Sometimes a challenging macro will mask, you know, either strategic or market share gains. I f in 2024, all of a sudden, let's say things normalize or things are getting better, you know, what do you think people will see more evidently in terms of what you've been working on this last couple of years?
Yeah, great question. I think everything we do on the commercial strategic side is to spur ad spend and on our platform in particular, not necessarily we have the clout to rekindle a macro spend. Y eah, so a lot of the initiatives that we've done of late have been working with the buy-side, so working with agencies, working with their marketing partners to try to be able to get their spend onto our platform as efficiently and effectively as possible. I think that that's kind of been the flavor of a lot of those commercial relationships, a nd so what we would expect is when spend improves, you'd see a disproportionate share going to Magnite.
T hen maybe if we shift to product, we'll start in the unexpected place. Let's start at DV+ to mix it up. You've seen tremendous strategic gains from last year, and then, obviously, there were some challenges that you went through, had to go through some reorganization. Can you talk about the changes you made in DV+ to right the ship? W hat's working now, and if there's any more tweaking and change that you still see to go?
Yeah, I think what we've learned is there's—it's always work in progress. To step back a little bit, our business is, you know, 60% is DV+, which is display video, that's mobile, digital out of home, audio, and then there's the CTV business, that's roughly 40% of the CTV growing historically faster than the DV+ business. And in DV+, it's all in one platform, and 90% of that business is open auction. It's the classic bidding, in many cases, header bidding, so there's multiple exchanges bidding on the same inventory, running that unified auction, kicking it down to the final auction. V ery complex business, a lot of wiring. The scale is astounding.
We're talking about, you know, today won't be an unbelievably busy day for us, and we'll process 1 trillion ad requests on that platform, 1 trillion. And so many things can go wrong when you process 1 trillion of anything, right? A ll it takes is half of 1%, and it can hurt your numbers that day. I t is a constant job of devoting a team, which we did, devoting the resources to it, and honestly, just half of it is probably commercial, the team that goes and talks to the publishers. The other half is internal, just making sure that the wiring's right and the signals are clean, and you tweak that, and you tweak that.
Y ou know, I would love to say there was one or two things that we did that were no-brainers, and bam, that got us back to growth, but it's 100 little things, and it'll be 100 more little things next quarter and 100 more little things. I t's just a beast that keeps having to be fed, b ut, we're elated with the progress, taking share in a market that's down. CPMs are roughly down 30%, and our business is up this past quarter, 12%. Y ou do that by processing more volume, and that's where your CapEx investment comes into play, that's where your knowledge comes into play, and it's starting to get to a point where others just can't chase that.
You just can't justify, like, 40 million more machines to try to catch up with us, a nd so we love the idea that the volume continues to grow because it becomes a moat for us.
Yeah, and maybe repackaging those first three questions again for David. How do you go on the macro when you're looking at giving guidance? Obviously, it's a very, you know, challenging time to feel for sure about anything.
Yeah, exactly that. Y eah, as we look forward in our last call, we gave some guidance, kind of higher level around 2024. We definitely are taking a very modest and moderated view to the macro in 2024, s o sort of taking the muted environment that we see today and trying to maintain even some, you know, conservatism off of that. We do have politics kicking off next year, but even there, you know, trying to take a fairly, you know, modest approach as far as expectations there. And, you know, hopefully things turn around, and there's an upside in all of this.
Then maybe get to, to the main event of all your conversations, I'm sure, as we get to CTV here. Could you talk a little bit about how you see this market evolve, maybe since we were sitting here last year? Because it feels like there has been a lot of strategic changes.
Yeah, and I think you saw that evolution play itself out in, you know, the Q3 performance and Q4 guide. T his time last year, Disney had yet to run 1 ad on their Disney+ service. Netflix, I think, had 1 million ad households. Paramount was wrestling with how they were gonna go to market. Warner Bros. Discovery was wrestling with what they were gonna name their service and, you know, came up with Max. A ll of a sudden, that inventory hits the market the exact same time. And it's this big wish, and it's the brands that advertisers are used to paying, buying, in linear broadcast.
