Magnite, Inc. (MGNI) September 9, 2026 Conference Transcript

Okay, all right, this is Omar Dusoki, I'm a senior analyst at Bank of America covering U.S. Internet, focusing on advertising technology and video games and consumer subscriptions. Some of the stocks that I cover in the ad tech space are Magnite, who's CEO we have today, Michael Barrett, Applovin, Unity, Double Verify, and a couple of others. So, Magnite, we've been following Magnite for a long time. Well, actually, sorry, since 2024, we turned bullish, you know, sort of late 24. And we've been bullish ever since. And it seems like much of the thesis that we had expected has just simply played out. So, one thing, for example, is CTV, they have two businesses, is the CTV business and the DV Plus business, which is focused on open web. You know, we've always been looking at CTV, you know, as your business, as a business that could outgrow the CTV market in general. You know, which seems to be, CTV advertising seems to be growing, I think, kind of like mid to high teens, expected long term. Yet you grew 36% year on year in the third quarter. your top 10 advertisers grew 40% year-on-year?

Yes.

So that's even faster than the rest of your business, probably because of upsells. Your upsell process is working.

Yeah, no question. I mean, one of the benefits we do, or most of our clients are global in nature, and so when they expand to different countries, we get to go along for the ride and help them monetize their inventory in those areas. So you find that it's kind of a land and expand with the big media owners.

Got it. And I think what I'm most interested in is you talked about the potential for kind of 25% annualized growth long term. And I was hoping to double click into that. So are you saying that it's going to be sort of like a 25% caker or a minimum 25% year on year growth, what you would look for, even with years where there are tough comps?

Yeah, no, I think it's very fair. Again, it's a longer-term goal, but I do think that most of the industry experts peg the growth rate of streaming, the ad-supported streamers, in the mid to high teens. And we have every expectation, given the breadth of the relationships that we have in streaming, that we'd outpace the industry growth. And so I think if you look at a 25% growth rate compared to, like, our 36% versus low to mid-teens growth rate for the industry, we feel very comfortable pegging that as a not just aspirational but something that we feel is very likely to occur. Okay, got it.

So this year, I think, you know, there's potentially like a big political cycle coming up. And, you know, to us, I think your guidance looks kind of a little bit conservative. I just still want to kind of get back to that question. So, you know, even in the context of, like, a big year, which, for example, could be this year, you know, depending on how political does and other factors where you're doing well, you know, is that sort of – are we supposed to look at that 25% as sort of more of a floor or like a kegger?

Yeah, I think a kegger. Yeah, I mean, you mentioned political. Political is very interesting for us. We're new at it. We're not scripts, right? And so we have two cycles to go off of, a national election four years ago and a midterm most recently. And the challenge is that it was so new, streaming was so new as a media mix for political advertisers that it's a little unknown what to expect this year. To date, we've been pleasantly surprised with the amount of spend, but we know it's heavily backloaded and it really does come down to are the competitive races, are the races in the right market, et cetera.

And so I do absolutely think we'll be talking in 2027 about ex-political from a comp standpoint, But I think we feel very good about where we pegged it right now and and how how penetrated are your services you know in the CTV camp at this point they can the reason I ask that is because you know there's this push in this poll where you know you're more seasoned advertisers or see you more seasoned publishers you you know have up sold them and that's one of the reasons why for example your top ten accounts are growing faster than the rest so you know you've talked about you know, positive kind of pricing mix shifts very recently. And the 40% growth on your biggest publishers, you know, that's higher than the overall 36% growth. So I assume you're still penetrating the market, right? And the newly onboarded publishers are going to start off with lower take rate, you know, type services, which, you know, eventually will go up. So there's probably some latent pricing mix, you know, in your model, which I think analysts have to think about, you know, further off, eventually the take rate will improve, right, while the top line may – sorry, with total ad spend may, you know, may moderate.

