a little bit about kind of what they're doing in the space. There are definitely reports coming far and wide that they're being aggressive in trying to sell this product. So have you observed heightened competitive intensity from them? And why should investors not be uncured about it?

I think that my competitor, The Trade Desk publicly is every quarter, he says that Amazon is not a threat. And no, he doesn't see him and all that. We don't take the same -- we don't have the same take. It's Amazon. So why are they a threat? Yes. Why? Because they have the lowest cost of capital of anyone. So I don't care where they put that canon, they can point it anywhere and you have to be concerned about that. So that's just being honest. We don't really -- they have a DSP. Their goal of what I do believe about them is their entire existence is to sell ads on properties that they own. They are out offering 1%, what I do for 1%. Like I haven't seen any degradation in my business because they're offering 1%. And for what they're talking about, this 1%, we charge 3% for it. Trade Desk charges 3%. No one's making a change to save 2 points. To switch DSPs is an incredible effort for a brands. And I don't think anyone is moving for that. I think a lot of brands, we recently won a customer we announced Molson Coors. I said to them, you sell product on Amazon. Why were they not in the mix? Just curious. And they said, well, we do sell product with them. I've seen their DSP. They've been in here a lot. They do not have a good DSP, which I know, I hear that a lot. It's not built for us. And their whole goal is to sell more Prime Video. And we are going to buy Prime Video from them. The have a Thursday Night Football. They have the NBA, like we are going to spend with them. But we don't believe that they have our interest in mind. Why would I give them my customer data, and let them plan all my buys on Disney and Paramount and all the other ones. Like we're a little weary of them, was their point and their biggest fears is that they come out with Amazonia Light, which they said they've done in many categories. So no, we're not going to choose them. So I think that again, I think it's Amazon. So you got to watch them. But we haven't seen -- they've talked about a lot in the press. But they themselves have a long way to go from a feature set standpoint, to even get to a par with some of us -- the other four that are out there.

Got it. So you brought up Molson Coors, and I want to stick with that because I think historically, you've served mostly U.S.-based mid-market customers. But with Outcomes and this Molson Coors launch, it seems like you are kind of expanding there. So talk about that win, what it means for Viant and kind of how that relationship was established? And then what does that mean for your ability to go after customers that maybe investors would associate more with DV360 or with The Trade Desk?

Yes. Well, every time you go to market, and this has always been the case for, I don't know, 10, 15 years. Customers say, all right, I get you have your DSP, blah, blah, blah. What unique data do you have or inventory that you represent that would get me to switch? And like that's everything. What exclusive data do you have that I can't get anywhere else? Or what inventory do you own or have access to that I can't get through anyone else? Those are the two like headline value propositions. Molson selects us really based on -- and they had a whole press release around this, but scale of our Household ID. They have about 180 million plus people in their CRM system. They own a ton of different brands. They are big time and addressable advertising. I talked about our Household ID. But what they realized they're very sophisticated. A lot of people, what they wanted to do was not just take their 180 million CRM file, match it in a database. They want to see how many you actually find out on the Internet. How many people do you actually find. And that's actually the real test of addressability. It's not like you matched 180 million e-mails in a database. No, go find them on Disney+. Go find them on Paramount. Go find them on those apps. Because I want to serve them addressable advertising. That was the first one that we separated ourselves. We absolutely crushed The Trade Desk, we crushed their incumbent platform. We really stand out there. The second thing was -- so again, what unique data do you have? They realized it in Household ID. The next piece was around IRIS. This is a regulated company. They need to show ads at a certain level of 21-plus households. They can't be showing Coors Light ads on kids' content. They get sued for it all the time. IRIS_ID was huge in that. that when you buy today, if you buy CTV and any other platform, you don't get to know the content that's behind it. So if you buy Paramount+, they have 20,000 titles, which content is it? If they show ads on kids programming, they can't do that. IRIS_ID was an amazing solution for them because we know what the actual content is. We know the name of the show, and we know everything about it. And so that was big for them, not just from avoid getting sued, but if this is live NFL football, you're going to have a football theme ad. If this is -- and you're going to -- let's say, you're going to show a Coors Light or Miller Light that they own. If it's a romantic comedy, you may show a different one of your products, you may bid differently. They saw the value of that is being able to increase campaign performance. The last thing is, and another unique data signal that we have is something that we call the supply quality model. Beer sells are down, only 52% of drinking age Americans say that they actually drink alcohol. It's the lowest point in 70 years, probably why our birthrate is depleting. But what they have to do is get their ad dollars to work harder. When sales go down, marketing spend as a percentage of sales. And so they have decreasing budgets likely because of sales, yet their CEO is still calling for growth. And in that pitch this year because I've been pitching them for the last 3 years. And it's what are we going to do to get -- you have a lower ad spend, but how are you going to grow? And they're like, yes, that's like topic du jour internally, how are we going to do that with less money? I said, we are going to use not only our Household ID, you're going to be more addressable, you're going to show ads on more relevant content. But you're going to use and take advantage of our supply quality model. And I'll give you some facts that you may not like, but the truth is about the Internet that at least 30% to 40% of the ads that you buy are not seen by a human, are not shown to a device that's a real device or -- and it's not on a site that you think it is. And it's likely I made -- what we call a made for advertising site. That has a boatload of ads all over it. There's no content, and they get people to just click on the ads, and you think that they work. We audited and took them through their whole media plan. We used our supply quality model, showed them that for the same dollars, just by using our supply quality, we're going to get you another 30% of more working media, and this is how we're actually going to use a lower dollar amount of ad spend this year, but to actually drive more results. That is usually controversial for people because I'm not selling a brand safety solution. I'm not. I'm just selling, how do I get my clients' dollars to work harder so that they use advertising to grow their business more effectively. That's all I want. If we start growing sales at Molson Coors, that's good for me because sales go up, budgets go up, they spend more money with me. That -- The last piece that I told you, either gets us thrown out of the room or it's a big buying signal.

Got it. Well, I think that's our time. But Chris, thank you so much for being here.

Thanks for having me. Thank you, guys.