So not surprisingly, and probably could have been more thoughtfully explained earlier, that, a lot of the digital-first, CTV-first players, if you looked at our list, who were the top CTV clients two years ago, you'd recognize, you'd recognize a Pluto, you'd recognize a Tubi, but then you'd be like: "Wow, what's that guy? What's that guy? What's that guy?" Now, you look at the top list, and you know every one of them, right? A ll of those guys aren't, you know, CTV-first players. They have these linear sales teams, and not surprisingly, they want to continue the game of selling direct to the client. So even if the client wants to process the transaction programmatically with data overlaid and through pipes, the client wants to sell it. The broadcaster wants to sell it directly to the client.
T hat service that we offer, that programmatic, the pub-sold programmatic direct, is our lowest take rate offering, simply because we're just playing as a technology partner. Whereas before, you know, two years prior, those clients that were CTV-first didn't have huge sales teams, and they wanted us to bring demand as much as possible. When I bring demand or I run an auction, I get paid a lot more for that because the service I'm providing is of greater value. And so that, to me, is the biggest seminal change in 2024, 2023, all exacerbated by a tepid ad growth market.
Because we felt as though if it were gonna grow in at a normal pace, call it 20-25%, that it wouldn't be cannibalistic, that those folks would still get their money, and perhaps all that growth would go to the new guys that are coming on board. W e think that it's going to level itself out once ad spend returns to normal. And we also think that the trend line is that the Paramounts and Warner Bros. Discovery are getting more and more comfortable with biddable, more and more comfortable with programmatic, and they'll start to move up the food chain in terms of using other products of ours that carry with them higher take rates.
L ong-winded way of saying, we feel as though we're at the lowest point that we will be from a take rate standpoint, and that the reason why we got here was because a lot of popular inventory came on. Thank God they're using us. If you look at our ad spend growth, it's over 20% for the year. So it's far outpacing the industry growth, which is pegged around 9%-11%. So double the industry growth. That's our lifeblood, that's our oxygen. You know, we're elated with that. You know, we wish it was all using our higher-tier product, but we feel as though that's, that's, you know, in a nascent market, that's step one: Are they using us? And they are using us. Is, is someone cutting us out of the game? No one's cutting us out of the game.
We feel really good about where we're positioned longer term, and we'll just have to muscle through this product mix shift for the next couple quarters.
Yeah, and if we kinda double-click on how this kind of gets sorted out, like, what breaks the publisher direct? Is it too much volume for them to handle themselves? There's an advertiser pressure on how they would like to purchase advertisements, or is it something else?
No, probably more of the latter. That's how advertisers want to buy. I mean, they don't, they don't want to replicate the broadcast model in streaming. They want to use it for better targeting. They want to use it for automated workflow, and they'll get their way, especially in a marketplace where, you know, it's people aren't tripping over dollars. I think that on the advertiser side, another phenomenon will occur, and that will be an influx over time of non-traditional broadcast advertisers. T raditional broadcast advertisers, generally speaking, have grown up on the concept of reach. I just want to reach households with mouths in them because I sell soda.
T here's a whole slew of, you know, whether they're energy drink guys, whatever the case might be, a whole slew of beverage advertisers now that are like: "That's such a waste. I want to reach college students. I only want to reach ergo, you're gonna have far more micro-targeting, which will help with CPMs. They'll all want to do biddable, because the only way you can do that is biddable. Y ou fast-forward a couple of years from now, you're gonna be talking about 10,000 advertisers, not the 500 that we're talking about today. So today, it's all about broadcast advertisers shifting to streaming, and that's not unattractive. That's a $90 billion TAM, but that's not where it's gonna end.
