Hey, everyone. I'm Scott Devitt, Wedbush's consumer internet analyst, and very happy to have Nick Kormeluk, SVP of Investor Relations and Head of Real Estate at Magnite, with us today. Nick, it'd be great if you just give a little bit of background on yourself, and then, you know, introduce Magnite to us.
Sure. No, thanks, thanks a lot for hosting us and inviting us, Scott. I really appreciate, I really appreciate the, the interest and, chance to, to kind of talk to your audience here. Yeah, my background is I joined Magnite about, seven and a half years ago. I was, the CEO Michael Barrett's, you know, kind of first hire after he, he joined, you know, a couple of months before I did. Our core management team has been in place, for this kind of journey that started as a company. Rubicon Project, went through a merger, picked up three CTV assets, and, and really kind of landed on the- landed on the, you know, CTV map as, as now a market leader. My background prior to that is, you know, investor relations strategy.
I've actually been in line management positions, running full P&Ls. Ran an agency as well, been at companies like CBRE, Broadcom, Albertsons, as well as on the small cap side of the spectrum, and kind of all throughout in between. It's been a fun challenge. I, I've been doing IR here for the seven-and-a-half year duration. I picked up real estate, which Scott and I were chatting, is not the most natural kind of combination, but, I get a lot of leverage out of the cities that I go to, that we operate offices in, where, investors happen to be, you know, financial centers and advertising centers coexist in those same markets.
So we get a lot of leverage out of that, and it's been a fun journey as the company's gone through M&A and expansion and really does have a global footprint and presence. So for those of you that don't know Magnite, I'll give a quick intro to kind of where we fit in the ecosystem, and then I'll kind of talk about how the money flows, which kind of is the aha moment. You know, we're talking investors, right? So money flow is high and topical. We sit on the supply side, so effectively what we do is we support publishers, and our largest goal in life is to make publishers get paid fairly or the most that they can in a fair, well-lit, curated environment for their inventory.
So that's making sure that the most demand sees their inventory, values their inventory, uses it for the targeting that's possible to be able to come up with their ID matches and find exactly the audience that they're looking for and get a high return on their investment. At no point do we wanna make money for a publisher that doesn't return, you know, on investment for a buyer, because they'll never come back. So, you know, we have that joint mission to be very fair, but to also make sure the publishers see the most demand, which generally is done through a competitive process. You know, sometimes auction, sometimes it's private marketplaces, sometimes it's price discovery in one to set price floors in negotiated deals. Our business is largest in connected television.
Last quarter, that represented 43% of our revenue ex- TAC, followed by mobile. Mobile is in the mid-thirty-ish %, followed by desktop in the mid-teens. So that comprised our revenue mix. As you would expect, kind of the, the largest of them is growing the fastest. The middle, meaning mobile, is growing kind of the second fastest, and then desktop's in a, you know, relatively, you know, flattish, you know, at a slightly upmarket. I would say the mobile and desktop, for definitional purposes, that's what we call DV+, not because it's DoubleVerify plus, but it's, you know, display, video, and other in our nomenclature and our description, and CTV. So that's how we break out the two businesses and really speak to them on a, on a separate basis. DV+ is everything open web, right?
So it's on mobile and desktop, it's mobile video. It is actually some streaming video that doesn't come through a big screen or a large screen TV. We count that, you know, in the DV+ because it's a smaller device and it, you know, it's bought separate. You know, bought differently, and the CPMs tend to be smaller. We also have out-of-home, we have podcast, we have audio, we have native, we have display banners. All of those things kinda contribute our, you know, display, video, and other business. Margin profiles of both businesses are very comparable, a little bit different on how you process and how they operate, but they both operate at very high margins. One doesn't fund the other and hope for a better future in one or the other business.
