Greetings. Welcome to The Trade Desk fourth quarter 2019 earnings conference call. At this time, all participants are in a listen- only mode. A question -and- answer session will follow the formal presentation. If anyone requires operator assistance during the conference, please press star zero on your telephone keypad. Please note this conference is being recorded. I would now like to turn the conference over to your host, Mr. Chris Toth, Vice President of Investor Relations. Thank you. You may begin.
Thank you, operator. Hello, and good afternoon. Welcome to The Trade Desk fourth quarter and fiscal year 2019 earnings conference call. On the call today from our Ventura headquarters is Founder and CEO, Jeff Green, and on his first earnings call as Chief Financial Officer is Blake Grayson. A copy of our earnings press release can be found on our website at thetradedesk.com in the Investor Relations section. Before we begin, I'd like to remind you that except for historical information, the matters that we'll be describing will be forward-looking statements which are dependent upon certain risks and uncertainties. I encourage you to refer to the risk factors referenced in our press release and included in our most recent SEC filings. In addition to reporting our GAAP financial results, we present supplemental non-GAAP financial data.
A reconciliation of the GAAP to non-GAAP measures can be found on our earnings press release. We believe that providing non-GAAP measures combined with our GAAP results provides a more meaningful representation regarding the company's operational performance. I will now turn the call over to Founder and CEO, Jeff Green. Jeff?
Thanks, Chris, and thank you all for joining us. You've already seen the results, but let me just cover a couple of highlights that I think are particularly noteworthy. Total spend on our platform in 2019 was a record $3.1 billion. Spend in Q4 topped $1 billion. The first time we have ever crossed the $1 billion mark in a single quarter. This record spend helped drive revenue growth of 39% for the year. Our adjusted EBITDA margin was 32%, marking the fourth year in a row above 30%. Net income was $108 million, and we have been GAAP profitable every year since 2013. These results are highly encouraging. We continue to generate strong revenue growth while we also deliver profitability that is significantly higher than almost all other comparable software and internet companies of our size.
We will also continue to look for ways to invest that profitability to accelerate our growth. With all of that said about 2019, I'm even more excited about 2020. I've never felt more confident heading into a new year. That's because as the largest independent DSP in the world, I believe we are uniquely positioned to grow and grab share regardless of any changes to the regulatory or tech environment. I'm expecting spend growth to accelerate this year to 35.7% year-over-year versus 33% last year. We project that will drive revenue growth of at least 30.5% and EBITDA margin of at least 30%. Part of our confidence comes from the amazing progress we made in 2019.
Whether it's bringing more and more advertisers onto our platform or the tremendous growth we've seen across our inventory availability, or the explosion of connected TV and how that is transforming advertising, or the growth of the use of data on our platform, which grew 65% year-over-year, I'm amazed at how far we've come, and so quickly. The progress we made in 2019 puts us in a great position to grab share, not only in 2020, but well into the future. I want to spend a few minutes getting into these points in more detail. I believe that in many ways, we're at a watershed moment in our industry and for our business. I say that not simply because of the revolutionary impact of emerging trends such as CTV, but also because we're at a point where we no longer have to make the case for data-driven advertising.
Advertisers get it. They understand it, and they want to apply it. It's no longer a question of, what the hell is this thing? It's more a question of, how can I do more with you? While advertising is about $725 billion today, we expect it to pass $1 trillion in about seven years. We continue to assert what we have from the very beginning. Data-driven decisioning is the future, and the objective and more transparent companies who manage to gain scale will be the dominant force of the new advertising market that exists on the other side of digital transformation. On the inventory side, if you gave me a blank sheet of paper a year ago and told me to predict what a successful year could look like, I would have never come close to what we actually accomplished.
In 2019, we exceeded every one of our goals to bring on premium inventory. The mantra of our inventory team, led by Tim Sims, is to be first and go fast. This is so important because adding massive scale in areas such as CTV, audio, and mobile is a great leading indicator of future spend on our platform. If you look at CTV specifically, an advertiser can reach over 30 billion premium inventory impressions every day from nearly every major content provider. That means there was exponentially more inventory available in 2019 than the year before. Here are just a few of the inventory highlights from 2019. We became the first DSP to integrate with Amazon Publisher Services. We expect this relationship to grow in 2020. We were the first DSP to launch live sports with Disney programmatically. We were the first DSP to run programmatic guaranteed audio with Spotify globally.
In Europe, we integrated with Channel 4, Sky Germany, ProSieben, RTL, Mediaset, TF1, and Rakuten TV. That's like the Champions League of European media. It's a similar story in Asia, where we became one of the first DSPs to integrate with many of the region's top media providers, such as TVer in Japan. In Southeast Asia, we have added over 13 new premium TV partners, and these include VTV in Vietnam, TrueID TV in Thailand, and HOOQ across all of Southeast Asia. One last one, which I think is particularly insightful and important, we became the first DSP to go live with FreeWheel's Unified Yield products. This is effectively header bidding for TV. Let me explain that one for a minute. The ad ecosystem is not too dissimilar from the equities market.
Supply and demand isn't always measured completely, and when an order is placed, or in our case, an impression is sold, it may not be routed in a way that maximizes yield. Banks or walled gardens may route the impression in a way that doesn't narrow the gap between the market price and the transacted price. Header bidding is a technology that allows the demand side to access all ad impressions and bid on them. It collapses the siloed supply and demand so that the market is purer, cleaner, and fairer. No front running. It was pioneered in display advertising, but now CTV content providers working with us are starting to adopt something very similar. FreeWheel has progressed the quality of transactions in CTV by creating a header bidding-like product. This is important because traditionally, TV media sales have followed the classic waterfall method, which silos demand.
