Discover more from Yet Another Value Blog
B. Riley's Dan Day describes Supply Side Platform (SSP), CTV + Magnite $MGNI thesis (podcast #195)
Dan Day, CFA, Equity Research Analyst at B. Riley Securities, joins the podcast today to discuss the digital advertising business, SSPs and his thesis on Magnite, Inc. (NASDAQ: MGNI), the world’s largest independent sell-side advertising company (according the company's website).
***This podcast is brought to you by Stream by AlphaSense.***
Experience unparalleled ROI and efficiency through AlphaSense Expert Insights
Looking to optimize your primary research budget with comprehensive coverage and cost-effective expert insights?
We’ve got you covered. Combined with AlphaSense’s vast content set of company documents, sell-side research, news, and earnings transcripts, Expert Insights offers:
Broad Industry Coverage: Access to 30,000+ expert transcripts across sectors, including Technology, Media, Telecommunications (TMT), Healthcare, Consumer Goods, and Industrials.
AlphaSense’s AI-Powered Search: Provides quick navigation to key information, enhanced by summaries, headlines, and audio to find the data points you need fast.
Transparent Pricing: Enjoy a flat fee of $400 for 1:1 expert calls, no matter the expert's profile.
Looking to enhance your investment research returns? Get in touch and book a demo now!
Please follow the podcast on Spotify, iTunes, or most other podcast players, as well as on YouTube if you prefer video! And please be sure to rate / review the podcast if you enjoy it, or share it with someone else who would enjoy it (more listeners is a critical part of the flywheel that keeps this Substack and podcast going!).
Disclaimer: Nothing on this podcast or on this blog is investing or financial advice; please see our full disclaimer here. The transcript below is from a third party transcription service; it’s entirely possible there are some errors in the transcript
Transcript begins below
Andrew Walker: All right. Hello, and welcome to The Yet Another Value Podcast. I'm your host, Andrew Walker. If you like this podcast, it would mean a lot if you could rate, subscribe, review wherever you're watching or listening to it. With me today, I'm happy to have Dan Day. Dan is an analyst at B. Riley. Dan, how's it going?
Dan Day: Good, Andrew. How are you? Thanks for having me. [inaudible]
Andrew: Hey, really excited to have you on. Before we get started, let me just remind everyone, nothing on this podcast is investing advice. Always true, maybe particularly true today. Dan covers a wide swath of sectors, actually. We initially connected because you cover some of the really small quirky cable companies. We're going to spend most of the time today talking about the SSPs. But just a reminder, we'll talk about a lot of things everybody should remember not investing advice. So that out the way. Dan talked about a lot of stuff, but the companies I really want to talk about were particularly the ad tech companies and particularly the SSPs. The one that I'm probably most interested in is Magnite. The ticker there is MGNI, but they have a publicly-trended competitor that you probably need to talk about. When you mention them, PubMatic, the ticker there is PUBM, you cover both of them. So I'll pause there. I guess one more thing, I have a lot of friends who are long this, I've looked at Magnite and they used to be Rubicon multiple times over the year. Every time I look at them, my eyes start to bleed a little, and I'm like, there's a lot of acronyms. I know the very basics of what they do, but how they do it, the business, everything still remains a mystery to me. So I'll turn it over to you. What is Magnite, what are the SSPs and why are they so interesting?
Dan: So first of all, it's funny you say that because one of the rationales for pushing my coverage into the ad tech space is that I feel like the buy side needs a lot of help understanding an area where there was a bunch of IPOs and SPACs in 2021. A bunch of them are 70%, 80%, 90% off where they went public at. And a lot of people are trying to pick over it and see what deserves to be trading, where it's trading at, and what's interesting. So you're right, tons of acronyms. I started ramping on this space two, three years ago. It takes a lot to get going, but some really interesting names. And I'm sure we'll talk about the Google trial that's likely next year that's very relevant for the sector. But we'll start with, you mentioned the SSP. So SSP is a supply-side platform.In simple terms, let's say I'm browsing espn.com and you see a banner ad. That banner ad gets typically sold programmatically, which is really just real-time auction. So what happens is there's a demand-side platform that represents the advertiser, and there's a supply-side platform or SSP that represents the publisher. This is all dumbed down. It's a little more complicated than this, but this is enough for now. So on the DSP side, demand side platform, the number one you think of is the Trade Desk. The ticker on that is TTD. They are far and away the largest player in ad tech.
So if you're thinking DSPs, think the Trade Desk. The other big one is Google has one called DV360. So you're an agency, you're representing Ford. They want to blast out a bunch of banner ads. Someone at your agency logs into the Trade Desk, they set a bunch of parameters for the campaign and they just get blasted out to a bunch of websites. So in real time, the SSP is representing the publisher. So the DSP wants to get the best price for their advertiser. The SSP wants to get the best price for the publisher. They meet in the middle, they come at a price. And typically there are multiple SSPs involved in every auction. So the SSP that gets you the best price is going to win the auction, and everybody has their own take rate. So you start at a dollar. Let's say the DSP gets a 20% take rate, the Trade Desk is about 20%. And then the SSP side typically is a little bit lower than that. In an open auction programmatic, it would call it 10% to 15%. So off the top, that's 30%, 35% of the dollars going to the programmatic intermediaries from $1 spent by an advertiser to a dollar receipt by a publisher. So that just sets it up and what it does, and again, we're still talking about the open internet here. CTV is a whole nother interesting game that's very different than the open internet, and it's very important for Magnite. So let's just stay on the open internet for a bit because Magnite has two distinct segments. There's their DV plus segment, which is display video, and other. So think of what I said, mostly banner ads and obviously also video ads. If I'm on ESPN again, I click on a video. There's probably a pre-roll ad that comes up. Those get so programmatically.
Andrew: It's inescapable. I want to watch a seven-second clip of a sports highlight from that last night. ESPN insists, "Hey, you've got to watch this one-minute-long advertisement before the clip will show." It's absolutely inescapable.
Dan: I watched it yesterday. It's like a 20-second clip I want to watch. And you going to get through a 45-second ad first.
Andrew: It's absolutely insane.
Dan: So the only wrinkle here is, I keep saying the open internet. So just for somebody who's totally new to this, the open internet is the contrast to what you usually call the walled gardens. And there's three big ones. There's Google, there's Meta, and there's Amazon, and there's a couple of what you call Challenger Gardens. You could throw Snap and Pinterest and those kind of names in there. But when I say walled garden, basically those guys want nothing to do with DSPs and SSPs. Everything they do is through themselves. If you want to get by Google search, you're going to Google DV360. If you want to go to YouTube, you're going through Google DV360, and they're not using Magnite and PubMatic to sell it. Everything is them. They put these walls up and we don't play nice with ad tech intermediaries because we have the scale to say it's all us. So those are the two. So obviously YouTube is the juggernaut and online video. I think 40 billion-plus in ad revenue turn on 12 months. So just to be clear on that, there is no exposure to YouTube for Manite and PubMatic. It's all effectively everyone else who says, I don't have the scale and heft to put these draconian rules up and say... even ESPN isn't their competitive advantage isn't selling banner ads on the internet. They're willing to play with the programmatic intermediaries. Pretty much everybody is outside of really those three, I don't know, call it three to 10 smaller ones. And those three to 10 control 70% of the ad spend. So I don't mean to minimize it, but the rest of the open internet is what we're talking about with Magnite and PubMatic.
