I'm a senior analyst on the US Internet team. I cover two sets of companies, video games and interactive entertainment, as well as advertising technology. The ad tech stocks that I cover are AppLovin, DoubleVerify, Unity, and Magnite, which I've been very pleased to cover. And today I have the CEO of Magnite with me, Michael Barrett. So yeah, we've been a fan of the stock for a while now, and all seems to have been right. I think investors are very happy with the stock performance as well. But today, you know, we wanna catch up with Michael about two things. We're going to talk a little bit about, like, an update, you know, about the DV+ segment and the implications of the Department of Justice case.
And then second, we're gonna take a little bit of a different tack and talk about, you know, growth drivers for CTV. Things that Magnite, in particular with its sales force, you know, has the capability to drive high growth for a very long time. So with that, I guess I'll start with questions. Thank you.
Thanks so much.
Thanks, Mike, for coming. The result of the DOJ case came out last night, and they seem somewhat favorable to Google. Is there any read-through from that case to the ad tech antitrust case at all?
Not that we can piece together different courts, different judges, kind of different, different remedies that they were looking to seek. I mean, both are. The commonality is there's some structural remedies being proposed by the DOJ, but I wouldn't, we've been racking our brains for the better part since the ruling came out so now, and we don't see much of a read-through.
Okay.
Nor does our outside counsel, which probably is more important than what I think.
And you know, obviously since the last time we talked, it's now a couple of months later. I think the behavioral remedies, the case that pertains to Magnite and supply-side platforms, is going to be in September, like later in September, if I'm not mistaken.
Yeah, the 22nd it begins, right? The expectation there is that it's called the Rocket Docket for a reason. Very complex, murky world, ad tech. So, you know, it wouldn't be without reason to have a ruling come down. Obviously, it's been found guilty, so it's not about whether it's guilty or not guilty. It's whether or not the remedy should be structural or behavioral. DOJ would like structural. Google has countered with, "We don't have to break it up. We'll behave this way, and this should be fine for the industry." Frankly, we don't have a point of view because if the behavioral remedies are put in place, it's good news for Magnite.
But even if structural is recommended from the bench, we've been told that it's not without precedent, that behavioral remedies are put in place during the appeals process, which you can bet there'll be an appeals process. So instead of thinking about this as a two- to three-year payout when it finally settles, behavioral remedies light up the minute the ruling comes down, and we think that's very beneficial to Magnite.
Since we last talked in June, you know, have your thoughts about the timing of those behavioral remedies changed at all?
No, I think it's aggressive to think that it could be something that impacts the current year, but if behavioral remedies are put in place, it will definitely have an impact in 2026.
I've also noticed that there are a number of lawsuits now against Google, OpenX, for example, and then some civil litigation. You know, what is, what should investors read into from that?
That there's merit, that we've looked at it closely, that we believe there's merit in those suits, and that we reserve our judgment as to whether to proceed or not there.
Got it. Okay. So, DV+, your segment that focuses on the open web. There seems to have been some share gains recently, and I was wondering if the share gains are more a function of, you know, switching between vendors or more ad spend through Magnite, or both?
Yeah, it kind of begets one begets the other, right? You win big accounts like a Pinterest that excites advertisers to programmatic advertisers to run through Magnite because it's the only way they can get the inventory. More dollars run through it, the next big publisher, you know, a Spotify or a RE/MAX sees that, and so it's just kind of cycle that we've been through. We've been beneficiaries of this supply path optimization, where the big agencies, the holding companies, all which are here today, they basically wanna work with fewer vendors and wanna have more strategic relationships. And so with every one of the holding companies, we have that kind of preferred relationship, which we're not talking about a modest shift to spend, we're talking about hundreds of millions of dollars in shift.
So if all of a sudden, you have that dollar, if you have that, spending power, that big density on your platform, it just begets more supply. So it's this kind of, virtuous cycle that, we've been benefiting from over the last 24 months.
Got it. Very, very clear. Yeah. So it sounds to me like the industry structure is consolidating a little bit. The long tail is, you know-
At the
Kind of coming on.
Correct. Yeah. Yeah, there's no question. If you look at our growth rate and the growth rate of the industry, that we're taking share.