Direct sales teams sell the highest valued inventory to the largest buyers first, and then so on down the waterfall. As more TV content shifts to connected devices, header bidding augments this process. Buyers like The Trade Desk can now compete for every impression. That means we can bring more demand. Suppliers get to consider all the demand at once, which increases CPMs and improves forecasting for publishers. Where we work with FreeWheel's Unified Yield, we have seen spend increase up to 5x. We're convinced this will have a big impact on TV over the coming years. I know we've talked quite a bit about CTV on these calls in the past. We are in the middle of a once-in-a-lifetime consumer shift to connected devices and streaming content.
That's driving all providers to rapidly develop and deploy streaming services, many of them working with us to maximize ad demand. We are seeing that translate to our platform. CTV spend grew 137% in 2019, and current trends indicate it will double again in 2020. If any of you were at CES this year, you would have been impressed by the scale of this transformation. You would have also seen that we are only in the very early stages. There's a whole section of CES that is devoted to the future of marketing, and this year, CTV was the largest topic of conversation. What you heard was advertisers looking to devote more of their TV campaigns to CTV, and content providers racing to make more of their inventory available. I was on stage there with Linda Yaccarino, Chairman of Advertising and Partnerships at NBCUniversal.
We shared our common view that this is the year where the industry starts making that accelerated shift to connected TV, which will transform everything about TV advertising, whether it's the application of data, measurement, or the traditional upfront process. Of course, we're already seeing this shift as major new platforms come online, including NBC's Peacock. Live sports are also a major driver. As broadcasters look to CTV to optimize the unpredictability of live events, many said that sports would be the anchor for linear TV and would slow viewer migration to CTV consumption. In 2019, we disproved that thesis. Last year, we ran ads in nearly every major sporting event. While this merely scratched the surface, it provided very compelling proof and case studies for what data-driven advertising can produce for live content. Needless to say, in 2020, we expect this trend to continue. Switching gears.
On the demand side, there was a noticeable shift in 2019 in terms of how many advertisers are looking at the value we provide. As I said earlier, we have spent much of our 10-year history making the case for data-driven advertising. That started to change in 2019. Today, advertisers get it, and as more channels get digitized, such as TV, they want to apply more data across all their campaigns. They want to do more with us. As a result, in 2019, we saw more advertisers shift their budgets, and in the last 12 months, about 75% of the Ad Age top 200 advertisers spent at least $100,000 on our platform. Early in 2020, we are seeing that trend continue. To date, spend has been particularly strong in North America. In 2020, we expect more large brands on our platform than ever before.
That's one of the big reasons why we expect spend growth to accelerate in 2020. Advertisers are shifting spend from large search and social platforms onto our platform. Much of the spend is expected to go to newer channels, such as CTV and audio. For example, we recently signed a major new agreement with a large multinational life sciences company, and they're shifting a significant portion of their advertising spend away from the walled gardens. That's a common thread among many of our new or growing advertisers. In 2019, the number of MSAs with brands doubled. We expect similar trends in the future. I do want to point out that the overwhelming majority of brands that do sign MSAs with us directly continue to work with their agencies. As you know, we've always been close partners to the agencies, and this won't change.
We also expect more and more brands to sign with us directly so that they can control the activation of their data as their programmatic spend continues to grow. They will continue to lean on their ad agencies for guidance, strategy, and global execution. Why is this shift occurring? As advertisers commit more campaign dollars to data-driven advertising, attributes such as transparency and objectivity become more important. Advertisers understand The Trade Desk is the most transparent and objective platform in the market today. This is even more relevant in the context of issues such as privacy and identity. Let me take a few minutes to explain. First of all, the market has been testing the virtue of the walled garden strategy.
To be clear, the term walled garden describes platforms who do not provide or enable measurement of performance of their advertising outside of their own four walls. You can only access their inventory through them and only review results as measured by them. They get to grade their own homework. What the last 10 years has proven is that this strategy can work well if you have a majority market share. If you own most of the social media ads, or most of the search ads, or most of the UGC video ads, then a company can afford to say, "Buy it my way or not at all." That only works if you're big enough so that marketers are afraid not to buy from you. In TV, this market share concentration does not exist. No studio, no channel, no cable company, no MVPD.
No one has the leverage to pull that off in TV. Because we think video in all of its forms will be about half the trillion-dollar advertising pie, we predict that CTV will be the quantum leap forward that eventually forces all walled gardens to change course. While the economics of CTV are putting pressure on walled gardens, so is the state of the privacy debate. As you all know, there are regulations such as CCPA and GDPR, as well as actions from tech companies such as Apple, Google, and Firefox, to alter the role of cookies. All of these get reported, and over time, it starts to feel as if the advertising industry is lurking around every dark corner, looking to violate a consumer's privacy to steal their data or something like it. That's not what advertisers are trying to do.
It's not even how they think about it. I've spoken to hundreds of CMOs, many of those at the biggest brands in the world. I've yet to meet one that doesn't care deeply about building trust with consumers. They know that they're in the business of building relationships between them and consumers. I've never met one that is throwing caution to the wind on privacy. I've always been in the business of aligning my interests with the buy side, I also care deeply about consumer privacy. I care about it as a global citizen, as a consumer, and as the CEO of a data-driven tech company. In fact, we have thought about privacy from the very beginning when we founded The Trade Desk over 10 years ago. It has helped us prepare for GDPR and CCPA and browser changes without significantly disrupting our technology or business practices.
