The Trade Desk, Inc. (TTD) Earnings Call Transcript & Summary

If everybody can please take their seats, we're going to be starting in just about a minute. We'll start out the presentation should be live on the website. There's a card on your seat or on the table where you can also scan and go there may take a couple of minutes just with everyone going at once. So you'll have to be patient there. But we'll hope that the WiFi system here is not overloaded. I just wanted to thank you all for coming. I'm Chris Toth, I'm the Vice President of Investor Relations. I don't see Jake Graves is in the back. He's a Manager of Investor Relations. We really appreciate all of you traveling to get here or battling the elements to get here. We have a really good crowd. We were oversubscribed here. So I wanted to thank you for that. We're just going to get started right now with the presentations. Just real quick. We've talked about downloading the presentations. There is no Q&A during the presentations. We will have a Q&A session following Blake's presentation so that we can ask all the questions then. We have 2 10-minute breaks planned throughout the day specifity should take about 3, 3.5 hours in terms of the presentations. And with that, I'm going to have Jeff Green come up on stage, Founder and CEO of the Trade Desk.

Hello, everybody. Hope you're having a good day. Sorry, we're a couple of minutes late. I think everybody is a little rusty from not doing this for a little while. So we did the first one 5 years ago today and the second one 3 years ago. So I'm excited to talk about what's changed in that time because everything did really everything did. So today, I'm going to talk about our vision and our strategy for growth. And at the risk of skipping at the end of the book and making this boring, we have not changed anything. Our vision is exactly the same as it was before. No major changes. But again, the world has changed, our ability to influence has changed. And this is as much a discussion about the open Internet as it is about the Trade Desk. And we are more focused than ever on the open Internet than we've ever been. Honestly, when we started this 5 years ago, when we started talking about the possibilities of the Trade Desk, I mean we pointed to the same TAM we're going to talk about today. We pointed to some of the very same issues that we're going to talk about today. But in terms of our role in it and being a leader of the open Internet, and in some cases, the voice of the open Internet, I never really thought that it would shape up quite that way. I thought there would be more of a symphony than there has been and there's just such an opportunity for us to continue to lead and change the market. So we're just constantly looking for ways to improve the open Internet, and we think that's our way to compete with the biggest players on the Internet and arguably the most successful companies in the history of technology, and many of those are the most successful companies in the history of capitalism. So we're really excited about what it means to marshal the troops and the army of the open Internet to make it all better. So the pandemic change everything. I don't know to spend a lot of time on this, but you all know worldwide inefficiencies were essentially hiding in the Zoom shadow as the money supply increased by 40% in 3 years. U.S. monetary policy, fiscal policy, global supply chain issues, U.S.-China relationships and of course, everybody stayed home and streamed. Internet and arguably the most successful companies in the history of technology, and many of those are the most successful companies in the history of capitalism. So we're really excited about what it means to marshal the troops and the army of the open Internet to make it all better. So the pandemic change everything. I don't know to spend a lot of time on this, but you all know worldwide inefficiencies were essentially hiding in the Zoom shadows the money supply increased by 40% in 3 years. U.S. monetary policy, fiscal policy, global supply chain issues, U.S.-China relationships -- and of course, everybody stayed home and streamed. But as productivity changed around the world, the opportunity for us got bigger. And one of the things we're going to spend a lot of time talking about today is a secular tailwinds, a few headwinds, too, but secular tailwinds that are just unbelievable. It was interesting in January of 2020, I address our entire company at an event. And I said then, I am more worried that the wave of opportunity is too big and that we're not ready for it than I am, that it's smaller than we think. Again, that was in January of 2020. And then March of 2020 happened, we were uncertain. As you'll recall, that summer, there was a bit of a downturn for about seemingly 5 minutes. And then before you knew it, that wave was bigger. And what looked like a headwind for a moment actually turned into the biggest secular tailwind we've ever seen, maybe ever will. So we talked to them 3 years ago about how we fuel the world's growth Yes. As she said, it looks the same, yes, I may or may not be wearing the same suit. Fit better back then. No, I don't think it was the same. I don't think I would have to -- but a couple of things that we said from that Investor Day that I just wanted to talk about. So I'm going to go quick because you already heard this presentation or many of you did. I said back then, CTV and video will be our largest channel it already is. So we -- and we said roughly 5-year time horizon. Again, that was 3 years ago, it already is. We said that our ID footprint will be larger than any single company's login footprint. We have -- and Samantha is going to go over this in more detail and be more precise in the numbers. But we have over 3 billion devices that we see on UID. As you know, UID2 is an open source. So it doesn't mean that we see everything. Everybody has their own instance and can use UID their way. So -- but just the ones that we see over 3 billion devices, of course, many of you know that Google has about 3.2 billion log-ins. -- Logins and devices are a bit different, but having over 3 billion devices we're well on our way. So I am confident that this statement will be true in the time horizon that we set. Grow data usage at least twice the pace of media spend. We've lived this. The pandemic accelerated that. And honestly, economic slowdowns, recessions, make it so that people become more deliberate and more data driven in the choices that they make. So if that's what's ahead for us, people will become more deliberate, more careful about the way that they spend and expect more from every dollar. The only way to do that is with data. So I said then China will become a top 3 market for the Trade Desk. There are a number of themes that are -- my notes are different than the ones that I sent them. I'll speak to the notes that I wrote instead of the ones that are up here. But a couple of things on China. First of all, most marketers are afraid of it. They're afraid of it because it's different because the government plays a different role in the China market than the one that they have in almost every other part of the world. And so when you're looking to make investments in China, you tend to make -- underinvest compared to the opportunity that exists -- and so what they -- what most brand marketers want is somebody that they can trust, especially with their data. And we'll talk more about the way that we've won trust for the biggest brands in the world. But that represents an opportunity for us that is very different than other companies going into China because we bring money into China. So different than taking it out. Now there certainly are huge opportunities for us to do outbound like we have like a Facebook has, for instance, where you take Chinese companies and help them advertise the open Internet around the world. But then of course, there's opportunities to advertise domestically. Last year, this was our fastest-growing office around the world. The 30-ish that we have around the world, this was our fastest growing, both in terms of spend growth as well as just growing our team. So China continues to be a tremendous opportunity for us. Of course, there's been a bunch that changed because of COVID and geopolitical issues. But I think slow this down a little bit in terms of becoming a top 3, but I still have as much optimism as I ever had for the opportunity in China. So we said back then price discovery is foundational in healthy markets, that's even more true for us today, and we're even more focused on it than ever. We also said the data driven is better than guessing. That's fairly obvious. But there's a lot of guessing that goes on in advertising even today. It wasn't just in the Don Draper era. It continues today and that's essentially what we're replacing with what we do. We said then we spent a lot of time pointing to the $1 trillion TAM -- and that was before we diluted the dollar a lot and then build it back up. But one thing that we didn't really foresee and didn't quantify as well as I think we could have is just the opportunity that exists in shopper marketing that actually widens the TAM. So that $1 trillion TAM, we're about $750 billion today. We're on track for that to happen. But you can argue that of the $500-ish billion in shopper marketing that if half of that is available or part of what we could address today that we're already at $1 trillion TAM. So we said then that we want to empower agencies and get closer to brands, I don't know that we've lived up to any of the statements that we made in 2019 more than we did this one. We've just done this so well. The team has done this so incredibly well. And we'll talk more about how we have. But this is, again, something we've we said in the S-1, we said in the document that I wrote that was basically the preface to a business plan 13 years ago. The buy side will always be in the power position, -- so we said this then, I just think it's really important to note this because you can get confused by this. And especially as you look at CTV and some of the really strong position that quality CTV content is in, but then also some of the positions that other tech players are in. And sometimes when somebody that has a lot of supply doesn't do well on earnings, people assume that's going to affect us. And I think often it does not because we lose sight of this fact. In advertising, there is more supply than there is demand. There always has been and always will be. The reason why that's true is because you can put up now showing ad 1 of 3. And then during political season, you can make it now showing at 1 of 5. There's no reason why you can't add more supply, especially in a digital environment. And so given that fact, it is always a buyer's market because as soon as you get close to equilibrium, you can create more supply. So from our perspective, the way to create the very biggest company and capture the opportunity is to make certain that we align our interest with the buy side so that we're working on their behalf and using that power position to align and grow and do something different than those that are big trying to service both sides that are full of conflict of interest. We're going to spend a lot of time talking about conflict of vendors today. Decisioning is where the media is where the money is. So I actually don't think I need to spend time on this because all of you live that inside of equities markets, commodities markets, that's proven over and over every day. Those that decide what to buy and sell are the ones that make the money. So this is one where I feel like I was only half right, and I'm going to spend a lot of time explaining why today, but sell side will continue to specialize in fragment with the statement that I made 3 years ago. I believe that the specialization is happening everywhere. And the fragmentation has happened in CTV, but the fragmentation has not happened on most of the browsing web and in most of mobile, and we'll talk about why. We said then -- we didn't actually create a slide for this then, but we did say this then. -- that we're obsessing about supply chains. I think one way to summarize the success of Amazon is just that their executive is obsessed about supply chains. How do we make them more efficient? And if we make them sufficient enough, we can ship product at the best prices than in 2 days. And sometimes, at least at an aggregated level is just that simple. If you obsess about supply chains and make it more efficient for the people that you represent, you can create something that hasn't been done before. And that's, I believe, critical to our success is that we obsess about supply chains. I think there have been times in our company's history where we've said, if we focus on the buy side and the other parts will take care of itself, and that's not necessarily true. At times, we have to disrupt the inefficiencies of the supply chain as fast as possible in order to best represent our customers. So I said then that I believe the buy-side aggregates around 3 to 10 players because there's economies of scale. And I would -- I said then that objectivity matters more tomorrow than it did today, and it matters more today than it did yesterday. That's a complicated way of saying every day that goes by, objectivity matters more. So Google, Apple, Amazon, Meta, they all have a place in advertising. We said then that they have to get into this space, and there were some questions about like, well, Apple get into the space. Of course, they have to get into the space. multitrillion-dollar company has to be touching things that have trillion TAMs. And advertising is more orthogonal to their business or closer to their business than drilling for oil is a comment that I made back then. I'll stand by it today. So of course, they're going to get into the space. but none of them really have a place in objectively buying the open Internet. And most of them are trying to monetize supply of their own. And it actually creates a very complicated problem for all of them because much of the inventory that they have, they're swimming in it and their cost of goods sold or TAC is almost 0. And when you contrast that to things that are very expensive, like premium content and CTV, it becomes very hard to have that barbell strategy where something is very expensive and something it's got most free and then to monetize those well and especially make good trade-offs for the advertiser. So many of those, like Meta, for instance, at one point tried to monetize the open Internet. And as Sense said, we're just going to monetize our sites. And I think there's some wisdom in that, especially because I think there's some honesty in saying to your advertisers. Yes, I think mine is the best. And I'm going to try to help you monetize there, which is different than saying, "I'll objectively help you buy the Internet. And by the way, I think you should spend 90% of your spend on my stuff. So at least one winner will be independent. And I think it's likely to be the biggest as well because of that objectivity. And especially if there's only one objective player, which seemingly is the way that it's shaping up in the last 3 years, a lot has changed in terms of the competitive landscape of open Internet DSPs that I think we've clearly led the pack. We're the most likely to be the winning independent and objective player. And I think that also means that we're the most likely to be the biggest. So that was a look back. I told you it would be brief, I wasn't too long. And now I want to talk for a minute about some of the macro forces that affected our business. Some of these are fairly nuanced to talk about the landscape. So starting in March of 2020, the wave got a lot bigger. So as I mentioned, in January of 2020, I said I'm really concerned that the wave is too big or bigger than we can be ready for, and we just have to work as hard as possible to be ready for it. For instance, this year, we set a goal to hire 1,000 people I wouldn't have thought that we would have done that in early 2020. Incidentally, we're on track to hit that goal. And it's not been without lots of work to protect our culture to protect what we built and to be very organized, but to continue our focus. And thank God, we had a very clear vision and a very clear focus going into the pandemic because those that didn't really struggled, especially if your strategy was one where you say, hey, let's just try to respond to the times, and we'll figure it out, which a lot of ad tech companies have done that. Like every 6 months, they're just sort of reevaluating and they just kind of shift. And especially if you're a public company that does that every quarter, which is seemingly some do, then you're in big trouble when the world changes as much as this did. But our focus has always been the same. Our vision has been the same. We want to represent buyers objectively. And we want to help them make the decisions on the entire open Internet, and we believe you have to be global and omnichannel to do that. So the use of data-driven advertising surged. So it's actually, I think, difficult to explain holistically why this happened so much. I mean, some of it has to include cuts, the decks are out of sync, whoever is moving the decks. Some of it were because people had to make cuts early on in the pandemic. Of course, there were supply chain issues. Of course, people are expected to do more with less. Of course, there were some companies that had especially big companies that had weird comps and that meant that the advertising spend was lumpy. But really, everyone was expected to do more with less. And especially when you take advertising teams and advertising by its nature, is a very collaborative thing, right? And the agencies typically are very collaborative in their culture and the way that they operate. That was extremely difficult for them to do in a pandemic where everybody is working from home. So the only way that you can be disciplined in the way that you spend is to be data-driven. You put all of these things together and everyone had to become more data-driven, and that's exactly what they did. And it meant that programmatic moved up. So as I mentioned, global advertising continues to the path of $1 trillion. And then, of course, there's the expansion of shopper marketing where that is all coming together. When you -- shopper marketing is, of course, when you add loyalty programs and sponsored listings and all those things, you get to roughly a $500 billion TAM. And I think over time, most of that is available to the Trade Desk. Even things like sponsored listings where it's just optimal for those to get as much demand as possible. And again, back to the -- there's more supply than there is demand, welcoming as much demand as possible becomes increasingly important. So let's talk about some of the biggest players in tech and just some of the most controversial issues that are facing the landscape today. So rumor has it that the government is going to file suit against Google, any minute now, and that Google has really struck or stuck between privacy and antitrust. And honestly, they keep, in my view, keep oscillating between which one they're trying to respond to and which one is the bigger threat. And what I think this has done is made it so that it's very hard to get things done inside of Google. So we've had some tell us that it feels like Google is run by lawyers today as they're trying to navigate these 2 very different issues. We'll talk more about that. But that does make it so that Google is operating a bit more -- they're a bit more slow in the decisions that they have to make, and that represents an opportunity for us, which again, I'll talk more about in just a second. Let's talk about identity. So as long as we're talking about the big tech, it's impossible for us to talk about what's going on, on the Internet and the landscape of the Internet without talking about identity. Identity is just shorthand for the concept that there has to be a way to understand anonymously and in a privacy safe way who a user is consistently and by who a user is. I don't mean their name Social Security number or anything like that, but instead, in an anonymized way so that you can provide personalized content relative to the interest of an individual. And everyone is trying to figure out how to do this and everybody has different opinions about the way that it should work. And I just want to walk through what that war looks like. It really is the open Internet against, I would say, 2 companies, Google and Apple. And if it looks like Apple is a bit of a pack man heading towards meta, that is probably what it feels like to meta to where if you were trying to explain how Meta lost almost half of its market cap. I don't think you'd point to the macro trends nearly as much as you do the war on identity. So I don't think that it's well understood what's happening in identity. I don't think that the land grab is well articulated. And so I want to try to take a minute to talk about it today. I believe that there's really 2 primary approaches to identity. And when you abstract them, they're very philosophically different. One is saying, "I'm trying to aggregate all the power. And sometimes the rhetoric is, I'll take care of this for everyone. And the other one is trying to federate at all. It's almost like an authoritarian government versus a democratic one. How do you handle power. There is no question that there is power, but one is trying to aggregate at all. The other one is trying to distribute it. And the risk of this is that there's a bit of asymmetry in this. And it also is -- the risk reward, I think, is very different. So what -- actually, I'll come back to that point that I was just going to make. The reason I put this up here is because the war maybe -- maybe the past image is important because there's a sword and a shield and privacy in particular, is being used as a sword and a shield. It has been weaponized but it's also a defense to prevent big companies from being auditable or sharing any insider data. So it creates less transparency in the overall marketplace. But you can also say that it's a cold war and that it's very subtle in terms of the way that it's being interacted or the warfare is very subtle. And this is a great example of that. Where, on one side, you have that, would you like personalization? It's wonderful. On the other side, are you sure you trust them and they're both essentially doing the exact same thing. And the prompts often on the open Internet as the way Apple manages it are often not very clear. I would argue deliberately, ask the app not to track. I don't know what that means. If I ask them, what if they say no, how do -- what do we do then? This idea that on this one, we turned it on because it will make things better. On this one, are you sure you want to allow it. And sometimes the prompts become even more one-sided but really what Apple has done has created, I think, a misperception. I don't know that it was deliberate, but I think it's happened anyway. The misperception is really what I say there in 4, which is they created the perception that they manage privacy for the entire Internet. They like I think many consumers think that this is this is supposed to protect the entire Internet. And really what Apple is doing is they're protecting their ecosystem. As long as you stay inside of their ecosystem, they are taking privacy very seriously. I don't think you can dispute that they're taking it seriously. But the debate about what tactic is the best, I think, is a real one. Because you're aggregating all of this data and because you're -- I believe what Apple has inadvertently done has created themselves as sort of an intermediary for consumers on privacy, I think that comes with great risk. Because if the consumer misunderstands or if bad things happen when you're in the middle of everything in privacy, which I would argue they will. I think that is inevitable. Then you have set yourself up for a risk reward that I don't think is worth it which is why I argue at 5. I don't think any company should want this job. No one should set themselves up to be in the middle of all of it. So I believe the only answer. And this is, I believe, where we land, irrespective of how long it takes us to get there when the risk-reward gets weighed out, we will end at a place where consumers can clearly understand who they're taking risk with. They have a relationship directly with individual companies, and they choose where they decide to trust people. And then they have control over those settings and they can undo it. And only with this type of setting, do I believe that we can comply with things like GDPR and some -- all the rules that are basically include things like the right to be forgotten