Thank you. It is your stage. Well done. This is Michael Barrett. He's the CEO of Magnite, and we have the CFO, David Day, and we have the SVP of IR. Real estate. Real estate. He keeps adding titles. I keep up with his title. Empire under construction. Okay, I'm Laura Martin of Needham, the senior internet and media analyst. I'm here with Michael Barrett on stage. We just had Todd, who's in the front row here on CREDEO, and he said very nice things about Magnite. I want to start with this issue of curation, because I think when I was at CES, at CES a year ago, not 2025, but 2024, the big talking point was that Amazon had basically was going to tie connected television ad units through Prime Video to purchases. We'll get to that. That isn't the new thing. The new thing is, five months ago, when you and I were in the barbershop at CES, this issue of curation, where we're going to take data and instead of having it attached to the DSP, data is now moving to be attached to the SSP. Thank God you guys did not label it an acronym, but it is called curation. I think the idea is to take some of the fees away from DSPs and attach them to the SSPs to get a higher publisher economics and to protect the data, that more people are willing to attach data to the SSP because it feels like it is protected and the SSP protects the data more. Let us talk about that trend first, and specifically, Michael, what it is, because some people, this is such a new trend, some people do not know what curation is, but also how it adds to the economics of SSPs. Let us start there. Yeah, and I think it really got accelerated during the talk of cookies going away, right? Cookie audience creation, audience segmentation was really largely driven on the buy side at the DSP level, largely with the use of cookies as signals, right? They would compile the audiences, whether it's third-party information, first-party information. What we have seen during that period of time when they're like, oh, the rulebook's changing, we saw a whoosh in activity on the supply side because the signal strength from the end user, the publisher, and their relationship with their subscriber or visitor, that was a stronger signal. The first-party data started to be utilized as audience creation. When you decorate an audience that, generally speaking, did not have a signal, and you make it more valuable by saying, these are auto intenders or these are young moms that shop at big box stores and have kids under the age of four that need diapers, when you do that, you add value to the inventory. The economics of that can be somewhat substantial. You have seen that, generally speaking, on the DSP side, where folks will say, well, I do not pay 20%. I pay 8%. That is my fee for the DSP. Yet, generally speaking, it all winds up to be around 20% because when you put your audience segment into your shopping cart and you go to check out, if you layer on data on top of it to make it more valuable and to make the advertising more efficacious, you wind up paying economics on that. Generally speaking, on the buy side, that's where the economics have always been. Now you're starting to see through our curation tools or curation in general, the publisher and the supply side partner participating in those economics that really have been outside of the realm. Why now, Michael? Couldn't we have done this three years ago? Why now? Yeah, again, I think the acceleration was we were facing as an industry a radical change in the deprecation of cookies. Folks were incented to want to try it. There's also another underlying aspect to it, and that is simplifying the ecosystem. A lot of the curated partners that we work with, partners that come and bring different units or different data and put them on our platform for buyers to get access to, used to try to sell those direct to the buyers. Little by little, agencies and DSPs have been saying over the course of the last several years, I want to limit the number of partners I work with. I want to limit the platforms I buy from. Yeah, I like your stuff, but why don't you put it in Magnite, because Magnite's my primary partner and I'll buy it through the Magnite exchange. There's been supply path. Is that The Trade Desk too? I mean, wouldn't those people be using The Trade Desk and the data would be? They're more supply side oriented. If you have a unique unit on it, if you think of a TEEDS or you think of folks that have unique ad units, publishers, that's part of curation, right? You have a cool ad unit. Instead of trying to sell that yourself on your platform, it now, generally speaking, is more available on Magnite. Okay, that's interesting. Yeah, we've seen this long-term willingness to go to fewer and fewer players. Magnite keeps winning in that. Just when Todd was on stage, he said, really, there's like two ad servers for CTV. It's FreeWheel and Magnite. You guys say that all the time, but it's nice to have a completely. We say SpringServe and FreeWheel now. I think that's Magnite. Totally fair. Totally fair. I was voting. You still have some work to do. There you go. Let's do RMNs. RMNs are retail media networks, and it's really this tie-in of ad