Just because Facebook and Google gather your most personal information does not mean that advertisers have any interest in acting or trading on that information. Quite the opposite. In most cases, brands are servicing a long-lasting relationship that is incredibly important to them. They understand the risks inherent in compromising consumer privacy. What advertisers are squarely focused on is how to drive better and more relevant ads that help fund the great internet experience that we all enjoy while simultaneously protecting consumer privacy. Outside our industry, these motivations are often misunderstood. At the core of our value is the trust that advertisers have in us, at The Trade Desk, to put their first-party data to work. Through their first-party data, they can model who their most loyal customers are by product.
They can understand what kinds of characteristics are common across these loyal customers, and they can use those data models to find where those patterns exist elsewhere in the market in order to reach their next million customers. That kind of data modeling can be done on our platform without ever having to know any directly identifiable information about the consumer. Advertisers know that we can help them do their modeling work with a limited amount of data or a vast amount of data. If we operate in a world with less data, we're darn good at that. Our whole system is built around making objective choices with limited data. We do that with many of the impressions we see today. In these environments, we already deliver high ROIs for advertisers. If we get more data, fantastic.
Our platform can ingest as much data as you can possibly use in a privacy-safe way already. We do that now better than anyone. We are committed, responsible stewards of consumers, agencies, partners, and advertisers' data. In this way, we are completely aligned with what advertisers are doing. We only use data in a way that builds trust. With this in mind, let me give you my opinion regarding Google Chrome. Google is between a rock and a hard place right now. They are under pressure from regulators and from other tech companies such as Apple, who block third-party cookies. Even if Google creates an environment that only provides for a high level of targeting within the browser that they control, the value exchange of the internet does not change.
What that means is that advertisers and content providers will have to find new ways to fund content through relevant advertising. The L.A. Times cannot tolerate a 50% drop in CPMs just because of a change in policy at Google. In one scenario, a content provider may then ask for consent for a reader's email address on every visit, and then they'll work with the ecosystem to build new identity models. To The Trade Desk, it doesn't really matter what the identity model or approach is. We'll be successful regardless. In fact, I can make the case that we become indispensable in that latter scenario. The next two years will be fascinating, and we'll fully engage with Google and the rest of the ad ecosystem.
Even though the cookie-based internet does not represent the bulk of our business and is certainly not part of the faster-growing segments of our business, like CTV or mobile in-app, it's important, and we'll be fully engaged on behalf of our advertisers. Advertisers view us as their representatives, and we will continue to represent them in the discussion about identity technology. Because we only represent the buy side and because we don't own any media or a media platform, they know we are objective and represent their interests. They know we will protect their data, and they know that we won't allow anyone to trade in directly identifiable information on our platform. As advertisers make major platform decisions about where to run their campaigns, I hope this explains why our objective approach is a significant driver of demand for us.
By the way, this objectivity point is also critical in another fast-growing segment of our business, and that's in political advertising. The 2016 presidential election cycle saw candidates start to leverage data-driven advertising. For the 2020 cycle, nearly every candidate is using it. We are starting to see more spend in this segment, which we expect will increase as the year progresses. As you know, some tech platforms have decided to sit out or limit their role in the political advertising process. At The Trade Desk, it's a very different decision. We are not a free social media platform where bad actors are easily engaged and can, at times, spread misinformation. We only deal in paid advertising. That, combined with our vetting processes, micro-targeting limits, and our commitment to objectivity, mean that we're rapidly becoming a preferred platform for major political campaigns.
I know I've covered a lot of ground here, but I think it's all important to understand what's going on in our industry and our position within it. We really are firing on all cylinders as we enter 2020. First, and as I said earlier, I could not be more excited about our inventory growth in 2019. We have always operated under the inventory model that if you build it, they will come. The economic model of our business means that it cannot grow without growth in premium content. In 2019, we laid more groundwork for demand growth than in any year previously. I focus so much on CTV because, as we discussed, advertisers view CTV as a way to break their dependence on walled gardens. It's why, if current trends continue, we expect CTV spend to double again in 2020.
Second, we have more large advertisers on our platform than ever before. About 75% of the Ad Age top 200 advertisers spent at least $100,000 on our platform in the last 12 months. Even with this momentum, we still see significant growth potential as data-driven advertising becomes an ever larger element of their campaigns. As a result, their current spend with us is still small relative to what we expect them to do moving forward. Third is political. We expect The Trade Desk to be a preferred platform for all major campaigns as they ramp up spending throughout the year. Finally, global expansion. We've made significant investments outside the U.S. over the last two years and will continue to do so. We believe we are in a strong position to grow international again in 2020. The Trade Desk is very well-positioned as the advertising industry evolves.
Our business model hasn't changed since we began 10 years ago. It was founded on the belief that advertisers would demand an open and objective platform as they optimize their growing digital ad campaigns. With every passing year, as marketers embrace data-driven advertising, that belief gets validated further. We are executing well. We're poised for more growth. 2019 was an excellent year, and we expect 2020 to be even stronger. I'm going to turn the time over to Blake to discuss our financials.