They're both very margin-positive, high-leverage businesses, where, you know, you've got generally more fixed cost. And in our space, the way that funds flow is effectively a brand or an agency, it could be either one of those, puts dollars to work in, you know, in advertising, and they look and buy certain inventory. When they buy that inventory, they generally, not in all cases, but the vast majority of that comes through a demand-side platform. We work with, you know, 100 different demand-side platforms out there, the largest of which are like Google and The Trade Desk, and Amazon. And then through those dollars, we receive the dollars and the ad spend from a DSP, we take our fee out of that and then pass that along to the publisher.
So people always ask, you know, being closest to the money is most important in this ecosystem, and they're exactly correct. However, the demand-side platforms are closest to the money from where it's being spent. However, for publishers, we are critically important because we are their source of revenue. No longer is digital advertising programmatic a long tail of remnant unsold inventory. That was the market seven years ago. For those of you that may not have been very active in this space, now, we comprise, in many cases, a vast majority of publisher revenue and are critically important in a space that's consolidating. So, I'll pause there, and then we can kinda go into kind of what some of our... Kinda how the business model works, what our growth's been recently, or kind of go into different verticals, whichever you think is most appropriate, Scott.
Yeah, I think the first area may just be, you know, what is it that creates competitive advantage for an SSP? And, you know, so what's unique about Magnite in that regard? And what is the lay of the land competitively in terms of the extent that there's, you know, market share, other relevant players? Just be curious on that topic. It seems like there's a lot of players, but you've been able to differentiate yourself, Magnite has in terms of size and scale.
... Correct. So there's two vastly different markets, and I'll start with kind of the traditional, you know, SSP, which is the non-connected television side of the business, which is our what I call display, video, and other, or our mobile and desktop businesses. In that business, traditionally, what an SSP would do is either run an auction or participate in an auction, use data to try to identify highest-value users, gain great scale, pass along the right auctions to the right buyers, and then win auctions, and you get paid only when you win. And I would say that there was a period of time in which adding more and more bidders, it was very easy to do. You and I could download an SSP bidder online through Prebid, and you and I can be in business as an SSP the next day.
The barrier to entry was very low. You may not have scale, you may not be able to keep up with privacy and the fines that are associated with it. You may not have identification tools, but you're like, "Eh, I'll take some at-bats, and, like, what's the worst that can happen? I don't win much?", and for a while, that confused the bidding algorithms at demand-side platforms, and in many cases, the same buyer, maybe Mercedes, that was bidding against themselves through three or four DSPs, five or six SSPs, and they were driving the price against their own auction up against themselves, right? Wasn't really beneficial to the buyer, but it was beneficial to a publisher, well, that game ended four years ago.
Like, that bid confusion, bid density, and publishers were very nervous to cut off any SSPs because they thought, "Hey, somebody's sending me a check. I'm not gonna say no, don't send me money," right? So it was very difficult for them to kind of really mettle test this and say, "Hey, if I cut these folks off that are a true expense, that are consuming development resources, that have things that break, that call into me, is that revenue gonna, you know, make its way to me through another source?" We're well down that path now of consolidation, similar to what happened on the DSP side. So that kind of breath of life that expanded and made CTV... or, sorry, DV plus more democratic, and lots of people and very low barriers to entry, that has since gone away.
Privacy has made it much, much tougher for compliance and the fines that are being doled out if you're not actually passing consents. Sheerly having the scale to process the number of impressions that exist in the open internet, and being able to then filter through machine learning and AI tools, just the most relevant ones. As an example, we process 1.5 trillion ad requests per day. We only send to market the best 300 billion.
That scale of being able to comb the internet for all of the best possible inventory and send the best ones to only the best people, because DSPs don't sit there and say, "Give us all 1.5 trillion." They each give you a very small pipe, of which you have to send them the best of what you think they will buy, and you need to, in real time, dynamically change that if they gain a political ad, and they never did political advertising. And then when that spend is done, ramp it down immediately and send them something else. Or if they got a new auto client, fill that auto campaign, then replace it with something else.