Andrew: So let me just lob in a few questions. So I think a lot of people when they look at this, they start thinking about the video segment, that's the sexy segment. But as you said, I think more than 50% of the revenues still coming from the older banner ad questions. Let me just ask the first question. The SSP, if I'm ESPN, I work with multiple SSPs. I'm going to work with Dan's SSP, I'm going to work with Andrew's SSP, and it's all happening in real-time. But basically, it says, hey Andrew's visiting the website, whoever's giving me the highest value add, the ad I'm getting the most for is the thing that's going to go on the website. So I guess the question there is, why is the take rate for SSPs not a complete race to the bottom? Because the serving, the marginal ad there is pretty costless. And if you have an ad at a dollar, if we're both got a bid at a dollar and you just said, oh, I'll drop my take rate from 10% to 8%, then ESPN would go to you. And then I'm incentivized to say, oh, well, I'll go from 8% to 7% because there's no marginal clause. I know the take rates are lower than the demand side here, but why haven't they just raced all the way to zero?
Dan: So it is a fair question. This is effectively the bare case on SSPs is that they're just a commodity. And so you contrast the SSP side from the DSP side. The DSP side has fragmented down to three that matter and then a longer tail, the three that matter are the Trade Desk, Google DV360, and Amazon has a large DSP and then there's, I don't know, five or six others. Yahoo is...
Andrew: Amazon's DSP, is that only on their website and sites that are taking Amazon? Or is that all across? Are they actively bidding? So if I turned on advertisements on my website, could I start getting Amazon and certain DSP ads? Okay.
Dan: Yeah. And they all have their own quirks. You have to use Google DB360 to buy YouTube. You have to use Amazon to buy Amazon. So they almost have this lockin. Really the only independent one is the Trade Desk.
Andrew: Yeah. Okay, cool.
Dan: So on the DSP side, you've got these three, and this is why the Trade desk, in my opinion, gets a massive valuation premium to the SSPs because this side has fragmented down to call it an oligopoly. The SSP side, and again, we'll get into this with Google. Google still has 40% to 50% of the SSP side on market share. So they actually have more market share on the SSP side than they do on the DSP side. So outside of Google as an SSP, everybody else is just so fragmented. Magnite's probably the largest at less than 10%. PubMatic is the second largest publicly traded with under 5% market share and market shares for SSPs are pretty tough to back into. So take these with a grain of salt, but they're fairly rough numbers, but they're directionally accurate. And there's a handful of others. Criteo has an SSP with mid-single digits, and then there's like a million SSPs that are like, I'm good at monetizing mobile inventory in Germany and just things like that. So it just totally fragmented and there's no reason that a publisher wouldn't just add one more SSP. That's been the case to date. So really this programmatic header bidding auction started to gain traction around 2016. And initially, it wasn't too bad for the SSPs because there was only a handful of them, and then more and more and more. And every publisher said, yeah, sure, we'll add one more SSP. Yeah, sure, we'll add one more SSP. If you win some auctions and give us a higher bid, what's the problem?
So you've got publishers probably on average using five to 10 different SSPs plus other things like Taboola and Outbrain and some other ways that they monetize, but it's probably too many. And this has been an issue for them, and it's why the take rates have been depressed. There's other reasons. The DSPs get a bigger take rate because they're also selling data. Typically the data is bought on the DSP side. So, Trade Desk's actual take rate was 12%. They're brokering data for another 8%. So the take rate is actually the same. In an open programmatic, it's just the data tends to get transacted on the DSP side. But SSPs used to be like 20% plus take rates 10 years ago, five years... not five years ago, but 2016, call it. And that's gone down to more like 10%, 15% at most. Part of my thesis on why Magnite and PubMatic are interesting for open internet.
Andrew: That was actually going to be my next question. And just before you hop in there, there is a big Google antitrust case breakup, huge YOLO upside there. Let's set that case to the side. Because it is there, it is interesting. But I think there's a debate on the probability and everything, and if that was your only thesis, I think you're in YOLO mode. I'm not saying it can't be a good part of the thesis, but you're in YOLO mode if that's the only thesis. So let's just put that to the side and say, why are the SSPs interesting, excluding that huge bull case.
Dan: Because that long tail should die of SSPs. And there's a couple of reasons for that. The very simple nice way I put of it is an advertiser goes, DSP, SSP to publisher is actually not, it's actually way more complicated than that. And for no other reason than a bunch of companies figured out how to wedge their way in the middle and take a tiny fraction. Usually, it goes DSP to an SSP who sells to another SSP who might sell to another SSP who might sell to a publisher. And so I'll go back to 2021 when it was just a crazy year for digital advertising. Everything was bonkers. Everybody was growing 20%, 30%, 40%, 50%.
Andrew: Look at the stock prices for any of these guys. I'll use Magnite. Magnite was 40% to 60% at the start of 2021. They could do no wrong. They were buying SSPs. And as you and I are talking, Magnite is in the seven to eight range. So not that stock price is the indication of business value, but this was like ground zero for, oh my God, things were so good early 2021, everyone spent all their time online, things were going great. And now, there's a lot of other things, but partly COVID hangover, things have gotten worse. So absolutely.
Dan: If you were in the digital advertising business and you didn't have a banner year in 2021, you need to do something else because it was pretty hard to not. So the point being is that, this got way too complicated. And now you're talking that 35% of every dollar going into ad tech intermediaries, it's suddenly like 50%, 60%. And in 2021 the ad tech people called it the FOFO, fear of finding out. They knew this was happening, but they were like, oh, it's fine, everything's great. I know there's probably way too many intermediaries. When things went bad and things get started to get worse, all of the agencies suddenly said, I need to figure out exactly which of these intermediaries are adding value and which ones aren't.
So there's this trend called supply path optimization, another fancy ad tech acronym that will get thrown around a lot SPL. Basically what that is, is just making sure I understand from spending a dollar to that going to the publisher who's getting a cut of it. And so this is what's going to cut out all of these guys who wedged themselves in the middle. And what it should do is kill the long tail of SSPs that frankly don't add any value. They just popped up to arbitrage. And none of them are public. People are always like, oh, who's public that I can short on this? They're all very tiny. Like I said, some guy who popped up to sell, I don't know, some very niche type of inventory in Germany, that kind of thing. So Magnite and PubMatic, if you look at any SPO announcement by group M, any of the... and again, the programmatic pipes, they're bought by agencies on behalf of advertisers for the most part. Agencies are the ones logging into the Trade Desk, Google DV360 and they're the ones saying, you know what, I trust Magnite and PubMatic. I know what I'm getting. I know the fee structure. If I'm only going through them, I'm going to turn those up and I'm going to say, I'm going to spend X percent of my budgets just through those SSPs, and I'm going to turn all the other ones off because I don't know if they're doing all these weird reseller agreements. This may be too in the weeds, but are you familiar with the Made for Advertising ecosystem?
Dan: Okay. I'll get to it because I do think it's important. So there are basically these clickbait sites that exist for no other reason than to get people to click on them. And the content is awful. I'm sure you've seen them.
Andrew: Absolutely. Yeah.
Dan: They'll be like, here's what Jan Brady looks like now, and it'll be like 50 slideshows, and the ad density will be insane.