Thanks, Jeff. Good afternoon, everyone. I want to start out by saying that I am thrilled to be a part of The Trade Desk. I joined the company because I am so optimistic about the opportunity in front of us. It's rare, in my opinion, to find a company with such a unique value proposition in such a large and growing available market. It's also rare to find a company that can grow at a rapid rate while producing profitable growth, even as we try to invest in the right areas as quickly as we can. I also have been very impressed with the team Jeff has assembled. I'm excited to be part of this journey. I look forward to helping The Trade Desk scale to a much larger company.
As Jeff mentioned, 2019 was a strong year for The Trade Desk as we continued our growth trajectory, ending the year with more than $3.1 billion in spend on our platform and $661 million in revenue. We also ended the year with adjusted EBITDA of nearly $214 million and GAAP net income of over $108 million, marking the seventh straight year of positive GAAP net income. The Trade Desk's ability to give advertisers an alternative to large search and social media platforms was on full display in the fourth quarter. Q4 revenue expanded 35% year-over-year to $215.9 million. We saw strong spend growth during the holiday season in our audio, connected TV, and mobile in-app channels. In Q4, we saw spend in both North America and international accelerate versus Q3.
89% of our Q4 spend came from existing customers who have been with us for over one year. Turning now to our expenses. All comparisons are on a year-over-year basis unless otherwise noted. Total operating expenses in Q4 increased to $163 million, up 46%. Platform operations, or effectively our cost of revenue, increased 34%, driven primarily by the increase in QPS, our platform fees. As a percent of revenue, our platform operations is decreasing year-over-year. We expect more improvement in 2020, which I will talk about in a few minutes. Technology and development grew 36%, driven by increased investments in our platform. Sales and marketing grew 59%, driven primarily by an increase in account management and business development. Finally, G&A grew 61%. We expect to see the most leverage improvement in our G&A in 2020 and certainly over time as we scale the business.
Adjusted EBITDA was a record $83.5 million in Q4, representing a 39% margin. Adjusted net income for the quarter was $71.6 million, or $1.49 per fully diluted share. We closed the year with $255 million in cash and short-term investments. Our DSOs ended the year at 118 days, and our DPOs were 94 days. Finally, I would like to share our guidance for the first quarter in 2020 and preview our focus areas of investment. For Q1, we expect revenue to be $169 million and adjusted EBITDA to be $35 million. For 2020, we expect the full year revenue to be at least $863 million on total gross spend of at least $4.24 billion and adjusted EBITDA to be at least $259 million or about 30% of revenue.
For 2020, we expect total operating expenses of $752 million. We expect our full year tax rate to be about 25%. We expect stock-based compensation expense for the year to be about 13% of revenue. Finally, share count is expected to be about 50 million as we exit 2020. From over the decade I spent at Amazon, I believe in actively managing how and where we allocate capital within the company toward areas that can drive growth and efficiency, not only in 2020, but in the years beyond. Let me preview a few areas I am very excited about. First is in our people that can drive incremental spend on our platform. We will aggressively add more business development and account management employees. This includes both in North America and international.
With the massive available market in front of us and the generational shift to connected TV, you should expect sales and marketing to generally grow a little faster than revenue growth over the next few years. The second is product expansion on our platform. One thing our customers have come to expect from us is more product and functionality on our platform. It is why the company relaunched its platform in June of 2018. It is also why we will continue to invest heavily in technology and development as we keep our focus on the customer and build features and functionality that generate better ROIs for media buyers and their clients. That should spin our flywheel even faster. In the coming year and beyond, one should expect to see more capability in CTV and better tools for measurement and contextual targeting, for example. Third is in our platform operations.
As our QPS increases, our cost to serve also increases. However, we have the opportunity to improve our efficiency and lower our platform operation expenses even more as we scale. This includes improving the management of our data centers, managing the cost of our outside cloud computing partners, in addition to investing into efficiency initiatives that can reduce our cost to serve per impression over time. I can tell you from the relatively short time I have been here that we are 100% focused on building the best products and services we can for our customers, rapidly gaining share and programmatic while it's still in its infancy, and building for the long term, not the next quarter. We already invest today in the areas we believe have the most potential.
I look forward to helping the team move even faster and more efficiently than it does today, and I'm very excited about the opportunities we have ahead of us. With that, I will hand it back to Jeff for any final comments and of course, Q&A. Jeff?
In closing, let me reiterate that we are off to a great start to the year. Our expectations for spend acceleration in 2020 demonstrate the trust advertisers place in us. Our objectivity is more valuable to advertisers than ever. We are even more bullish on our future than we were a year ago. The opportunities in areas such as CTV give us confidence that we will continue to aggressively take share. Our investments are paying off. When we see surprises, they are typically to the upside, and we believe The Trade Desk is best positioned to realize continued growth in our industry for this year, next year, and beyond. That concludes our prepared remarks. Operator, let's open it up for questions.
Thank you. At this time, we will be conducting a question and answer session. If you would like to ask a question, please press star one on your telephone keypad. A confirmation tone will indicate your line is in the question queue. You may press star two if you would like to remove your question from the queue.
For participants using speaker equipment, it may be necessary to pick up your handset before pressing the star key. One moment please, as we poll for questions. Our first question comes to the line of Shyam Patil with Susquehanna Financial Group. Please proceed with your question.
Hey, guys. Congrats on the strong 2019, and welcome aboard, Blake. I had a couple of questions. As you guys mentioned, 1Q looks like it's off to a very strong start with 40% growth, which is an acceleration from 4Q. Annual growth looks to be in the low 30s% range, at least initially. Can you just talk about this and just the puts and takes that you're taking into consideration? I have a quick follow-up.