You are constantly shaping the type of traffic that goes to a buyer in order to get your win rate up, and the higher your win rate goes, and the better you are at that, the more they open up your, called QPS, or your bandwidth, to send them more of the impressions they want, versus your competitors that may be just randomly sending them things and hoping something sticks and hoping something wins. So it really has become a very heavily data science, AI-driven business to really get good at processing scale and be able to get the right things to the right people, all within the compliance of the global privacy standards that exist. So I think that market now, and Amazon's probably a great example of it. Trade Desk is another great example. Trade Desk, a couple of years ago, still worked with 130 SSPs.
They cut out exchange Bidding through Google. They dropped that number considerably. I would venture to say that even The Trade Desk might only work with 20 SSPs today, but I would say there's probably four that make up 80% of their volume. Similarly, Amazon just decided earlier this year to cut off all but three SSPs. They're like: We get all the inventory we need, we see everything that that's required. This is just a cost and expense without financial benefit, so Amazon consolidated down to three. So we think that will continue to play out, and again, the advantages are really process, it's scale, it's compliance, it's efficiency, it's identification, it's making sure you're ingesting everybody's ID. We don't play favorites.
UID to us is just as good as Ramp is, just as good as Human is, just as good as first-party data is. So we'll ingest all of those, 'cause the more that you put into the auction, the better off somebody's gonna identify it and pay up for that inventory, even a Criteo or an Amazon, which tend to pay more than anybody else. So that's the advantage that we bring to that market, and we've been growing nicely, even in a muted ad spend environment for the last two to three years, we've been growing high singles, low double digits in that business, even when the CPM, there's been CPM pressure, and the market has not been robust for broad macro advertising spend.
Now, I'll pause and say connected TV is radically different, and in connected TV, the landscape had six viable competitors going back about, you know, three years ago. And really, like, probably the most important, you know, recent win that or business that was up for grabs was Netflix, and there were really only three viable competitors that competed for the Netflix business. That was, that was FreeWheel, which is owned by NBC Comcast, that was us, and that was Google. FreeWheel, you know, legacy got in the business from an ad-serving perspective. Just about everything they do is still ad-serving as their market share. We, you know. They, they probably have 40-plus% share of the total, you know, of the total TV programmatic connected TV business. We then sit at number two in the mid-twenties%.
We were probably in the teens, you know, after we finished our acquisitions and have nicely grown market share up into the mid-twenties, and Google has been losing share in the marketplace as well. They probably sit in the low teens as, in terms of percentage of market, and then you get some cats and dogs, really kind of internal efforts and kind of a hodgepodge of people that play a small role or two in, in making up the rest of the market share. That's been getting lesser and lesser, partially because we bought, you know, three of the players and consolidated those together with Magnite. We merged with Telaria. We bought a company called SpotX, and then we bought an ad server, which was called SpringServe.
In that business, your role is radically different than what I described our purpose for existence is in non-CTV. In CTV, if we said, "Hey, we'll run an auction, and if we win, we'll get paid," Netflix would look at us like we're nuts, like, "That's not what we need." What they need is they want an enterprise partner to run their ad operations. It's not simply a downloadable bidder like it is in non-CTV.
You need an ad server to be able to run and execute all the complex rules that come along from TV, like competitive separations, like frequency capping, like pacing, like rule auction dynamics that happen in bidding, that take different ways that you wanna be able to sell your inventory, either directly, spot market, in an upfront, or actually put it in a private marketplace with millions of different possibilities and rules and pricing floors throughout that auction, and follow the rules that are required for only running one auto ad or only running one alcohol beverage ad. All of those things, and then mix it out and generate the highest revenue possible. So we really do function in those cases when we're hired to be that ad operations partner, we do function as enterprise software. You only choose one ERP, you only choose one CRM.
It creates mess to your workflow if you're using multiple. But the role that we play is not simply being invited to bid on something, the role is actually running their ad operations. So that's why you need an ad server. You need those to be very tightly integrated, and that's why in programmatic, when people vote to make a choice on a technology platform with programmatic, we've been winning almost every account that comes up for grabs, and have been gaining share, and FreeWheel and Google have been losing share.