and control and consent. So one analogy that I've used every analogy is imperfect in this, but I think it helps. It is -- there are companies who are saying, why don't I create the biggest safe in the world, and we can put all the data in there and especially the sense of identity, which centralizes the way that data gets used we can put in this central database and all have eye scanners and really thick walls, and we'll talk about all the things that we do for security, and we'll put it all in these 2 amazing safes or 3 or 5. I think that poses a too big to fail risk that is as great, very different, but as great as the 2008 global financial crisis, where we say we've created something that's too big to fail except for the entire Internet is based on it and including the way that the entire Internet is monetized. You look at the trickle-down effect, for instance, of at Facebook being less effective in their targeting since identity has been weakened a little bit. But the trickle-down effects include lots of small businesses that are much less effective than they were before. And if you're less effective and yet still taxed at the same rates, then it has a devastating effect on the economy. Now if we've done that inside of one company, what if we do that in regard to the entire Internet. So essentially one way, and we're going to talk about it in more detail and Samantha is going to cover this better than I will. But one way to describe what UID does is instead of having 2 safes or 3, we have 1 million of them, millions of them. And now what some will argue is that by having 1 million safe, you have more security risk. And you're right. You are right, you have more security risk, but it doesn't mean a bigger one. And it's okay if you put $100 in the hotel safe and use the passcode 11111 because you're risking $100 I think the greatest risk to the system, to the Internet is to put it all in one place. And especially when you consider the fact that every database in the world has seemingly been hacked, whether we're talking about NSA or FBI or Microsoft or wherever, like everyone has had some security breach. And if you put everything in one place, that just creates a risk. The very first rule we created when we laid down the gauntlet for engineers to create UID was to not create any honey pots is what we call them. Don't create a central place where everything can be stored and everything can be rocked. That is the problem. And do not move data unless you have to. So when a company has data, let's not say step one, put it in our system. We think that is inherently dangerous. And that is different than what many of the biggest companies in the world are asking of brands and advertisers and credit bureaus and so many other companies. So much more to come on UID. But I find this to be extremely interesting. And I probably spent more time thinking about Google in the last year than I did in the last 10 and it's interesting because I think that might also be different in the way that Google thinks about us. When Sundar was a code a few months ago and was asked by Kara Swisher, who's a very direct interviewer, you say you have competition, but where is your competition really and he pointed to TikTok and the Trade Desk. I thought it was very interesting in the appointed the Trade Desk in part because only 3 months before that, I was sitting at the dinner next to Eric Schmidt and I introduced myself, and I said I'm the CEO of the Trade Desk and said, Oh, what do you do? So I've spent a fair amount of time thinking about how did Eric Schmidt not know what the trade -- and by the way, we then went on to talk about what the Trade Desk was and it was all new to him. And if you weren't following dates, he was CEO of Google for 10 years, then he was Chairman and has been an adviser up until 2020. So it's not that we were nonexistent, but that -- it is also true that even today, we're only 2% of Google's market cap. So to point to us as one of the biggest indications of competition, I think, is interesting. So let me take a stab at explaining why I think those can coexist. And this is just me giving you my theory and spending a lot of time watching what's happening inside of Google. So we say all the time, we don't compete with Google, big Google. We don't have a search engine. We don't have fiber in the Pacific Ocean, where they don't have a space program. They don't have self-driving cars. We don't have user-generated content video platform. We don't have a publisher ad server. We don't have a browser. And all of those are some of the, what I would call, priorities 1 through 36. We compete with priority number 37. I believe that priority number 37 doesn't really move the needle on a Google P&L. And that's one of the things I think is beautiful about competing with Google and what we do, especially because we do it objectively. But I think the combination of us doing it objectively as well as in order for the 37th highest priority to compete with us, they have had to make the market less fair. They've had to use tactics and approaches that are at times seemingly draconian to the marketplace. And I believe that has created a risk for big Google that is not worth it. So I believe that, that has likely been escalated to Sundar. And because those decisions have had to go under great scrutiny, it seems like Jedi Blue and Bernanke and all those things that have had lots of headlines have made it so that what was previously a part of Google that was ignored by the CEO is now a part of Google that represents a risk. And also I think if we do really well against priority #37, it likely has a domino effect into the rest of the business. By the way, therein lies the problem that so much of this tech is integrated into each other at Google that unraveling it and making certain that it's fair, is extremely difficult to do at Google. It seemed to me that the strategy before was to let every division operate independently but there became a moment where they started to work together and started to create incentives across different assets. And I think that has created a risk for Google. That has also made it so that Google has created the illusion of competition on the sell side. So I said before that I was only half right about the sell-side wood fragment. The sell side has fragmented especially in connected television because connected television is extremely competitive. And by the way, it's perfectly concentrated, which I'll reiterate in a few minutes. But the rest of the Internet is not as concentrated. And Google has a bunch of tools and techniques to make it look like there's more competition than there is. I actually had contemplated going into great detail about how this happens. I think I'll save it for another day. There might be a time where we'll publish an invitation. You're all welcome to come it will likely be after earnings. And we'll walk through what exactly has happened to the sell side. I believe that backdrop is really important to understand the state of the landscape but also to understand the opportunity. So is probably the most tax dense slide that we have and certainly one of the more controversial ones. But I believe that Google has had a negative impact on the ecosystem and competition. And that's, in large part, for those of you that have been with us since the very beginning, in order for us to take a meeting at the IPO, we asked people to watch a 12-minute video on header bidding. Without header bidding, I don't know if the Trade Desk would have had the opportunity that it did because what header bidding essentially did is it made it so that there was an if statement in front of Google's ads server. It made it to the open Internet created better price discovery and a publisher could manage yield without relying on Google, the same way that it had in the past. DFP was still in the middle but some innovative companies like Index Exchange created hacks to make it so that Google's decisioning was weakened. Google fired back by creating a product called Open Bidding. I think it's the most deceptively named product, perhaps in recent ad tech, it was not open or transparent at all. I would argue that its biggest victory was pacifying SSPs to stop doing publisher development and just rely on Google. None of them could afford to interact with publishers on a regular basis and win client relationships with them for less than 5%. So just letting Google pay them to do it meant that Google secured their place in the middle. And it makes it extremely difficult, I would argue, even impossible. I've heard SSPs argue impossible for them to disrupt Google as the yield manager. And then there's other programs that created things that are more one-sided. All of this is built on top of the search demand. So if you don't enable Google to use these tools as a publisher, then you do not get access to the millions of advertisers that come from search. I think it's telling that Google has had to rename their products over and over again. They don't use terms like exchange anymore. I think that's because what they do isn't an exchange. It doesn't create transparency or price discovery. It's not built on fairness, which is what I believe is at the core of what Google does. So why take the time to talk about this today at Investor Day. I would say it one to understand the landscape and to understand that if we have won and done so well, in a market that is unfair. So we found ways to do that. And if Google is under greater scrutiny than they've ever been from governments all over the world, not just in the United States, but all over the world. You would think, especially when you really -- if you really understand the risk that's being taken there and the reward that is not likely -- I don't believe it's rational for Google to continue that from a strategic standpoint. The only way that it gets there is if you let those 37 priorities or 50 or whatever it is, operate independently. But as soon as you start thinking holistically, protecting the company, then the risk/reward is no longer worth it. I think it's highly likely that the Trade Desk will be competing in a much fairer market as time goes on. And if we've won in an unfair market, imagine what we can do in a fair one. And I believe -- and I said this at the last Investor Day that it is highly likely that the connected television is the way that walled gardens come down and at end state, there will be no walled gardens. Understanding the current state of the landscape, I think, helps you understand the dominoes are falling in our direction. The pressure that has been building on Google has made them at times, act more aggressively in order to preserve what they have. And I believe that creates a risk reward that is not worth it for them. And you don't have to go very far to not be looking -- or to see headlines that are scrutinizing that. Incidentally, and I'll talk about this in a second, one of the things that we've done is we turned off open bidding. It's had a very positive effect on what we did, but it also was another instance of the Trade Desk leading the open Internet because a lot of other DSPs, including Yahoo and Amobi and others turned off open bidding themselves. And when we did, we publicly said that we did not believe that it was a fair -- sufficiently fair or transparent market for us to continue to transact it. Because of this backdrop, I would also argue that journalism is struggling more than ever. And what I mean by that is not that there isn't quality journalism, of course, there is. What I mean is that it's very hard to pay for it. And journalism, as you all know, has been struggling to monetize properly to really understand the Internet and transition to a digital world. And that has been exacerbated and worsened during the pandemic. I think, of course, because that plays a critical role, especially in more democratic governments, I think this adds other pressure to the walled gardens and those that have very high take rates to do something different because of the effect on the system. But there is also an opportunity because that is very fragmented by the very nature of journalism, it's very fragmented. That creates an opportunity for the open Internet. So as the open Internet is looking for a way to rally all of these disparate companies and different interests, journalism becomes one very important binding entity in that cause. And we'll talk more about some of the things that we've done to begin helping journalism. So this is one of the best things that happened in our ecosystem during the pandemic, and it happened without any fanfare, any attention at all. More words will be spent on this, but the data ecosystem has been cleaned up and our very best days are ahead. So we'll talk more about exactly what we did there. But lastly, everyone stayed home and stream during the pandemic. So we did TV landscape was shifting and 10 years of evolution compressed into 3 years basically. So what that shift did was it made it so that COVID accelerated, of course, the consumption of CTV, but it also accelerated the decline of linear or cable television. And sports, which if you listen to surveys in early 2020 or even mid-2020, you ask people what's the #1 reason you hang on to cable. Sports was #1 reason. And then when most of sports got canceled, the reason to hang on to it also went away, and so it accelerated the decline of linear and cable. I said incidentally at the last Investor Day, the traditional television and cable television is a ticking time bomb that we have mistakenly created like 3% declines and that, that will go on for 20 years. That's not what's going to happen. It's going to get to a place and then fall off a cliff. And it does accelerate as it becomes less and less economically viable. We're already at that point where it's accelerated. The pandemic accelerated that, but then it's lack of economic viability. Most of the big traditional media companies at the beginning of 2020, said they have 7 years. There's pretty much consensus among them now that they have less than 3 to make the transition to digital. And by the way, I'm not suggesting that they won't do well. Many of them have become very focused on CTV. I've loved our partnership with Disney. And it's been a market leader, one reason for that, aside from them, just having very smart people that are thinking about this all the right ways, is that they have been valued on Disney+, it was just pretty amazing to hear all the discussion from investors around Disney+ and less discussion about theme parks and just all the other parts of their business. When you couple that with the other 2, Netflix and HBO, who all -- I think it was within a 90-day window that all 3 of them said they were going to introduce AVOD offerings. That represents the biggest surge in CTV inventory and opportunity that we've ever seen. So I keep talking about how the wave of opportunity keeps getting bigger. Well, that wave, I believe is it continues to grow. And I think what's happening -- what will happen in Q1 when all 3 of those really expand their inventory in AVOD represents a game change and I think is really important to the healthy evolution of the open Internet. So incidentally, it might not be lost on some of you and might be lost on to my view. But the privacy was of some of the companies like Facebook that have been really negative affected by social or by the privacy, all of social really has been affected by it makes it so that when that media is slightly less effective, people are looking more towards areas that are 100% authenticated and don't have the same issues. So when you're looking for signal and have quite literally an identity crisis throughout all of social, and then -- and then inside of connected television, almost everything is logged in. It just creates a very stark contrast for a marketer. So the pressure on Google continues to make brands look for alternatives. So one other takeaway of just all the things we were talking about with Google is that brands more than ever are looking for alternatives. So they're often coming to us. And we used to start discussion saying, how can we win budget from Google? And we started that. And they're often now brands saying, how do I move budget off of Google and on to you? Where they're initiating the conversation. The asymmetry has become more intense for brands than it's ever been. As I mentioned, the data pipes have been replumbed, we'll talk about that in a second. The inventory has gone up in CTV and Q1 will be the biggest influx that we've seen is my prediction. And incidentally, any economic headwinds to the consumer will accelerate that because at a moment where all of them are looking for more subscribers saying, would you like to pay more, so that I can continue to fund the content is not really an option. So AVOD becomes a more important aspect to every one of their businesses. The bottom line is CTV is enabling the open Internet to get the very first dollar. That's never been the case before. So when I stood here 3 years ago, programmatic and the open Internet was pretty much getting the left over. That was the majority of the time. The first dollars went to Google and Facebook. We got what was left. Even in big brands that we're trying to be progressive and move towards the open Internet, many of them were saying, I want that to be true. It's not true yet. We have work to do on the open Internet before we can. I want it to be true, let's keep plugging away as the way I would summarize the sentiment of 2019. Now more and more, the very first dollar is going into connected television, and that's creating an opportunity for everybody in CTV and especially those of us that are objective and don't own any content. It's hard for me to overstate the significance of Netflix entering this space, in part because so many of the content companies are afraid of them because they've done such an amazing job and I've even said this to read Hastings. I'm just blown away at how they managed to recreate their business again and again, right? I mean it started with DVDs, then they started repping other people's content then they became one of the most successful content creating machines ever. And I think because of that ability to pivot all 3 of those times, I think it would be a mistake to bet against them when they pivot for a fourth to welcome ads, which is, of course, different than what they were saying a couple of years ago. Of course, I said at our last Investor Day, and I think the 1 before that, certainly, I said it before then, that Netflix will eventually have to show ads. And that's largely just because we created models that showed that at some point, you have to create so many rate hikes that it becomes cost prohibitive. If you look at the history of television, even if you just look at TV from 15 years ago, you had 500 stations on your cable package and 495 of them had ads and even then you were paying more than $100 a month. When you put all of that together now and the content is so much better. And the content is more expensive, the average consumer can't pay $300 or $400 a month to get rid of the ads. And when you look at companies like Hulu, soon to be wholly owned by Disney, they make more on the ad-funded users than they do the others with 80% of their consumers roughly saying that they would prefer to see the ads then to pay more for the content. And so if you believe that economic pressure and competition is even more likely to make that true, you can see why I'm making such bold predictions about Q1 being a place where inventory searches. So I think it would be a mistake to bet against Netflix. I think it's also important to note that because Disney has done such a phenomenal job of thinking about personalization and privacy and balancing those 2 things. I think they have also created case studies that Netflix will borrow from. Of course, Netflix hasn't been in the ad business for long at all. You can even argue they're not in it yet. But the -- I believe they're thinking about it the right way. I think Greg and Reed are doing a great job of getting up to speed. And I believe that they and Disney and a number of others will represent a faster clip of change just because we'll have more horses in the race, if you will. I do think they also understand the value exchange for our consumers, which is they want fewer ads and they want them to be more relevant. And if they don't have one of the best ad experiences in the space, it will be detrimental to their company's future. And so they have to provide relevance, light ad load. And of course, they also have to respect the consumers' privacy. But both of those things have to be respected. If you just respect privacy, and then show tons and tons of ads because they're irrelevant. It's not good enough to maintain the amazing experience and be the leader that Netflix has been historically. So advertising fuels and leads growth around the world, and I want to talk about the ways that the Trade Desk has been doing this and what we're going to do going forward. So from the last 3 years, you all know this. You all have been a part of it. I want to thank you for your belief in us. And I just also just wanted to just take a minute to just say how happy I am that we've done this together. And not only did we, of course, make the stock price go up in that time, but we've done, I think, an amazing job together of making the open Internet more viable, and it's not possible without our investors. We have become closer to advertisers and brands and agencies than ever before. And so we have gone in shoulder to shoulder with the agencies to get closer to the brands. 