units to purchase and outcomes. I would say we're going back one CES to the day that Amazon announced that it was going to make all of Prime Video connected television units. Of course, that sort of created a new benchmark of performance television, right? We'll have a performance TV guy later on today. That's a whole new area. Wall Street's decided that this is like the pole position, highest RPU, highest CPM is somebody that can tie connected television to performance. When you think about the future and your connected television ad units, how are you tying them to performance or to specifically purchase? I'd like to hear more. Can you talk about this trend? Yeah, and I think where we draw the line on our role in that nascent but growing trend is we'll just be the supply partner. In other words, let's take an example like Walmart, okay? A lot of the inventory that Walmart owns, pre-Vizio, but even with Vizio, right, it's more display oriented on the Walmart app or the walmart.com site. A lot of their partners, like a Procter, are big TV spenders. They'd prefer to have their ad unit through the retail media network to be broadcast quality. They'd also like it to be on Disney or on Netflix. In other words, they'll still have the data because they may be using a Walmart DSP. To get to the supply, they'll come to Magnite to buy those 30-second spots, those 15-second spots. The data is still within the DSP, so they're able to track the performance of it and look at how the impact is on Walmart because they are tracking a Walmart user on Disney. They use us as a supply partner because we have the inventory that they're looking to buy from. We participate in retail media networks as a principle, like in the case of we announced some of our hotel and travel partners, United, et cetera. Usually speaking, we're the vendor that they come to buy the supply from. The economics of that are higher CPMs, and we participate in our take rate with our publisher in that respect. Okay, so what's interesting, Michael, is that you must not think that this is where the world goes. You must not think that this drive to purchase is where the world goes because you don't have a product for this. I definitely think it's heading in that direction. It's hard to believe that every ad that's sold will be attributable to a shopping cart. We are big believers in shoppable ads on TV. We just don't feel as though it's Magnite's role to create the data to support the effectiveness of it. That is a first-party role with the retailer. It's a first-party role with the commerce partner. If you're a travel partner like United Airlines, they have all the data. They are the ones that are going to put it together. We're their media partner. They come buy the media from us, attach the data to it, and they're able to explain to their client how it was effective and why it increased the purchase cycle of that individual. We don't feel as though it's our job to be a principal in that data creation because it's data that's owned by the advertiser. Okay, okay. One of the things that Todd made the point, and I think Wall Street agrees, that the sides are softening. Where it used to be Trade Desk says, we only do demand. Well, they're doing a lot of stuff that looks like supply. And you guys with your ClearLine are doing stuff you used to be were just supply. Now you're softening as you move towards direct integrations to some of the ad holding companies. Let's say the ecosystem is becoming more fluid. My question is, is that good for you or bad for you over time? I think it's important to kind of pan back and see what the market wants. Advertisers have been pretty clear that they do not necessarily want their partner at the DSP level to also own their inventory so that they are convinced that they're getting the best possible deal and they're not working through a trading partner that owns the supply as well and double dipping per se. They have been very vocal about that, like the big agencies. One of the biggest complaints about Google's DSP, DV360, is that $0.85 of every dollar that you put in goes to AdX. Most folks do not believe that that is the best use of their dollars. I think that if you pan back, there is not a crying need from the buy side, the advertiser, the one with the actual money, the marketer, asking for a complete collapse that this same person that I buy from should be the supply side partner. I think that we're very happy with the position that we're in. We think that folks value the independence. In the instance, as you pointed out, when folks want to do very simple buying and not use sophisticated DSP type capabilities, we have this buy side interface that folks increasingly are taking advantage of for CTV when the publisher, the media owner, someone like a Warner Bros. Discovery, they create the audience segment, they price the audience segment, they know who the buyer is, they give that buyer permission to buy programmatically. That's all very simple transaction kind of stuff. Using something like ClearLine in that instance makes a ton of sense because we're not talking about a huge auction. We're not talking about multiple sites. Just