How do you think, you know, from here forward, is there further consolidation or opportunity for you inorganically to drive growth or does it come more through market share gains going forward, and you kinda have the assets in place that you need?
Yeah, you know, I think we do have all the assets that we need. There's really nothing else out there to buy. I think a comment that, you know, as we were going through the process with Netflix, that they made to us as they were doing their, you know, buy versus build, or, you know, partner and hire folks to do what they needed. The comment to us was, "You bought all three people we would've considered buying in the ecosystem that existed." So there's really no standalone independence that exist, unless somebody that had an asset decided to spin it out.
Then, you know, if NBC, you know, NBC Comcast decided, hey, they need to take care of things on their balance sheet, they don't wanna be in the third-party ad tech business, would they spin out FreeWheel? Like, that's the only thing that, you know, that-
Mm
... that would move the needle, but it's obviously, you know, not within our purview and not something that they've even, you know, remotely considered at this point. Which is also very nice, like, having that scarcity and sitting, you know, sitting there as the only independent that can play that role, it also helps us win business. So we don't, we get invited to the party. People would much prefer to work with somebody that doesn't have their own and operated inventory. Like, obviously, FreeWheel would preferentially treat NBC as a customer and client, both from a roadmap perspective, a feature development perspective, as well as ad dollars that flow into their sales force, just like a Google would.
They'd prefer YouTube versus sending those dollars to a third party who may make 5%-10%, you know, of a take rate on something versus capture 100% of those revenue dollars. But I would tell you that if you know, especially in, like, the Netflix example, if our tech didn't work, if they didn't believe that we were the best suited to deliver revenue for them over the next five years, they would've chosen one of the other partners if they would've made more revenue and been able to deliver a better, you know, a better picture to Wall Street. So we're truly chosen on performance, and being independent is a very nice, you know, add-on, you know, add-on, complement to the tech stack.
Is there a house view, you know, that's shared publicly on how what Google's going through now, you know, affects Magnite and the opportunities it can create for the company? It would seem like, yeah, who knows what's gonna happen and, you know, what they choose to divest or are forced to do so, but it would seem like it could have implications for your business.
Yeah, no, it's a pretty exciting time to kinda level set. You know, this really does, in our view, apply to the display video and other part of our business, the all other and not connected TV, because they don't have-
Yeah
... a dominant presence in CTV. So CTV is really unaffected by this. On the non-CTV side, and I think they've been careful in CTV because they don't wanna put YouTube at risk, right? So if they had that same type of dominance or... and then they haven't been sought after as a partner to kinda replicate what they did in open internet in connected TV. But on the DV plus side, you know, it's not just the ad tech trial in Texas, right? It's, you know, the search trial and others that have kinda kicked up that, you know, some type of divestiture seems to be the path of, you know, and least common denominator to be able to globally settle whatever claims and suits have gone against Google.
What that looks like, you know, is it the ad server that's spun? Is it the ad server plus the SSP? Is it the DSP? Is it the search business? Who knows, right? It can take on a whole lot of different flavors, but increasingly looking like something's gonna happen. In display video and other, the share concentration of Google is massive. They have, you know, probably 199% of the ad-serving market. In the SSP side, they probably have 50% share, and DV, you know, on the DSP side, they're probably, you know, somewhere in the 40% range, 'cause Trade Desk has done a nice job at least of carving out some share. So our view there is that something will happen, and the playing field is anything but level right now.
So when do the rule changes occur as either part of a divestiture or part of, you know, a, you know, a verdict and remedies to really open up, you know, what today is a heavy, heavy bias that's, you know, not playing by similar rules by which Google benefits and dictates their share? An example of that is something like, you know, we've seen display banners, and for certain publishers, we would increase what a normal clearing price of $1 would be. We decided, "Hey, let's see at what price we'll be required to win that auction," and we shared it with publishers.