5 years ago, in particular, if we wanted to talk to a brand, often, our agencies were resistant. They were afraid of what that meant. Now in part because of the backdrop of pressures from big tech, they often ask us to do it. We go in shoulder to shoulder, trying to help the brand do the right thing. Incidentally, when you bring all 3 to the table, occasionally, brands and agencies incentives get misaligned and agencies aren't necessarily always thinking long term about the brand's best interest. When you bring them all together, that stops and you get much better and that has happened over and over again where because of the 3 of us being at the table together, brands and agencies are more aligned than ever, and we're more aligned with both of them than we've ever been. We've been moving a lot of MSAs to JBPs, or joint business partnerships where we're creating incentives for growth. So it's not just here's an MSA and here's your rates. But let's create a partnership where if you reach certain growth measures, if we do certain things, if we do the right thing and focus on decisioning and better data usage then we create incentives in all forms for them to do the right thing in which because we've aligned our interest so well, these JVPs are just a beautiful part of our business. Of course, we've had a 34% 3-year CAGR over the last 3 years and premium video has become our largest channel. We transitioned UID to UID2 with just amazing pace and clip I'll talk more about that in a second. We launched OpenPath, and we stopped participating in Google's open bidding. Solimar launched on 7.7 without a doubt, our most successful release in the history of the company. Where in less than one year, we moved over 100% of our client base. So it's not very often that you do that. I mean I still have to update my Apple operating system to, I think, the last 2, right? It just takes so much time for me to do it when 2 back up, then you really need some time to do it. So moving over 100% of the client base is something that doesn't happen very easily in one year. It's a commentary on the quality of the product. Shipping quality product during the pandemic was admittedly harder. And as I look at the bugs and other software all over the Internet over the pandemic it makes me more proud of our team for what they accomplished and what they shipped in Solimar. Again, we did more transformation in the data marketplace in the last year than we have in the last 12. We'll unpack that in just a second and then we also manage our culture through. So let me talk about our path forward. So first, the night before we went public, I've shared this with some of you before, I was just agonizing about if this was the right decision. Some people think that going public is like an exit and it laughable because it's exactly the opposite, right? You're signing up to do this for a long time. And we were signing up to get big. That's the way that I looked at it. We were going to be a very big company, and we were going to continue to fight for a very long time. So why did I think that we could win? I mean here, we are going out less than $1 billion company at least at the time of the IPO. That changed in a matter of days. But why did I think that we could win and literally, the night before we flipped our S1 public, and I told our entire company, which for me is kind of like going home and telling your mom and dad, you're getting married at Thanksgiving or something where once you tell the team, it's over. It's happening. And as soon as we told the team, we also just pushed the button to flip it public. But the theme that I was ruminating on over and over again was really here we go at this point which was the objectivity wins trust with advertisers' first-party data. So I had a conversation with one of the biggest advertisers in the world shortly after we went public. And they -- because it was the CMO, even though the team knew us really well, he didn't. And I said, we built this whole business for you to align with you. Really, our whole business was designed so that this conversation would go really well. And that's a good way to get people's attention, by the way. He responded by saying unpack that. And what I said was I believe the only way to compete with a company like Google, who has just unbelievable data in search, I mean we share our thoughts with them but the only data set that is more valuable to you, then all that search data is your data. If you know the 100,000 people that buy your product on an annual basis, and then you can look at like model against those, and you can find other people that look just like them. That is way better than anything you can get in search or social. So really, what I have to do in order to have a data advantage is win trust. The only way that I can win trust is if I promise to never own media to never compromise my objectivity and always align my interest with you. By the way, there's economic incentive for me to do that anyway, simply because there's more supply than there is demand, and it will forever be a buyer's market. So if we align our interest, then I have the very best chance to compete against the biggest companies in the world, not because we're as big as them, but because we have better tools than them, which are the data of the biggest brands in the world to help the biggest brands in the world. And one additional risk, aside from all that, if that isn't enough, is that if they were to take that data and send it to somebody who owns inventory, whether that's in CTV or in something else, then you're essentially saying, here's all the users to gouge me on. Let me give you all the data so that you can know which impressions I want the most. And that's, of course, better handled by somebody that aligns their interest with you as well as has created one of the biggest data marketplaces on the Internet so that you can join your data with other third-party data and use that really well. That is only possible if there is a common currency on the Internet like Unified ID. So if we were to go back and redesign the Internet, no one ever would have set up cookies to be what they were. Cookies -- the way cookies have been used in advertising have been something of a hack. It wasn't ever designed to be used that way. No one ever expected that you would have like sinking happening with 27 pixels at the bottom of the page, every time you land on a page. And of course, the Internet will be much better as it moves to an opt-in Internet, and that's one of the things that UID is aiming to do. But I don't think enough has been said. I don't think we've said enough, and it's only in forms like this where we ever really have enough time to explain. But the way that you create a common currency and change the infrastructure of the Internet is to start by partnering with the infrastructure players of the Internet. And so even in parts of Google, we've created partnership because it's in the best interest of those priority 23 and 31 at Google, it's in their best interest to partner with those common currencies, especially when AWS has adopted it or Snowflake has adopted it. So -- and I don't mean to minimize the 5 or 600 companies that have adopted UID. Those are significant. And the fact that we've done that in such a short amount of time, like unbelievable that this team has done that. During the pandemic to create these sorts of partnerships. And the level of personalization that this creates inside of a company like Disney, where you across everything you do can create better personalization. Which Disney is very clear. Their vision is very clear that personalization is an essential part of a premium experience. Digital personalization is a very important part of a premium experience. And of course, so is privacy. So leveraging this alongside with all of these other companies, it makes it inevitable that thousands of other companies will adopt it. And I believe all the risk all -- the biggest risks to UID as it relates to adoption have been addressed. And as we're seeing the infrastructure players of the Internet adopt it and implement it. Our biggest problem has instead moved towards how do we bring people on quickly enough rather than just will they buy into the concepts? So the Internet is moving to opt in. It's not there yet, but it is moving to an opt-in Internet, and that is a better Internet. And in order for it to really happen, it has to happen across the open Internet. It can't just happen inside of a single ecosystem. So I want to introduce you to a new product that I don't think many of you have heard about yet. This is in a private beta right now, which is a pass for the entire open Internet. And so what this enables publishers to do it is join their user bases. And so one of the challenges, for instance, that address -- that faces journalism. Yes. One of the challenges that faces journalism is that in almost all of them, than 5 actually, I think it is in all of them. Less than 5% of their ad impressions are during a log-in. So in other words, if you're on Apple News or on Facebook or wherever you find a news article, you click on it, you go to the other page, you're not logged in, typically about 99% of the time. So what that does is that in a world without cookies, that would mean that you have less addressable advertising, 99% of the time. So in a world where journalism is struggling more than it's ever been taking $10 CPMs that are based on personalization, turning them into $5 CPMs because they're not, is economically devastating. So instead of creating a different log-in for every website so that you go log in everywhere or instead of creating that sort of annoying pop-up that happens on seemingly every e-commerce side or every furniture site where you get there and I just wanted to look at a couch, right do I have to give you my e-mail address, and then you give it to them and then, well, now I'm going to give you my phone number like you tricked me. You said I only had to give you your e-mail address. And how can I trust you to deliver my couch on time if you're going to do this? Instead of that game, what can happen is you authenticate your e-mail address one time similar to another SSO. SSO is not tied to creating a new e-mail account or anything. You use the e-mail that you already have. You authenticate it. And then when you go to other sites, all you have to do is click to opt in for that site, and you can opt out of any of the individual sites that you've opted into. So one by one, you can sign in to every site. And with a click, you can take advantage of the authentication that you've already done. The easiest comparison to this is actually in Shop Pay from Shopify, where to most consumers, in fact, the first time I signed up, you fill out all the stuff, you go to there and then you go to another site, you see Shop Pay you plug it in and you're sort of blown away the second time by how easy it was to do. And that's essentially what we're doing with Open Pass. And what that does is it makes it so the 1% of this news outlet can be joined with the 1% at this news outlet and consumers are actually more likely to visit the sites, the same way that, by the way, I am at ShopPay simply because I know that it's easier to transact there. So if I know that a merchant uses ShopPay I'm more likely to buy there from than from somewhere else that doesn't because I know that it's easier. The same thing will be true with this. It gives an independence to the publishers that they don't have with the current SSOs. So as you know, many of them you go to, and you have options to log in with companies like Google but they don't incentivize to share that addressability. It's actually to create more dependence on Google, not to give them independence. And because we're doing this for the open Internet, it becomes a fairly compelling pitch to a publisher who wants to use that and then it creates more independent. So we have a queue of large companies wanting to implement this. Like I said, it's in a private beta at this point. But it's one of the things I'm most excited about. I wanted to share that with you today. So inside the data marketplace, and this deserves more time than I have today. But there have been a number of problems that have made it so data marketplaces have been really anemic across the entire open Internet. Let me take a quick stab at trying to describe what each of those are. So one is matching math. So what used to happen is like let's say that Trade Desk has 80% cookie coverage. And then we get the ad request from an SSP, let's call it, Open-X they have a 40% cookie coverage because they don't look at -- they don't touch consumers nearly as often. We look at 11 million QPS. They provide, let's say, a few hundred thousand QPS. So their coverage would never be as great. So meaning of the commercial Internet or of users that are commercial on the Internet, they have cookies on 40% of them. If you take our 80% with their 40%, then we fall Power cap 30% covered between the 2 of us. Now if you take a data company like an Experian or an Oracle who's trying to sell third-party data, they often have even fewer touches than those SSPs. So then their data coverage will be something like 10%. You put the 30% with the 10%, and now we have like 4% coverage. So. This has been -- has made the data ecosystem on the open Internet, anemic for the entire existence of the Trade Desk. And what it does is it makes it so that you're trying to widen the pools. And in so doing, you make them less effective because if you said I can only get to 3%, and you were saying, Oh, these are the people that are in market to buy a Porsche for the last 30 days. That segment is so small when you only have 3% of that segment, then it's no longer worth doing. So instead of doing that, you then widen it and say, you were interested in a car. And then that's less helpful if you're trying to sell luxury automobiles because the majority of the consumers that are in that segment are not eligible to buy your product. So the biggest way to solve that is to solve the matching problem so that you have 100% overlap. So instead of the 80, 40, 10, you have 100, 100. And the only way to do that is with a common currency like UID, so one of the first places we start is actually not in running to create a campaign to put in front of 1 billion consumers, which some have wrongly assumed that that's what is required in order to change the infrastructure of the Internet. Instead, it's to go to data companies and say, if you adopt this, we will have 100% match rate with whatever gets onboarded first from advertisers so that you can layer your data with theirs 100% of the time. So even if that audience is small, uploaded from an auto manufacturer, and they say, here's my 5,000. Now you can layer on top of that 100% of the time because you both have exactly the same currency. As a result, the segment can change. So the incentive before this was for the segment to get bigger and less relevant. Now the incentive is to get smaller and more relevant and more impactful. Most of third-party data has also been bought by scrolling and clicking. So instead of doing look-alike modeling partly because of that math problem, we have made things automated. So one other thing that I just skipped over on the fractional pricing is that when you segment and you are constantly ending. So like let's say that you buy age gender income from 3 different places. If you buy one and then you buy the other, usually when you add a second data element, it doesn't create as much impact as the other. The curve is asymptotic. It flattens but if you just make the cost linear because the second data element isn't interested in saying, I would love to take half as much as the first one. But that's what you would have to do in order to pay them in a way that made it cost effective for the advertiser. So instead of doing that, why don't we look at the relevance of every segment, make certain that we layer them on the media decision as often as possible and then pay them in direct proportion to the relevance that they provide for that particular campaign. We've created the system, though, while more complicated, 100% automated. And what it means for the data companies is their data gets used way more often. What it means for the advertisers, they make data-driven decisions way more often. But it also means that the pricing becomes cost-effective so that you're dividing it up instead of multiplying it as it becomes less effective because of course, if you layer 37 data elements, it's not 37 more effective necessarily. And of course, it's not linear. So we've essentially solved the matching problems, the scrolling problems, the efficacy problems, next up is price discovery. So what we will do in our product over the next year is make it possible for the data companies to see what's being used and create price discovery for their data segments. What most don't know is that there has been no price discovery for data owners in the third-party data ecosystem, and that has also contributed to the data ecosystem being anemic. So let me touch quickly on just the supply chain. So what happened when we turned off open bidding. So of course, because walled gardens don't have a monopoly in CTV or audio or immersive gaming. They, of course, are trying to get in. But because authentication is so common in all of those experiences, this created a huge advantage for those formats and the spend that we were spending on open bidding move to CTV. That was actually a little bit of a surprise to us in the sense that the like-for-like would have meant that it went to other forms of display, which is often where that -- where was being monetized but it didn't. It moved towards those things with more authentication and of course, because CTV is perfectly fragmented in the right sizes, this became a significant way to reduce the advantage of walled gardens. So there's another product that I want to just talk about for a second, and that's Open Path. And if it bothers you like it does some of our employees that these are so similar to each other, open path and open path. Just remember that Open Path is something for consumers and open path is something for publishers. So this is a B2B product. And what this is, is it makes the same day that we announced that we were no longer using open bidding, we introduced OpenPath, which is a way for publishers to plug in directly to us. So many of the biggest publishers, I met with one of them yesterday, it's so fascinating how different the dialogue is now versus 3 years ago. Three years ago, they were saying hey, we'll work with SSPs or you're on the buy side, we're on the sell side, there was a gap. And now there's how can we work together. And we start every discussion by saying we're here representing the buyers. We are on the buy side. But to the extent that you want to do your own yield management, we are happy to have you plug in directly. And again, for those of you that want to attend a session after our earnings where we'll talk about the supply chain we went over some of those very same concepts with some of the biggest publishers recently to get their take on how opaque things can be on the other side of that it makes it so there's a strong incentive for them to take control over some of their yield management and plug it in directly with us. So I do believe that shopper marketing can and will be one of the largest drivers of incremental growth over a 5-year period. Again, their shopper marketing, which includes loyalty programs and sponsored listings as well as retail media and then there's a smaller subset of retail media. Sometimes, we use those terms interchangeably and we shouldn't because one is a subset of the other. But that's in part because many of our retail partnerships and we've just made so many retail partnerships since we last met, including companies like Walmart and Target and Walgreens and Home Depot and just so many others. Those partnerships are largely about fixing a measurement problem that is partly brought by the demise of cookies, but it's also because measurement has not been very effective in the open Internet. And now to make it so that you can show an ad to a consumer and then see when that same consumer actually buys it in a brick-and-mortar store, and connect those dots end to end, creates a level of efficacy in the open Internet that we've never had before at a scale that we haven't had before. And so we have very similar dynamics happening in retail media that we have happening in CTV, which is when you get the biggest in the world competing with each other to make it more efficient and better and to win share, competition actually fuels the innovation and that's what makes the ecosystem better. So there are parts of the ecosystem where there's adequate amounts of competition, and it is going faster than ever. And retail media is one of the places where that's happening. We launched in Solimar, what we call our measurement marketplace. And in fact, in our in the release that we'll do in the middle of this next year, one of the things that we have to improve on is making it easy to sign up because it seems like every marketplace, either it's somewhat anemic and you're trying to get things like apps to participate so that people will use the marketplace and join it. Or you're overwhelmed with so many people participating that you have to do curation and try to prevent crap from being loaded in there. And we've gone from having the first problem we want people to join, please come join our marketplace to while there's so many choices here, it's overwhelming. And that is what we have to address in our next release, which I'm just so excited about the ideas that we have as well as what that means. It's the sort of problems you hope for. So one thing I just want to just prime is that it is not lost on us that inside of that $1 trillion TAM, there's a bunch of spend that comes from SS&M. At times don't like the use of the acronym SMB in our own 4 walls, just because SS&M are very different. And because most SS&M market on a few websites that are massive, the YouTubes and Facebooks and Instagrams of the world. But of course, there's just so much that can be done for them on the open Internet as well. Just the difficulty there is in making it simple. Like arguably the greatest thing that Facebook has done is creating very simple on-ramps for ESSENCE and there is an opportunity for us to do something like that. I do believe it requires us to get closer to the consumer, and we're spending more and more of our time and thinking about what that means for many years from now. So I don't think it's anything we have to do to address the opportunity in front of us. In fact, I'm daunted by the massive tidal wave in front of us of opportunity that is largely about the largest players. And when we talk about OpenPath, we're largely talking about many of the biggest publishers. So if you think about the top 500 advertisers and the top 500 publishers that represents the majority of the head in both of those curves, where there's a big head or that had a very long tail, but the tail is also very long and covers as much or more surface area simply by being almost infinitely long. It's not lost on us that there's opportunity in that and especially when you're in the business of powering and enhancing the open Internet. That's very different than trying to help a single destination. So it's something we're very interested in. So as I've said before, to be one of the 10 at the finish line, you have to be global, you have to be omni-channel, and you have to have a unique data advantage. I think we spent some time talking about what those are for us. So we will create a forward market. This is very important for transitioning the upfronts. But into something more sophisticated, and we call that the forward market. Our partners have to be able to build to us. We have to continue to make CTV decisioning enhancements, especially as more and more inventory comes online, it makes it possible for us to use data more and more and be more selective so our value proposition in CTV is going up related as it relates to that and then, of course, to continue to make data enrichment. So we've talked about this already, but there's new supply and partnerships coming online. There will be better measurement. There will be more data layered on this, and there will be a much better supply chain. To just think about what we're after here all of TV is roughly $250 billion, depending on whose numbers you use. CTV is ballpark $16 billion. But if you divide that up and look at what's decisioned, which, by the way, Tim's going to talk a lot about why the future of CTV is biddable you'll see that we've merely scratched the surface of where we're going. And we've managed to win the lion's share of that decision spend. So the last thing that I just want to touch on is investing in our culture. So what most I don't think a lot of people have talked about what happened to corporate cultures during the pandemic. It was the hardest time of my career to lead. It's hard to lead remotely. And I've said internally, like I feel like I used to make decisions on 97 data points, and now I make it unlike 11 because it's just so much harder to interact. It's harder to take 5 minutes and talk to somebody instead you have to schedule something over Zoom. And then the person over Zoom isn't always as open, especially if you have a very collaborative culture but our culture was always built on hiring great people who are very collaborative that get along that have high EQ that also share common values and common vision. And because we've had the singular vision, it's been something where I believe what every company has done during the pandemic is make withdrawals from that cultural piggy bank. I think it's been almost 0 deposits. So during that time, just making the withdrawals, it was something that because we had so much saved, and we had so much focus I believe we grab more land during the pandemic than any other time. And I believe the same thing can happen and will happen if there's economic downturn that we will once again, rally around a very common vision and the impact that we can have on the open Internet. Wave of opportunity continues to get bigger. CTV is getting the first dollar and it is moving up the prioritization for advertisers. And what we're doing fuels and leads global growth. So in the same way that tech often does that, I believe advertising does that for tech and for everything else. I just want to say thank you so much for the time today. Thank you for being our investors, for believing in us. And for taking the time to understand our business, there are so many people here that I've spent a lot of time with over the years. And I just really appreciate your support and your willingness to learn this esoteric complicated ecosystem. That complication is one of the things that actually is a bit of a moat around this ecosystem because it prevents people from getting them it's very difficult to understand, but it is an opportunity that continues to grow, and it will only continue to grow. So thank you so much. I'm going to introduce all the speakers today. So we're going to hear a short presentation from our CTO, who recorded this because he couldn't be here today. But if you don't know, Dave Pickles and I founded this company almost 13 years ago, Dave is one of the best technologists I've ever met in my life. I often tell people that to understand, Dave, you have to understand that his mother had a PhD in psychology and his father was had a PhD in engineering. It was a nuclear physicist. And so the 2 of them created this person who understands people and machines and boil things down to their essence. And at times, when you're having a conversation that is focused on technology, he realizes that what it's really about is people and so as a result, Dave is just an unbelievable partner unbelievable CTO. So I'm excited for you to hear from him now. Thank you.