Where it's gonna end is it's gonna be, you know, the 10,000 new advertisers coming in, and you winding up getting the ad that is as personalized on your Instagram feed, on your TV, targeted to you.
I s what's holding up that second leg? Because, I mean, first of all, that makes total sense. When you watch a show-
I hope so.
. You're watching Monday Night Football. It's a dozen advertisers, right?
Yeah, correct.
It's insurance companies, it's cars, it's repeat, b ut when we're talking about that long tail, I mean, traditionally, it would go to something like social or somewhere where they have that. Is it data? Is that the gap of what will keep the, like, what makes that 10,000 advertisers come to CTV? Or is it being able to figure out top-of-funnel performance?
It, it's a little bit of both, right? You're going to, but the units that are coming on board now, they're shoppable ad units, or they're really performant units. So it's not just to sit back and watch a 30-second ad. So, you know, you got guys like, you know, LG, Samsung, you know, tying in ads to the remote clicker. You know, you always have the QR codes. You have Paramount doing some really creative carousel ads. So I think you're gonna see this explosion in creativity that meets that need. It'll be fascinating to see how Amazon Prime rolls out, right? So overnight, those guys are going all ads, January 2024.
It's over 100 million households, and one of the things they're doing is taking their non-broadcast clients that advertise in the retail media network and saying, "Hey, this is new inventory. Let us help you create a TV ad." So now, all of a sudden, you're getting thousands of these guys with TV creative being able to test the medium, and that's gonna, I think, really kick it off. I don't have access to that Amazon ad inventory, so it's not a direct positive for Magnite, but I think a huge positive for the whole ecosystem in teaching folks how you do it, right? And then, you know, next thing you know, you're gonna be competitive, and Netflix will have to do it, and Disney will have to do it, and I think that that'll really spur it.
Yeah, and I know we've talked about this before, so even in areas you don't have access to an Amazon or a Netflix currently, them dislodging those linear dollars does make it easier for the flow across the TV, right?
100%.
That's the hard part.
Yeah, dragging it into the ecosystem, that's a victory because it then it's there for good.
Speaking now to partners that you do have access to, right? We've seen a lot from whether it be, you know, Fox, Paramount, Disney, Hulu, Disney. I'm gonna stop right there before I just mess it up. But a very common question I get from investors is how to think about how these partnerships, like, quantitatively impact, who's gonna be the most impactful? You know, it's almost like there's been so many good press releases.
Yeah.
It's trying to figure out which one is like, should I really be paying attention to. And if you could just speak to that a little bit and thinking about, you know, 2023, 2024, you know, what partnerships are you the most excited about?
Well, on a commercial side, from you know, working with publishers, we're, you know, fully deployed with every player that's out there. So I think of that more as a, you know, kind of rising tide. As their success goes, as Max gets more distributed, as Max has more households with ads in it, you know, our boat goes up with the tide. I don't think there's one you would point to that is gonna do a seminal event like at Amazon Prime, where overnight they, they go ad. I think they'll be forced to. I mean, think about the competition. You know, Netflix and Disney have been really very careful about rolling out that ad tier, even knowing that that ad tier is the highest ARPU.
You know, Hulu's published numbers that, that $4 tier with the ad load of 8 minutes or so yields a higher revenue per user than the highest tier with no ads. So it is the way to go. The concern is, how do you get there and mitigate churn as you're working these $17 guys down to $4? T hey've been very concerned about that and very, you know, cautious. I think Amazon flipping a switch overnight and having 100 million US ad households is gonna be quite interesting in what it does from a competitive dynamic. I think you might have one of those type of events where, boom, overnight, one of the guys has 60 million. They're gonna have to if they're gonna compete, right? So that's on the commercial side.