But how do we – Yeah, you'll get – it'll be closer. So to date – and, again, our take rate is a contractual agreement with publishers, so it's not as if – if we talk about scenarios by which there's lower take rates or higher take rates, sometimes people immediately go to oh they're under pressure the market is forcing a collapse and take rates what it is is it's a product mix um so generally speaking when someone that has a legacy advertising business gets into streaming and hires a magnate um they're very much inclined to kind of go at it on a walking pace and say hey we've been selling advertising our you know for 100 years we're going to continue to sell the exact same way we sell linear we're going to do the same with streaming because we want complete control we've always had complete control and we just want you magnate to operate as our plumber so we'll go find the advertiser we'll negotiate the price and then we'll just execute it automated so programmatically and so that carries with it a lower take rate. It's a product that we offer. It's publisher sold programmatic. And not surprisingly, every large media company starts off that way because they're a little concerned, you know, is the programmatic ads going to ruin the consumer experience? I don't know that advertiser. Is it going to be cheaper CPMs? And so it's just this journey that most folks have to go along till they get comfort.

And of course, at the same time, the buyers coming at the media owner saying go programmatic or you're gonna lose dollars and so that accelerates the uptake of the service by which we bring the demand of which we get compensated higher so this is super interesting because I remember speaking with you about a year ago after can the can conference you know and I asked you you know it's a mega bull case here just simply you know the entire industry moving to bit of a programmatic and I recall your response which you know it's not necessarily the Holy Grail, right? Now, recently you did kind of cite that there was a shift towards biddable programmatic from, I think, programmatic guaranteed. So, you know, is that still how you think about it? Or like, you know, is adoption picking up and can you kind of see a path to biddable programmatic, you know, this very high value added service, you know, being, you know, the, I don't know, the majority or eventually the entire market? Has any of your thinking changed on that?

No, I think that if I were to, I'm going to say definitely I think biddable programmatic is where the whole market's moving to. I think sometimes when we say biddable programmatic, we think of the open web, which is trillions of ad requests, billions of impressions going to auction, and random advertisers winning and showing up, but no one really cares because it's ephemeral, it's a banner, what's the big deal, where there's a big difference between a 30-second spot running and a million-dollar produced streaming series. So I think what you're going to see as the majority for the top tier streamers is invite-only auctions. So it'll still be biddable. It'll still be an auction. But they're going to want to know who those advertisers are. They still have two floors at Disney of people who review advertising creative. So it's not for the faint of heart to be able to get an ad on TV and have it approved. So we've done this automated through technology that allows them to feel comfortable about the ad and the quality and whether it meets standards to appear on their shows. So I think invite only, and that's for what we think of as the centerpiece major streamers. As you go further afield and you talk about device manufacturers like Roku, LG, Vizio, Samsung, they have tremendous amounts of inventory. Even the Disneys have inventory in fast channels that they don't consider the same as something like Handmaid's Tale or whatever the case might be. So when you deal with that type of inventory, that will be much more democratic, much more open to biddable. And that's where you see the world of the small to medium-sized advertiser that is the bedrock of Instagram, the bedrock of Alphabet, start to come on to the CTV environment.

And I should have been more clear. There's like open auction biddable and then there's invite-only biddable. Just clarified. And when you mentioned to me last year that it's not the Holy Grail, you were talking about open auction program. Okay, so that would be, you know, a very interesting, you know, world we would live in for Magnite if that were the case. And who knows? Maybe it'll kind of happen someday. So then let's maybe move on to your margin profile, which has just basically continued to get better. You know, part of it I think we already alluded to was kind of the mix shift to higher value services, right? Right. You know, so I think you had talked about potential for the opportunity. You always talk about 30, 35 to 40 percent.

Correct. And the potential to exceed 40 percent at some point, you know.

Yeah, that's certainly not a cap. Yeah. Right. So is that mainly a function of all the businesses you have in place today, that potential to exceed that 40 percent? Or is it also or is it mainly a function of what's going on with DOJ and DBplus?

Yeah, no, all those statements about the aspirational margin profile of the company have been stated long before any DOJ, Google investigation. So we feel, listen, seasonally, almost every fourth quarter, we exceed 40% in adjusted EBITDA. So it's really a question of just revenue growth on top of a pretty consistent cost model. And we're our biggest cost, like most technology companies, is people. And I think we've done a very judicious growth in that area. We haven't outpaced what we needed. We'll always add on, you know, 50 or so bodies a year. but in order for us to drop another billion onto the platform, that doesn't require another 1,000 people added to the company. So we've had a very stable profile of employees. And then the other cost is the servant cost, the cloud cost. Our volumes continue to surpass every expectation that we had going into budgeting, But our teams have been doing a great job of processing three times the amount of impressions at the same cost that we did the previous year. So in order to do that, we pulled forward some capex into last year, fourth quarter, to build out another data center, taking as much traffic as we can off the cloud onto our own boxes and modulating costs that way. So I think keeping our cost structure where it's at and just dropping more revenue on top of it naturally just gets you a forehandle in terms of margin profile.