Yep.
folks who said, "Oh, it's already happening and taking share from Google," but that hasn't budged. Google share is still at 60%, we're still at 6%, so the share is coming from the longer tail, as you pointed out.
So let me shift now to CTV, where I think a big, a really big part of the thesis here, the long-term growth thesis is, although certainly your comments on open web are very positive. So I think on the last call, you continued to mention that CTV ad spend still grows faster than revenue. Okay? But you're seeing the gap narrow. So I wanna maybe think out a little bit further, a couple of years. Do you see a world where CTV revenue growth could greatly exceed CTV ad spend?
I could see a world where it would be on par. I don't know if I wanna see a world where our growth rate in ad spend is flattening and we're making it back up on margin. I think that what we would look for in an ideal scenario is comparative growth rate, that if ad spend is growing at 15%, ad spend is growing at 15%, so that it's equalized in that respect.
Okay. And correct me if I'm wrong, but a big part of the revenue growth would be upselling, right? It's moving to higher take rate services and functionalities. So, I'd like to dig into that a little bit. You know, this leveling off or potentially slight outgrowth, you know, of revenue versus ad spend, there are endogenous and exogenous factors that, you know, would compel your growth. And I'd like to dig into those a little bit. For example, these would be things like sales motion and innovation. So could you describe to us a little bit the customer upsell process, you know, both from your perspective as well as the customer's perspective?
Yeah, sure. So, we're fortunate enough to have relationships with just about everyone other than YouTube and this space, we're their in most cases their primary or if not exclusive programmatic partner. And you have to think that the programmatic universe isn't kind of as homogeneous as the open web. You have your linear broadcast, legacy broadcast guys with streaming services. They have a different kind of complexion in terms of how they go to market, what they prioritize. And then you have some of the digital first, more of the OEM guys, the Roku, Samsungs, LG, Vizios. People I don't think appreciate how much inventory that they have that they've brought to market.
And then, of course, you have Amazon and Netflix that are kind of digital first, but much more of a consumer-oriented streaming service. With the big streamers, the Disneys, the Paramounts, the Warner Bros., they very much early days of programmatic, and they probably think it's early days right now, wanna emulate the sales strategy that they have deployed for the last 50 years. And that is, they have a very talented team of salespeople. They have deep relationships with the top 500 leading national advertisers, and they wanna be able to pitch their dwindling linear, along with the growing streaming, to be able to yield maximize across the entire Disney portfolio, not necessarily top tick on this impression, on this show, on this programmatic channel.
Where do we get involved in that in terms of helping them through that life cycle? Well, first off, the initial stage is enabling programmatic, so plumbing. We're a plumber. We come in, we make it happen so that if they wanna do programmatic and they wanna do it, publisher sold, they are able to talk to their agency counterpart, set the price, set the targeting parameters, and then we execute for them. And that's largely where we find ourselves with those, that cohort right now. But if you look at some of them, like for instance, you know, Tubi at Fox, Pluto at Paramount, even Hulu at Disney, kind of different brands than the mothership, very much digital first and very fast-growing. So much more open to, "Hey, what can programmatic do for me more than me trying to direct sell?
Biddable." And they're like, "No, we're not doing the open Wild West like the web. I'm not, I'm not having some schlocky ad run in my high-produced content." So generally speaking, the first step is invitation only biddable, so invitation only auction. Those 80 advertisers, I'm comfortable with those 80. Let's see if bidding real time on impressions yields greater than me selling at $15, because that's historically what I sold it for. And lo and behold, what they're saying is, "Yes, it does." That there are some $70 impressions in there because that buyer knows that that person's in market for a European luxury automobile, so they're gonna bid higher. And then there are some that are below $15, but when the dust settles, there's better return on ARPU.
Getting customers comfortable to be able to break what has been a fifty-year cycle of selling as much as I possibly can in the Upfront, and then worrying about the leftovers. Now, holding back from the Upfront deliberately to yield maximize downstream, that's a journey. I think our sales team does a wonderful job with data, statistics, with the help of the DSP community, because they wanna do that as well, trying to educate them on what the right yield management is in terms of inventory management. Contrast that to the digital first OEM guys, they're very big and biddable. Our role there is, you know, as their principal programmatic partner, is to make sure that we bring as much demand as possible for them.