Andrew: Sometimes on this podcast, I'll click and my computer, like somebody will be saying something and I'll be looking it up, and I'll click on a site and it'll be one. I know exactly what you're saying. It's at the bottom of every Yahoo article. You won't believe what Jan Brady looks like now, and you do it and your computer goes slow, just advertisements everywhere. Absolutely.
Dan: Yeah, this is a big debate in the ad tech world whether they should be turning these off because they actually love them because if you're at Group M and your goal is to get super cheap reach, like, oh, you go to your boss and you say, I got a million impressions for like $0.25 CPMs. Isn't that awesome? And it was like, it showed up on one of those crappy made-for-advertising sites and nobody even saw it or as a video...
Andrew: That actually relates to another question I was going to ask. So you mentioned that there's a huge long tail of Dan and Andrew have 10 niche websites in Germany that they're serving as the SSP. When you're going through Magnite or PubMatic, how much is there a brand integrity question where like, hey, I'm a buyer, I go through them. I'm sure everybody's got a little bit, but it's not going to be next to... if I go through them, my stuff's not going to be next to porn or that type of stuff. Obviously, the Jan Brady, that's not something you love your content to be, but it's ultimately harmless. You could imagine a much more harmful version of it. How much is working with the big guys? Maybe it's not quite in the number, but you're getting a little bit of brand safety there.
Dan: Yeah, there's two companies that this is their job, double verify DV, and they both went public in 2021. IAS, Integral Ad Science. The brand safety suitability thing is their problem. Like that's their job. I'm a GroupM, I want to make sure... and they do work with the walled gardens too. So if there's some crazy video on YouTube, you would not like to show up before, their job is to block that from happening. So outside of the SSPs, there is this whole ecosystem of brand safety and suitability that almost all large advertisers are using one of DV or IAS, to make sure that stuff... and when it goes wrong, which it does go wrong, they're the throat that gets choked, not the DSPs [inaudible]
Andrew: So let's turn a little bit more to the opportunity. And part of the opportunity, I think as you said, hey, all of these have been cut down 80%. And there's still a lot of growth to come, as you said, we'll probably talk about the video segment in a second, and that's a really interesting, potentially sexy growth story. But let's just talk on valuation. Because I think people look at these on a headline valuation. I'm looking at your initiation. So the numbers might be slightly stale, but your initiation prom, I think it was April has, hey I've got a near-term price target of about $15 10X 2024 EBITDA. I've got about $15. But longer term, if you look at the growth and you looked at... I could do a DCF and say $30 per share. So could you just help me frame the upside and then I want to provide a little pushback on the valuation?
Dan: Yeah, sure. So really it's like I said, a lot of the long-tail SSPs die and the big ones that had the SPO agreements with the agency buyers eat up that share. There's so much upside just for meeting up share. And then obviously, you layer on a potential ability to take share from Google and it's turbocharged. I think that's where a lot of the growth comes from. And then second, it's like right now we're in a downturn. And I put a note out the other day that was tried to frame that up because some people are like, oh, I see Meta at 10% growth. I see Google search at solid growth. What are you talking about a downturn? If you think about it, there's brand advertising and then there's performance marketing and performance marketing is what's holding up really well. The brand advertising side, which is what a lot of the programmatic stuff is. It's not even really meant to be clicked on. It's just meant to like, I'm scrolling through an article, I see, oh, the show's on HBO Max hits me, hits me, hits me every time.
Andrew: The performances, it's the famous stuff Google got. You search for life insurance, the top thing is life insurance. Somebody will pay $50 if you get that click because they're likely to buy life insurance. As you're saying, the brand advertising is the traditional, if you're watching TV and you see an ad for Jeep, you're probably not going to go buy Jeep tomorrow. You can't really track that, but you're building that brand. So when you are ready to buy a car, Jeep is top of mind. And we mentioned ESPN multiple times. Anybody who goes on ESPN, the front page is generally brand building, not smash this monkey and go buy this thing because that doesn't really work unless you're actively searching for it. That's why Google's got such a good business, but yeah.
Dan: Yeah. And so you mentioned TV, obviously the big brand advertising medium is the TV. Radio podcasts, CTV people talk about a lot. So programmatic is often complimentary to that. It's like, hey, you want two things. You want reach, and you want frequency. You want to hit a lot of people and you want to hit them multiple times. So what a lot of advertisers think about programmatic is like, TV is pretty expensive. Let's get some really incremental, pretty cheap reach and frequency. So when those TV budgets draw down, which they have and it's not just because of people are cutting the cord and stuff, CTV is also significantly decelerating on the growth rate. Obviously, radio and podcasts have also pretty meaningful radios and decline podcast growth has derated. So it's programmatic. So I think people are mistaking that and it's like, oh, I look at the walled gardens growing again, and I look at the programmatic schmoes back in declines, what's going on there? I think it's more the brand versus performance dichotomy that people underappreciate on that.
Andrew: So just to lay out the question a little further, it is a question I had. How much of the stocks going down 80% from 2021 to 2023? Obviously, there was a little bit of mania and results were probably like two turbocharged by every spac getting money in and growth at all costs and stuff. But how much is simply 2022 and 2023? Performance marketing has held up great. If you've got something that's going to drive sales, it, there's still a budget for that. There's probably always going to be budget for that. But companies were worried about recession. A lot of things are influx, people were over inventoried. How much of the drawdown do you think is just, hey, like we are in, not an economic recession, but a programmatic recession for all of those reasons versus, hey, the stocks were way overvalued. The competitive environment is a disaster or that type of stuff.
Dan: I think it is a combination. I don't know. I think that there's a lot of bear cases on the SSPs that feel like they're playing out right now. And that's why you saw them sell off after the last earnings call, both Magnite and PubMatic were off.
Andrew: Because I do think a lot of my friends who are long it are like, hey, you need to look that this as a cyclical. And the times by cyclical is when things are at their worst, and this is probably than nadir. Nobody can call the absolute bottom, but it seems like this is in the near... it doesn't seem like we're going to a recession. It seems like the brand-building should come back. If you're a brand, you can cut brand building off for a little bit and not hurt yourself, but eventually, you hurt your long-term results. So I think a lot of my friends who are long think, hey, this is the year you want to buy the nadir versus, I guess Magnite's down 40% to 50% on their Q2 earnings results. So what's going on that's driving that versus is it just cyclical, I guess.
Dan: That's a good time to get into CTV.
Andrew: Okay, great.
Dan: CTV is very important for Magnite.
Andrew: You describe it before we hop into it?