All right. Hi, everyone. I'm Dave Pickles. I'm the CTO and Co-Founder the Trade Desk. You heard Jeff talked about all these amazing growth drivers that we have, how we're getting closer to our clients and this amazing tailwind that we have in the explosion of connected television, but I'm here to remind you that none of that would matter if we had a bad product. And what we do in technology is just critical to our success. So I want to walk you through what we've been doing since the last time we talked and where we're going in the future. I'm sorry, I couldn't be there in person this time. I remember last time, I was amazed by how deep everyone was on ad tech and what we do. So it was very exciting to talk to you all. I'll have to look forward to seeing you all in person next time. So I'm going to start by talking through some of the fundamentals of our business that have stood the test of time. Understanding our fundamentals helps you understand the significance of Solimar, all the product releases that we're doing. I think it's important to set the stage with that, and it will also help frame the investments that we're making in the future. So let's start with simplicity. I always try to frame our goals as simply as possible. I think that simple, powerful ideas win. And so if you need a ton of nuance, to explain what you're doing I tend to think you're going to lose. Ad tech can be very esoteric. It doesn't have to be -- I could throw out every buzzword in the book of anybody that wants to ask what we do or I can tell them, look, we provide a platform for buyers of advertising to make the incredible range of options and possibilities understandable and help them grow the brands. That's what we do as efficiently as possible, help them grow their brands and hit their goals is by making the complex simple that I think we've achieved a lot of our success. So the most fundamental principle of the Trade Desk is the concept of decisioning -- if the -- the dots here represent different advertising opportunities you have. Blue dots and yellow dots could be good or great impressions for this campaign, the great dots are not so great impressions for this campaign. It's still important to look at everything and be holistic across all the possibilities because what you want to do is fill the campaign with the good stuff with the blues and the yellows. And if you're looking at less of the universe or if you're only looking at one channel or something like that, you're going to just see many fewer of those opportunities and so you'll end up spending the budget on a bunch of grab on a bunch of average stuff, you'll be blending it in, and that's going to decrease your performance and just will never be the most efficient way, I hope the brand. So we always believe in omnichannel platform when you have to be able to look across all kinds of supply. We add value by not only choosing which impression to buy within the channel, but also by choosing which channels to be in and how much to be in those channels. You've always heard Jeff say there will be no mobile DSPs that you no display DSP. If you want to be able to compete then you want to be able to perform the way we can. You have to look at everything, and that also involves understanding all the identity landscape involved in all of this and all the different kinds of integrations. So there's a tremendous amount of work just to get to the table and understand what are the possibilities even are. Next Critical tenet for us is human and machine. So we've always believed that machines and AI are great at some things. Humans are great at other things, allowing them both to come to the table and do their best work is probably the greatest challenge we have solved. I think we've done it better than anyone. It can only improve. It's an incredibly difficult problem to do, how much do you let AI take over certain areas? Which areas do you let to run with? And how do you create insights into what the AI is doing so that people can use these incredible is keeping up in their heads to create value. This is really why Solimar is such a big deal. It's the UX around the man and the team and the opportunities that makes a more so powerful. So covering the open Internet. So this is another core tenet of what we've always done is making sure that we're fair. We're looking across everything. We don't have any conflict of interest in what we do. So this helps our buyers by making sure that we're always only investing in the opportunities that are the best for them, but also helps the open Internet by making sure that publishers who create great spots and get the most money. We're not steering any inventory during any dollars towards inventory that we control and we're really helping power us thriving open Internet where the green gets to win and the best content gets to make the most money. And we're built for growth. So one thing about our business that in the early days was really, really hard, but now it's really great is the bigger it gets, the better it gets. And so there's a huge cost or upfront cost to build the platform and then to look at all of the QPS to look -- just to look at the opportunities, it requires a lot of computing power. And so you've probably seen this if you've been following the company for a while, when we outperform our expectations on top line a lot of that drops to the bottom line because we have this large cost to do the looking and then it costs relatively less to buy impressions and to pre-report to do rest of things that we do. And so now it's a great advantage for us that we're large. The business continues to get better and better. So we've got more and more capital to invest in more and more ambitious engineering broadband, which leads us to Solimar. So Solimar is the culmination of years of effort in innovation, upgrades across all of our platforms focused on the UX. We won't run you through some of the feedback that we've gotten on this. So we get lots of comments all the time from our customers. These are some of my favorites, and I really like the one where they say how simple but powerful it is. That's what we're always going to be for with these upgrades is how do we expand the power of the platform, how do we leverage AI more effectively, but we got to make it simple. We got to make it easy to use. It's been a huge challenge. One of the things we're doing in Solimar, probably I think the most important thing we're doing is all more is doubling down on the goals experience. You've always been able to set goals in the Trade Desk obviously. But in Solimar, we make it much more central. And what we're trying to do is make it so the traders, the people using the platform spend more of their time thinking about what is the goal? What should the goal be? Do I have multiple goal and have a good way to express that. Instead of fidgeting with bids, trying to achieve something that they've never told us they're trying to achieve. Let's have a better conversation about what we're going after. Once we've gotten the goals, that puts more pressure on us. It puts more responsibility on us to do a really good job with data science to unlock better performance and to live up to that conversation that we had. So they told us about our role they need to make sure that we can do a very good job going to hit those goals and hit them more often. A few examples of things we've done here is we're making pacing and performance more aware of each other. It's a difficult thing to do. You're trying to hit the spend goals. They want to spend a certain amount over the campaign, they want do it fairly smoothly, but then also you want to treat spending money on the good stuff. So those processes need to be aware of each other. And we're moving some processes that used to be off-line processes into online processes in the bidder to be able to make decisions more quickly and in real time. And then improved insights and recommendations. Obviously, as AI does more and more, we have more work to do to help build trust with traders and to give them insights into what's happening and then also to surface in these opportunities. We always want to be having that conversation, particularly about like is there any channel that you should be in that you've turned off could perform better than what you're doing now. So we always have to level up the conversation whenever we have a chance. Results pretty much big for themselves on this. So just a few exciting months here. We've had a doubling of the number of data elements being used for impression, which I'll actually speak a little bit more out in a minute. 50% increase in the number of channels used on average. We got to 100% adoption over the course of one year. Very hard bought here just managing through a lot of change. Nobody loves it when you change the UI on them. They're used to doing things in a certain way, but we were able to show just a tremendous amount of volume and get all of our customers to move over. And then to usage go is our AI product is at 95%. So there's just very few people now who don't want any help from the AI, which I think speaks the trust that it is. Identity Alliance in Solimar is one of the other big winners in all of this. So we've been continuing to invest in our products to just help the buyers not have to worry so much about the identity landscape and to be able to take whatever data you've collected in whatever environment and put it to work across everything that you do and just it makes sense. So we've simplified this product. We've improved this product. It has more range, 20% more. It reaches 96% of person-level devices in the United States. And then we've seen 10% increase in usage of this product. So it's just another huge win for everyone. The campaigns performed better and the trades are able to focus on other things and it makes us more able to put the spend where it should be inefficiently. So as we focus more on goals and those being more central with measurement also becomes more important, measurement needs to evolve advertisers are moving away from last click metrics. And so we've done a lot of work to partner with all kinds of companies that can provide data sources to provide better measurement, real-world results, whatever that means for the brand. So that can be brand awareness that can be in-store sales. You want to have a good metric that we're measuring to a metric that actually correlates to the outcomes for the business so that we can have the better conversations, and we can do a better job of allocating spend across channels. So talk for a minute about areas of investment that we're going after in the coming years. Some of these are fairly but just -- so the forward market, a data-driven guaranteed marketplace, which we have literally been talking about since the inception of the company. The ability for buyers to express a desire to pay a premium to have something in the future that's guaranteed and also the willingness of sellers maybe to give a discount for the guaranteed revenue that might come in the future. But to do those things in a way where we don't give up decisioning and don't give up everything that's great about the new way of volume that we've created over the last decade. So we want guarantees. We want to way to execute upfront, but we want to maintain decisioning and control when we do that in a way that's agreeable for all parties. So we're getting more and more traction on. This is definitely helped by are getting closer with sellers, which is happening for a number of reasons. So we're able to have direct design conversations with all of them and just try to facilitate what the buyers and the sellers of media want, which is this driving forward market. We've been investing a lot in understanding the relationship between linear and connected television because this has rapidly evolved over the last 2 years, most buyers are very interested in incremental reach. So I have a campaign have been running a new in here forever, but I know there's a whole bunch of households that can't reach that way anymore. How can I extend this campaign into CTV efficiently and know that I'm doing the right thing and not doubling up a whole lot. So we've done a lot to help that. We had a couple of agencies to make these tools central to their planning process for 2022. Yes, really a lot for optimal allocation across both of them. Instead of sort of during a pin at it and saying, well, let's just divide the budget by x based on ratios, we can do a lot better than that or customers to be appreciated. We can also do targeting based on content attributes. So even within the incremental reach plan, you can use the rest of these attributes to further tailor where you want the spend to go within CTV. We can really achieve things in CTV, we never could in linear, which is very exciting customers are starting to really embrace it. We've invested a lot in the data marketplace. I touched on this a little bit earlier. But so we are absolute data junkies, and we think the data should be driving everything you should be using all of the data, all of the time to do a great decision. But it's been hard it's been more anemic than it should have been, mostly because pricing of data didn't have a lot of correlation with results. Which meant the average number of data elements used in decisioning was lower than we wanted it to and data partners were incentivized to optimize for scale rather than quality. So they would just give you these huge data sets at this price and kind of take the earlier which worked okay. But we're in a position now where we can lead the marketplace, and we're really driving the conversation of how data should be priced to enable it to be used all the time and in a much more rich ways. So we aren't just changing pricing. We're laying groundwork for a thriving data marketplace, and we've seen great results so far. So hybrid pricing, we've used to better align the pricing with the results. So here's a sort of a strong example, but it's something that can happen where you're spending $1 on media and you can't allocate $0.80 to data, right? Because then there's no money left to buy media with. It's crazy. You spent $40,000 or $50,000 budget on media. So with an 8% fee and an 80% cap your data costs, in this case be much lower. The cap insist because in more premium environments, I think we all acknowledge that you're paying more for that premium placement. That really great spot and that really great show. So the data cost shouldn't be just a straight percentage of all of that media. There's a place in between where everybody gets a fair payout and it allows the advertiser to use data all of the time on everything. We've also made it possible for them to now use multiple data elements and to split the payout between the providers. And we've got everyone to in, that's the right way to do that. So if you know 10 data points about a consumer that are relevant then you use all 10 of them, you don't pick one of them because you didn't want to pay for all of them. And the lift in performance there are something that all of the mall the data owners get to share with that kind of pricing. Open Path is another huge area of investment and really a long time coming. It's really something that actually other people have done sooner than us. But as the supply chain has become more complex and more okay. And particularly with the advent of header bidding, it's made -- so it makes more sense to have different ways for demand from the Trade Desk to get into pre-bid and so I think it's easier to just look at a little bit of an eye chart. But generally speaking, publishers run ahead of wrapper like pre-bid, they'll collect demand from SSPs from Google. And then the Trade Desk demand flows through all of that and we see those impressions through all of those. For a publisher who is willing to do the work to incorporate our demands directly into the header wrapper, they can do that with open path. And so that's not us providing an SSP. We're not doing any of the things that an SSP does. It's really just allowing a publisher who wants to take that function in-house or part of that function in-house to see what the Trade Desk demand looks like and to incorporate that directly with our header wrapper. This also leads to better and more direct relationship between us and the publisher. But we want to be super clear, we're not building an SSP. We're not servicing publishers. This is an active buying. This is us bringing demand to the table in a more direct way. Yes. So the takeaways of our approach, we look for powerful simplicity in a complex and Ecotech space. We bring the richest set of tools and features, we continue to innovate. We ship product every week, all the time, we're always pushing. And that's been a huge part of our success is just staying ahead with the curve on all of the things that we've always been excited about in this space and there's so much more to do. So we just keep pushing hard and bringing innovation to the customers about as quickly as they can handle absorbing it. We have years of trust that's been developed between our product team and our clients. And like I said earlier about the size of the business, only a massive scale platform can win. So the players may already be in the game, most likely players already in the game. It's a real tall hill to client to get to the point where you can just keep the lights on and pay the bills while looking at everything that we do and then to then catch up with the innovation and the features of would be pretty tough. So thanks for your time today. I hope to see you all in person next time we do this, and I'll take it back to Jeff.