an automation. That's exactly right. Let's stay on Google, which is funny, you and I. Todd called it Google instead of Alphabet, even though they changed their name. No one calls it their roommate. Stay on Google. I would say it is controversial, and you guys come down on one side of the controversy very strongly about what it means. Let's assume that the government in September forces them to spin off their ad server and their SSP. You guys have said on your call, you think that's a big deal. I would say a lot of people think it's not a big deal. Tell us why you think it's a big deal and why it's positive for Magnite. It is a big deal. What is more than likely going to happen is in the September trial, it has been clear that the DOJ's case is divestiture. They want structural remedy. That will almost definitely lead to an appeal process. In the lengthy appeal process, the judge has signaled that they will impose behavioral remedies in the interim. Google themselves has come forward with a solution, behavioral remedy, that frankly is pretty fair. All we are asking for is a fair shot. We are not asking for more inventory. We see all the same inventory Google sees. We are just asking for a better shot at winning the auction. In that case, the slightest change in win rate is incredibly accretive to our business because we are already processing the inventory. We are already bringing that inventory to auction. We're just saying that if the playing field was more level, we'd win more auctions. In that case, everything's sunk cost, and that's just gravy on top of it. It's quite substantial when you look at Google's market share of 60% plus. We are by far and away the second largest player. Our market share could be as low as 5%, maybe as high as 8%. If 50% of Google's market share goes up for grabs, we feel really confident that we have a big shot. Let me build on what he just said. What he's implying is Google has always had the right of last bid, which means when they bid for an impression or when they're selling an impression, they have the ability to look at all the other bids and then beat it by half of a penny or one penny. If behaviorally they are prevented from doing that, Magnite would win more auctions because that really is unfair. Magnite would win more auctions. They only get paid if they win an auction because they get paid in their take rate, a% of ad revenue spent on their platform. If the winner was Google, Magnite doesn't get paid. That's what he's saying is if the behavioral remedy while awaiting the appeal on the structural remedy spin-off or not spin-off, if even the behavioral remedy is to not have lost look for the Google, you'll win more. Every 100 basis points they said on their call is $50 million to them. Pragmatic said the same number, $50 million-$75 million is what Pragmatic said. Because Google's just so big. I want to push back something Todd brought up that I thought I had not thought of that I want to bring up with you. Sorry, it just made me think. Google right now, ad tech, is this big. What Google focuses on 90% of its time is search and YouTube, right? That's where its value is. This ad tech thing doesn't have the best people in Google. It doesn't have the most focus. I mean, it has unlimited resources because of search, but they really do not care, right? Whether it even makes money. It is losing share, as you know, every quarter for the last six quarters, ad tech for them has been losing revenue as they put it over in their first-party YouTube or search. If it really is spun off, is that really better for you? I understand it would lose its data advantage of being tied to this excellent data. Suddenly you would have people that for the first time, they only get paid on competing in that entity, and it will be big. It will be a $30 billion entity. It will be three times the size of Trade Desk versus Magnite. Is that really a better outcome for you? You know, I think is it a better outcome for the industry? What does spin-off mean? Does that mean the ad server gets spun off? Does that mean the exchange gets spun off? How does that work for the publisher? I think this is what the judge is really wrestling with. It might be, given the findings of their monopolistic practices, a breakup might be the right thing in terms of a ruling. Is that the best thing for the publisher? That would really strand the publisher if all of a sudden they're ripping out an ad server, having to find a new ad server. I think she seems to be leaning into the idea that if it's enforceable, behavioral might be the smartest way to go down. We certainly feel that way as well. That's the best case for you. I think it's the best case for the industry. That said, if you spin off both the ad server and the exchange, one person can't buy both because that's the reason why you're spinning it off. The question then becomes, they're two separate entities. We really like our chances in that scenario. I think any way you spin it, it's a generational opportunity for Magnite. Very interesting. Okay, very interesting. Pricing power and differentiation. We have the rise of these two new, I'm going to call them two new walled