Sure. First of all, I can't express loudly enough how excited I am for the year, and also how excited I am about Q1. I've never been in this position where I'm this confident going into a new year, and especially have the momentum that we do from Q1 with 40% growth year-over-year. That's just a surprise. We knew it would be strong, but we didn't know it would be that strong. Of course, connected TV is a huge driver of that. We're also seeing more than expected from political. While in 2016 presidential election year, it was low single digits. It's low single digits again, but they're higher than last time. Of course, on the other side of Super Tuesday, certain candidates might spend more or less as a result of that, and it's really hard to predict what that'll do.
Of course, we're also watching the virus around the world, the coronavirus, and trying to measure the impact that that will have on everything globally. While we don't think it will have a significant impact, in the current environment, it's prudent to be measured. We're more comfortable moving expectations up as we go, but our optimism is as strong as it could be.
Awesome. Then I just had a quick follow-up. The initial EBITDA margin guide for the year of about 30%, I think that's the highest initial guide that you guys have given in your history. I assume some of this is just natural operating leverage and some of it's efficiency that's identified by Blake and the team, but I was just wondering if you could talk about this a bit more.
Yeah. I'll give my two cents, and I'd love Blake to give his sort of fresh opinion. I feel like I've been apologizing to investors since we've been a public company where we say we know that at maturity we will be a very profitable business, our hope is to reinvest as much as possible for now. We still believe that this market is in land grab phase. As I have always said, global advertising will soon be a trillion-dollar industry, we want to capture as much of that as possible, that means we need to grow as fast as possible. We will not do that irresponsibly, we're going to continue to look for ways to invest as aggressively as possible.
After multiple years of investing as aggressively as we can and never wanting to hire the wrong people or hire at a rate that will negatively impact our culture, we've consistently outperformed on EBITDA, almost apologetically. This was our first time of looking at it and saying, even though we're investing as aggressively as we can, I think it's highly likely that we'll produce something close to 30% EBITDA margins yet again. That's how we get to this place, and I'm really excited for what it represents. Blake, I'd love your fresh opinion.
Thanks, Shyam. You're correct that we have guided in prior years to somewhere between 27% and 29%. Like Jeff said, we feel with our scale, we can deliver 30% in 2020, but still investing in growth areas as quickly as we're able to. The only additional real point I'd add is remember that we don't manage to an EBITDA target. If we see great opportunities to invest that can spin our flywheel, we're going to take that option. Still looking at the investment opportunities in front of us, we feel like 30% is realistic.
Cool. Thanks, Shyam. Next question, Devin.
Our next question comes to the line of Michael Levine with Pivotal Research. Please proceed with your question.
Great quarter, guys. Thanks for the question. Wanted to talk a little bit more about Chrome, the recent update that they made on February 4th. It's probably the biggest question that we get from investors is just skepticism regarding how you guys can navigate through it. I mean, it was great what you gave us in the prepared remarks, Jeff. Maybe you could help clarify and maybe give a little high level about cookies because, obviously it seems like the guide looks particularly strong.
Yeah. While I said it in the guide, there's certain pieces of uncertainty out there in the market. Admittedly, this is not one of them that we're including in any conservatism. We have pretty strong conviction that the value exchange of the internet is not changing at all, and that targeting is not going away. Google has done a good job of starting a good debate. They haven't necessarily given clarity in their most recent blog post, and let's just remember that that is all that it was that sort of started the recent discussion. It's simply a blog post where they said they expect to make changes in two years. There's been lots of supposition about what that means and what the impact will be.
Let me start by just talking about maybe the worst plausible scenario, and then when we see that that one's not that bad, we can then walk back and talk about where I think is much more likely to land. Some are afraid that Google will do exactly what Safari did, which is they'll eliminate third-party cookies altogether. What Google had announced only four-ish months earlier, or maybe it was eight, or earlier in the year, they had announced that they were going to handle third-party cookies differently, and we gave them lots of public praise for threading the needle between privacy and relevance. If they were to take a Safari-like approach, which just remember that Apple is not directly dependent on advertising really at all, and Google's entire business is built on top of advertising.
It seems unlikely, but let's say that they go there. What I would do is point you to the part of our business that is most dependent on cookies, which is our display business. While not our fastest-growing channel, if you look at that closely, and we have in just anticipating what would happen here, Safari spend in Q4 grew faster for display than the rest of it. The reason why I think that's really important is because it's indicative of the fact that Apple's policy changes and getting rid of third-party cookies did not stop us from growing. In fact, we're growing faster there. I'll also just reinforce that display is ballpark 20% of our business today. There's not that much of our business that is directly dependent on cookies at this point.
Even in a situation where Google goes to a policy like Apple, I don't think that we're negatively impacted much at all, if at all. What I think is far more likely than the scenario that I just described is that Google will find a way to do a better job of threading the needle between privacy and relevance than what Apple did. I suspect they'll end up landing at something closer to the policy that they talked about just months earlier. We have had great discussions with people at Google who are front and center in this discussion. The discussion inside of Google alone is far from over. It will be really interesting to see where they go. In all plausible scenarios, we are as well off as we are today or better, the way that we calculate it.
Thanks, Michael. Next question, Devin?
Our next question comes from the line of Tim Nollen with Macquarie Group. Please proceed with your question.