Okay, very interesting, so let me make sure I think I understood that correctly. So, you know, there's a risk trade-off, essentially, in the mind of the publisher. You know, we're-- they're used to doing things a certain way. They're used to having, you know, a certain price knowing beforehand, some kind of a guarantee. You know, yet they still reserve some, and they're experimenting, essentially.
Correct.
You're enabling that experimentation, and then there's also a consultative process where, you know, you're kind of trying to move them along into more programmatic direction.
Yeah, that's exactly right.
Got it. And, you know, is there, I guess, any way that, you know, how would we, how would we, think about the pace of that evolution of thought? You know, maybe any particular publishers that, you know, you would consider leaders or laggards if you don't want to mention them? Right, 'cause I, as a sell-side analyst, I want to, you know, try to track the pace of movement here to try to figure out, like, eventually, like, how long it's gonna be and eventually what kind of growth rate we can see.
Yeah, it's a great question, and without getting too much into specifics on individual media owners, you definitely see more experimentation when they have some flexibility on the brand. So in other words, I cited before Disney's Hulu, which isn't exactly equal to Pluto or Tubi, being fast channels and Hulu being subscription, but certainly not. It certainly doesn't have the legacy media brand associated with it, which I think gives them, from a go-to-market, more flexibility to say, "Okay, you wanna do that? You can do that here." And they can learn from that, and they can see, oh, my lord, we used to sell it for $15, now we're getting $17.50 for it if we do two-thirds billable, one-third. And so I think the evolution is in place.
You saw programmatic play a huge role in the upfront this year. Every upfront presentation, programmatic was front and center 'cause that's what the buyers want. The buyers will accelerate this as well, and that is we're really just talking in this conversation, the traditional broadcast buyer. What's super exciting is there's a world of advertisers that have never been able to advertise on linear, on broadcast or cable because of the barriers to entry. The spots were too expensive. The creating creative that could be accepted by the big broadcasters, way too prohibitive. And so now you see with the utilization of technology, AI tools for creative, for tracking, for measurement, for attribution, you're starting to see the growth of DSPs like MNTN, tvScientific, bringing social advertisers to TV.
And that's really where the rubber hits the road because they only are biddable. They don't know how else to do it. That's their playbook from Instagram, so they want to take it over and apply it to TV. And because of the excess supply of inventory out there right now, the price points are low enough for those performance advertisers to be able to have a good experience. And so I think one of the stories of the next several years is gonna be moving from an arena of 500 national advertisers to ten thousand advertisers, many of them SMBs that are advertising on TV for the first time.
So I come back once again to, you know, to sales. You know, I guess, how is Magnite kind of staffed for that opportunity? Like, yeah, I think, when we look at your results, I think that the margin guidance you gave implied that you were gonna be investing, you know, in the second half. You know, how do you think about the sizing of your sales force and, you know, basically to kind of potentially push this opportunity and accelerate growth for you?
Yeah, I think we are fully leveraged on our sales team, and that it's not a question of adding more bodies. In rare instances, we recently opened an office in India, and we opened an office in Norway. Those are kind of rare examples because we're very global as it is, so those are just some spots where we wanted boots on the ground. But generally speaking, we have a buy-side team, we call it demand facilitation, that talks to agencies, that talks to marketers that have in-house their programmatic. They talk to mid-tier agencies in kind of the flyover states, and they talk to the DSPs. And so we are able to bring that demand with the current sales force that we have.
That world that we're talking about in three years from now, that 10,000 of advertisers, they're going to be aggregated by DSPs. They're gonna be aggregated by merchant aggregators, and so it's our job to make sure the plumbing is there for them, that it works, that they're getting the right signals. We are not going to deploy an army of people chasing 10,000 advertisers. As it relates to increased investment in the second half of the year, and we really haven't signaled a number, but that relates more to infrastructure, pulling forward infrastructure investments because of what we are seeing when we are able to take some of the traffic off the cloud and put it on on-prem.
So if we know it works, and we know we have budgeted for 2025x amount, why wouldn't we pull over some that if that means the payoff is that much quicker for us?