Taking 10 minutes right now. We'll reconvene in a few minutes past the hour. [Break]

Good everyone, please take their seats. We're going to start with our next speaker, please. Everyone please take their seats, and we will start with our next key group.

Hi, everybody. Welcome back. First of all, thank you again for -- thank you again for being here. I am very excited to introduce you to someone who's never spoken at an Investor Day before because she joined the Trade Desk since then. She's quickly risen through the ranks and she and I have been incredibly close partners as it relates to UID and changing the ecosystem. Samantha joined us from Oracle Data Cloud, where she led strategy and business development there. But really, I just want to introduce her by telling you about a conversation that she and I had about 1.5 years ago, which was -- we were talking about UID and the potential for it to just change the entire Internet. And I remember both of us starting to feel overwhelm and daunted by the significance of how big the Internet is and especially the open Internet and just how many different players there are. And just how much change it represents to have a currency, making it possible to move from opt out to opt in as well as to just make the Internet replace this prehistoric technology of cookies. And our fears transition in the course of the conversation to what if it works, what if we do it, what if we change it? And if -- when we really started to think about, well, you don't really have to change that many people or that many companies. You don't have to win over that many hearts and minds to create momentum behind all of this. You fast forward to today and some of those logos that I put on my slide saying that underscore how the infrastructure of the Internet has changed. I don't think that there's a person that deserves more credit for that change in the world than Samantha Jacobson. So I'm super excited to introduce you to our Chief Strategy Officer, Samantha Jacobson.