Oh, thanks very much. I'd like to ask about a couple more things on CTV, if I could, Jeff. You've been ebullient about this for a long time now, and the numbers are all coming through very, very strongly. I'm curious if you could give us a bit more color on where you are buying CTV ads. I'm thinking, is it network apps like an ESPN app? Is it a virtual bundle like a Sling TV? Is it an AVOD platform like The Roku Channel or an Amazon Fire TV? Is it something open ubiquitous like Pluto TV or something like that? I'm sure it's all of that, but anything you could sort of give us a sort of layout, where the activity is. Then you've got Peacock coming this year from NBCU, which is probably going to be a pretty big ad-supported service.
What are your thoughts on where that gets you to? If I could ask a follow-on regarding FreeWheel, your comments on FreeWheel and the header bidding. Is that entirely within the NBCU/Comcast ecosystem, given that they own FreeWheel? Or is this FreeWheel operating on behalf of the industry as a whole? Thanks a lot.
Okay, great. If I skip any part of your question, call me out on it, because I definitely want to address all three of those questions. First, at a high level on connected TV, just want to remind everybody that connected TV spend increased in 2019 by 137%. Shockingly, that is our lowest growth year to date at 137%. We're in the middle of what I think is a generational shift from traditional television to internet-driven television. While many people look at it and say it's going to be a linear decline, no pun intended, I actually think that we're going to hit a cliff at some point where linear television is not able to do what it does without a certain amount of economics, and we're not that far away from it.
I've always said that TV is something of a ticking time bomb, and I think we're getting closer to that. We're much, much closer to that, especially as we look at our scale and see 80 million households available through our platform just in the U.S. Of course, we expect to double again in 2020. With all of that as backdrop, let me dive in a little bit to your first question more specifically, which is: Where does the inventory come from? First of all, it's coming from the large networks, first and foremost. Those are the content owners themselves. Whether that's an NBC or CBS or ABC or Disney or Discovery or Fox and ESPN, and I could just keep going down the list. The large networks, working with them directly is the biggest and perhaps the fastest-growing, if memory serves.
Secondly, the content on all different connected devices, things like Roku and, of course, our Amazon partnership. We're also seeing inventory from channels that don't exist in traditional television, so Hulu and Sony Crackle of the world, for instance. Last, just new MVPDs. That's where Pluto TV of the world and Sling TV of the world come in. All of those together are fueling that massive growth, and just really exciting to be a part of it, especially when the content owners themselves have changed so drastically in the last couple of years. That's the first question. As it relates to Peacock, in phase I, Peacock is not going to be ad-funded. Which, by the way, if I were launching it, if I had the job of deciding how to launch that, I would do the same thing.
Phase I, I would try to win over consumers. I would likely add as much quality content as possible. I think Disney+ used this exact same playbook. Then after you gain the loyalty of customers, I would give them the power of choice so that they could choose whether they wanted to pay for their content by seeing ads primarily or by paying additional money. I suspect that that's where Peacock and NBC is going, so that represents, I think, a huge opportunity for us in the future, even though when Peacock first launches, it won't be ad-funded. Lastly, as it relates to your FreeWheel question, the Unified Yield is, I think, a very important move for the industry and impact on the industry, that is not just isolated to Comcast inventory. It affects all of the inventory that FreeWheel powers.
FreeWheel is a significant ad server and yield management platform for content way beyond just Comcast, and they've done a good job of maintaining business outside of Comcast, even though they were acquired several years ago. I'm really excited about what they're doing, and that does represent yield across everything.
Our next question comes from the line of Nick Jones with Citi. Please proceed with your question.
Hi. Thanks for taking my questions. First, you're talking a lot about walled gardens inevitably coming down, and you're pretty vocal about it. If Google kind of hears this and maybe sees it coming as well, what do you think the push and pull is they're considering in any column changes? Does it potentially benefit them to make it harder to attribute or target to further entrench them in their walled garden to try to postpone the inevitable? I'll have one quick follow-up after.
Yeah. I think in broad strokes, and this is all conjecture, but it's not by accident that in the most recent blog post, which has created all the discussion, that they put a timeline of two years. That's on the other side of an election. That gives them a lot of time to figure things out. You're seeing way more consideration and conscientiousness about both privacy and antitrust inside of Google. I think that they're looking at both of those things and trying to make certain that they don't step in it. That makes it really hard and just having spent a lot of time at Microsoft on the other side of antitrust, you just get reluctant to make choices and you try to be deliberate and you take longer. That's what I see happening there.
If they were to get rid of targeted advertising for everyone else, then I think they have an antitrust problem. If they were to not do anything, then I think they have a privacy problem. I think they're just trying to figure out what's the right thing to do with both of those, and in a very Google-like way, they're going to try to engineer their way into something that's better. I'm really optimistic about it. We're having more conversations with Google now than we ever have on this. I'm really optimistic about where we'll land.
Great. Thanks. Then Blake, just one quick one. Can you talk about the free cash flow results on the year?
Sure. You can see that in the release, the cash flow from operations. Let me pull it up real quick. About $60 million. Free cash flow is at about $19 million. The difference there is the CapEx. We've got PP&E and capitalized software of about $40 million. The vast majority of that's for the facilities and the build-out as we scale offices around the globe, so both in North America and international.
Nick, one thing I just want to clarify on one of my previous responses. I misspoke when I said that Peacock is not ad-funded. What I meant to say is that Peacock is not using programmatic ads today. It's ad-funded. Those are all just sold directly through NBC. As a result, we haven't announced anything between us and Peacock on this. For all the reasons I stated, I'm optimistic that at some point that'll happen.