W e've talked a bit about our buy-side tool, ClearLine, and that's a bit of a game changer for us in that it's kind of a product that was requested from the larger agencies to solve a problem that they had, which was the existing ad tech infrastructure wasn't able to accommodate some of these frozen linear dollars, largely because of fee structure, largely because of just what their needs were. T hey say: "Can you build just a simple buy-side tool that's connected to the top publishers, that doesn't even have to be biddable? These are deals that we've already cut with these publishers. We wanna process them programmatically, but there's no way we can justify the existing fee structure in the, you know, end-to-end ad tech ecosystem.
“C an you build us a more simple product?” I think, you know, we’ll be able to talk about that product as it matures throughout 2024, and that could be a needle mover for us.
Yeah. O n CTV, if you think of it more broadly, right, the two big variables is the impression growth and then the take rate, right? Kind of a fee and the volume. Right now we've seen revenue growth lag impression growth as that take rate mix changes from your perspective. Could you just kind of walk us through how you think of those two metrics, going into 2024, what the major variables are, and then if you know have any assumptions or guesses for us about how those things converge?
Yeah, that's a hard question. David, you wanna handle this?
Yeah, sure. S tarting at the top of the funnel there. Yeah, I think, because you see, this huge, huge whoosh of premium inventory coming on the market, you're gonna see, you know, obviously continued large, significant growth in impressions, puts some dampening on the actual CPMs that are being paid. Fr om an ad spend perspective, you know, we mentioned in our last, quarterly call that our CTV ad spend grew by 20, you know, greater than 20%. And so, and so we're seeing pretty significant, you know, dollars flowing into the, to the CTV, ecosystem. W e'd expect, you know, higher, you know, highest impressions, and high ad spend growth going into 2024.
From our revenue perspective, we've talked a little bit about some of the driver of that growth being this premium inventory at the lowest take rate for us. And it's a good thing because that's those are our at bats that establishes our relationship with those premium players. O ur revenue, direct revenue, will lag that ad spend growth in 2024, but we see that narrowing the gap over the course of the year.
Yeah, that's extremely helpful. Maybe staying on David. Thinking about just kind of the balance of revenue growth and profitability, especially in a challenging macro, right? And how that impacts saving now, but need to invest for the future. Can you just kind of talk about that dynamic in 2023 to 2024, kind of those key areas of investment?
Yeah, I think. Y ou know, and the answer is yes, both, both are super important. And so, you know, we're focused on maintaining some margin and cash flow, and that's one of the things that we really like about our business, is even in a depressed economy, even with some depressed CTV revenue growth, you know, we're growing our EBITDA, our free cash flow. We'll generate, you know, $100 million of free cash flow this year. I t's certainly an important component. From an investment perspective, you know, for the most part, we have the resources that we need, so we don't need to add a lot of people.
What you're seeing from a product and feature development perspective is we finished up a really important migration this past year, and so we brought our two streaming platforms together that we acquired from Telaria and from SpotX. W e had a lot of energy going into, you know, that migration, and that's sort of... Now that that's behind us, we're really leaning in heavily into new product. So you'll see, I think, an acceleration on that front. Michael's talked about, you know, ClearLine. On the DV+ front, we developed. There's a standard that enables us to facilitate native advertising in a much more effective way, and so we've developed that standard, implemented that.
W e're excited about our prospects in DV+ to continue to grow and take share, as that native, you know, continues to grow. W e'll, you know, we'll keep a view on our adjusted EBITDA margin expansion, and feel like we're, we've sort of now taken all these assets that we've acquired, and we're, you know, just ready to rock and roll now.
You basically answered this question already. It was next on the list. Speaking of all these assets you've acquired, do you have an interest in acquiring any more? ...
We just cleaned the kitchen and put it all together. So I think we've got what we need. We're super excited. We don't need to add anything to be really successful. The CTV potential that we have is huge. You know, if there was an opportunistic tuck-in of some kind that would accelerate some of our product development, you know, we'd assess that on a case-by-case basis, but nothing looming out there.