Hey, everybody. I am so excited to be here with you guys today. One of my favorite parts of my job is that I get to think about which bets we want to place for the future and not just focus on this quarter. And so with that lens, I'm really excited to take some time today to talk to you guys about identity, about the data marketplace that Jeff and Dave talked a little bit about and about shopper marketing. But I wanted to start -- assuming my slides catch up. So I going to start by sharing a reflection I had. So I came home last week and I was exhausted and my 6-year-old immediately asked to play Candy Crush on my phone. Many of you guys have kids, I just did not have it in meals, here you go. What am I going to make for dinner? And my version of making dinner is ordering in because I am not a cook. So we decided to order Papa John's and then we're watching SpongeBob SquarePants on Hulu. As you can imagine, I did not get much of a vote in what we chose to watch -- but when I was reflecting on that experience, something that I found really interesting is that in each of those instances, we were using a device connected to the Internet, right? My son is on my phone, I'm using my computer to order dinner, and we're watching TV via Hulu, and yet none of them had a cookie on me. So why would we think about cookies as a consumer sense of identity and that didn't even take into consideration the fact that I was listening to Spotify and my drive home or the fact that I was going to go binge-watch Below Deck as soon as they put the kids to bed, all of those are points of interaction. And yet none of them had a cookie. But you know what they all did have, they all had my e-mail address. I logged in the Candy Crush using my e-mail address. I ordered pizza and Papa John's website using my e-mail address. I logged into Hulu using my e-mail address. My e-mail address was always present. And what we find is that consumers expect to receive tailored advertising in exchange for free or reduced content. So it is critical that marketers can understand the whole consumer and not just cookies. Dave Pickles was on this stage 5 years ago in 2017, and he talked a lot about the fact that cookies are an outdated technology. That has never been more true than it is today. Safari and Firefox already block cookies, CTV doesn't support it at all. And so when we think about identity, we need to make sure that we have a sense of identifiers that truly are present wherever the consumers are consuming media. What would that look like? I'm so glad you guys asked UID 2.0 is really centered around this sense of identity, right, whether it's keyed off of a consumer phone number or e-mail. And it was built with the concept of consumer privacy trust and transparency at the epicenter. Jeff talked a little bit about the fact that it's a distributed technology, but I think that is so core to its success. Because UID2 can be used by anybody in any ecosystem as long as they're adhering to those principles. And so it's a fact that it's open source that you can access that code and that everyone is able to use it to support their business that I believe will make it so successful. So if you ask me what a UID2 token is, I would tell you, it is a hash salted encrypted ID. And if you are anything like my husband, he looks at me like I have 3 heads when I say something like that. He has no clue what I'm talking about. So I just wanted to take a minute to break it down a little bit. So when I interact with a brand, whether that's Peloton or Paramount+, they ask to collect my e-mail address or my phone number in exchange for tailored marketing, targeted advertising, et cetera. So I input my e-mail address right? samanta@thetradedesk.com. From there, they would then overlay the UID2 algorithm. And again, it can be within their ecosystem using our integrations from Snowflake, from Salesforce whomever or they can send it to the UID2 operator. And that ecosystem takes it and hashes it. It turns it into a binary large object or a blob, which is really just a fancy way of saying a bunch of numbers and letters. From there, we then also salt it. So this is specific to UID2, but that's the ABCDE at the end. So it makes it this unique identifier present in this ecosystem. And then we encrypt it from there. So it adds another layer of security. And so when companies are using UID 2s to talk to each other, they're not doing so by sending my e-mail address back and forth, but they're, in fact, all the way down here at step 4, which is why it's such a secure ecosystem. And I did want to take a little -- a minute to talk a little bit about identity versus data. I know Jeff touched on this a little bit, but an analogy that I've heard that I found really helpful is that of a safe deposit box. Identity is the identifier on the outside of the safe deposit box. It allows 2 companies where I've given both of them my e-mail address to communicate about who I am. But it doesn't tell either of them anything about the content of that safe deposit box that is data, right? These companies may have the same UID2 on me, but they don't know that I'm a mom, that I live in Colorado, that I'm a dog owner, all of that, those are pieces of data. And so when we talk about UID2, it is truly just the sense of identity, it is not including any of those pieces of data. And we also add another layer, so we talked about kind of that token that's created at the end we also allow companies to use double encryption protection. And what that means is the company takes their list of UID 2s and add another layer of encryption specific to themselves. And then transfers it to the end user. And that entity has to receive the decryption key from that publisher. So it's another customer-specific level of encryption that's added and the origination point of data has control over who receives those decryption keys. So there's all sorts of layers of protection and ability to control the receipt of that data as well as the creation of that data. And what we find is that it has been so widely adopted because it allows advertisers to still do the same things they need to optimize their business, right? They can use it for targeting, for measurement, for optimization and best of all, it works across all channels. We've also made sure that UID2 is interoperable with other ID solutions. So we're not saying it has to be this or nothing else. You can leverage the accuracy and scale of UID2 in conjunction with everything else you're doing today. So it provides superior business results while also creating a better, more privacy safe consumer experience. This is how we're going to get on that path towards the opt-in open Internet. And what we've seen is that it generates incredible results. So Maiden is a high-end cookware company. For me, again, cooking is putting ramen noodles in the microwave. So I wasn't familiar with them. But if you've eaten at Gramercy Tavern or line in Chicago, your meal was prepared using their dishes. And they want to target at-home chefs. And as you can imagine, it's at a very high price point. So they look to UID2 to understand how they could more effectively market to their customers. And what they found was that by connecting data more accurately, they saw a 22% increase in conversions. So people are more likely to buy because they're talking to the correct consumers. And they were more likely to convert more quickly, right, a 33% increase in conversions while spending less on their marketing dollars. So hugely beneficial for their business. This is just one example of many that we've seen across our advertisers. And it doesn't just benefit advertisers. We recently released a case study with Fubo and they adopted UID2 in their platform. And what they found is they had a 62% increase in ad spend year-over-year. And the reason for this is because advertisers are willing to pay more when they know accurately who they're able to talk to. And they then see better results, just like we saw a 14% increase rollouts for these advertisers. But this is much bigger than any given campaign or even the Trade Desk. This is critical for the success of the open Internet. The market needs to have access to data across all of these different ecosystems. We can't just silo the data and ask an advertiser to say, Great, you're going to spend x dollars on this ecosystem and y dollars there. Imagine if a marketer only knew who I was on Facebook, they would have missed out on all of those touch points I talked about at the outset and they wouldn't understand that they've already marketed to me in multiple areas, and they're going to continue sending me the same message over and over again. It's an inefficient way for them to use dollars. And candidly, it's a terrible consumer experience. Which is why it's so important to have access to UID2 to work across these ecosystems and to be able to measure across all of these ecosystems. And as we've talked about, the success has been overwhelmingly positive. The thing that I am so -- when we look at adoption is the array of types of companies that have adopted this. So Disney, early adopter of UID2 when we think about on the advertiser side, P&G recently announced their endorsement of it. We think about the infrastructure. You've got AWS, and you've got Snowflake and you've got Salesforce. And we have numerous competitive DSPs also using this technology because this is not a Trade Desk specific thing. This is the fabric of the open Internet, and it's how we'll continue to connect the open Internet together. But of course, we do not want to rest on our laurels. We are going to continue to focus on making sure that we activate this incredible demand that we've seen. And that's why we're so focused on making sure that nearly every advertiser is transacting on UID2 by the end of the year. As we've talked a lot about why it's important to have an accurate sense of identity. But of course, marketers also want to be able to overlay data when they are targeting those users. And that's why we've spent so much time improving the data marketplace. So one of the biggest challenges that we have in the data marketplace is just the sheer volume, right? We've had such an arcade and way of buying and overlaying data on the campaigns. And it's understandable. It's daunting. There are hundreds of thousands of segments from which to choose in the data marketplace. I mean imagine you guys had to decide if you want to invest in a company. And somebody said, Great, there's 500,000 research papers available on this company. How are you going to pick which ones you're going to read. It wouldn't make sense from a cost benefit perspective to read all of them so you've got to have something for how you're making that selection. And what we found in the data marketplace is that advertisers were sometimes saying, Well, this company was in talking to me last week. They had an awesome PowerPoint. I will invest my dollars there. Or they may say, yes, I recognize his brand name. Therefore, I'm going to pick that one. Or it could be size, right? I need to reach a lot of people and make sure it's really effective. I'm going to pick the segment that has the most people in it. Or maybe they're looking at price, like, this one costs a lot. It's got to be really good, right? I mean surely, other people know what they're doing. You'll notice that none of those pieces are selecting based on the performance that's being delivered. And so that is where we have focused on driving change. By adding multi-element bidding or the ability for advertisers to layer on data to the extent that it adds value and pay based on performance, we are now allowing advertisers to pick data that best delivers against the cause of data -- against the cause that they've identified for their campaign rather than asking them to pick it out of a drop-down menu across hundreds of thousands of segments. And as Dave talked about, we've switched over to a percentage of media pricing so that it's not cost prohibitive to apply data to your full campaign because what we're finding is before on a CPM basis, many advertisers would want to use the same data across their campaign, but they just couldn't justify that cost on certain forms of media. And so we've rightsized that. And it's no surprise that the results have been so positive, what we've seen is -- so we launched this in audience predictor in April. And what we've seen is there's a 150% increase in data elements being used. So advertisers are adding more data and it's performing better, right? They've had a 22% increase in conversions. So it surely is that win-win ecosystem for advertisers, for data providers and for consumers who are able to get more targeted tailored messages. And of course, it wouldn't be a conversation about data if we didn't talk about shopper marketing, the latest buzzword. So everyone seems to have all sorts of different definitions for what it is, what it isn't. Shopper marketing is really just connecting with the consumer when they're ready to buy something. But as we all know, that has changed drastically, especially with COVID. So take my family, for example. We live near a Super Target, and we buy everything there. We buy groceries there, you need backpacks for back to school time, Target has them. You need formal dresses for weddings Target has them. You name it, we get it at Target. And what we used to do is we go Saturday morning and walk down the aisles and we found all sorts of things. We just couldn't imagine living without thanks to the Endcap displays or the middle of the aisle, advertisements for chips, whatever it may be. And that's how we filled up our cart. But with COVID, I joined the other 263 million Americans who shop online. And so it's no surprise that marketers will spend $100 billion, $100 billion on retail marketing next year according to GroupM. But as you can imagine, my trips to Target now look very different. So we frequently won't even go into the store, we'll pull in and do the curbside pickup or order something in the app and have it delivered to my home. And yet, the amount of money I have spent a target has not gone down at all much to the secret of my husband. And so we are still making those purchases, but now I am influenced at very different points in my purchase cycle. Instead of waiting until I get to the store and walking through the aisles, my decision to buy something comes when I'm looking online or when I hear an ad or where do I see something on TV. And so it's critical that marketers adapt their messaging to meet consumers in the moment where they are. So I want to bring this to life a little bit with an example I had recently. So I don't know about you as I am not a toothpaste loyalist. I used to go to the store, and I would pick based on whichever had the shiniest box or if this one was 3x more effective and this other one was 5x more effective, 5 is bigger than 3. That must mean that better toothpaste, and that's how I picked it. And so now what I find is when I go to the Target app, I type in toothpaste. And even if they have the same Crest scope that's sitting on my bathroom counter, if it's on the fifth page of results, I'm never going to get there. And instead, to me, it doesn't matter if it's Crest, Colgate, ARM & HAMMER. I'm going to pick whichever one is not the chalky white stuff and whichever is like the minty green, that's the first one to come up. That is such a shift in how marketers need to advertise to consumers and so it's because of that behavior or because consumers are now looking at the 4U coupon area, and that's how they're making their impulse purchases, that marketers need to be really smart about the messages they give and they need to think about data in a whole new way. But what we found is that marketers are the -- or sorry, the retailers are in the perfect position to leverage UID2, they have a direct relationship with consumers, they are able to get their permission to use their data for targeted marketing, and they are then able to protect that data that they have gathered from the consumer using UID2, right? Their relationship with their customers, their crown jewel. They want to make sure that it's protected at all costs. And UID2 does that. but it allows them to leverage that signal to be able to make sure that they're marketing to them really effectively and that they're allowing their partners, whether that's a Procter & Gamble or a Unilever to also access that data and advertise to their customers as effectively as possible. So they're plugging in that data in a consistent way using UID2 across the ecosystem for both targeting and measurement. And what we found is that there is a huge amount of demand from retailers that want to be able to leverage this data, both for themselves and for their partners across the ecosystem. And that's why we have such an incredible array of retailers available on our retail marketplace that David talked about. And we have hundreds of advertisers that have come to use this data. And it's not just the 80-plus percent of CPG companies we work with are the major food and beverage brands that we have partnerships with, but it's all sorts of advertisers. So let me bring this to life for you with an example. Home Depot, for example, has great data on people that are coming into the store because they're buying all sorts of items if they've just moved homes. You could take another company like State Farm, they don't have any presence in a retail store. And yet that information around new homebuyers is incredibly useful for them because they want to be able to market their -- they want to be able to market insurance to those entities or to those customers. And so it's not just CPG companies or food and beverage companies that are able to benefit from access to this data, but it's truly the hundreds and hundreds of advertisers on our platform that are finding utility in it. So whether they're coming in with their first-party data, whether they want to leverage data from retailers directly or whether they want to use the more traditional third-party assets, the way that advertisers are able to gather the maximum amount of benefit is by keying it all off of UID2, so they've got a consistent sense of identity that they can leverage for their initiatives. And as you can imagine, we've seen incredible results. So BIC is a French company. They make everything from ballpoint pens to Lighters. And recently, they ran a campaign for male and female razors. They wanted to understand how they were performing and how to optimize their campaign. So they use the data available from Walmart to understand who to target. And then they were able to measure on the back end to understand how it was performing and had nearly a 500% return on ad spend. So for every dollar they're spending to market, they get $5 in transactions. Now to put that into context a little bit, Nielsen says a strong campaign would be a 270% return on ad spend for this category. So that's nearly twice as twice the performance that Nielsen sees is good. and the consistent presence is the ability to use UID2 to understand how to connect data for both targeting and measurement. With that said, I am so excited about all of the momentum that we have. We have seen incredible adoption of UID2 we're able to see it in a privacy-centric consumer-first way that allows us to plug in across the ecosystem. Advertisers can then layer this with data that's now accessible to them and is able to help them achieve their campaign objectives. While providing access to new types of data assets in the retail marketplace. And the thing that I am most excited about is that we are just getting started. The wind is truly our back, and it's going to be an incredible 2023. Thank you guys so much.

You can see why I'm so excited to work with Samantha and just all the progress. When I look at all those logos and I think about the infrastructure partnerships that we have, it's impossible not to be excited. It's a bit of a dream to actually have all those logos because I don't think we could have made up a better group of partners to have and what we have in UID2 so thank you again for all the work and amazing success. At our last Investor Day, Tim Sims was representing the inventory side of our business, and that's where he's grown up and spent most of his time in our business, managing our inventory partnerships. You can imagine that, that job has changed dramatically as CTV has gone from almost 0 percentage of our business to our largest channel. But of course, that's also changed as the relationship with SSPs has changed because their businesses have changed and ad exchanges and of course, complicated relationships with partners like Google. But Tim also has understood the advertiser side of our business. So we made a very bold move during the pandemic, which was to join our inventory and our revenue side of our business. So that Tim could speak to our customers about the opportunities in inventory and especially in CTV. That's proven to be incredibly effective, and it's largely because inside of Tim's head is an understanding of both sides of the ecosystem. So it's my pleasure to introduce Tim Sims.