gardens. For sure, Amazon is new. Walled garden, CTV tied to purchase. All of your growth trajectory, or a lot of it, is CTV. TikTok, I would say, is a big competitor in the, let's call it non-CTV, the DV+ space. Those are to me two new walled garden kind of opportunities. In the open internet space, it feels like we're getting more of this pressure to limit the number of players. We're getting SPO, which is supply path optimization, is squeezing some of these intermediaries out. I feel like even the open internet a little bit is turning into a walled garden in the sense of scale, that the players in the open internet are going to be bigger. Because they're competing against, they have to compete. All those guys also compete against the new TikTok and the new Amazon. Talk to me about industry structure and what competitive advantage you have to retain pricing power and differentiation. I think you've seen in the numbers and particularly our growth in the non-streaming business, the DV+ business over the past year. We put up great numbers in 2024 that far exceeded industry growth rates. That shows that we're benefiting from this supply path optimization because we're gaining market share. That's why you're growing so fast and not just connected television. No question. Yes. Because sellers have said, we don't want to deal with 50 guys. We want five guys because there's so much fraud. That's exactly right. No, it's just too complicated. Why work with more partners when if you can go to one partner like Magnite, and everything that's bought and sold programmatically is on our platform, whether it's a banner ad, whether it's a mobile app, whether it's mobile web, whether it's digital out of home, whether it's streaming, everything is there that you can buy. Absolutely. Correct. Yeah, Correct. Yeah, exactly. Streaming. Therefore, why deal with, there was a time five years ago where maybe six players had to complete that portfolio for you. Little by little, the scale players are just picking up. There used to be point solutions. Oh yeah. There used to be the mobile guy you had to go to buy your mobile stuff from. There was the digital out of home guy that you had to go to buy your digital out of home. Oh, the sell side too. There was the streaming guy. There was the display video guy. There was the native guy. All of a sudden, that's kind of collapsing into mega platforms, which were the leading. Okay. You are saying that your DV+ business is benefiting, I just want to make sure I understand your answer, is benefiting from consolidation outside of connected television, or it is benefiting from the fact that you have added that you are the connected television player and therefore it benefits because people bring their DV+ business also? No question at the big agency holding company level. Two years. I'm talking about the sell side. I'm talking about the publisher side. It is just this virtuous cycle because if you have all the buyers gravitating to your platform, then the sellers are like, well, that's where the dollars are. When they go to a GroupM or WPP Media and say, hey, I'd like you to buy my stuff, they turn to them and say, yeah, put it on Magnite. I buy through Magnite. It just creates a cycle. It is global in nature because our biggest streaming partners are all global. We go along for the ride. When you go to those markets, little by little, the local broadcasters go to the agencies, the agencies say, hey, put it on the platform where Paramount is. It really does create the cycle for us. Okay. I'm going to go to questions next. I want to ask about live. Sure. Tell us what's going on with live and why it's a growth driver for Magnite, live television. It is just fascinating. Live television kind of almost equals sports, right? You are going to see in the upfronts, we were with some folks last night at the Knicks game. Yeah. Sorry, Knicks. The upfronts are quickly becoming this kind of tale of two cities. That is, if you have sports inventory, it is going to be a good upfront. If you do not have sports inventory, it is going to be a brutal upfront. The smart guys are bundling the sports with the entertainment and getting the deals that they are looking for, where the guys without sports probably will not wrap up their upfronts until late June. Everyone is just looking at sports right now. All sports are streamed now. Every major league has cut streaming deals. Consumers have voted. They like to have their sports streamed. It has been 100% the purvey of direct sold. There has been no programmatic in it. In fact, even the stream stuff that you see, it is passed through. It is the direct sold ads on the linear broadcast that is passed through. None of it's decisioned. None of the pricing is decisioned. It's all sold by direct sales forces. Not only to a list of approved guys, it's to Coke or to BMW. Correct. Correct. That's exactly right. It's so interesting to me. Fast forward five years from now, it'd be inconceivable to believe that live sports would be so different that they wouldn't want targeted advertising or data-driven advertising or automated advertising. It is going to be a huge component of our media