This is kind of a required question during this year in tech. How are you thinking about generative AI and whether or not there's something specific, but, like, how it might affect the Ad Tech ecosystem? We've obviously heard a lot recently about made-for-advertising content, maybe as an ecosystem for that.
Yeah, I mean, we have a whole data science group, had for years, we're, you know, refer to it as machine learning, which is definitely generative in the sense that, given the volume, we talked before about the 1 trillion requests a day. W e don't want to bring a 1 trillion ad requests to market and auction them off if no one wants to buy them. P art of the trick is understanding who buys what, what are their buying patterns have been yesterday, last hour, last month, and making sure that only those impressions come to auction, and then they get auctioned to the right buyer.
T hat's a huge, complex task in and of itself that isn't exactly ChatGPT, but it shows you that the bedrock of our business is machines and machine learning, run by, you know, a powerful data science group. I really do see a huge application in a post-cookie world from targeting. I think you're already starting to hear rhetoric from some DSPs that are like: Hey, we have this AI script that's better than any cookie that's ever been used. Magnite, feed us with that trillion, we can cycle through it really quickly, and we can tell you, you're killing, you know, 100,000 ads that you should have been sending me because I like those ads.
Y ou're gonna see a lot more of that using huge data sets that we have and being able to have scripts that can rip through them and do targeting based upon other parameters other than a cookie. Y ou're gonna see a huge explosion there. On the downside, we are already seeing a huge explosion in content, right? It's pretty easy to create content now with those tools. The good news is, we, generally speaking, work directly with the publisher. I f all of a sudden a publisher comes out of the clear blue and claims that they have, you know, 100 million impressions, it goes through a review process on our platform.
W e do our best from a curation standpoint to make sure that only the brand name good stuff is on the platform, and not this wellspring of dubious sites that are just being created as click farms.
We have about three minutes left. Are there any questions from the audience? I've been kinda combining two kind of macro-related questions. One is, you know, what are you most excited for, for Magnite over the next 3-5 years? T hen the other is, what is the biggest management decision that you think you have to make strategically to kind of make that happen?
A s David said, in 3-5 years, you know, there's, we always give our DV+ business, you know, kind of the short side in terms of conversation, just because of the allure, this generational shift of consumer behavior from primarily taking in their entertainment, news, and media through a cable box, to now having it streamed. T hat's inexorable, it's gonna continue. Will ads catch up sooner or later? Sure, they will, because it's the best, the brightest, the most affluent people that are doing this.
O ver 3-5 years, you're just gonna see this come to fruition, but much faster than everyone thinks, because that has always been the case with anyone that's trying to prognosticate on the growth of digital, anytime, ever since digital began. When they do the look back, they're like: "Wow, it happened in 3 years, not 5 years. It happened in 2.5." So I just think that the next 3-5 years, the size of the CTV business, the growth of the CTV business, is just all there for the taking. W e did those acquisitions to put ourselves in this position. So for us, to me, just executing ferociously, is a payoff on, on all that.
I don't see a crossroads where we're at, where we have to say we have to stop all investing on DV+ 'cause we don't have enough resources to do CTV. I think we're, you know, close to, you know, a little over 900 people. It's a good-sized company. We don't have to grow to 2,000 people to pull this off. And I just think that we've been building all this in kind of choppy environments and, you know, if you can tell me the next 3-5 years will be normal environments, I'll sign up.
Our big strategic decision was the build or buy in CTV three years ago, and we're really happy with our buy decision and where we've ended up.
Down under a minute. CTV streaming, we spent a lot of time on it today. Any recommendations, good shows you guys were watching?
Yeah. Did you watch the Beckham one on Netflix?
Oh, I'm married to a Brit.
Oh, there you go.
She's watched it.
It's a great show.
Yeah.
It's good. The guy's crazy.
I will, with that, super excited to have Magnite back here again, and look forward to welcoming you back next year to see where CTV is going.
There you go.
Thank you.