All right. Hi, everybody. As Jeff mentioned, my name is Tim Sims, I'm the Chief Revenue Officer of the Trade Desk. And as Jeff just mentioned, in that role, I do have a little bit of a unique remit in that I oversee all of our go-to-market partnerships on the client side of our business. And then I also oversee all of our partnerships on the supply side of our business, which includes how we work with all of the global media companies that we buy inventory from. So one way to think about inventory on our platform is if your company sells ads for a living, that's part of that team's remit. So another way of saying all of that in a sentence is that my role in -- I am responsible for all of our go-to-market partnerships across both our clients and our inventory partners. So today, I want to spend the time just taking a little bit of a deep dive into how we continue to win share and how we are accelerating our position as a leader in CTV by leveraging what I believe is the most robust CTV platform on the planet. But first, I want to frame the conversation around the current state of the marketplace because as many have pointed out today, a lot has changed since the last time we all stood on this stage. So the marketplace right now in the media landscape is in flux. The land is kind of shifting under our feet. There's been massive change over the last couple of years, especially in CTV. We're seeing the shift from linear to digital as represented by CTV. We're shifting to lighter ad loads in CTV, which means that those ad positions have to come in to a higher CPM, which comes from data-driven advertising, by the way. And all of that change, all of the shift that's happening in the landscape is great for companies like ours because we thrive in moments of change because when we take the focus and the vision of our company and apply it to moments like this, it gives us an opportunity to add value to our clients. And when we're able to add value to our clients, that's when we can grab land. So as a result, we continue to outpace the market. So the data that we're looking at here comes from one of our clients, IPG and also e-marketer. But regardless of the data source that you look at or the benchmarks that you use over the last couple of years, we continue to outpace the market. But most importantly, as we kind of look out into the future and we think about the next several years, what do we see? And how are we going to continue to lead the marketplace? So for the next half hour or so, I'm going to spend most of that time explaining why I think we are better positioned than any company in the space to continue winning and to keep leading the way, especially in marketplaces like CTV. But all of that starts with this, which is a strong go-to-market strategy and a strong way that we take our offering to our clients. So a lot of time and energy is spent internally talking about just first-order principles, right? Like what are the core pieces of our business that are just absolutely central to our success. So I'm going to unpack 3 of those today. The first principle is objectivity. Jeff talked about that. The second is our people. And the third is our obsession with accelerating change in our industry. So I'm going to go deeper into each of those. But I think what is amazing about all of those principles is that those principles are the same today as they were when I walked in the door at the Trade Desk almost 9 years ago. We've been so consistent in our vision for the last 13 years since the day, Jeff started the company 9 years ago when I walked in the door, that all of these principles that we're going to talk about today, while continuing to be central to our business have also been consistent. So I'll start with what I believe is the foundation. So if all of our central philosophies of everything that we talk about, I don't think there is a single stronger central core than this one, which is our objectivity. So this is our biggest advantage in the market. And we won't put this at risk. And the reason we won't put this at risk is for a relatively simple fact, which is that -- and this happens very often. I just sat with a brand yesterday and had this conversation because we regularly sit across the table from the biggest brands, the biggest content owners, the biggest agencies in the world. And we can honestly look them in the eye and say, our interests are aligned. I can help you decide between this impression and that impression without any conflict of interest. And that is extremely powerful for advertisers. Because if you think about it from a marketer's perspective, they are constantly getting pitched something along the lines of, "Hey, buy my inventory. Here's why". And they're constantly getting pitched a nonobjective point of view on the marketplace. Instead, we say, "Hey, let's look at the open Internet. Let's look at those 11 million, 12 million ad opportunities per second, and let's help you decide which 1 is best for you. And let's use your data to do that". That objective position in the market is one of the biggest foundations of our business. Now interestingly, that objectivity that we talk about a lot on the advertiser side cuts both ways because, as Jeff mentioned, when I also have the opportunity or privilege to sit across the table from the largest media companies in the world, that objectivity cuts in both directions. Because when I sit across the table from a large content owner, when they look back from their vantage point at the landscape of people who do what we do on the demand side, there's a lot of complicated relationships out there. There's people who produce original content. There's people who distribute their content. There are people who wish the future looked like a world where an impression on a video filmed on your cell phone was the equivalent to an impression in the winning drama. So when they look back at the landscape, it's very complicated for all of their potential partnerships. But when they look at The Trade Desk, they see a beacon of independence. And they see a beacon of objectivity because they know our only goal as a company is to connect them and their content and the ad opportunities against their content to the demand on our platform. And as a result, they lean in. So the second foundation that I want to spend time on is our people. So I think a lot of people probably stand on stages like this and say things like, our people are our greatest asset. Now I agree with that. However, I want to look at this through a slightly different lens because I think what is more powerful is a group of people who believe in something. And this is why I think our people are so important. Jeff touched on this a little bit. But if a group of people organize themselves to achieve a goal against something that they believe in, that's the true power of a group of people. So I think what differentiates the group of folks who work at The Trade Desk is that we are all united behind 1 simple mission, and that is to empower the open Internet. So we don't believe anyone wants an Internet controlled by just a select few companies. We believe in empowering the entire open Internet. That's our mission. And when you get a collection of people and you point them all at something that they believe in, that's the true power of our people. And in addition to that, we've maintained our commitment to bringing on more people into the company. So we've continued our pace of hiring even in the face or even in an environment where others have not, because we believe our ability to invest in our people is our engine for growth. It drives and it fuels that collective inertia towards that goal. And we organize ourselves that focuses on our clients. So we've created distinct roles that really focus on different parts of our clients' business. We have sales and business development people who bring new opportunities to our clients. We have account managers who help our clients understand and leverage the best of our platform. We have traders who bring detailed insights and optimization tactics to our clients so that they can get the most out of every dollar that they invest in the platform. And lastly, as our clients become more and more sophisticated, we have a technical account management team to help guide them as they build technical solutions on top of our platform. So in addition to all of that, in addition to the roles and responsibilities that we've created, I believe one of the most important things we do is this, is that we build trust with all of our clients. Now that trust, as Jeff pointed out, is built on this foundation of objectivity, but we earn trust with our clients with every phone call, every engagement, every product release. And our clients, as a result, don't view us as vendors. They view us as partners because when we have all of these touch points across every engagement that we have with our clients, we're creating a collaborative environment and we're advising them on what to do. And so through that, our goal always in every engagement, whether it's a client or someone who's an inventory partner of ours, the goal is always partnership. So as a result, that leads us to what I believe is one of our other core strengths as a business. We have created the world's largest portal into the open Internet. So we've brought on hundreds of inventory and data partners. We've created a marketplace of measurement, as Samantha just pointed out. We laid the foundation of addressability, targeting and measurement and all of that is so that we can maximize value for our clients and make sure that we're doing something that's going to help their businesses. Another way to look at that is like this. Because all of those things that we do in the platform connect from our first engagement with the customer to the products that we build that Dave talked about, to the obsession we have about the supply chain and finding efficiencies in ways that we reach data and inventory, to the partnerships that we bring into the platform across measurement and data and inventory. All of those things connect. And all of this at every stage builds trust and creates value for our clients. Another way to look at that same paradigm is like this, in that we've created a flywheel for our customers' business. So when a client strategy collides with the inventory, data measurement and all of these things that we bring to bear in the platform, that creates a virtuous cycle for our clients. where everything that they're doing is informing how they reinvest their media dollar on the next campaign. And our clients agree. So we continue to expand our client base and grab land. I think the reason we do this is because our clients value a trusted partner that is focused on the open Internet. And really important to this is that our agency partners remain central to our strategy. Now we've evolved and we've started to talk more to brands. But as Jeff pointed out, we do that arm in arm because the agencies are still a critical component. But often, when we go into those brand conversations, we're doing that arm in arm with the agencies because at the end of the day, they might be the ones executing the campaigns, despite the fact that we might have an MSA or JBP with that brand, the agency might still be involved as a hands-on keyboard element to execute the campaign. So that partnership today between the brand and the agency and the Trade Desk has never been stronger. So as a result, we continue to go upstream. We continue to engage with more and more global brands and we're structuring more long-term partnership frameworks to keep those partnerships accountable. We want to reach milestones together. We want to set goals for ourselves. We want to set ambitious goals that we want to do as partners, both on the business and on the product side. As Jeff mentioned, those often take the form of joint business plans or JBPs. We've done a fantastic job of growing that part of our business over the last several years. But let me unpack what a JBP or joint business plan actually is for just a second. So as I mentioned, it's really a framework for a partnership that often is done over a multiyear engagement where we set milestones for each other for our partnership. Those milestones could look like spin targets on our platform, they could look like things like how do we adjust the mix of spend over time to move it into a much more data-driven decision execution methodology. We could look at things like feature adoption and say, "Hey, you're a little light on certain features. How do we figure out how to migrate your business into using these features that are going to be good for your business?" But we always want to do what I just said, which is that we want to structure these engagements, all through the lens of what is right for the clients business. These are nearly all custom, right, because we're looking at it through the lens of their business. There's no copy paste, take the JBP we did for somebody else and give it to another brand. We want to make sure that all of these are designed specifically to bring value to that business because of what I said earlier, we're partners, we're collaborators. We're advisers. We want to be there with these clients to help their business succeed. Now you might ask, well, what does that actually look like in practice? So here's an example. So this is a large financial services company. And this is a company that we didn't work with a few years back. So they came to us, and CTV was one of the big drivers of why they wanted to come and partner with the Trade Desk. They had amassed a pretty significant data asset, activating that data asset on CTV was really, really important to them. So these new engagements often start with test budgets. So they'll come on for a couple of quarters and start to build then they'll start to find the value in that, and they'll start to ramp. Then we'll start to see growth over the course of a year. And at that point, once we've done everything that I just talked about, we've engaged with the business on all levels. We've understood exactly what they're trying to achieve. We've built things potentially for them that are accustomed to their business. At that point is when we start to think about what are the next 2 years look like? What do the next 3 years look like? And at that point, we engage with a joint business plan to set milestones and goals for our partnership, including accountability around what those next 2 or 3 years are going to look like. Another example looks like this, and this happens a lot, which is that a client will come to us with a test budget. In this case, they were really focused on decision CTV. Then they're likely doing this with other DSPs. So they're also testing a few others to see, hey, where should I place my bets as I think about CTV in this particular case? Then they'll take a quarter or so to kind of go through their evaluation process to determine who is the right partner to use. And in this case, they came back to the Trade Desk because of the performance of CTV and decision data-driven CTV and they said, you're the partner that we want to lean in with and that's what's represented in the first quarter and second quarter of this year. And then from there on, we start to look longer term, and we start to figure out what's the right framework for that business, how do we create long-term accountability, long-term milestones for all of us in the form of JBPs. But everything is designed with this in mind. All of those JBPs, all of those client engagements are designed with this, how are we going to create a flywheel for our customers' business because if we help them succeed, then we succeed because our interests are aligned because we're objective and because we build trust and as a result, we continue to grab share and we continue winning. Now One of the amazing things about these core principles that we've talked about today and even some of the things that I'm going to talk about later, is that all of these scale globally. When you think about concepts like objectivity, empowering the open Internet, hiring great people to build trust and collaborate with our clients, that is true everywhere in the world. Now that we're able to kind of get back out into some of these markets around the world. Everywhere I go, I'm just reminded of this fact constantly, which is that everything that we are doing as a platform is universal. Everywhere we go, these same principles, the same first order principles are true because all of our partnerships are global. When we look at things like our holding company relationships and many of these JBPs, they are global in their structure. And so are we. Because we have this commitment to objectivity, data-driven media buying, building a scaled marketplace and a world-class product, all of those strengths are universal and that foundation is strong. And so we continue to add at offices where we think it's prudent. So just in the last 12 months, we've added an office in Milan, Stockholm, India, and we've expanded our operations in Australia to include an office in Melbourne. Now in each of these cases, we often incubate from a hub. So if you think about somewhere like Melbourne or even Milan or Stockholm, we'll incubate that opportunity from Sydney or from London. And then once we believe it's the right time, we'll put feet on the ground. So we're always very deliberate in the choices we make. We're not throwing darts at the map. We're really being very deliberate where we know that there's a foundation of our business that we can continue to expand. And we do that because of this fact because our partnerships have never been stronger on a global basis. I mentioned our JBPs and our global holding company agreements. But if you think about the supply side of our platform, we're engaging with people like Disney, NBCUniversal, not just here in the United States, but globally as well. Disney owns Hotstar in India. So we're figuring out how that partnership can expand there. NBCUniversal owns Sky. So what does that look like for our partnership with Sky across the whole of Europe. All of those opportunities are there because our partners think the same way we do. And this is something that we've always said, which is that we're a global company that happens to be based in the United States. So everything we do, we look at through a global lens. All right. Now my favorite topic. So shifting gears for a minute. I think one of the most exciting parts of the media business is the shifting landscape that's happening in television. So I'm going to spend the next half hour or so, maybe 20 some-odd minutes, if I watch the clicker out there, focusing on how we're continuing to drive change and add value to our clients and partners in what I believe is one of the greatest moments in the history of the media business, which is what's happening in television. But first, we talked about first principles. On the inventory team, this is how we look at the world. We hold this really near and dear to our hearts, which is that our mantra is to move fast and be first. We think about this all the time. And in practice, what this looks like is through an engagement with the CTV partner like HBO Max. HBO is arguably the gold standard of content out there. I think we've all probably recently watched a show on HBO. So HBO launched HBO Max in May of 2020, and they launched the AVOD service in the summer of 2021. And when they launched the AVOD service, they launched with some insertion orders, but when they were ready to launch programmatic, we were first. And why were we first? Why did they pick us as the launch partner? Because of all of those principles that we just talked about, because we have been great partners to them all along the way. We've helped them test their pipes early on before they were ready to launch. We were the first in the door. So when it was time for them to go to market with programmatic, we were their first partner. This is exactly how our inventory team thinks about every engagement that we have, not just in television but in audio, digital out-of-home, every new channel that we expand into, we're always thinking about moving fast and being first. And all of that serves this mission, which is we believe when our clients log into the Trade Desk, they should have access to every impression on every screen everywhere in the world. Now that job will never be done. But it does serve our belief that at in-state all things will be digital and all things digital will be programmatic. So as long as we keep this out there as our North Star, and we believe those other 2 things, this will continue to be our north star for the inventory team, coupled with move fast and be first. So I mentioned one of my favorite things to talk about, which is television. So nowhere are those 2 facts playing out more distinctly and more quickly than in the marketplace for television. So in the past life, I used to read a lot of scripts different story, different time, different presentation. But we could not have scripted the moment that's happening right now to create what I believe is the most powerful tailwind for our business that possibly could have happened. And that is what's happening in CTV right now. We anticipated this shift years ago. So as you guys may recall, we launched our CTV platform in the fall of 2017, almost 5 years ago. Now we've had 2 years of spend prior to that on the platform in CTV, but we kind of formally launched our CTV platform in the fall of 2017. Now while we anticipated the shift of CTV, we couldn't have possibly predicted or anticipated the global pandemic. And that accelerated the consumer shift in CTV by probably 2 or 3 years. So in terms of what I was just talking about, all things will be digital. That shift is well underway in CTV, and it's not showing any signs of slowing down. I think at this point, the consumer has voted. We're now watching TV this way. So in other words, CTV is now the dominant form of TV. And that's not just me saying that, that's Nielsen and other companies saying that CTV is now the #1 way that we watch TV in the United States. And the rest of the world is following quickly. So when you look outside the United States, Europe and other markets are also embracing CTV and programmatic, but I believe there is a big accelerator coming into these markets in the form of new entrants into the AVOD market. When you look at people like Disney+ and Netflix, their footprints are global, right? They address homes well beyond the United States. So I think this moment right now is putting everyone in the European and broader non-U.S. television market on notice right now to start to figure out what their business is going to be as it relates to digital as it relates to CTV and as it relates to programmatic. Because the kind of duopoly triopoly makeup of many of the markets in Europe, now is about to be broken with these 2 entrants into the marketplace. So they are all going to have to figure out how to participate in the open Internet and many of them are well on their way. But I think this fact is going to accelerate the CTV curve outside the United States. One other area that's really interesting is what's happening in sports. So if you rewind the clock, sports was one of the reasons to keep your cable subscription or keep subscribing to linear. I think a lot of that's changed. And while sports is still kind of the stalwart of traditional television, if you look at like the highest rated shows of last year, like 23 out of the 25 highest-rated shows were like NFL football and that is still true on linear television. But it's also all moving into CTV. I still think it's incredible that you can log into the Trade Desk today. You can set up a campaign and go live with a campaign this Sunday in NFL football. 3 years ago, that would have been crazy. If I had stood on a stage like this and said, "You know what, I log into our platform and running NFL football. Fox and all the rights holders of NFL football been like "never" that's never going to happen. That can happen today. Livesport has moved into digital. Livesport has moved into streaming and the monetization of Livesport has moved into programmatic. And everything we just talked about leads to this, which is an incredible opportunity for advertisers, 90 million households, billions of impressions, the ability to use data advertising are embracing this opportunity as they should because I think this is the greatest moment if you're an advertiser, this is the greatest moment in the history of the media business. I can now take my data, and I can point it at the television screen. It's incredible. I can break free of age and gender. I can do so many more things now because of what's happened in the landscape. And the pool of supply is going to continue to grow. We talked about Netflix. We talked about Disney+. But as that pool of supply continues to grow, that just creates more opportunities for advertisers to do just that, to point their data at the television screen across a broader swath of inventory. And what's amazing here is that it also creates an extended and expanded opportunity for advertisers to manage reach and frequency holistically. This is a really important point. Managing recent reach and frequency holistically, is in many ways, the holy grail for a lot of advertisers. And even as this pie expands, the math still works with higher CPMs because we're being more data-driven in our buying. So I'm going to walk through an example. I wish I had these slides with me yesterday by the way because we were sitting with a large brand yesterday, having this exact conversation, which is contrasting the kind of old way or even in some cases, current way of buying television, which is that I buy television from this broadcaster, this broadcaster and that broadcaster. In doing so, I create frequency silos. And so this broadcaster has no idea how many ads I served on another broadcaster because they're independent of one another, right? So in creating those frequency silos, this creates a problem for advertisers because as we experience this as consumers, this is the -- I'm seeing the same ad over and over and over again, that's because of this because of all of these frequency silos. And so in this case, the number to focus on there is 40x, right? So I'm getting 40x frequency at 10x per broadcast network. But because we look at the entire universe of CTV, we're seeing from all of those broadcasters, we're seeing all of it we can manage reach and frequency holistically, where all of this continues to be more efficient for the advertiser because I used to want to just talk to everybody 10 times a week or 8 times a week, I can't do that when I create all these frequency silos, but when I can look at everything holistically, I can do that and I can manage everything much more efficiently and effectively. And as I mentioned before, even if the CPM is higher because I'm leveraging my data or I'm bidding into an environment, the math still works because the savings that I elicit on managing that frequency holistically, still make it worth it. And we see this over and over and over again with every single advertiser we work with, every case study we run with everybody that this continues to work. Holistic campaign management biddable data-driven media buying. This is the direction of travel for television. And I'm going to spend a bunch more time on this in a minute. On top of that, Samantha went deeper on this, but we're connecting the entire media landscape to retail data. So just to give a real-world example, what this means in the context of television is that my ads on Hulu or my ads on Peacock can now be linked to a sale at Walmart or a sale at Albertsons or at sale at Walgreens. When we start to illuminate that fact to a brand, you can just see their eyes light up like you've never seen like, "wait, my television investment. I can now link it directly to selling product at Walmart", yes. That's exactly what we mean. This is a powerful combination that's never been possible at scale that is now possible today. There's a lot more work to do to wire up all of these retailers, but the snowball is starting to roll. So with the last 15 minutes, almost exactly that I have, I want to zoom out for a second, and I want to talk about the state of the CTV marketplace, because it is, as we've talked about, evolving under our feet. But I want to examine what I think are some of the dynamics that are underpinning the CTV marketplace right now and define what I think is the north star for the TV business going forward. So first, a little bit of backdrop. So when I sit here several years ago, I talked about this concept, which is that we have major content consolidation happening. And that content consolidation even 3 years ago had already happened, and now it's continuing to happen even most recently with Discovery's purchase of WarnerMedia. So as we predicted, this content consolidation continues. And interestingly, part of the strategy with many of these guys is to kind of fold distribution into kind of like super apps or whatever, but Disney and WarnerMedia have talked about or Discovery and Warner have talked about this. But despite all of that kind of content consolidation, they're still is kind of this other challenge, which is that the landscape can still feel kind of fragmented, especially to people who don't spend their waking days obsessing over this like I do, the landscape can still seem pretty fragmented, especially as consumers when we're like, how come season 1 of Yellow Stones on Peacock, but season 5 is on Paramount? Like I don't understand how that works. But it's really from a consumer perspective, complex, but from an advertising perspective, it's really straightforward. And as I walk through this next bit, I think this will really underscore what our strategy is as it relates to inventory partnerships in CTV. So First, you've got this. So content is kind of king in all of this, right? Content owners control what I believe is the greatest content that we've ever had, probably in my lifetime, maybe even ever. And so the content owners are kind of the center of the universe in all of this. Now they distribute that content downstream to a bunch of different partners, right? They work with OEMs and OS operators. They have their own apps, which also get distributed through here. They distribute content through virtual MVPDs. So there's a bunch of different outlets that they distribute content at. And then downstream from all that are a whole bunch of ad opportunities. And this is where it starts to get really interesting because when you look at all of these ad opportunities, one thing that often gets talked about is that, well, aren't some of those kind of off limits. And that is, to some extent, true in that sometimes the operating systems get a couple of ads or a couple of percentage points on the ads that they can sell. But the universe that we play in is this one, because of a very simple fact, which is that the content owners will today and forever, control the majority of the ad rights downstream from them. Even with their negotiations with the operating systems, those inventory shares are coming down and down and down. There was even a dispute last week, I think, where somebody got turned off from one of these. But we play in this larger world because of that simple fact that I just mentioned. So as a result, our strategy remains unchanged, which is that we work directly with the content owners. We want to create the most efficient and effective route for our demand to reach a Disney or an NBCUniversal. Now Oftentimes, that will include engagements on things like Open Path that Jeff talked about, or UID2, as you've seen with Disney. But what we're after here is we want to create the most effective and efficient path to that content owner because they control the majority of the ad rights. And when they sell us that ad, they sell it in a device agnostic way across their entire footprint. That's a really important point. So when we buy ads from Disney, we buy it across Roku and Amazon Fire and Apple TV in every distribution point that they have because they control the majority of those rights downstream. So our strategy, again, remains unchanged. We partnered directly with the content owners. And one of the benefits of those relationships, by the way, is that when we have those conversations with the content owners, we get to go deep with them. We get to think about the future. We get to think about where is all of this going? And we get to align on the future of the marketplace. And I believe that the future of the marketplace in TV is both ad-supported and biddable. I think the ad-supported things like not even controversial anymore, now that the announcements from Disney+ and Netflix have happened. But I believe that the future of the CTV market is biddable. And the reason for that is that it allows advertisers to use data and it's good for the content owners. And I'm going to explain why. So right now, everyone kind of agrees with that statement. So this is kind of the typical life cycle of somebody launching a streaming app, which is that streaming app will launch, sometimes they'll do some insertion orders or category exclusives, then they'll slowly move into programmatic, sometimes with PG and then ultimately move to biddable. So I think this is pretty typical. What's been fascinating to see over the last couple of years is that time line has shrunk. So the distance between launch and then running biddable marketplace has gotten a lot shorter over the last couple of years. But this fact still remains, which is that when we look at all of those distribution points and all of those ad opportunities and you look at that CTV TAM that Jeff showed in the pie chart, there's still quite a bit of it stuck in this direct sales insertion order mode of bidding with only about 1/4 of it being unbiddable. So in other words, despite a lot of momentum, most of CTV is still kind of stuck on this on ramp. So I'm going to take the next few minutes and explain why I'm pretty optimistic that the entire industry is eventually going to merge onto that freeway. Hopefully, with less traffic than that one. So first, a little bit of backdrop in programmatic. So just to provide Dave actually showed a similar thing, but this is kind of the old world, right? This is when the publisher gets to pick the ads. There's no buy-side decisioning, everything just kind of runs, hope I get the ads I wanted and then I get some reporting later on. That old world is kind of a pre-programmatic nondecision world. This is what programmatic brought to the table. Data-driven, I get to choose as a buyer, which ads I want, I get to use my own data. So over the next couple of slides, I'm going to use some actual data from our platform to illustrate why I think it is inevitable that this is the future of CTV, biddable, not the old way. So first, let me just orient us here on what we're going to look at. So on the vertical access, this is price. And on the horizontal access, this is volume of bids. So we took a look in our platform basically to see what kind of the makeup of that demand is. And for the PG dollars that we have on our platform, typically, they rest in a price range on average of about $18 to $23. Now again, those are averages. Sometimes they're a little bit more, sometimes they're not. But on average, that's where a PG campaign will transact from a CPM perspective. So then we said, well, outside of PG, let's just take a sample of bids across our platform into CTV over several hours on a single day and see what that looks like against this kind of price band of PG. And this is what that looks like. So again, this is a distribution of bids from our platform. They're bucketed in ends because otherwise, this would be impossible to read because they also include volume. And so all of this is biddable demand on our platform from a single day. Now you'll note pretty obviously, most of that biddable demand, if not all of it, is coming in at a price point pretty significantly higher than what PG transacts at. Now the tragedy of this and all these blue dots is, we didn't win all of them. And so that is just crazy to me. So the buyer missed an opportunity to serve an ad to a high-value customer and the publisher missed an opportunity to monetize. Now the incredible thing about that is that that's just an inefficiency in the marketplace. That's going to get worked out, right? Because that will get solved. So when you zoom into this a little bit more. This is where the data to me gets -- the data shows something that's very exciting to me about the future. Because when you look at this, it's millions of bids at $60. It's millions of bids at $45, significant premiums on top of what PG typically transacts at that missed opportunity, which by the way, if you do the rough math on that, it's like $1 million just in those 2 dots. And so that inefficiency in the marketplace, that's going to resolve because when those high bids come into the auction, the buyer is bringing their data. They're willing to pay that price, and it's waiting to be captured by an efficient marketplace. So to put a finer point on this, auction-based selling for CTV is the way of the future. There is no way that Netflix will be able to command the CPMs that they're talking about right now in marketplace in the future without an auction and without an underlying framework of identity to enable buyers to activate their data. The end state for premium content is an auction. And let's double click on that and explain why. So when we look at that price be of $60 and 4 million bids, how does that actually happen? So first of all, an advertiser brings their data into the platform. We then look alike model that data and also leverage third-party sources and other value-added services in the client -- or in the platform to expand to that audience and then we point that audience at the universe of CTV. That right there is how $60 bids and 4 million of those happen because we continue to bid and bid and bid on these high-value audiences on behalf of our clients, and they're willing to pay the price because they're going to talk to an audience that they care about. Buyers care way less about audiences defined by the seller, audience is defined by the broadcaster or someone else. They want to talk to the audience as they understand it. And this is why I'm so optimistic that the future of CTV is biddable because the demand is there. There's just inefficiencies in the marketplace that have to be sorted out. Incidentally, this is also where the forward market is built. You heard Jeff talk about this, not only on our last earnings call, but also earlier today. Dave mentioned this as well, a guaranteed framework of buying in a biddable environment, this is where it's going to be created. It's going to be created right here. where all of these bids exist. All of this demand exists at a premium price point where the buyer is willing to pay it. We just need to create the market for that right there to replace the upfront and that's the future that we're building for in television. Now here's the good news, and this is why I'm optimistic on top of everything we just saw, which should also give all of us a lot of optimism that, that's the direction of travel. But when you look at what's happening in the marketplace, a lot of the broadcasters are already thinking about things in these terms. They have reconciled with all of that data that I just showed, and they have said, "I have got to figure out how to capture that opportunity". So when you look at somebody like Disney, they've integrated UID2. They're on a mission publicly to automate over half their business in the next couple of years, they're reshaping their tech stack to capture that opportunity that we just looked at. Netflix has the opportunity to build this, by the way, from the ground up from day 1, they can build the marketplace for their inventory, the right way from day 1. And again, the only way that Netflix is going to be able to command the CPMs that they hope to get to continue to monetize their content is through an auction model just like that because that's where those CPMs live. They live in a biddable auction-based environment. So that's why I believe the future is bright for data-driven auction-based media buying of CTV at in-state because, like I said a few minutes ago, all things eventually will be digital and all things digital will be programmatic. That's the direction of travel of CTV. And ironically, it's also the reason for this, which is why I don't think any CTV or broadcast company can form a walled garden strategy in connected television. The landscape is nearly perfectly fragmented where nobody has dominant market share in television. So as a result, unlike marketplaces for search and social and things like that, nobody can be too draconian in the way that they approach the marketplace. And Furthermore, premium content is expensive. And in order to monetize that content, you basically got 2 options, subscriptions or ads. And the best way to monetize the ad part is what we just saw, which is a biddable auction-based environment for your inventory because that's where the highest CPMs are that are going to continue to fund your content and continue to allow you to offer these AVOD services so that you can drive more subscribers. So I think it's going to be really difficult for anyone to swim against that tide because I don't think it will work that way, and no one has the market share to demand you to buy it another way. So as we look out ahead, I think the marketplace for CTV is going to thrive. I think the future of CTV is open. I think the future of CTV is biddable and the future of CTV is global. And for all of those reasons, at in-state, I think, as Jeff said earlier, CTV in the open Internet are going to command the first dollar of every media plan. And then the walled gardens as we know them today, are going to be competing for the proverbial leftovers. Now to end on one of those first principles I talked about. So we have a tradition at the Trade Desk that's been here since the day I started, which is that everybody who starts with the company sends out an intro e-mail, and they just introduced themselves to the company. Now I love reading these things because back to something we talked about earlier, it just reminds me that we've all come from different places. We've all come from different walks of life. We've all come from different backgrounds to all be here right now for this moment. And as long as we keep challenging ourselves as a company to be thinking about things at in-state, and we stay focused on where we're heading, how we add value to our clients, through our partnerships and the products that we build and we continue to focus on building that future of the TV landscape that we just talked about. And we never lose sight of those first principles of objectivity, moving fast and being first and accelerating change wherever we can. I think that is going to be the core of how we continue to drive the business forward because we are all powering this collective inertia to move our industry forward and focus on that 1 simple fact to empower the open Internet. I feel lucky every day to be on this mission with everybody that works at the Trade Desk and every member of our team. So thank you guys so much for the time. And yes, talk to you later. Bye.