mix. It will be a big player in streaming. One of the things that folks kind of, you know, they say, you know, they look at NFL games and they're like, oh yeah, you guys should be doing NFL games. The reality is there's only 283 NFL games a year. That was 18 weeks times. I guess that probably doesn't matter. Therefore, you know, that's really not. It's 18 weeks. Yeah. Not a whole 52-week game. Yeah, exactly. That probably will never be programmatically available. That's fine. There is so much sports out there that's streaming. Every regional sports network that's collapsing, all those teams are creating their own streaming services, their own fast services. They all need help. Think about what an SSP does in the open web. What we do is we take 10,000 little sites, bring them all together as one big site, and buyers buy it because there is scale and there is a price. All these little regional efforts, we are going to bring together, make it look like one live event, and buyers will buy that nationally. These guys will benefit from national ad dollars that they will never be able to get on their own. The national advertisers will benefit because we have just opened up a huge package for them in terms of sports. It's a lot cheaper than national sports. Bingo. That's exactly right. Okay. Questions from the audience for Magnite? No? Okay. Let's go to SPO. Supply path optimization so far has been really benefiting you because like you keep saying, people want simplicity. They want it all in one place on both the buy and the sell side. When do we end that? When do we get to hit stable state? When are all the intermediaries that aren't adding value or weak siblings, when do they all go out of business? When do we hit steady state on what this industry is going to look like for the next decade? I think you already see it in streaming, right? It is obviously scaled by the complexity of the open web DV+. I do not know if we will ever get there. It is just so much supply. There is just so much chicanery that goes on at the really tail end of it. Shenanigans. Shenanigans. I love it. Tom Foolery. You know you're never going to police all of it. I mean, we curate our platform pretty well. We know every publisher on it. We know it's high quality. That other stuff, we don't really concern ourselves with, even though you can make a buck at it. It's not something we want to do. If you look at streaming, you know the vast majority of our clients work with two SSPs. That isn't where DV+ is going. Oh, no, no, no. I'm saying if you look at streaming, it's already there in terms of SPO. It's clean. It's CTV done. Right. It's well lit. They know exactly where their dollar is going through on the vast majority of the streaming players. That's the future. I have to push on this. So I'm sitting at Possible, and one of the CEOs I was meeting with, not you, said, hey, the big problem right now in streaming is there's a lot of fraud where people tell me, oh, I just bought a Hulu unit, 1,000 Hulu loons for $10. Let me guess, it was a verification guy. $10. It can't possibly be well lit. There can't possibly be two because otherwise guys wouldn't be out there saying they're selling Hulu. That's just a corner. You're buying through the wrong people. You know I have a relationship with Hulu. You can only buy Hulu through me. I'm the only guy that can sell Hulu. If you're getting Hulu from someone else, you got a problem as a buyer. You got a problem as a buyer. Buyer beware. Yeah. There's stupid people everywhere. Exactly. Not to put words in your mouth. Okay, so CTV done. Let's talk about SPO, supply path optimization in DV+, the open internet. When, if, never? You know, look at it. When header bidding first started, the number of slots reserved for SSPs was probably north of 10, you know, south of 20. These days, I don't think our normal publisher setup that we see in header bidding is six to eight. Does six to eight go to four? Possibly, but it's never going to go to one. There's always going to be the need to want to have competition between a universal auction. It's been severely rationalized, but it's always going to be a little messy. I mean, again, the scale is enormous. We process a trillion and a half ad requests a day, of which only 400 billion of those go to auction. We are seeing the world's inventory. How many of those auctions, how many actually do you win? Therein lies the secret of the business, right? Not a lot. That is why it is so attractive that if we were able to just move that needle by a tenth of a percentage point. It's the Google thing. Exactly. Okay. Right. One of the points you're making is there's a real capital barrier to entry here that a lot of people don't realize. Huge. Not only tech, and you got to have the right people. Like when you bought the CTV, remind me what his name was. Telaria and SpotX. SpotX. SpringServe, yes. You bought SpotX. You bought a bunch of people. Yeah. I think SpringServe arguably was your best tech investment, but the other two bought you people, which are equally hard to. Yeah. It's not to the liking of the folks that sold us SpringServe. I know. I met some of them. They're all wonderful, but