We would start on a $1 auction clearing price and take that value, take our bid up to $2, take it up to 3, take it up to 5, take it up to 10, take it up to 20, and still lose. You're like, "Wait a second," like, there's no way that a $20 bid on a $1 display banner CPM should not win and not clear.
You can do things in your auctions like send the Google SSP from the ad server the bid request immediately, and then return back the winning bid that Google runs within their auction, and only after that, send it out to all other SSPs to submit a bid and be like, "Sorry, guys, you by the time you actually returned your winning bids, which were higher, we'll justify that they were higher, but the auction's timed out, and you ran out of time. We're sorry." So those are small, tiny, little rules that have led to Google having 50% market share and 50% take rates, right? They charge 5% for the ad serving. They charge 20% for their SSP. They charge 20% for the DSP. They charge another 5% for running header bidding.
So now you're sitting like an unlevel playing field and 50% take rates on top of it. That, that will change, right? It's not something to hold your breath for. It's not something that, you know, we're gonna be building into a 2025 guide. But to the extent that this is... You know, you've got a longer-term horizon, you've got a three- to five-year, you know, view on the space or on a company, it's, it's something that we think will manifest itself in our results over time. It's just gonna take a little bit past appeals, past what that exactly looks like. And really, to us, there's no cost. We're processing those ad requests. We're passing along the most efficient bid request. In some cases, when it's Google doing this, we may not even pass it along as a bid request.
But to us, there's no tech lift. There's no additional work. We're already playing there, and your win rate simply goes up. You incur no additional expense if your win rate goes up, so this falls through at 95-plus% incremental margins to the degree that any of their market share redistributes. And we sit there as the number two share player. I'd call it 8-9% market share of open internet, ex search and social. Below us, there's probably still Xandr. Microsoft still has a business. Maybe they're not gonna invest in this relative to other investments they choose to make, and initiatives. And then you've got, like, an Index Exchange private company and a PubMatic that kind of round out the top five in that, in that, in terms of market share.
Okay. That's helpful. And then, you know, so that's a growth opportunity, certainly. The greatest organic, you know, growth opportunity, just given the makeup of the business, the growth rates now, and we're in the fastest segment of advertising is CTV. And so be interested in the...
Your view of, you know, as CTV continues to develop, I mean, it's a pretty established and maturing industry now, how you think of, you know, walled garden versus open internet as it relates to CTV, and there's a few players that make up quite a bit of the viewership time, and how that affects or doesn't affect your business, should that viewership be aggregated among a very few number of players over time versus being as spread out as it is here in the earlier stage of the development?
Yeah, no, it's a great point, and that kind of dictated our approach of entry into the business versus, like, building a product and hope to then sell, you know, sell a good product and crack lots of long-standing relationships. We went about it from an M&A perspective. To your point, you know, we quote that 30 market participants in CTV control 85% of the world's CTV inventory, right? So it's very concentrated. It's not like the open internet with millions of websites and hundreds of thousands of apps, right? It's very fragmented, and even if you chose to specialize in one group, you know, you couldn't specialize in them all, right? So there's degrees of specialty where, you know, where somebody can have, you know, a lead or cozy up to certain publishers and do better or worse for some.
But in CTV, it's vastly different than that. And the agencies also, you know, are upset about relinquishing a lot of what their control was in media buying previously in CTV by building marketplaces, by still being involved in those transactions and those media buys. So even agencies want to try to play a bigger role. So I would say that two, three years ago, we would've been sitting here saying, "Yeah, you know, I really hope that we can win some of these large accounts." Our revenue composition would've been folks like Tubi and Pluto and Fubo and Crackle and Dish and Sling. Now our revenue is comprised of folks like Netflix, Warner, Paramount, Disney, Roku, LG, Samsung, Vizio, so all the largest market participants in the ecosystem, and we've won them over playing a role.