Did the potential to have market share gains in DVplus factor into your decision to make that additional investment to move away from the cloud and build out more of your on-prem?

In the back of our minds, most of the investment that we're doing in taking cloud off is almost all streaming. So streaming is a 100% cloud business that we are now doing a hybrid model to. There are certain aspects from real-time reporting that cloud is absolutely essential, but there are other low-value things that we're doing in the cloud that can easily be done on boxes at a much less expense. Now, to answer your question with the DOJ in the recent release of the opinion or the closed opinion, there's no question that those boxes will come in handy for that. If there is any increased volume of bidding from a Google that we weren't seeing previous, we're going to have to absorb that capacity, and those boxes will come in handy for that for sure. And then that's the whole idea of being able to burst to the cloud. So if all of a sudden you have this profitable traffic coming in that's crimping your bandwidth on your own proprietary boxes, you can jump to the cloud and buy bandwidth there.

Got it. And, right, so I think if we just move really quickly, I do want to talk about mobile on DBplus, but since we started talking about the DOJ and the potential outcome there. So first of all, you know, there is potentially a redacted version of the ruling coming out at some point.

Yeah, so the ruling came out, the opinion came out this week. It was closed, sealed only for DOJ and Google to look at. They have a 14-day window where they can redact any sensitive. They obviously can't redact the remedies, but they can redact any sensitive information that was used in the findings. That probably then leads to another 14-day journey with the judge looking at it and saying, you're a little heavy with the black pen there, open that up, open that up, and then maybe in a 30-day window, it gets posted and everyone knows what the remedies are. We know that the remedies aren't structural. We've known that for some time. The judge signaled that that wasn't going to be on the table. And the behavioral remedies are real. And if she said that most of them are going to be adopted, and in that case it will make a more fair and efficient auction at AdEx. It will make a more fair and open ad server from Google, their DFP. and all that should benefit the ecosystem, including Magnet.

So I think this is really interesting because, you know, we've, like I said, you know, we at B of A have been bullish on your CTV business for a while now, and that's playing out. The DV plus business, you know, has kind of been a laggard, you know, more growing somewhat inconsistently. And I think, you know, probably on average, like mid single digits or so, you know, Open web advertising is not known as a growth business. So, you know, for us, we're interested in figuring out a couple of things. You know, obviously, you're not going to tell us all the answers, but let's talk about it a little bit. So there's a margin impact, and there's also a potential for, you know, this entire industry to become a growth industry again, okay, because of additional competition. You know, similar to, I think, what happened, you know, in mobile when a lot of these kind of mobile ad tech players started innovating and improving their models to drive growth for mobile game developers. So I think, you know, first of all, about, maybe about a year ago, I remember putting out a piece where we sized the, sort of the ad spend, you know, through the SSP market, you know, on open web around 25 billion. And this is all, you know, from research from 2025. And we sort of estimated that a 1% market share gain, you know, which would be like, you know, 250 million, I guess, you know, could result in a really significant increase, you know, in net revenue. Like, you know, I think would increase EBITDA by 67%, right? So can you just walk us through your thinking on what that sensitivity in terms of market share gain, if you have kind of those numbers handy in your mind?

Sure. So it was through the discovery process. It was estimated that Magnite had the second largest share of programmatic open web spend that Google's share was close to 60 percent and ours was 6 percent. And we were by far and away the largest of the – we're the second. Everyone else after that, way long. And we estimated that for every one market share gain, so we eat into Google's 60%, that could represent $50 million in margin to the company. And so I think that it's safe to say that we're not going to lose share post-remedy. I think it's safe to say that we'd win it based upon the share that we already enjoy in the market, so we'd be the outsized beneficiary of the ruling, given our second-place status. You know, we have deep relationships. One of the things we've always invested in, even in the early days when we were Rubicon and became Magnite, was our publisher sales team. It's incredibly well-respected, incredibly knowledgeable, and a lot of our peers have cut down their sales team in that area because in the open web, as you know, it's kind of commoditized. It's this header bidding. You're always going to get slotted in the head, so why do you have to have a deep relationship with the publisher anyway? But we've always maintained those deep relationships, largely because of our streaming business. Also, long story short, I think one of the biggest unlocks is going to be if they will be allowed. I can't imagine the judge wouldn't rule in this capacity right now your data is locked within the ad server the Google ad server you can't bring it anywhere and they desperately want to work with us to do their private marketplace deals they don't want to have to put them in attics they don't want to have to keep them in GAM but they can't it's just such a heavy lift they've even tried it and they've tried to cut and paste and it's just a mess so we think that magnet could really be an outsized beneficiary there because we have deep publisher relationships. They want to do more business with us. It's probably going to be at better terms than they're doing currently, and we're, you know, sized to catch that business.