yeah, there's no question. A couple of years from now, people look back and say, I can't believe you were able to get SpringServe for $1 billion. Even at the time. I mean, we spent $30 million on it. At any price. It's still like the Irish idea now of buying. Everyone's like, this is such a smart move. Yeah. That is what people said, even on the day of AdServer. You guys have turned, look, to me, it is the WhatsApp problem of Mark Zuckerberg. He turned 18 people. There were 18 employees at WhatsApp. Now they are saying, oh, you bought it because you did not want to compete with them. There were 18 guys. Without Mark's, you know, moving on. Okay, CTV. Let us talk about one of the things you brought up with sports. Oh, questions from the audience. No. Going. One of the things you brought up with live sports is this issue of programmatic guaranteed PG, so programmatic guaranteed private marketplace, all of which is priced ahead. There is no decisioning. It is really hard for you to make, you have scales of take rates, right? The highest take rate is, let me call it decisioned, which is you're figuring out, you're sending it out to the open web and getting the highest rate on an ad unit. What% are we now of that? Where do you see that going over three years? We kind of look, the way we've talked about our revenue buckets in CTV, we really talk about three buckets. It's publisher sold programmatic. Okay, which I would call PG or PMP. Yeah. I would not even worry about those terms. It is just if the publisher has a relationship with the advertiser, they cut a deal with the advertiser. They discuss price. They agree upon price. They agree upon the audience segment that they are going to buy. They turn to us and say, can you execute that for us? Obviously, we provided value in that, but not ultimate value. That's a 3%-4% take rate. That's not an unfair range. That publisher sold programmatic piece is of the business. Have we given enough updated numbers? Two-thirds. Okay. The one-third is I go, I have a relationship with the buyer. I said, hey, would you like some good streaming inventory? I turn to my publishers and I say to multiple publishers, hey, we've got a campaign. It's $1 million. Would you like some of those dollars? They're like, awesome. I'm like, okay, we're all set up. We're just going to flip the switch. I found the dollars. I brought the dollars to the publisher. I made the introduction. That carries with it a much higher take rate because I instituted the demand. That's Magnite sold programmatic. Publisher sold programmatic, Magnite sold programmatic. That's the vast majority of our business. We used to have a slice of the business that was the managed service business. That's drifting to zero. In that equation, two-thirds publisher sold, one-third Magnite sold. That's not unusual at all in the early stages of programmatic. The big streamers want to have more control. They want to know who the advertiser is. They want to know what the customer experience is going to be like, the viewer experience. Little by little, they start to look at the cogs and they're like, whoa, I'm having really expensive. Sales costs. Yeah, the cost of goods sold. I'm having these really expensive salespeople go out and sell 30-second spots, which can be bought and traded with data. Little by little, what you start to see is publishers relinquishing control. You'll see an explosion of DSPs. You talked about shoppable ads. There are all these specialty DSPs that are focused on the small to medium-sized businesses that run on the Instagrams of the world, the TikToks of the world. They're taking their creative through AI and making it a 15-second spot, a 30-second spot and buying TV inventory from us because it's cheaper now because CPMs have come down because supply has gone up. All of those folks carry with it a much higher take rate because we're making the introduction. They've never had a relationship with those advertisers before. They're plugging into Magnite. We are that partner that is the kind of Magnite-sold programmatic in that scenario. That's totally fascinating. I want to do take rate. If 3%-4%, which was my suggestion, is in the ballpark, not unfair for a publisher sold, can I say 8%, which was the old Telaria as the take rate for Magnite sold? Yeah, 8% would be the lower range of that. 8% to 10%. Okay. Fine. And then the managed service today is of total non-take rate, but because I know that's a huge take rate, but it's like 2% to 4% on its way to zero. Yeah, to Minimus. Yeah. Yeah, to minimus. Okay. So we're going to get new DSP demand because supply of connected television units is up thanks to Netflix and Amazon, which is driving CPMs cost per thousand down, which allows more, let me call it, performance-based CTV advertising. Exactly. It's allowing them to test it, right? I have a really important question. My opinion from being at the upfronts is that the guys that control the most premium ad units are very focused on selling content, whether it is sports or whether it is Law & Order or whether it is some new show they announced on stage this week, which is not an audience-sold segment. It is a content-sold