Thanks, Nick.
Our next question comes from the line of Mark Mahaney with RBC. Please proceed with your question.
Okay, thanks. Let me throw off three quick ones. Anything on China? Is there any materiality in your 2020 outlook related to China? I know it's still early days for you, but you've been in there for over a year, so just comment on that, even though I know there's a lot of issues going on in that market now. Secondly, I want to call you out, Jeff, on your answer to the first question I didn't quite get. The question is your guidance for the first quarter has this very strong revenue growth for the full year, implies this real material deceleration as you go through the year, kind of implies that you exit 2020, kind of like mid or high twenties revenue growth. Maybe that's the law of large numbers. Maybe there's some tough comp issues.
Just talk through the, and maybe to you or Blake, but just talk through whether are there reasons for us to be thoughtful about expecting to see from a very high level in Q1, a market deceleration in revenue growth as we go through the year. Finally, sorry, Jeff, one high-level question. I thought about all these targeting changes that have occurred, not just the cookies, but really just a clampdown, I guess, on the commingling of third-party data. My interpretation of all this was that this gave a huge advantage to the walled gardens because of all the great first-party data. You're kind of making a different pitch. You're making the argument that there may be less data or more data, it doesn't really matter. It'll come down, the winners are going to be the companies that best work with the data.
I think that's your pitch. I just want to just ask you just maybe just take another turn at that. I think that the companies with large amounts of first-party data win, and then your pitch is that Trade Desk still wins because in any data environment, restricted or expansive, the best data, not manipulators, but the best data integrators, those are the ones that win. Am I stating that your position right on that? Thanks a lot, Jeff.
Sure. I'll answer one and three, and then I'll give Blake first stab at number two, and I'll come back and just add color on the guide. In terms of the materiality of China, we haven't forecasted anything material in our forecast for the year from China. It's one of those places where we could get surprised to the upside. We've done so much groundwork laying there that I'm very hopeful of what can happen there. Of course, there's lots of unforeseen things that they're managing right now, namely around the virus. Of course, a significant amount of our operations are headquartered in Hong Kong, and it had its fair share of political challenges last year. Despite all of that, I'm incredibly optimistic about what it could contribute in 2020, even though we haven't put it into our plan.
The third thing that you asked on targeting changes. I think you mostly characterized that correctly, which is that I don't believe that there will be a huge advantage to the walled gardens, in large part because they don't have the most valuable asset in the ecosystem today, which is the data assets of the buyers. The people who are funding the Internet, the advertisers, they have their own data, and they're more and more, and this is why we spend so much time in the prepared remarks talking about the mindset of a CMO and the way they're thinking about their data. They are looking for someone that they trust, not a media company, someone that is trying to help them deploy their data in the most objective way possible.
We're just continuing to win more and more favor from advertisers as it relates to their most precious data asset, which is about their customers. Because we have that partnership with them, and because we've always been focused on the buy side, I just believe having that advantage, plus always doing the right thing as it relates to first and third-party data, even if there is an environment where there's less of it, I think all that will do is make the quality go up. I don't think data-driven decisions are going anywhere. I don't think that that advantage that you hypothesize is actually real, in large part because as they get bigger in their own data asset, people become more reluctant to share their data with them. With that, Blake, anything you want to say about the guide before I add?
Sure. Yeah. As we talked about on the call, overall we're very happy with the full year guidance. I would say that you can definitely see that Q1's gotten off to a great start for us. Political spending's a part of that. We have a lot of candidates involved right now, and it's going to be lumpy. We don't see any major headwinds today, but like everyone knows, we're going to have to watch how the macro environment unfolds and just the confidence that we'll raise as we go, we perform to the upside is how we're going to operate.
Yeah. The other thing I'll just say to add is, you'd said it must therefore be going from 40% to 20% something or low 20s. That's not how we see it going from 40% something to 30% something. That's largely because we think in the current environment, it's prudent to be measured, and we're more comfortable moving expectations up as we go.
Thanks, Mark. Next question, Devin.
Our next question comes from the line of Mark Kelley with Nomura. Please proceed with your question.
Hey, great. Thanks, guys. Not to keep harping on the guide, I just wanted to get a little bit more clarity on the mix of the gross spend outlook for 2020 and then also the revenue growth. I think this is the first time in, I think maybe three or four years where it looks like gross spend will grow faster than revenue. That implies that the take rate will come off a little bit. Well, help us think through that. When I think of CTV and data and some of the big drivers that you guys have talked about as being material this year, I think CTV is kind of a higher take rate. What's the right way to interpret just the mix of the gross spend outlook versus the revenue outlook? Thank you.
I love that you've asked this question, especially because I think if you understand how we approach this philosophically and strategically, then it can help you with modeling, because this is really a strategic question that you're asking. The strategic part of this is, what do you care most about? Making money or grabbing land? We've essentially said we would rather grab land than anything. How can we adjust what I call consumer surplus, what we're giving back to our customers and adding more value, that just translates to more client retention and then makes the flywheel spin faster. It is true, we're giving up a little bit of take rate to help the flywheel spin faster to get more spend, and that's why I'm so excited by the gross spend numbers that we're sharing, because that, to me, represents what land grab is.
That will have a trickle-down effect. This isn't something that we're sort of subject to, it's more that we're making a deliberate decision to give more back, to get more loyalty from our customers, just what consumer surplus is.
Thanks, Mark. Next question.