Yeah, there are several levers, I think, that could potentially help your business. I mean, the one that I had pointed out a year ago was just, you know, simply improving your win rate because your infrastructure is already in place, you know, so I don't think there's much else to spend, but simply, you know, having a higher win rate because Google doesn't get first look or last look, you know, would flow directly through, like, any kind of revenue there would flow directly to EBITDA, and you'd have 100%. That would be the fastest one, yeah, because we don't have to process one more impression. We take $2 trillion ad requests a day.

It's simply a huge efficiency-like increase. You're doing the exact same thing you were doing. You're just winning more. You're taking $500 billion to auction. You're auctioning $500 billion a day. Your win rate goes up at half a percent. It just drops to the bottom line.

So that's just one lever, though.

Yeah, correct. Yeah, that's right.

That's kind of what seems to be like the lowest hanging fruit. But I think the one that I'm most interested in in terms of potentially that the market may reprice the DBplus business you know, is, you know, can innovation, you know, kind of re-accelerate growth, you know, and help the open web, you know, grow again? Because if indeed, you know, Google was anti-competitive, then, you know, it would, I would understand that they were probably underserving their customers, the customers didn't get as much revenue, and they couldn't invest in content, you know, and that virtuous cycle that we've seen in mobile gaming recently could potentially happen in the open web. What are your thoughts on that topic, you know, in that breaking, you know, the monopoly, you know, could actually re-accelerate and rejuvenate this open web ecosystem?

Well, yeah, a couple of things. Yeah, a couple of things there. So, just for the sake of everyone, DV Plus is kind of this portfolio of non-streaming media, right? And so, it's open web browser-based traffic, which has, as we know, significant headwinds from these generative answer engines that aren't sending search referral any longer. It was a time when, at the height of it, Google would scour your pages twice to feed their search engine but send you one referral for every two times. A pretty fair balance. I think Entropic now is 900 times they scour your site and they only send one referral. So that's not going to change. Nothing the DOJ does, nothing. that is just consumer behavior changing and the balance between sharing content and getting referral in back is broken. Our business, that's a portion of our DBplus business. As you pointed out, in that bucket is mobile app, which isn't browser-based, which isn't search referral-based. So in-app, digital out-of-home, that isn't browser-based. Audio, which is a big growing piece of our business, podcasts, et cetera. None of that has anything to do with the open web. So it's a pretty healthy, diversified bucket. So with the DOJ ruling and changing growth profile of the open web, perhaps, it would change what you would think about as DVplus as a grower because we've pretty much told everyone, think of it as a flattish, which a lot of people were elated about because they were afraid it might be going down 30 percent so now all of a sudden this could be a growth profile it would change a lot of the dynamics of the company as this being a savior for the open lab as I cited before those terrifying numbers about lack of referral I personally would love to think that it could help but I'm not so sure if what we're talking about here is redistributing where the money current money goes so in other words the current money all sits with Google if we reshuffle that current money and don't expand that money the publishers kind of left in the same plate that they are today in terms of the amount of revenue they're gonna get their partners their vendors might have a different revenue profile but I'm not certain I've seen anything that changes the dynamics of the open web from that particular role. Okay.

Now, is, for example, open web potentially the structure of open web changing? Obviously, it already sort of has changed because of the, you know, how the LLM's direct traffic, you know, but is potentially the sort of end state of that structure, does that change the market dynamics in favor of you? For example, you know, in CTV, I think we pointed out that, you know, whereas open web has traditionally been fragmented, CTV is highly concentrated, you know, therefore the market power in our view is like more on the supply side, you know, which helps for you as a supply side partner. So would you potentially see something like that, you know, happening in the open web where it becomes more CTV-like, there's more concentration among publishers, you know, which I think would ultimately sort of be better for your distribution model?