The only inventory we don't touch, and you may not even define it as CTV, is YouTube, so obviously Google serves YouTube. And then on the NBC side, NBC Comcast is served by FreeWheel. So out of the you know largest 30, we have 28 of them that we transact with. So that really is helpful for us. They're all at different stages of their journey, some that have had large broadcast businesses started off by very lightly putting this into programmatic, meaning they'll have us execute deals, but it'll still run through their sales force.
Like a Disney, for example, or a Warner or a Paramount still moves most of their inventory through the sales force, but they call it programmatic, and we say programmatic light, because we'll allow you to target and buy and segment audiences to allow them to execute it, versus just running the same ad for everybody to see, like a cable or a satellite ad would. So they're in those journeys. Roku is one that, that's in a transition stage. Roku is moving more to programmatic, opening up to more DSPs, bringing in different demand sources, and as they make that journey, our take rate goes up, right? If we're being used instead of a sales force, instead of paying a sales commission, we have a very nice lift to using machines to sell versus using salespeople to sell.
We truly believe that the market does turn into not just the top 200 media buyers that bought on linear, but it expands to five, 10, 15, 20 thousand buyers, and now 100,000. A $20,000 check, if you're trying to reach just your audience, matters because you don't have to hire a salesperson to get that person to execute their inventory. AI is also disrupting the cost of an ad. The cost of an ad three years ago was $500,000 to run a television-quality ad. Nobody. If somebody had a $20,000 marketing budget, they couldn't even make an ad. Today, you can create one using AI tools for a couple of hundred bucks. Amazon will create your free ad for you if you spend $70,000, and then you can take that ad, download it, and use it on any other platform.
So the hurdles and the barriers to entry to be able to, first, be a relevant buyer, two, to get your creative for a reasonable cost, is all opening up this market to the way it should transact and bringing in those new dollars. And all the broadcasters talk the right, you know, language. All the TV OEMs want those dollars. So there is going to be, and is currently underway, a movement to bring all these new dollars into the ecosystem and really expand the buying pool for television.
Great. Maybe the last question is, it would be, you know, how do you think a Generative AI affects, you know, the industry and the company over time?
Yeah, it's, I would say that, TBD on kinda how it affects and impacts the search market. So I think the search market, you know, is changing for how people, you know, explore, look, or seek guidance and counsel, so I think it will be transformative in search. I think, you know, for us, we've had some machine learning and tools that we've used that now are getting greater adoption, where people are like: "Great, let the machines make the decision. Auto-configure these new ad pods and campaign." The savings is you can launch it today versus a week from now, right? So you're generating new revenue faster. A lot of those auto configurations that people would want control over are being adopted quicker, and I think that on the creative side is the biggest disruptor.
I think Gen AI. I was on an investor roadshow two weeks ago in Toronto, and the person, you know, the investor that I met with, his nineteen-year-old son built a Discovery ad for all the Discovery kind of Naked and Afraid, and lumber, you know, lumber haulers and everything else. He tells me his son came up to him and he's like: "Sorry, Dad, I maxed out your card, and I created this ad. You wanna see the ad?" He's like: "No, tell me first, you know, what card you maxed out." And he's like: "Oh, the $50 Apple gift card you gave me." And he said he ran the ad, and it looked better than what Discovery runs for all their, you know, extreme, you know, wilderness, you know, series for $50.
Mm.
I think that's a perfect example of Gen AI really, really disrupting and bringing down a hurdle to entry into this space. And these ads are passing creative tools, ingestion tools, that we now build so that a live person at ABC doesn't have to sit and review every single ad that's gonna run on ABC. Now we have creative review tools that process that, kick out the exceptions, and only the exceptions, the 1%, get reviewed, versus the 99% plus the 1 get reviewed manually. I think that's the biggest disruption that AI is gonna have in our space.
Mm, perfect. A very good intro to the business and a lot of great insights, and really appreciate your time, Nick.
Yep.
Thank you.
Thank you. Appreciate it.
Bye-bye. Bye.