Got it. And, you know, so there are two sides to this market, the buy side and the supply side. I think you ran through a little bit of both. I still would like to better understand, you know, the connection, like, you know, you don't actually make any money off of the, actually I shouldn't say it that way, but you don't-
It's true. They don't pay us.
Right.
On the buy, I have a lot of people calling in, people that don't pay us.
Right.
That's right. But I guess
Good business model? No, I'm trying to get at, you know, you invest where you invest against which customers, and you consider both sides of the market your customers to some sense, right?
100%.
Right.
Yeah. You have to... In order to create this marketplace for our publishers to get the, our job is to make sure our publisher has access to all the demand that's out there.
Right.
And if we're not doing our job, people aren't going to just be in path to Magnite. And so we make sure that when our publisher lights up, the world's demand is bidding on their inventory.
Got it. So the investment you make in, you know, facilitating demand, you know, helps increase the value of the SSP to the publisher.
100%, and increases our new wins with new clients because they look at it and say: "Wow, all that spend is..." You know, they see public company, how much spend goes through our pipes on any given year, like, wow, that's three times more than anyone else in your space. We can't not be there.
Got it. Well, another driver of growth is innovation, and maybe we could switch to that. So I noticed we had a quick chat in the hallway here, and I was asking you about the general availability of SpringServe, and you branded, rebranded it, it sounded like. And I thought that was an interesting anecdote about you know, kind of customer perceptions in the sales process, but touches a little bit on innovation as well. You know, why did you guys decide to do that?
Yeah, there were a number of reasons, and just to back up a bit, SpringServe is our ad server in connected television, so it's not a general ad server. It doesn't compete with GAM, it competes with FreeWheel. So it just does streaming, high-value video, and when we first brought it to market, we were like: "Hey, we have an ad server, and we have an SSP. If you don't need an ad server, just pick this. If you need an ad server, pick that. If you need an ad server and an SSP, pick them both," and it started to confuse people because a lot of the capabilities in the ad server were also some of the capabilities that we had in the streaming platform.
So what we said was, "Why don't we just bring it together, call it all SpringServe? So it's an SSP with ad serving capabilities, and if you don't want it as a primary ad server, just don't flip that button. Just use it as an SSP." But what we found is, generally speaking, people use elements of the ad server even if they have an ad server, because of what it can do, how it can increase monetization. And so it just made all the sense in the world to bring it together. It's also a huge competitive advantage. As I cited, FreeWheel is our largest competitor in that market, but if you look at any of the other SSPs, they don't have ad-serving capabilities.
And so by incorporating that into the SSP, it kind of cuts the legs out from underneath them because now they have SSPs that are... do not have the feature set that we do, and we don't have to argue about whether it's ad serving or not any longer.
That's really interesting. You're putting a more powerful tool in the hands of your customers, whether or not they use the entire functionality-
Exactly.
It sounds like.
Yeah.
which, probably puts you in a better competitive position because they can-
Very unique and competitive position.
Yeah.
But by going out there and selling as such, it really makes them ask harder questions of the competition as to why they would use them if they don't have X, Y, Z feature.
Very interesting. So maybe another one that I heard you talk about recently is, you know, some of your investments in artificial intelligence. And I'd like to maybe open a discussion about how these new AI capabilities and basically make your products more sticky. You know, that's essentially a very huge theme. Could you walk us through your thinking there?
Yeah, I think, and there's a slew more to come, but the ones that we've kind of centered on, kind of all fall in the same kind of neighborhood, and that is, making sense of this vast amount of traffic that we have on the platform, usually through the lens of, hey, I'm looking for a mom of young kids, help me do that. And so AI tools, agentic tools, are able to spider the tens of thousands of sites that we have on the platform and be able to look for context. Oh, there's an article about parenting. That's a good environment to put an ad in or a signal from that publisher that says, "Hey, it's a mom of young kids," 'cause we're a registered site, and we know that profile.