Dan: Sure. So connected TV. So basically it's the same thing I explained on the open internet. It's just selling for connected TV publishers. And this spans from your fast and VOD channels, Pluto TV, and the random long tail ones that you'll see on the Roku channel. Anything to Hulu and HBO, I guess max, not HBO Max. So they do everything and I can't emphasize enough because people think of them as the same. CTV is very different than the open internet. Think about what a typical commercial ad break looks like versus when you're browsing the internet. And an ad break on TV, Coke and Pepsi can't be in the same pod. Mcdonald's and Wendy's can't be in the same pod. You have to normalize the sound, don't you hate it when the ad comes on and it's like 10 times louder than the show you were watching? This is a problem in CTV because the ad tech isn't frankly all that built out yet. And then there's the problem of, I hear this all the time, you see the same ad every time and I'm watching Hulu and it's the same ad every single time and it's a frequency capping problem. So a lot of these people just tried to take programmatic advertising and just poured it into CTV and it's a very different ecosystem over here. It's a very different way you need to advertise. You need to understand that hitting people too many times with an ad is going to turn them off. And publishers need to do things with their ad server and SSPs that limit that if you're going to sell programmatically. So Magnite, which was formerly the Rubicon project, as you mentioned. Magnite was formed with the merger of Rubicon project, which was all open internet stuff. And Telaria, which was a CTV specialist, they were purpose-built for CTV really premium video, but eventually pivoted to connected TV. This was 2019, the two of those got together and I actually think... it was an all-stock merger, so it wasn't really a price paid, but talk about the dilution. I think that was smart to go after someone who designed their SSP specifically for CTV. And I think that was the right move rather than other SSPs that just said, "Hey, we're going to take what we do on the open internet and try to do it for CTV publishers." And two, there's a lot of people who know CTV Ad Tech well at Telaria what they then did was they bought SpotX in 2021.
SpotX was similar to Telaria a CTV-focused SSP. It was Telaria and SpotX were the two that had emerged as the best independent SSPs for CTV. So they effectively cornered the independent SSP market for CTV with those two acquisitions. Again, we can quibbly pay too much for SpotX. They probably did. Everybody who did an ad tech acquisition in 2021 probably paid too much. They did do it with some of their own stock, which was over, I think $45 at the time. So that helps. Their stock was arguably overvalued, probably overvalued. So just to to frame it up, this is how Magnite became 40% plus of our revenue is CTV is they merged with Telaria and then they bought SpotX. So part of my thesis again on Magnite is that, in order to win as a CTV SSP, you need to have been purpose-built for it. And you need to have the ad tech talent which by buying Telaria and SpotX, they effectively have cornered.
Andrew: Two questions on CTV...
Dan: [inaudible] pause here. Go ahead.
Andrew: So two questions on CTV. So CTV is you're watching, you mentioned many of them, everybody knows Hulu Max, Netflix ad-supported, you're watching them and this is going to programmatically insert the ads. So I guess on Magnite, who are they working with right now on the CTV side?
Dan: So they effectively helped Disney build their programmatic ad tech stack. And that goes back to Hulu. Disney Plus only has had ads since recently, very recently. So they worked very closely with Hulu when they were Telaria to help them build their programmatic ad tech stack. They pretty much work with everyone. I don't know. And probably not Peacock, and the reason that they don't do Peacock is because their biggest competitor is FreeWheel, who's owned by NBC. FreeWheel is probably the biggest SSP competitor out there. And their argument against them all the time is, you know us more than them is we're independent and they're not.
Andrew: Can you say that one more time? You just cut out for a second.
Dan: Sorry. So NBC owned by Comcast obviously owns SSP and Ad Server called FreeWheel.
Andrew: So the Magnite argument is why work with FreeWheel when they're owned by your competitor? Work with us. Okay. So I think people can see why this would be a huge business. Every single Netflix is rolling. In the past year, Netflix rolled out the ad-supported, everyone's rolling out the ad-supported, and everyone when they roll out the ad-supported, they've said, hey, we are... I don't follow it quite as closely as I used to, but everyone who's wrote on ad support has basically said, "Hey, the results from the ad-supported are great. People love it. People come in, they save money. Our rates on ad-supported are better than we thought. Our engagement are better." Wveryone's focusing on the ad-supported. You look at the Disney charter dispute that just ended. Which I know you and I have talked about offline for different reasons, but you look and one of the results is Charter gets to wholesale Disney's ad-supported thing. So obviously, everyone's pushing people into the ad support funnel. So you see this huge growth. I've got a lot of questions here, but the first, they do seem to have an advantage. They are custom focus and everything you said with programmatic makes sense. But it's not like there are no competitors. Netflix I think is working with Microsoft on their thing. Who are there other competitors who actually matter on this side?
Dan: Yeah. It is FreeWheel really, is the big one. And then there are a couple of others. PubMatic does CTV. All the SSPs do CTV, they're all going to say they do it. And PubMatic has done a decent job, for example, but they don't even break out their CTV revenue and it's hard to tell exactly how big it is, and it was like doubling year over year last year, but it's like, oh, I'm not sure exactly what the base is on that. And now PubMatic in a piece of revenue that's not large enough to specifically break out is actually declining in CTV revenue. So the thought there is, there's been a first-mover advantage for the guys who specifically built it for CTV. So Telaria and SpotX was one, FreeWheel was another. There's a smaller one called Unruly that was bought by a company called Tremor. I've heard there...
Andrew: I know Tremor.
Dan: Tremor rebranding is Nexen. So they have an SSP and they're going to rebrand Unruly and I think that was actually previously owned by Fox, so I'm sure that Fox does a lot of their...
Andrew: Let me ask you. So it does make sense. Everybody's pushing to ad-supported, these guys have worked with Disney, they've worked with big guys. You can see the upside as everyone goes ad-supported. And I do think [inaudible] unwinding. There's just going to be a lot of growth in the sector. So I think you could frame that really easily as a bullcase. Let me push it back with a bear case. Look, there's going to be six online video services that matter. Take YouTube out of it. But if you're really thinking about the programmatic, hey, we make high-end quality and then we sell it, it's going to be Netflix, it's going to be Disney. Amazon Prime, you're probably not getting. I don't know, do you know who's doing Amazon Prime... I don't think they have full ad-supported yet.
Dan: I imagine they do it themselves. I have a hard time believing they're using anyone.
Andrew: Well, there's going to be six. NBC, Netflix, Amazon, you name it. There'll be a few. And I do wonder, you mentioned in your note, Netflix will face a buy versus build decision. It doesn't seem like Netflix really likes the Microsoft thing that they've done. I don't know too much about that, but if I'm the six of then, why am I going to work with Magnite instead of just saying, hey, why don't I build this myself? If I'm going to have 10% to 20% of the advertising market, no matter what, advertisers are going to work to work with me, I'm going to have a great brand product, build it myself. I don't have to give Magnite this thing. I can control it all internally. I can work with the advertisers directly on maybe more bespoke things. Why don't all of them just look at it and say, I've got the budget, this is important to me. Ad-supported is the future for my thing. I can't have someone else controlling my ads, just cut off the middleman and build it myself.
Dan: Yeah. No, it's another I'd say one of the more common bear cases I hear is one that they just build it themselves. It's a lot harder to build it themselves than you'd think. And it takes a lot longer and it hasn't worked for open web publishers that have tried to build some of these themselves. So some of them have tried to build SSPs themselves and they end up just doing that and having their own thing with monetization from the others on the side and realizing it's worth it to pay the take rate. So it's been tried before. And for a lot of these guys who have new ad-supported tiers, it's going to take you a while to build it. You're going to have some bugs when you build it. I don't know. They're under pressure to see results there. Magnite can come in and say, hey, right now, today we will start monetizing it for you programmatically. And they've been successful with that pitch. And I don't know, maybe outside of Amazon, they do everything themselves unsurprisingly. YouTube does everything themselves unsurprisingly. Outside of those, everybody's working with SSPs and CTV. And then the other piece of it, and this is the layer behind the SSP is the ad server. So the ad server is the one that says, and again, not all CTV inventory is sold programmatically. Some of it is sold direct, just like it's sold in linear TV. Actually, a lot of it is still sold direct.