Thank you, Tim. Thank you, Samantha, before that. One housekeeping item for those on the webcast, we ask that if you want to submit any questions, you can do that now, probably over the next 20 minutes, you'll have that window open. And for everyone else here in the room and at home or in the office on the webcast, we're going to take a short 5-minute break right now and then come back and work on getting it wrapped up. [Break]

If everyone complete return to their seats, we're going to get started with our last presentation.

For rounding third almost there. So thanks again, everybody, for hanging in there. So it's my privilege right now to introduce you to our CFO. I know because, of course, the CFO handles IR more of you in the room know him than you know some of the other speakers. But Blake hasn't been with us that long, although it's been a couple of years now. Hard to keep track of time during the global pandemic. But it's such an honor to work with Blake and to bring the expertise that he brought with him from Amazon. Of course, at Amazon, they live through scale that very few companies in the history of the markets have ever lived through. And as well as they instill a discipline that is something, honestly, we at the Trade Desk have to continue to instill in our people and develop and grow as an institution, while at the same time, maintaining a culture that is very different from some of the biggest companies in technology. So it takes somebody really special to bring that discipline, but at the same time, adapt the style that they have to really help the Trade Desk uniquely succeed. So it's just such an honor to introduce to you, Blake Grayson.

This is it. This is what you've been waiting for, right? All day, it couldn't wait. Seriously, thank you all for taking the time. I know it's a long day for all of you, and I really appreciate it. And so I just want to say thanks. Again, I'm the CFO for the Trade Desk. I've been at I've been here for almost 3 years. I joined about 90 days before the March of 2020 pandemic. I joke with Jeff that this was not in the brochure when I was evaluating the role but anyways, let's just get started right away. Today, I'm going to cover 4 different areas for you. Talk a little bit about growth and profitability. We take a balanced approach at the Trade Desk. I'll run through some performance results. We'll also take a look at stacking ourselves up against other high-growth software peers, there's some other ad-funded companies as well. And then we'll talk about the opportunity a little bit. You've heard a lot about that today. I'll just try to put my perspective on that a little bit. Then we'll talk about investments in leverage, how we think about that, the Trade Desk and something and areas that we invest in, but also how can we get operating leverage. This is not a business model where we have to choose between growth or profitability, we actually pursue both. And then finally, I'll talk a little bit about how we think about capital management, whether that's working capital, free cash flow and then the overall health of our balance sheet. So moving first on to the growth and profitability area. So you've heard about this today. I'm really proud of our ability to grow our top line spend throughout the history of this company. We have a 43% compound annual growth rate or CAGR since 2016. We generated over $6 billion of gross spend in the full year of '21. The thing I find interesting about this chart is by looking at it, you wouldn't notice that there was a pandemic that affected us, but it did. In Q2 of 2020, our gross spend went negative year-over-year. Programmatic decision spend is easy to turn down, it's easier to turn up. We did see and the reason why you don't see any of that impact on the chart is we saw a pretty nice rebound in the second half of 2020. As advertisers become more deliberate about their advertising spend and sometimes dealing with smaller budgets overall, they had to prioritize the spend that mattered. And so in so doing, we saw that rebound now the other thing in 2020, we also had a tailwind, mostly in the second half of the year around the U.S. presidential election. And also like House and Senate races that were hotly contested, and we built out a pretty good political program as well for that. From an adjusted EBITDA perspective, we've made a concerted effort on driving profitable growth. So our adjusted EBITDA compound annual growth rate of 50% is actually a little bit higher than what you saw in the previous slide of 43% for our gross spend. And we ended 2021 with over $500 million in adjusted EBITDA. And I'm proud of our ability to gain productivity as we scale. Now I do have to caveat this a little bit because in 2020 and 2021, companies and the Trade Desk was one of them did have lower expenses. We were a bit more methodical about our hiring and some of our expenses during the pandemic. And so we got some extra tailwind from that. But even regardless of that, I'm really proud of the company's focus and ability to grow but also do so profitably and do it in a smart way. So when you combine those 2 things together, there's a lot of vectors that we look at in order to compare ourselves against you all do it in this room and you do it better than I do. But this is a group of high-growth software companies to compare ourselves against. Essentially, trying to get some companies with relatively similar market cap stars, all that grew at least 20% in revenue in the full year of 2021. So I mean, as you can see, we're delivering adjusted EBITDA margins well above the average of these high-growth software peers. This graph is a bit of an eye chart, but just stick with me for a bit. This stacks the revenue growth on top of 2021's adjusted EBITDA margin. Sometimes you'll hear about this kind of nuance referred to as the rule of 40, where if you can get a combination of your revenue growth plus your adjusted EBITDA margin of at least 40%, that's highly desirable. Now it's obvious every company on here is either 40% or higher. Jeff, I think might have referred to this as the rule of 80 now that's changed, like 40%. But we've obviously -- I think we've done well against our high-growth software peers. And what we've also shown on this chart to the right is some very notable ad-funded peers as well. We stack up very favorably against some of the largest companies around that are in the ad-funded space as well. Now the interesting thing about that chart that I showed that was from full year of '21 is if you look at the first half of this year, the first half of 2022, there's a little bit of a profile change. We've managed the uncertainty well so far through the first half of the year. I think it is important to note that, like I said earlier, programmatic spend can be dialed down end up quickly. Like we saw in 2020, we can certainly be impacted by the macro environment. Sometimes I get investor questions, which because of the performance that we have up here, it's, oh, you're not affected by the macro. We certainly are. We certainly can be. We did in 2020. It's a perfect example of it for me when I go back and talk about it. But I think the nuance is, and you'll hear us as a company talk about this quite a bit, which is that we believe we can gain share in any market, in any environment. So while there may be short-term volatility up or down, over the long term, we think that provides us the ability to have conversations with our customers that will allow us to grab land and grab share quickly. And so I'm really happy with how we've managed through the uncertainty so far. It's still uncertain times today, but really excited about how we performed to date. From the North American international perspective, from 2019 to 2021, we've grown at a CAGR pretty similar in both regions. North America, both of these periods represents about 87% of our gross spend, international, about 13% of our spend. Most of this growth rate in here has been driven by connected television. That's probably no surprise. Both regions have connected television, grow spend CAGR rates of over 100% over this period. So CTV is really what's growing the vast majority of this growth. I mean all of our other channels are growing as well, but CTV is by and far away our largest and our fastest-growing channel. Now the thing that is interesting for those of you that followed us for the last portion of this year, there's been a bit of a separation. International, particularly in Europe, has felt a little bit more of a macro impact for us than other offices. And so that split that I referred to before, that 87/13, it's more like 90/10-ish for us right now. And that's okay. I think that for us, we believe very strongly that as the international macro conditions improve, we're in a position to, like I talked about before, when conditions improve and people make more deliberate decisions, we're in a position that we like internationally. But to date, the international business is a relatively small portion for us. Now when I talk to investors, I get a lot of questions about take rate. It comes up super frequently. So take rates just defined as our revenue divided by our gross spend. Questions tend to focus on our ability to maintain that take rate over time as we scale. And the take rate is essentially the share of the value that we keep as a company. And questions about whether or not as we grow and as we get larger, can we maintain that? You'll see on this chart, our take rate has been very, very consistent over the past 8 years. Our take rates ranged very tightly around 20%-ish every single year. Another question I get on take rate usually relates around CTV. And the question is, well, can you maintain your take rate as CTV grows and share? Well, the answer is you can see it right here, that from when CTV spend first started showing up on our platform back in 2016, the first full year of spend to today where video represents the -- it's like the low 40s percent share of our business. CTV makes up the majority of that. Our take rates have been relatively consistent for that whole time. I think if you look out over these last 5 years here from 2016 to 2021, our take rate has gone up 3x. It's gone down 2x. So there can be short-term volatility in it, but it really just indicates the history of the performance that the take rate has been relatively consistent. And then the other thing we do is we share these numbers annually to highlight that we essentially -- we provide a consumer surplus. We extract less -- or sorry, we add more value than we extract. Since 2016, we've made a lot of updates to our platform. You've heard about it consistently from all the folks who have been up here today. We've created the data and the value-added services marketplace. It's got over 250 partners. We've added products that drive value, whether that's cross-device or predictive clearing. We've got unified ID. You heard about fractional pricing as well. So we've done all these things. We've launched a new platform in Solimar. We're launching new products like OpenPath, that drive supply path optimization. So we do all these things, but we do them while maintaining the take rate, the value that we extract from customers. The reason we do that is we believe it spins our flywheel. And we can provide more return and more value to our customers, they'll come back, they'll spend more with us, they'll use more campaigns, and we're happy to do that if we can continue to be their partner and spin our growth. So -- and after all of this, we're able to generate efficiency with how we deploy our resources. This chart shows the full year 2021 revenue and adjusted EBITDA per employee. It shows we've got a desirable model. To be fair though, again, especially in 2021, we were pretty methodical and pretty deliberate about our hiring in 2021 with regards to the uncertainty. So that played a role in both of the kind of the performance here, I mean, we'd still be -- still very confident in our ability, even if we were hiring at our normal rate. But I think the benefit of us doing that in the situation that we did is that we're not in a position today that I read about that we all probably read about, about so many companies that are talking about layoffs or hiring freezes that maybe overextending themselves. It's -- of course, if things change, there's a number of levers we have available to us to make those kind of hard decisions. But I think those being methodical and deliberate with our investments over time has really set us up really nicely for us as we continue to grow out of the uncertainty. I'm also excited about the diversification across our business. There are some companies out there that are highly dependent on a few verticals, a few industries, things like that. We have a breadth, I think, that I really like as a CFO. Here, you can see the mix of our spend and the change from it between 2019 and 2021. I find this interesting because it's essentially what you might call as pre-pandemic, so 2019 into 2021 and seeing some post effects. So some verticals have grown. You'll see health and fitness up there has grown a bit. That's a pretty big vertical for us. It includes things like CPG products, health care, pharma, fitness-related activities. The one that I've highlighted on here that I find interesting is one that's actually declined in mix, and that's automotive. It's declined a bit, and that's an industry obviously, that spend a lot on advertising because they're very active in it. Supply chain issues that started for that industry in mid of 2020 still continue to persist a bit. I'm excited about it. We've actually won business from automotive companies over the past 2 years. And when supply issues, supply chain issues and the other issues resolve themselves, I'm really optimistic that this can be essentially a vertical that's still got recovery opportunity for us moving forward. I mean, I just know the pent-up demand for new vehicles is super high, and so we're really excited about that. Other metrics you probably have heard us comment. So we talked about our retention rate quite a bit. We continue to exhibit retention rates of over 95%. We've done that over 8 years in a row. We have a pretty reliable customer base that we get to work with. We worked with over 1,000 active clients we represent over 50,000 advertisers. It provides a level of scale that I'm excited about, helps us drive more leverage and operating efficiency, too, as we grow. You also heard about joint business plans or joint business partnerships that continue to increase as a percentage of spend. You've heard about this, I don't need to go into too much detail, just multiyear endeavors with large advertisers, the main benefits as our team gets to have multiyear strategic conversations with clients, not as much the tactical 6-month and under type conversations or even less sometimes you have with customers. For me, from my perspective, one thing to point out, these are not like standard SaaS contracts where you amortize revenue or anything like that. There are endeavors to spend that we join. We lock arms with our customers. We provide incentives over a host of different areas, and we deploy multiyear strategies with them. It leads to much more interesting and fruitful discussions with our customers. It just -- I think it really helps. I can hear it from Tim and [ Jed ] and other folks on the team just our sales, account management and client development teams really get the opportunity to have a much larger, longer like term conversations with these customers, and we think that's going to spin our flywheel faster and longer because of it. So real quick about the opportunity, and you've heard about this a lot. This slide just represents the open tenet is huge. It's that blue that section, the video, audio and display. This is about us just getting started. The entire ad industry is projected to be about $800 billion, $820 billion. I bet if I ask 3 of you in this room, you'll all give me different answers. So directionally, that's about how large we think it is. But to be fair, there are some channels in here that we don't really participate in today, like search. Today, we don't really actively participate in. Now even if you take search out, there's 500 -- over $500 billion of TAM or opportunity that remaining. Now while we really don't participate in social, we are winning spend from them. I think that social platforms have struggled a bit, the Apple IDFA changes and such are providing opportunities for us to have conversations with advertisers about spend that they otherwise might have put to those platforms. But what I'm really excited about is how big that traditional media numbers are below, the majority of that is linear television advertising. And so that's going to move to digital, and we believe it's going to mostly move to CTV. Of that, $820 billion, you see the $250 billion there in linear. We've got total TV advertising spend. You see CTV there and TTD is that tiny little slice that you can kind of barely see there. I know it barely shows up on the chart. I believe, just like Jeff does it, it's not a question of if that linear advertising spend is going to move, but how fast. And so we all have to make decisions like do you think it's going to go down in a very slow linear fashion? Or do you think it's going to follow the streaming hours, right? Which is already -- it's such a different viewpoint for me when I look at this chart that streaming hours are quickly becoming similar in size to linear, it's going to overtake it in the near future. And that number -- that $250 billion number doesn't even include what Samantha came up and talked about with regards to shopper marketing. Retail Media loan, depending on who you ask, could be $100 billion opportunity just in 2022. So -- it's just the size of the pie, the available TAM for us is so large. I actually showed a version of this to all of our employees. And it was a way to provide some reality in the situation where we are. We've been very excited about our growth today -- and we should be really grateful about how well we've done, but we are really just scratching the surface of the opportunity. And so it's really just to show everyone that if we can execute and do this right, the opportunity for us is multiples larger than what we are dealing with today. So a few quick slides about investments and how we think about operating leverage. So when we look at operating expenses, excluding stock-based comp as a percentage of our revenue, and you can find all this in our filings. Between 2019 and 2021, we made a concerted effort over the past few years to be really thoughtful about productivity, especially in areas that don't impact the customer experience. And so what you can see here is between those 2 years, we gained the most leverage actually in our platform operations bucket. And now that was our largest investment area in 2019. We've reduced that materially. Predominantly, that's the infrastructure needed to run the platform. So whether it's becoming more efficient with our on-prem data centers or more efficient with third-party cloud arrangements and structures that we have with a number of providers. I'm really proud of the team and just the diligence went into this. I really enjoy engaging with them on these discussions. I hope they do too. But it's something where it's -- I'm really proud of the work we've done there. And then you can also tell on this chart that we've gained leverage as well in G&A. And it's -- we're still investing. It's really just a question of how fast you're growing relative to your revenue and we still want to make sure we invest in these areas because we have to scale. Our support functions either scale up to support the growth we need, but I'm a believer that can we do so efficiently and be able to essentially drive leverage while we're also driving growth. And then we also track productivity in areas that we invest very hard in. Sales and marketing doesn't get a pass just because we think it's really important, right? It is important, but we also want to make sure we hold ourselves accountable to the amount of expense that we're investing and we're getting the return for ourselves that we think is realistic. Now this doesn't mean that we won't invest ahead if we have an opportunity available for us. But over the long term, we believe we have a very efficient model, and it stacks up pretty favorably to these high-growth software peers. And then over the long term, we believe that we can drive growth, gain share and generate adjusted EBITDA margins of 40%. Now in my earlier slide, some of you -- if you're watched closely, you might say like, "Wow, 40% seems like an underwhelming goal. We did a 42% EBITDA margin in the full year of 2021. Now my answer to that would be, I agree only if that was the only goal we went for. And that if we were -- if it was all about adjusted EBITDA margins, if that was the one thing that mattered to us more than anything, we can optimize our business for it and we could beat that, target that we have up there on the screen. The reason why it's on there is because we're all focused about that balance between growth and profitability. And you saw that in the previous slides when I was talking about full year revenue growth and on top of the full year adjusted EBITDA margin is that we have that balance. And I think it's really important for us to be able to manage that. So over the long term, I think we can drive profitable growth and continue to gain share. And so finally, just a couple of thoughts about how we manage our capital and what the thought process that goes into that. So working capital is top of mind for me, it's top of mind for my team. It's top of mind for the company at The Trade Desk. In general, we pay for media in advance of receiving payments from our customers. So that creates what I refer to as a working capital gap. Our days sales outstanding is larger. It's higher than our days payable outstanding. So when you're buying billions of dollars of media in a given quarter, having a gap of, call it, 15 to 25 days of working capital gap means there's hundreds of millions of dollars that are outstanding and float, and we have to be able to make sure that we're managing that gap well. You'll also note in the chart that there was a pretty marked improvement that happened right around Q2 of 2020. When COVID hit and the uncertainty that occurred, we spent a lot of time looking at our working capital process. And by inspecting our collections and payables processes, we drove some improvements, and we'll be able to maintain those, generally speaking, and it showed up in our free cash flow results. In this graph that shows trailing 12-month free cash flow, you can see that improvement that occurred in that back half of 2020 from those efforts. I mean if you look at our -- in 2020, our free cash flow was actually higher than our adjusted EBITDA for the year. That's pretty unique. 2021 also improved as well. But again, to be fair, we were still benefiting from the virtual environment, both in slower hiring and a bit lower than normal kind of overall, mostly employee-related around that. But even regardless of those 2 things, we have a business model that generates strong free cash flow. It allows us to self-fund our working capital as we grow, and it also adds cash to the balance sheet. Over the last 2.5 years, we've added approximately $1 billion in cash to the balance sheet. We continue to have zero debt outstanding. We've generated positive cash flow in every one of these years, 2019, 2020 and 2021. And at scale, I would expect all things being equal, this would continue to improve. Having the capital provides us flexibility, and we've had a number of discussions about it from a capital management perspective. And finally, just how we think about the options for capital allocation. Investing in the business working -- investing in the business and working capital is how to date, we've managed that. We've deployed it that way. We're always looking to reinvest and drive growth. Managing working capital is investing in the business. That's a function of it. I would also say that having that capital in challenging times, we have more flexibility to use this leverage as we can win more spend. As it relates to M&A, M&A is not something you hear much from us about at all. But it could -- it might surprise you to know we have a very active M&A program at The Trade Desk. Valuations are down significantly. Our team is constantly looking, but we're really -- we're super deliberate about it. What Jeff and I like to say is that we're very good at saying no. And we say no a lot more than anybody knows. But what I can also say on top of that is that if it is the right choice. We wouldn't spend the time on it. If it wasn't something we were actively looking at. But if it's the right choice, we'll make that choice, but we are picky and we have the luxury, I think, to do that. And our history shows that we try to make decisions not just right for the product or our customers, but also for our culture. And so we're thinking about all those things as we consider M&A. As far as share repurchases go, that's also an option for us to consider as we evaluate that. And if we ever did one, we prefer to be opportunistic about it and offsetting dilution, if possible. That's kind of at least the way that I think about it. Dividends are also obviously an option as well. But it's likely a little bit too early, to be honest, in our company's journey for that. But these are the options that we consider and I think about all the time. And so with that, I just want to say thank you, everybody, for sticking around. I think we're going to do a Q&A, but Chris will come up. Thanks a lot.

Awesome. Thank you, Blake. We're going to take just 5 minutes while the speakers, you can come up to the stage. We'll get the chairs, get the mics, and we'll get ready for the Q&A. [Break]

All right. We're going to begin our Q&A session. [Operator Instructions] So with that, I'm going to start off with Laura Martin from Needham.

So Jeff, one for you. love the shopper marketing initiatives. You're spending a lot of money on that and a lot of money on data. To me, they feel like they're at odds. Shopper marketing closes the loop between an ad buy and a clear return on purchases. So doesn't data become less valuable over time, the more you can close that loop because data is a proxy for the return on ad spend. So it feels to me like you're spending money on 2 things that work against each other. Could you speak to that, please?

Sure. Honestly, I've never thought of it that way. In large part because we're using the data -- and incidentally, Samantha, I'll pass it to you after I take a stab at this. But we're using the day to make more informed decisions so that people can get a higher return, right? So you want to advertise to the people that are the most likely to convert. That data helps you make a more informed decision about what -- where you're advertising. Without that -- and I think over time, incidentally, I think more and more companies are going to give you the measurement for free and then charge you for the targeting capabilities. So if you look at it that way, where the data is being invested sort of in the middle to incentivize people to be more effective and optimize their entire marketing spend to sell product in those few retailers. But from the retailer's perspective, they want that, of course, because they would love to have a CPG spending billions to sell more product in their stores, which creates obviously a flywheel for them. And incidentally, that's one metaphor that we use when we're talking to the retailers, which is you ought to think more about the flywheel and the way that Amazon does because, of course, they have that benefit. But there is, of course, the same opportunity to do simply because it's offline doesn't mean you can't do it. And with the online to offline capabilities that obviously can capture that. But then on the back end, you're just proving the efficacy of the data that you paid a small amount for in order to increase efficacy by far more than the cost of the data. It just creates a more efficient supply chain across the board. I don't know if you want to?

Yes. The only other piece I would add is I think that different retailers have different levels of coverage or access. And so for example, McDonald's, some people may have their app and provide their information or pay using a credit card. But a huge portion of their revenue is cash driven. And so when McDonald's thinks about leveraging data, there's portions of shopper data that may be of use -- but I think there's also a lot of other data-driven components that they want to leverage from demographics to lifestyle to behavioral that I think will be really important to continue to offer in the platform.

Our next question will be from Tim Nollen from Macquarie.

A question for Jeff, for Blake or both, I guess. We've got an interesting scenario in front of us where we know the ad market looks like it should be slowing into the back half of this year, at least into next year. And yet you've got all these great structural growth drivers. I wonder if you could help us sort of suss out what we might look forward to in 2023, I mean, in past recessions, you've had ad dollars shift from traditional to digital. Now that big shift is linear TV to CTV. So that could be a very powerful structure driver. I know you haven't called out a Netflix direct partnership yet, but could be Disney+, you do have retail media is great. So is there a way to think about 2023 cyclical versus structural?

So I'll take a stab at this. We'll both have our own crystal balls that we'll try to look into deeply and -- you can tell me the mines are wrong. I think that -- the way I think about it is -- this shift from linear to CTV, I think is going to be faster than anybody believes it will be. Will it all manifest that quickly in 2023? I don't know if it comes out quickly. But I think what we have shown from our history, and I've talked about this quite a few times, is that volatility creates discussion opportunities for Tim and his team that I don't know exist as well at other companies. I don't -- I mean this is just my anecdotal personal opinion. But I do think that the linear shift is going to happen faster than people expected to, just like the streaming fixed shift happened now. COVID helped that, but it's also sticking with these kind of viewing hours. And I think it's just a question of macro uncertainty, we'll just have to adapt accordingly. However, that turns out with itself. We've adapted, like I talked about with international and issues in Europe. But -- like I said, we have had the ability to forward invest in those places that can be ready for it, and I'm very comfortable doing that because when you look at the global advertising market, I think 2/3 of it is outside of North America. And so just knowing that fact alone would say if you believe you have an unbiased product where we can come in, represent the buy side and make that progress, we're in the best situation possible. But macro conditions is always going to be subject to that.

The only thing I'll add is just that part of what you're asking is how strong is the headwind of economic sort of macro stuff that all of you know better than I do. And then -- and then how strong is the secular tailwind behind CTV and growth and move to digital. And the thing that's kind of interesting about the interplay between those 2 things is that particularly in connected television, if there's stronger pressure on the consumer, both in the United States as well as outside of the United States, I think you could argue we're already in a recession in Europe. -- that it puts pressure to move to AVOD. Consumers are interested in paying for it with ads. And that makes it so that everybody has to move faster into AVOD, especially when you really consider the content machines that all of them are and how quickly they need cash to continue to accelerate and grab land. So even if there is sort of stronger headwinds than people thought weeks ago and who knows like it's so hard to predict at this point. And I would argue, by the way, that and I've made this argument in multiple forums before that I believe there should be a VIX for advertising. It should -- it will never move quite as fast as the VIX does for the equities markets. But if we did have one, I think it would be at an all-time high right now. And that's largely because they're all looking at the same tea leaves you are trying to figure out what the future holds. And there's just more uncertainty for them. So the average CMO is in a scan the landscape head on a swivel mode right now. And I believe no matter what happens, the secular tailwind of television in the first half of next year will be something we've probably never seen before and we'll never see again. And that is -- is going to be accelerated in the event that there are headwinds, especially un-predictive ones.

I'm Mike Morris from Guggenheim. I wanted to ask you about that sort of 16% CTV share and the 9 that's still sold direct. And what you see as the catalyst to moving that forward to. I think you referenced inefficiencies in the marketplace that need to be worked out from the media company perspective or the publisher perspective, it does feel like they want to like really like hold onto it tightly. So I'm curious as you look over the next, I don't know, 1, 2 years, what do you see as the kind of tactical catalyst to help them see the light, if you will.

Yes, there's a couple of things. So one is structural. So -- and I think most of them have taken care of this already in that they've organized their -- like if I look at it through the lens of a content owner, they've organized their sales team now where pretty much everybody now has a sales team that gets -- that is organized in a way that regardless of like the door that the money comes through, they're all compensated the same way. This was not the case a long time ago where it was kind of like the linear team ruled all. Now I think they've organized their team where it's a little bit more, I guess, fair. And then the other is technology. And so the slide that I showed earlier, one of the things that's happening right now in the marketplace that I think is going to solve for this problem is that when you look at all that opportunity in the biddable marketplace, because of the way that some of the legacy technology is structured, it's kind of unable to capture that. And that's the efficiency in the marketplace that I referenced earlier. A lot of these guys are building for that [ in ] state. Like if you look at how Disney is starting to think about their technology stack. It's all designed to capture that opportunity because we've showed that same data to all of our broadcast partners. And when they see that opportunity and that missed opportunity, moving as fast as they can to figure out how they break that log jam. And I think once that's solved both technologically and structurally from a sales team perspective, I think all of that direct kind of insertion order-type money starts to move really, really quickly over into biddable.

The only way to get the CPMs that they need to fund their content is through an auction. And especially when you have something that's really unique and scarce an auction is the very best way to monetize it. That's what they do in art and everything else that is unique and special. But if you look at, for instance, Netflix talking about $50, $60, $70 CPMs, that is only available in a biddable environment, in part, because on the advertiser side, that's the only way that you can bring your data to the equation and control for reach and frequency. You put all that together and it's inevitable.

Jeff, you talked a lot recently about just gaining share from social media budgets. Are you starting to see that shift happening in real time? Or is this more of a hypothesis and something that you see playing out over time?

Yes, I'll take a first stab at it, and then Tim, if you want to add any color. So I don't know necessarily what you mean by real time or how real time in the sense that is it happening right now? Or does it happen immediately when something changes. What I would, in general, say is that as efficacy as a result of Google and Apple changes to social. And that's not just Facebook, that's the entire category. As efficacy has changed, they all reassess. And as they've reassessed, they put more dollars into things like CTV and programmatic, where they have the opportunity to choose. It's also given them the ability to look more holistically instead of just in specific destinations and the paradigm that does take a little bit of time to understand why it's better to buy through an aggregator than to buy directly because that's a bit counterintuitive to lots of other supply chains. But because we earn our keep in decisioning that creates a much more effective media plan for the advertiser that then gets continued and perpetuated.

Yes. The only thing I would add is like that paradigm that I put up earlier today where a lot of the ways that social is being bought is kind of this like, well, I'll hand over my money and then let me know how it goes. And so that paradigm that I put up earlier where you kind of got the old way of doing things, the publisher decides because the social media companies are publishers at their core. That's what they do. And so that paradigm on the left side of the slide that I showed earlier, that's how they sell media. What advertisers want is what was on the other side of that slide, which is they want to choose and they want to bring their own data to bear, and that's been really difficult and social, especially for all the reasons Jeff talked about.