And being able to assemble that audience segment and put it in front of the buyer, as fast as we can, leads to a better buying experience, leads to better monetization for our disparate publishers that couldn't sell that segment on their own-
Mm-hmm.
until we brought it together as one segment across 10,000 publishers, and we participate in the economics of an enhanced CPM, if not an outright economics of a data deal that if we imported, like, ACR data and put it on it. Generally speaking, the agentic work that we're doing right now is in; we call it curation. It's in building audience segments, discovering audiences for buyers to help them find it, buy more from us, and we become that trusted source of where they want to buy from.
Does this apply to both CTV and open web?
Yeah. Last quarter alone, we added 50 curators, which we would have never... These are folks that specialize in doing that, and then they go out to an advertiser and say, "Hey, would you like that?" and we actually process the transaction and participate in those economics. We couldn't have. It wouldn't have been possible to do that on just one of the two platforms. It's across both.
Okay, and when you say you added fifty curators, I'm not sure I follow, what, because I just don't know the industry as-
Yeah. It's a funny term. Maybe think of it as like this. Think of them as old school ad networks, where they used to have to go door to door to sign up a publisher to accept this special ad unit. Like, our ad unit is this, it's non-standard, and it dances and jiggles, and then publisher's like: Okay, sure, I'll sign up to be part of your network. Then, they'd have to go to the agencies and say, "Hey, I got this great network of this great ad. Would you like to buy it?" And that's how it used to work. Today, that same person that has that special creative will just go to Magnite, put it in the marketplace. Publishers will say, "Yeah, it looks pretty cool." Click, click, click, click, click.
They'll sign up a thousand publishers, and then we'll merchandise it to the buy side, and they're the technology piece to it. They made it happen because they came with the innovative technology, sound, motion, whatever it was.
Mm-hmm.
And they don't have to hire a sales team for the supply. They don't have to hire a sales team for the demand, because all the agencies are saying, "Hey, hey, hey, great idea. Just go over to Magnite, and we'll buy it off of Magnite," because they've collapsed our partners to two partners. So that's a big growing business for us, which is nice, but even more, I think, impactful is that all that used to take place in the DSP, right? That's what the DSP used to do in a third-party cookie world, assemble your audiences in your DSP and then buy those audiences from an SSP. Now, you're buying the audience from an SSP and assembling the audience on the SSP, which we've been talking about for the last 24 months.
That is a trend that is significant for Magnite and significant for the SSPs, because we were always, 24 months ago, we were the dumb pipes, and the DSP was the brains, and we are commodities, and I think little by little, what you're seeing is: Wow, those guys are a lot more strategic than I thought, and I think you're starting to see that reflected in the energy around the story.
So, when I was listening to you now, I think, you know, I heard you talk mostly about creatives, and what these curators create is creatives, but why does that make them curators? Because when I hear the word curator, help me if I'm thinking about it incorrectly. It sounds to me like curation means, you know, you're looking at the audience and assembling audiences on the supply side.
Correct.
So that was the part that I missed.
Yeah, and some of them are technology companies that allow for, they're your video player, so they come in, and they don't, "Don't pay me for the video player, give me a slug of your inventory." Where the curation comes in is, you have this cool ad unit. It's across 10,000 publishers now. They've adopted it. But I don't want to just buy a cool ad unit. I want to buy a cool ad unit that reaches moms with kids. So then the curation occurs on top of it that says-
That's a collaborative process-
Correct.
with the curator and the publisher.
Yeah, they use our tools, exactly, to then slice it in and say, "Okay, for $100,000, I got you all these moms," and-
Very interesting.
Yeah.
Okay, that seems pretty powerful, and you know, you said you signed up fifty.
Yeah.
Right? In a very short period of time.
Yeah.
So, how long is this ramp eventually?
You know, there's always gonna be a diminishing scale. I mean, our goal is to sign up a thousand and have, you know, five hundred of them not delivering revenue, 'cause we're literally getting people into business. We're putting them into business. So, we'll obviously focus on the folks that have real demand behind them. Like, hey, we proved this. The agency wants it. If enough publishers adopt it, they will spend on your platform. So we're prioritizing it based upon how much spend we feel they can bring to Magnite. So the number isn't finite, but it's not infinite either.