Our next question comes to the line of Tom White with DA Davidson. Please proceed with your question.
Great. Thanks for taking my question. Just on international, Jeff, you sounded optimistic about growth there, but I'm not sure I heard you sort of affirm what you said last quarter about accelerating international in 2020. Does kind of coronavirus maybe factor into your view there? Just a quick follow-up on the free cash flow question. If my arithmetic's right, I think free cash flow conversion of EBITDA was kind of 9%, 10%, which is lower than we tend to see at other sort of tech or SaaS-esque businesses. Can you maybe just explain why that is? Is it just sort of maybe the timing vagaries of cash collection that can happen kind of between publishers, agencies, and advertisers, or is there something else there? Thanks.
Great. I'll take the first question, and I'll have Blake take the second. As it relates to international, admittedly, international hasn't been growing as fast as we want it to and as fast as we expect it to in the future. I'm expecting them to have revenue acceleration at some point in the not-distant future. We have great momentum in many of the markets. We had record spend in Madrid and Paris, and the U.K. in Q4. We also had record spend in Seoul, Korea. As you point out, in some of our hubs, like in Singapore and in Hong Kong and Shanghai, we've had just viruses and political issues and whatnot that definitely create a little bit of a speed bump.
We also are scaling in those businesses where there's a little bit of a stairstep function that we expect to see in 2019 and 2020. The great news is that we've done this so many times now that we have 25 offices around the world that we're seeing the same patterns over and over again, where we invest in the team, and then we see that stairstep happen. That's exactly what is happening in Seoul. That's exactly what is happening in Paris and Madrid. We expect that to happen in all of our international markets at some point in the not-distant future. We're definitely still at a place where we're investing.
To follow up on your cash flow question. It's an accurate statement to say that when you look at our DSOs and DPOs for the quarter, you get less flow-through from EBITDA because DSOs are, as we said on the call, 118 in Q4, DPOs at 94. The gap widened a little bit. I would just caution folks that just one or two days can affect that ARAP balance quite a bit. Just keep in mind, closing that gap is top of mind for us. Strategically, we want to bring it down. We feel very good about the quality of our receivables, but the timing of this can swing things quite a bit.
Our next question comes to the line of Youssef Squali with SunTrust Robinson Humphrey. Please proceed with your question.
Great. Thank you very much. I have two. Starting with you, Jeff. The data business, I think you said, was up about 65% for the quarter. I think that compares with somewhere like 80%-90% in the prior several quarters. Just given the changes to the data collection you talked about, the cookie elimination, et cetera, has there been any material change either in data availability or in the quality of data that advertisers can access on the platform? Blake, on your 2020 guide, I think it implies a couple of hundred basis points contraction in margin year 2020 over 2019. I think you called out negative leverage in sales and marketing as one. Is that the primary one? If not, maybe just help us how to think through it. Thank you.
Let me take the first one. I think your number is right at 65%, which is ballpark 150% of the growth rate of our media spend. I am really excited by the fact that it is growing at a faster rate in data than it is media, because that just means people are making more data-driven choices in our platform than ever before, and data-driven is the trend. I am really optimistic about that. I would not read too much into the 65% Q4 versus the 80% in previous quarters. You will recall that we launched Next Wave the year before. As a result of getting adoption from Next Wave in Q3 and Q4, we actually created pretty tough comps for ourselves in Q4. The fact that that is at 65% after all the momentum that we had in years previous, I am really excited and optimistic about it.
I believe our data teams and our inventory teams were the all-stars of our 2019.
Related to your 2020 guide question, correct, for 2020, where we're looking to invest the most would be in sales and marketing and technology and development. Essentially those components that either get us closer to customers, grab share, or build products and services that customers want. Like I said earlier, G&A, you're going to find will decelerate in growth year-over-year the most of any of those buckets. Then we expect to continue to get leverage in platform ops.
Youssef, one other thing I just want to add on the data side that I should have made certain that I mentioned is in Q4, we tested actually giving the same way that we did with consumer surplus and all other aspects of our business and we're talking about in 2020, we tested giving a bunch of data away to see what impact that would have. Additionally, that 65% difference is largely fueled by us experimenting with giving data away.
Got it. Thank you.
We have one more question from the line of Brian Schwartz with Oppenheimer. Please proceed with your question.
Yes. Thank you for taking my question this afternoon. Jeff, I just wanted to follow up more granularly on the European business, because it did look like it rebounded strongly in the quarter, which I actually thought was an impressive result considering all the macro. You had Brexit uncertainty. You even had the U.K. election late in the quarter. Question I wanted to ask you, is it just a better environment than we're thinking about over in Europe? Or is the priority from the brands, the agencies for your business just sort of overtaking any uncertainty people may have about that region's spending environment? Thanks.
Yeah. In terms of what's happening in that region, we have amazing momentum in Europe. Of course, it's a little bit more agency-centric. Of course, we've also been sort of managing the sort of post GDPR world, which is, I think, a really interesting case study for the rest of the world to look at. The fact that we're seeing as much data-driven decisioning, and we haven't seen any material changes to our business as a result of that, we think is indicative of what's going to happen everywhere else in the world. Overall, we're really excited about what's happening in EMEA, and we're continuing to make investments. When Blake talked about we're just constantly looking for places to invest, and places where we are going to make big bets and have confidence that they'll pay off, investing in Europe is one of them.
We have reached the end of our question -and- answer session, as well as today's conference call. We thank you for your participation, and have a wonderful day. This does conclude.