And this shift from direct insertion orders to programmatic is a whole nother tailwind we can talk about. So the ad server is the one that says, okay, there's a direct ad. You sold Disney, you directly negotiated this with Ford, this goes here. We're not even going to throw it to programmatic. Next ad slot, we've got another direct one. Next one. Okay, the next two are open market and we'll just auction off. So the ad server piece of it is the infrastructure layer. The ad server doesn't make much money, but it's super strategic. And Magnite has one called Springs Serve that came at a very cheap valuation when they bought SpotX. I actually think Springs Serve is one of the more underappreciated assets within Magnite's portfolio. FUBU uses Springs Serve. None of the big ones use it. And again, FreeWheel is really their big competitor. Freewheel is also an ad server and SSP. So the ad serving component of it to me in Magnite, it doesn't make a whole lot of money for them, but it is a very strategic asset that they got at a very cheap price that helps lock in their SSP to a lot of these publishers.
Andrew: You said that none of the big ones, so Disney who has worked with Magnite on the... Disney's not using the Springs Serve ad serving product? I'm just wondering, I hear you on, it's a strategic product, but if nobody's using it, then it seems like everybody else has already built their own or using someone else's.
Dan: Yeah. FreeWheel is probably the one a lot of them are using. And now they're regretting that from what I hear from some of them. Just FreeWheel plays a lot of games.
Andrew: Yeah. I do hear you on this just. I know symbols and I think you mentioned it like Netflix seems unhappy with the Microsoft Partnership they've got. When the Microsoft partnership ends next year or it's up for renewal next year?
Dan: Yeah, next year or it may end earlier than that.
Andrew: Okay. But so let's just say next year it's up and then Netflix will have an instant decision, hey, if we're kicking Microsoft out, do we build our own ad server or do we buy, and I know a lot of people think, oh, buy Magnite, and then you've got a functioning ad server, you've got it across like you've built out an advertising. And in the same way Amazon like, hey, when we launch a new business, Amazon retail is the core business for Amazon Web Service, they're the core function. And then you start selling it to everyone else. Netflix can be the core demand load or the core, I guess, supply side for the advertisements, and then they can go sell it to everyone else. You could see the strategic there. I guess the other pushback would be, again, Netflix history is build versus buy. I think they've done two acquisitions and they were very small IP acquisitions in their entire history. So this is a tech-focused company with great tech. Why don't they just go build it themselves.
Dan: Yeah, they might, and I think they might build the ad server themselves. I think they'll still lean on SSPs to sell inventory programmatically and like any big publisher, they'll sell a lot of it direct. And this is what I say to people, direct is not the enemy of programmatic. You can run direct through the programmatic pipes, it's going to be a lower take rate. But the enemy of programmatic right now is insertion orders in CTV. Insertion orders are just the exact same way you bought linear tv. You're just buying it on the connected TV stream service. So as long as they can take some of those insertion orders and migrate them to some sort of programmatic, whether it's a programmatic guarantee, which is really just like a digitized insertion order and then there's like private marketplaces. I mentioned open is like, hey, anybody can bid. Private is like, hey, I'm Disney, I negotiated with 30 and you guys are the 30 that are going to bid in this private marketplace. So these are the lower take rate things that is mostly going to be in CTV. I've said the open auction programmatic that you see in the Open web, the FAST guys, the AVOD services, they might use those and lean on those heavily because they don't have the heft to negotiate directly with advertisers.
Andrew: My concern was Disney and NBC, these guys, especially the sports content that drives huge things as they go online. My concern was, oh, they take it into house. The next thing I was going to say was, why doesn't Disney... the top 50 advertisers put the top three beer companies, seven companies are going to do almost all the advertising. Why don't they go negotiate direct? And it sounds to me like what you're saying is that is a threat. In fact, that might be the most likely thing that ends up happening if it doesn't, great for Magnite because they can land a big customer. But what you're saying is, hey, there is still a longer tail of companies that I maybe was discounting, but Roku is big, Fubu is big, there's going to be other services. There'll still be a lot of services and a niche streaming services is not going to go do direct deals. They're not going to be able to build this and Magnite, they can build a serious business winning the... maybe not the one through 20. It'd be nice if they land one through 20, but winning the 20 to 2000.
Dan: Yeah, that's a good way to summarize it. They'll still win the one to 20, it's just going to come at a low take rate because they're doing less for them. That's the easiest way to put it. And again, it's so different than open web programmatic because all of these big... the top eight CTV publishers, they've done ad sales forever. It's what they're good at. A lot of them have a lot of old ways of thinking on the buy and sell side, advertisers are just like, hey, I just want to buy the bear on Hulu. And they're like, well, you could do all these cool targeting things if you want to pay a take rate, you can target here, you can target there. And they're like, no, I just used to buy shows on TV and that's what I want to do.
Andrew: I do hear that. So I did a lot of work on Shark Ninja recently, and one thing that I heard from people after I published is they were like, look, you don't understand. Shark Ninja, I don't know how familiar you're with them.
Dan: I don't know them actually.
Andrew: The shark vacuum cleaners and the Ninja, if you heard of the Ninja Creamy, the ice cream maker, they do innovative appliances basically. And my pushback is like, hey, you can only... and vacuum cleaners have been around for 100 years. How much can you really innovate and not get cut out by Chinese stock offs basically that are going to copy them in 18 months and it's all going to go to zero. And one point somebody made was like, look, a lot of their competitors are... it's run by a middle manager who's 50 years old. He's been there for 25 years. He's always bought advertisements on TV. He's always done the same thing. And you've got Shark Ninjas that's young and hungry and they're dominating TikTok advertising and everything. He's like, they're just competing with an older, slower breed. And eventually, if you run this forward 20 years, eventually yes, the middlemen will retire and everyone will be advertising on TikTok. But as you're saying, if you're the GM car media buyer. Yes, you're underlings are probably experimenting with people, but right now you're probably just like, I've always bought on TV. I just go to my buddy at NBC and I say, I want 100 ads on Sunday Night Football, and here you go. And it takes time, but it does evolve over time.
Dan: Yeah. And it's all contextual. You're just making assumptions about the people that watch Sunday Night Football and saying, okay, I'm a beer company and I want to do that. Whereas what programmatic is telling you is like, hey, now that it's not over a cable wire, now that it's through the internet, we can do a lot of cool things in turn targeting demos. We can frequency cap, we can say, hey, you were hit with these ads on the open internet. You want to hit these people, we can bring in purchasing data. We can do all sorts of cool things now that TV is going from over cable to over the internet.
Andrew: I'm about to have my first kid and if you were advertising to me, I would watch football games and you'd see Huggies diaper advertisements and be like, guess what? Three years ago that was the stupidest advertisement you could ever serve me. And today it makes a lot of sense. Or if you bought a car two months ago, there's no need to hit the guy with a car advertisement. Whereas if you had, hey, he bought a car four years ago, now's the time. And you could imagine those types of things. It's one of the reasons for a long time it was so bullish on podcast advertisings like getting inserted programmatically because it's coming off your phone, you know where the person is. You can insert a time of day. It hasn't quite played out there, but if you've ran it really far forward, eventually it will get there. I want to talk add-backs and then I want to talk Google Ads trust. But we've talked about a lot. I just want to make sure, is there anything on the bull case or the bear case, aside from add-backs and Google that we haven't touched that you think people should be thinking about?