segment. The only way for them to protect content-sold segment is to not make it audience-based, which is commoditizable because audience is going to be substituted. When you say a third is Magnite sold, I assume none of that is sold based on the content. It is all sold based on audiences. That's exactly right. Yeah. Okay. That also means to me that it does not include most of these big content players. That's a fair assumption as well. Yeah. Because those guys are the ones that are the two-thirds that are selling the stuff directly. They have the big sales teams. They have the legacy broadcasts that they have to tie together. What you're talking about in terms of access to inventory is probably a lot more like virtual MVPDs, the device guys, Roku, LG, Vizio, Samsung. They all have a ton of inventory themselves. It could be winding up on a Hulu, but it's not being sold to the buyer as a Hulu because the Disneys of the world hate to play with the Paramounts of the world hate to play with Warner Bros. Discovery. They do not want multi-publisher tags where State Farm buys them all. They regulate their reach and frequency and say, whoops, too much frequency on Disney. I got to move some over here. Disney wants a direct relationship with State Farm. There is so much inventory out there from the streaming stuff that you are able to do these multi-publisher bundles and not disclose who the publishers are to the buyer. You see all price points, and I want to go to CPMs. You see all the price points for the two-thirds that you're seeing from Disney and Hulu. What's the CPM premium that they get direct sold versus what you're the third you're selling Magnite sold? Not on your take rate, but on the CPMs that's going to the publisher. Generally speaking, 50% higher. Wow. Yeah. The reason I ask is, as you know, we're having Jeff Green on stage later. He's always said that once you apply data, the CPMs are going to be higher, which has not been borne out, which is why, right? Because they haven't been, in fairness to Jeff, the TV guys haven't, they've been just worried about the 500 national advertisers on broadcast and making sure they catch every dollar that falls to streaming. These guys have never really concerned themselves with targeting. Like Mercedes doesn't care about advertising to a 14-year-old because they believe in their mind that they're creating a future customer at 14 and that it's all about brand. I'm not being derisive. That's just the whole broadcast mentality. The new wave of advertisers that are coming in, the new wave of brand managers, the new wave of CMOs abhor the idea of waste. They're going to be much more targeted. We're just early days. It's the same broadcast advertisers using the same broadcast playbook where the new guys come in. You see it with like a Tubi and a Pluto. If you talk to, like, Paramount, they're up to, they're up over 2,000 advertisers. 2,000 advertisers that have never advertised on any Paramount products from a linear standpoint, advertising on Pluto. You're already seeing it with the made for the medium programs. Like Tubi's the same. Like, I think Fox touted during the Super Bowl that was streamed live on Tubi that they had like 40 new advertisers that didn't show up on their broadcast that were able to advertise on Super Bowl quality. You're already seeing it. It's happening. Okay. We have two minutes left. Anybody have a last question? Okay. Then I'll ask one more and then we'll call it. Let's do Gen AI. Tell us how you're using Gen AI and AI in products and how important you think the role of Gen AI is over the next three years for Magnite. It is going to be just like everyone else in this space. It is going to be the heart and soul of everything we do. From a product roadmap, we have a product out right now. It is the search tool. It is Gen AI search for the curation. One of the things is our curation business has gotten so big, it is kind of hard sometimes to find the audience that you are looking for. It is able to generatively search the packages and create the package for you. Customize your package. We have three more products that we will be announcing next quarter. All of our efforts from that standpoint are taking generative AI front and center. Obviously, in our business, I gave that stat before, a trillion and a half impressions come in the front door and only 400 billion go to auction. That's all machine learning as to determine which 400 are the most likely to be bought, 400 billion are most likely to be bought. It is at the core of our business, machine learning, artificial intelligence. Obviously, from a productivity standpoint, our coders are using it to program. Our sales team is using it to transcribe notes, record calls, summarize proposals. Our marketing team is using it. We are very nascent in it. We only know what we do know. Our default position is there's someone out there right now with 12 guys that's looking at Magnite, saying we can take it down through AI. We have to think like that internally to protect ourselves from those 12 guys. Oh, we buy them. All right. That sounds great. I will call it right against time. Thanks, Laura. Thank you.