So, you know, we've covered a lot of things here, and it seems to me that there's a pretty strong trend of kind of value converging on the supply side and specifically on Magnite. And, you know, when a solution becomes industry standard, and it becomes more sticky, like it seems like you're becoming, you know, you would think that at some point you would have more pricing power, and I wonder, like, how that potentially looks. Is it purely like kind of an upsell mixed process? Or, you know, what elements of pricing power does Magnite have once it becomes such an entrenched player?
Yeah, so, great question, and I think that, if you look at the two businesses, again, the DV+ business, very mature, legacy Rubicon, started in 2008, public in 2012 or 2013. So it's, I think we celebrate in that business stable take rates. The idea is that our value has been earned over the years, and that we don't go through this savage cycle every year where people are trying to pound down the take rates. Like, it is, it's a given.
Okay.
So that to me is great because with all the spend that we're adding on to it, it's just once you get to a certain point, it's ripping at 90% margin.
Right.
CTV, it's a story of upsell and not necessarily pricing power, like just turning to Paramount and say: Hey, your 3% is now 8%, take it or leave it. It is Paramount. We've talked about the journey. You're selling a lot of stuff programmatically, direct. Why don't you try biddable auction? We'll run it for you. We'll only invite the same advertisers you want. For that service, we charge you X. So I think it's more walking them up the programmatic food chain so that this, at the ultimate level, the service that they'll be using from us is significantly higher take rate than the service they're using today.
You know, do take rates ever change at all? Right, I mean, like you said, at one time in the past, I guess people tried to pound them down, but, you know, is there a potential that they would ever go up, even if for like a small, minuscule amount, or is the goal generally just to kind of, you know, you know, set a certain service level at a certain take rate and leave it there so it becomes industry standard and easy for people to transact on?
I think for our strategic relationships, we're very pleased with where we are today, and we're not looking to, through the marketplace power of Magnite, walk it up much higher. You know, a lot of publishers in our world are hurt. We don't want-
Yeah.
to be the guy to put the other nail in the coffin, if you will. But when we aren't used as a strategic partner and we're just thrown in the mix as just another SSP, you know, the traditional header bidding, web display, browser-based, I'm always gonna keep eight guys in there. Magnite, I don't care what you say to me, you're just another SSP. I'm gonna run you in competition in a unified auction against everyone else. That's when we say to them, "Well, there's no take rate then." Because if we win the auction, what do you really care if I'm taking 50% or not? It's a unified auction, and so we'll monitor the take rate on our side with a variable take rate.
Mm-hmm.
If it's an auction with low liquidity, we can take a higher take rate. If it's an auction with high demand, we'll take a lower take rate, but we'll vary. And you'd be surprised at how many people are indifferent to take rates in that world. They're just like, "You're right. It's a unified auction. You win, you win. I wish I had the 50% take rate you took, but you still won, and so it's the highest price that I could possibly get. So the marketplace is telling me, take that offer.
Interesting. Okay, well, we have three minutes left here. Maybe I'll just ask, the-
Yeah.
Last question here. You know, Netflix has been, you know, doing programmatic now for a couple of months, I think.
Yes.
Any early learnings that you have, you know, from your partnership there, that you can let us know about?
Yeah. So we're very careful when we talk about specific partners, and particularly in Netflix. So we always kind of say it's their story to tell. But parroting what they've already said is, I think, fair game. And, you know, it took a while for them to move off of the previous partner, Microsoft, onto our stack and their stack because they built their own ad server. Focused on North America first, then went to EMEA, and now in APAC. And it's been a tremendous relationship. They've hired some amazing talent on the technical side. We work extremely well and close to them. We're onboarding more supply, more demand partners in international markets that aren't necessarily just the same big guys that are in North America. And I think that they're new to programmatic, too.
Their ad business is, you know, two-plus years old, but was never programmatic. They didn't do any. It was all direct, and so they're kind of new to the programmatic rhythms and are being very careful about how they proceed, the consumer experience, all that kind of stuff. But, as we've said, based upon what we're seeing, they're gonna exit the year as our largest client, if not one of our largest clients.
On a run rate basis?
On a run rate basis, correct.
Okay, well, look, a lot to look forward to. You know, great potentially structural opportunity here. So thanks again, Michael, for coming-
Appreciate it.
And I really enjoyed the interesting conversation.
Outstanding. Thanks, man.
All right.