Dan: I just think on the CTV side, the important point for Magnite is that, most of CTV today is still insertion order. So people talk about the growth as long as that's all insertion orders, who caress because the programmatic guys don't see it. Magnite put the product out called clearline, and PubMatic had a similar one that's basically like, listen, do that. Just digitize it and they'll make the workflow easier, make the billing easier. We'll take a 3% to 5% take rate for that, but start doing this programmatically. That point that about CTV and how old school it's still done is important because you've got the growth of CTV just generally budget shifting linear to digital. But then you've also got this tailwind of like, the dollars now still are just linear insertion orders that need to migrate over.
Andrew: You initiated with I think a 15 price target. I think your price target's still in the mid-teens. If I could put words in your mouth. The way your base case plays out here is, again, the stock is under $8 per share. So we're talking almost a double [inaudible] interest stock price. But the way this would play out is over the next 12 months, you get CTV, it continues to grow. People may be digitized a little bit more, so you get the growth from CTV and then the brand business basically stops declining because the recession fears have ended people, hey, we haven't advertised brand in 18 months, we've got to come back, we've got to start getting people. Brand business stabilizes, grows, whatever. And as that happens, people get confident and again, I'm looking, it might be slightly old numbers, but you've got about $250 million in EBITDA in 2024, 10X multiple. That gets you to the 15. All that happens.
Andrew: Say again.
Dan: That number $250 is probably a little high. Now I'm probably lower than that.
Andrew: It was from April. But I don't think I'm throwing out crazy numbers. So that would get you to 15 and that's 10X multiple, none of that's crazy. And by the way, they've been buying back that convertible debt at $0.80 on the dollar. So maybe your EBITDA is lower, but the debt is lower as well because it's always pretty accretive when you can buy back debt at $0.80 on the dollar. So that would be the base case and that frames why I've got friends who are interested in it, why you've got to buy rate. Let me start with the last bear case I would have, and that would be the add backs. I threw out the $200 million in EBITDA number. You can look at their reported EBITDA numbers. It sounds great when I say $200 million in EBITDA, but if I look at their add-backs, $40 million per year in DNA and that DNA does include, they capitalize internal tech, which is always a pet peeve of mine when people capitalize their internal tech and then add it back on DNA, I guess that's fine. But then they also have $300 million plus in annualized amortized acquirable.
Dan: Yeah, we can talk about that.
Andrew: I understand that, but it is just a big, big number. $80 million annualized run rate of stock comp and then they also have some merger and restructuring charges in there. So have I seen messier add-backs? Yes, absolutely. I used to work in private equity, I've seen way messier add-backs. But this is not exactly a clean story. I think when you say $200 million in adjusted EBITDA and just $80 million in stock comp right there, people say, well, that stock comp's a real expense, especially when the stock's at $0.8. Like that's eating up a lot of the equity cap right there. So I just want to ask about add-backs real quick.
Dan: Sure. So just the way I think about stock comp is like one, when I do a five-year DCF, I just look at the five years of stock comp, assume a price and dilute the count by that amount. And I get that of course stock comp is a real expense, but it's also not a cash expense. As long as you dilute the share count by an appropriate amount. I think you're capturing it.
Andrew: No, I hear you. It's one of those things, $80 million of stock comp when you're a $5 billion company is one thing, but $80 million of stock comp when you're an $800 million market cap company, all of a sudden it gets a lot more dilutive.
Dan: Yeah. I'd like to know how much of that stock comp is marked at 2021 prices. And if you marked it at $12, I don't know off the top of my head, but I imagine it's inflated to an extent by options and RSUs granted when the stock was at $40, $50, $60 and the actual dilution is much less than that. So again, I don't have the numbers right there, but I think everybody has that problem right now. So you look at Snap's, stock based comp, it's egregious, but it's all at $80 a share, which is like, okay.
Andrew: What is Snap now I don't even know where they're trading now, to be honest with you.
Dan: I think they're at like $10 or something probably under that.
Andrew: Okay. Cool.
Dan: So yeah, I'd have to do the math. It is certainly not the least stock-based comp of my coverage, but again, I'd imagine that there is some impact from them granting stock at prices. And again, how the accounting works is you don't change the price per share, it's all from the grant date.
Andrew: No, though the counter that would be, hey, if you're paying your people in 2021 $80 million a year and all the stock comp is at $20, and I've got friends in tech, all the stock hops struck at $20 and then the stock goes to $10. Well investors hate stock option repricing and rightly so. But if it's the CEO, it's one thing. He really is responsible for the stock price. If it's the lead engineers and stuff, they're not really responsible for the stock. And if you don't make them whole, I can tell you, my friends, they look to go elsewhere real quick if you don't make them whole. Because you generally get bonuses when you leave and it's just tough. Okay. Go ahead.
Dan: No. I was going to just say it's a fair point.
Andrew: So let's add back, I do think it's like, hey, if you're going to get the growth, if you're going to get the... you will leverage all of this, I'm fine adding back the amortized acquisition intangibles. I'm a little funky on the capitalized internal tech, but whatever, if you get the growth, I think that all of that's doable. Let's talk the big bull case. The big thing that some of my friends get so excited about. And that's the Google antitrust case. So for those who don't know, the Google antitrust case, it's about to go to trial. Is the trial March, 2024 or is the ruling March, 2024?
Dan: I'm not sure if it's January or March of 2024.
Andrew: I think it's March 2024.
Dan: Well, we don't know when the ruling will be, but people are saying it could be pretty quick, like six months is pretty quick for...
Andrew: If I remember correctly, and stop me if I say anything wrong because I'll ramble a little bit here, but right now, I said it right now? Because there's so much of the pretrial motions happening and all the discoveries coming out and everything, that's why I was a little confused, but I believe it's March, 2024. I believe it's in Virginia Federal court, and this is called the Rocket Docket Court. So people think this ruling could come as you said, very, very quickly. And a very quick ruling is actually interesting because there could be a change of administration in November. And I don't think a new administration, which if a new administration came in would probably be a Republican administration. I don't think they're very friendly to tech, but it also would be tough to have a administration less friendly to big tech than this one. So getting the ruling early could influence there if there was a change. But anyway, big ruling could come. It's antitrust case. Government's suing Google.And basically saying you're an antitrust issue and a lot of the things that Magnite is competing with Google in are square in focus here. So we don't have to opine on the specifics of the antitrust case. But I do want to ask if Google loses this antitrust case, there is a huge bulk case for Magnite, so maybe you can frame that.