Yeah, I think what you're seeing is a real bifurcation in terms of the type of media sites that are doing okay on the open web. And the ones that are doing okay are, generally speaking, destination sites in other than themselves. Like New York Times. So more well-known, more of a logged-in user base, bookmarked user base. and those are the types of publishers that we thrive on that we've always historically worked with. So I think you're going to see a shrinking of the longer tail and a consolidation among more premium-type publishers in the open web.

Are you already engaged with all of those?

We have deep relationships for decades with those players. That's the core profile of a magnate publisher. thus requiring the need for an expensive and super talented sales team to be able to work with them as their partner to increase their monetization.

And I guess just getting back to the technology question, I think, for example, the agentic stack that you have recently commercialized is an example of stuff that's going on. But once again, just thinking about, and we'll move to mobile after this, but just thinking about, you know, how this sort of unlock, you know, where there can be competition again will spur innovation. What are some, you know, some of those innovations that, like, could have occurred in the past but haven't that you maybe foresee over the next, like, three years, let's say, that could really make open web and mobile a better place to be for publishers and specifics?

Yeah, no, so I think that discoverability will play a huge role in it, and so I think that when you're dealing with the current world that exists and you're a buyer and you're hopping through 18 different dashboards to try to find the profile that you're looking for, the, you know, the young family with kids under the age of four that buy diapers at a big box store, it's sometimes not that easy. You're going to your DSP, you're bringing data, they're bringing data, each publisher has a slightly different definition, you're going to an SSP to get that, now you think about an agentic world where you can just natural language put that in to the interface of your agent that already has your algorithm in it that already has your data your first party data in it and its ability to go through something like a magnet and communicate with 10,000 seller agents to be able to pull all that together and make it easier to be able to target make it more efficient make it more working media going towards that, I think Argentic really is a huge unlock in terms of the ease of use that a trader, what the trader has to go through today and what a seller has to go through today to make this all work. We think of programmatic as this easy button. It's anything but. That really takes a lot of the operation frictions out of it, and I think you'll see more and more dollars going to that media.

So I was, you know, struck by the – I think there were suggestions made by the DOJ, and then there were also suggestions made by Google themselves. One of them was to make auction data available – or data in general available to rival SSDs, right? And I was kind of interested in how that specifically would lead to more innovation and a more efficient market and basically drive yield and returns for publishers, whereas it hasn't before because of that monopoly.

Yeah, well, there's many ways to make it more available and make it more fair. But just essentially, the dynamic that was hurting publishers was the fact that Google gave away the ad server for Frey. But in return, you had to use their ad exchange as your primary ad exchange, which was AdEx. Then you had the Google Buy side that had unfair advantage because they were getting all this information that other folks weren't getting. So I wasn't getting the same information from the publishers that AdEx had that they were sharing with DV360 or AdWords or whatever the case might be. So what it resulted in was a pattern of, we have absolute knowledge about floor pricing. We have absolute knowledge about where the last bid came in, and now we can beat that bid by a penny if we want to win it. Or we're the only bidder because there was a floor on it, and we can bid below the floor, and the publisher is going to take it because they'd rather take a dollar than have it burn. And so there was all sorts of price manipulation that they had access to that we didn't have access to that very same knowledge to be able to bid more fairly and accurately into it. So a lot of that's been taken away already, but by open sourcing the data, it would make our decisions far more informed about how we would go about through our bidding strategies to win this inventory. And so in theory, it should raise pricing on the margins for publishers. Got it.

And you, I think, Magnite made its own kind of statements to the DOJ public, I think, last September or so. There's a pretty big document if investors want to read that. Quite interesting.

Okay, so then let's move on to mobile for DV+.

Again, kind of thinking about, you know, how DV+, could be a long-term growth business. You said that mobile is bringing DV+, back to growth, I think, on your last call. And, you know, we've covered mobile ad networks for quite a while now, for five years, I guess, if that's a while. You know, AppLove and Unity, Moloco. Sorry, Moloco is private. But definitely AppLove and Unity, we used to cover Digital Turbine. What, I guess, in mobile ad tech space, you know, first of all, what inventory do you work with on the mobile side? Is it mobile games or is it non-mobile games or both? And, you know, how specifically do you enhance, you know, the sort of mobile publisher and the mobile.