Dan: Yeah, so again, the bulk case is mostly on the open internet stuff. I think Google intentionally is like, I think we should probably stay away from CTV Ad Tech. We're already the 800 pound gorilla everywhere. They have an ad server that a lot of people use, and they have an SSP too, but I hear mostly Magnite and FreeWheel is the big ones. So it would mostly be Magnite DV plus segment and PubMatic as the big open internet resting Google's grip on the banner and, and video ads that you see when you're browsing around the internet. So look, I think it's hard to ignore the fact that right now we're in the search antitrust trial. Next year is the ad tech antitrust trial. If Google had to pick which one of these to win and lose, there's not even a question. They would lose the ad tech antitrust one, and they would win the search one. What do you think of anything of Google search, YouTube, the cloud, that's it. Nobody even talks about the Google's network excitement. When's the last time you had a cell site analyst ask about Google's network segment on an earnings call? It's the DSP, it's the SSP, it's very esoteric. Nobody even cares. And by the way, most of their privacy problems are on the SSP and ad serving side. There's an argument that they might say, you know what, we definitely don't want to lose searching feature one. Maybe we should play nice and just split off the SSP and ad server and give them what they want on that one.
Because when we really think about it, I don't think anybody's going to care that much if we have to spin this out to shareholders. That's, I think, something that can happen. I don't know, just me throwing things out there, but I do think that if there was a spin out, which by the way, nobody's buying Google's sell side, and I'm making a lot of assumptions here. I'm making the assumption that the preferred remedy from the DOJ is to split off Google's sell side ad tech assets. So it's SSP and ad server. And they would get to keep the DSP. In the worst case for them, they would try to split the DSP off. So YouTube would have to open up to other DSPs. That's something I think would probably be a bridge too far. And maybe something they're like, oh man, what if they try to do that to us? Why don't we just split the sell side from the buy side? If you read that, the DOJ filing, it's all like, oh, this is like if Goldman Sachs owned the New York Stock Exchange, which is not apples to oranges to me, but it is a nice headline. But anyway, that's it. You split off the buy side from the sell side and theoretically that removes your conflict of interest. So I think there's the chance of this happening where the SSP and Ad server gets split off from the rest of Google is higher than a lot of people think. And PubMatic and Magnite are easily the two biggest public beneficiaries of that.
Andrew: Let me just drive why are they the public beneficiaries. Is it because, hey, if you're just splitting off the Google SSP, you've got another SSP. Previously they were tied into Google, but it's not like the competitor has gone away. So is it a beneficiary of just, hey, there is something to... if there's only two publicly traded players in a market and they're small, not a lot of analysts, not a lot of coverage, it could be more inefficient. Is it just, hey, it's bringing more eyeballs. Or if and when Google splits off the SSP, does the legacy business of Magnite have this huge beneficiary where, hey everybody had to work with Google because they wanted the DSP. Now it's free for all, 30% of the market's up in the area and we're going to grab 10% of it in huge growth, margins go up. There's less competition. Is it just a free for all? So how do you look at that?
Dan: So the easiest way to frame it up is if you split DV360 from the SSP and ad server, and DV360 has the most market share among all the DSPs. Go and look up Project [inaudible] from the DOJ filing, basically what it was is Google intentionally turning on a function that forced all of that advertiser spend on DV360 through their own SSP and ad server. And they didn't tell them, and it was just like a harmless looking little thing that was automatically turned on for everyone, basically that would put better bids through their own SSP and win more. Again, these are the things where people just don't trust them anymore to act in the best interest of these as an advertiser as a publisher because they play both sides and you never know which side's getting screwed, but Google's always winning. So that's the easiest way to think about it, is a significantly higher percentage of DV360 spend would go through non-Google SSPs. So that's one.
And then two is that again, I guess it would depend if you could... so there would be a further, could you split the ad server away from the SSP and I doubt it. But if you could split Google's ad server from its SSP as well, then it would be huge. Because like I mentioned how Magnite has an ad server and CTV, Google has like 90% ad serving share in the open web, and you can do a lot of things with the ad server that are sneaky to make sure that your SSP is winning a lot of the auctions. So if we were to go further than that, which I think it's unlikely, that would be massive. Because one, it would open up, you could pick who your SSP is. Google's played all these games where you have to use their ad server to get access to their demand and illegal time and this is the whole antitrust argument. So there's a couple of ways it could win. And again, I don't think the SSPs need this to happen but it's a huge call option in stock
Andrew: Let's say March comes around, DOJ press released with the government, "Hey, we've settled. We're we're splitting off our SSP." Or ruling comes in September, "Google, you legally must split off your SSP." It it's tough because we don't know the impact. It's not one of those things where you can say, oh, Magnite signed a new customer. I think they'll get $50 million. We literally don't know the impact, but what would your gut say, the Magnite stock would go up if Google agreed to do that.
Dan: Well, I have to imagine this. The stock would've anticipated it if the trial's going well, but let's just say it didn't.
Andrew: Then let's go with the March, 2024 before the trial starts, you would get some filings and stuff, but before the trial starts, there's a settlement. Google says we're splitting off the SSP, stock says, all right, it's happening, game time. What do you think the stock does?
Dan: Yeah, I think most likely it would've probably already run up, but I think Magnite could easily be where my price target's at in that scenario, at least over $10. I think it would rerate considerably because it would just be a lot of the bear argument coming out of the stock.
Andrew: What do you think...
Dan: And a non macro, obviously there's the macro side and then there's the more fundamental bear argument. If the macro is turning and some of that bear argument's coming out, we've seen what happens when ad tech rerate, it could be very aggressive and very quick.
Andrew: What do you think the odds are that Google loses this trial and or splits out SSP?
Dan: Oh man. I'd caveat this with... I haven't analyzed a whole lot of DOJ cases. All I would say, I don't know if I want to put a specific percentage on. I think it's greater than 50%, which I think is a lot higher than people think. I get a lot of people who I talk to about this and are generalists and they almost dismiss it. It's almost just like Google never loses these things. They get fined all the time. They just take it from all the regulators. They're bulletproof, they're not going to actually lose any of these things. And I don't think they're going to lose the search monopoly trial. I actually think they have a pretty bad argument on that. I think they have a very good argument on the ad tech antitrust trial because you can tie it specifically to acquisitions that they've made, double click and ad meld and all of these different things. You can tie it specifically to illegal tying between products, the ad server to the SSP, to the demand side. I don't know. I think it's a very good argument that they have combined with the fact that Google's facing two big antitrust trials and there's one that they would far and away prefer to win. And I don't know how this stuff works behind the scenes, but like I said, there's a scenario where it's like, listen, let's just come to terms on this other one. We'll give you what you want on that one. Maybe we get what we want on this one and you got a big win here. Chalk it up.
Andrew: Look, I'll admit I have not read the DOJ complaint here, so I'm talking a little bit out of my butt. But I have a friend who has a big position at Magnite and his argument was, hey even if you ignore the cyclicality stuff we talked about, everything that you and I have discussed currently, I think there is a really good chance Google loses this SSP trial. And my pushback was exactly what you said. I was like, Google's bulletproof. And he was like, "Dude, read the complaint. There's a real theory of harm. It's very easy to cleave off. And if you think there's any chance of that, Magnite is going to win." And his other interesting point, and I I'm just talking about him loosely. He might come on the pod and talk about it at some point, but his other interest point is, if you look at the history of companies getting broken up for antitrust, all of a sudden you've got the Google engineers who've been very protected. They've got the search engine gushing, people have to work with them, all this sort of stuff. All of a sudden they're subject to the really competitive whims of the open market that Magnite has been subject to the whole time they've done. And the Google SSP which... how much market share would you say the Google SSP has?
Dan: 40% to 50%.
Andrew: It's got 40% to 50% and all of a sudden they're subject to the whims of the market and the things collapse because it turns out they've gotten fat and lazy and the people who've always been scrappy take share really rapidly. Will that happen? I don't know. What he says sounds reasonable. Also, everyone I know who works at Google is also really smart and really good. So maybe they haven't gotten fat and lazy and they'll subside. But as you said, I think the market, if I had to guess is pricing in a sub 5% chance of it happening.
Dan: I think it's a lot higher.
Andrew: You said over 50%, I think my friend thinks over 50%. I don't know. It's something I'm going to have to work on, but I definitely don't think it's sub five and there is a lot of optionality there. Cool.
Dan: I'd say it's better than you. And again, I'd say that is maybe a dumb beter here. He's never bet on DOJ trial outcomes. But just talking to people or talking to a lot of people about this, I definitely think it's a much higher chance than generalist investors described to it.
Andrew: We could talk about DOJ all day. I think there is something where the DOJ and particularly the FTC have run up a string of antitrust losses. So I think a lot of people are dismissive a little bit of the cases. Because it used to be the base rate for the DOJ bringing a case was very good. But this case as I understand them and again, not having done crazy amounts of work, this case has real merits to it. And I think I'm going to have to get smart on it. Maybe as the trial gets closer, we'll do some expert calls on and stuff. All right, last question and then I'll let you go. because we've run over time. Ignoring the Google case, there's the bear case on cyclicality, there's the bull case on growth. What do you think the one thing is that the overall market generalists who look at this miss that you think is important to the story. It can be on the bull or bear side.
Dan: Yeah. I think they think that the DSPs are just going to trample the SSPs and you're going to have a world where SSPs don't exist. It's like they say the Trade Desk has made all these crafty announcements like going directly to publishers, undercutting SSPs take rates. There is this argument that the DSPs have all the power that will trample the SSPs that, maybe PubMatic and Magnite consolidate down. My pushback on that and having talked to publishers is that, at the very least for the open market internet, they very much value having someone on their side who's maximizing their yield that suffered from the Trade Desk. One thing that they would absolutely hate is if Google gets broken up and then all of a sudden the Trade Desk is the new Google and the Trade Desk can say, oh, go to direct to us, we'll give you a better rate than the SSPs will get. Well guess what? Maybe they can now, but in five to 10 years, you just killed all the SSPs who are representing you, and now there's really nobody it's just the Trade Desk, it's the new end to end. I don't think any publishers want to end up in an ecosystem where that's the...
Andrew: Do you think publishers are strategic enough for that? Because this is almost the predatory pricing thing where Trade Desk goes and says, "Hey, Dan and Andrew's SSP is yielding you a dollar. We'll yield you $1.25." And they can eat their take rate for years and years. And then once they [inaudible], they say, "Hey, now you're getting $0.75." And do you think media companies, internet company, are they strategic enough to say, "Hey, no, we need to make sure we're funneling some business towards Dan and Andrew to make sure we've got a competitor." Or managers run on a short term if they can boost their earnings and my example by 25% by going with Trade Desk and this clearly predatory pricing scenario. But go with that and let your replacement worry about the dealing with a monopolist three years down the line, five years down the line.
Dan: Yeah. So there are a couple of points there. One, there's considerations beyond the, CPM you're getting. For example, it's pretty crazy. We went over an hour without talking about third-party cookies in an ad tech conversation. But anyway...
Andrew: Didn't do any ATT talking or anything. Look, I'm a generalist. I told you, you could tell me I was stupid at any point.
Dan: No. It's actually I think a good thing. But anyway, with third party cookies going away, everyone says publishers have to have their own data. And publishers want to put that data in the SSPs, they don't want to hand it over to the DSPs. And so when you pass along that ad request, it's like, hey I'm someone who's logged into ESPN with my email. I give them the teams I like. I don't want to pass that on to the Trade Desk. I want to keep that close to the vest with my SSP and I'll let them pass that on in the bid stream and then the DSPs can bid on that information as they'd like, but there are other considerations beyond just like the price you're getting that SSPs can add value and...
Andrew: Why would you not want to pass it on to the DSP?
Dan: There's a worry that they'll use the data to just create their own data lake and then they won't need you.
Andrew: I hear you on create your own data lake, but you're still the person, the consumer is getting served the ad to. So even if they've got all the thing like, cool, you can build a better profile of the user, you still need to serve the user the ad through me.
Dan: Yeah. I think that's fair. I think there's also like a privacy argument that you're keeping it closer to an SSP that represents you specifically. And this came from a publisher that I talked to about open internet.
Andrew: As somebody who uses the ESPN website all the time and thinks they're targeting and everything is terrible and they always forget my username, I almost wish they would be less sensitive with my privacy and just to remember things. Let me ask one more question then I actually will let you go. You mentioned a lot of time you talk to publishers. I'm just curious because again, I'm a generalist in this industry. Every time I looked at it, my eyes bleed. Who do you talk to for the most part? Obviously you talk to Magnite, you talk to Trade Desk, you cover them. But who do you talk to in the industry when you're getting industry scoops? Are you talking to medium sized websites who are some of your channel checks when you're developing your view of the industry?
Dan: Yeah, so I go to the ad tech conferences and next week I'll be at programmatic IO, which is the big one in New York. So I just build context out that way. Again, as someone who two, three years ago knew nothing about ad tech and had to ramp up very quickly talking to the, I use this term very nicely the ad tech nerds which have been very helpful. And if any of those ad tech nerds listen to this and feel like I misstated something, please reach out to me.
Andrew: The nerds rule the world now. As a nerd, people who read the website know pretty much the only thing I read are hard fantasy books. So as a nerd, I'm very happy for the nerds to rule the world now.
Dan: Yeah. And again, I've built out some context, private SSPs that give me some... and they can talk more candidly then.
Andrew: Oh yeah. When there's no Reg FD and there's no publicly traded stock price.
Dan: Just to get the scoop on what the DSPs are doing and what their views on it are. And then just generally people at publishers are always helpful. And then obviously, the agency buyers, the group M's of the world who are the ones who are literally logging into their DSPs and doing these buys and just talking to them about what they're doing and how they're evolving the way that they spend and supply path optimization and all that.
Andrew: Dan, you're a little bit different than most of the people I have on the pod because I've had some sell side on before, but you're a sell side, so there's no Twitter account. How can people who are interested here, obviously if they've got a B. Riley account, they can trade through you guys and stuff, but if people who are interested in following up with you, asking questions about Magnite and everything. How can they get in touch with you?
Dan: Sure. So email's the best way unfortunately. We can tweet, but you're pretty limited and I just stay away from it. I don't feel like dealing with compliance on social media. So I lurk and I find some value out of it, but I stay away from actually tweeting anything about stocks just because I always worry compliance is going to be on me or something. So you can get me my email. It's email@example.com.
Andrew: Cool. firstname.lastname@example.org. Cool. Dan, this was awesome. We might have to have you back on, because there are a small cable company here too that you and I have chatted about in the past you might need to talk about at some point.
Dan: Yeah. Wide up west. Let's do it.
Andrew: Cool. Maybe after Q3 earnings we'll follow up, but this has been really great. You can probably tell because we went an hour 15. I really enjoyed this.
Dan: This is good.
Andrew: Yeah. Cool.
Dan: Thanks for having me.
Andrew: Thanks, man.