Trade Desk Inc (TTD) 2021 Q2 法說會逐字稿

Good day, ladies and gentlemen, and welcome to The Trade Desk Second Quarter 2021 Earnings Conference Call. (Operator Instructions) It is now my pleasure to turn the floor over to your host, Chris Toth, Vice President of Investor Relations at The Trade Desk. Sir, the floor is yours.

Thank you, operator. Hello, and good afternoon to everyone. Welcome to The Trade Desk Second Quarter 2021 Earnings Conference Call. On the call today are Founder and CEO, Jeff Green; and Chief Financial Officer, 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 would like to remind you that except for historical information, some of the discussion and our responses in Q&A may contain forward-looking statements, which are dependent upon certain risks and uncertainties. In particular, our expectations around the impact of the COVID-19 pandemic on our business and results of operations are subject to change. Should any of these risks materialize or should our assumptions prove to be incorrect, actual financial results could differ materially from our projections or those implied by these forward-looking statements. 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 in our earnings press release. We believe that providing non-GAAP measures combined with our GAAP results provides a more meaningful representation of 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. I'm pleased to report that The Trade Desk had a very strong second quarter this year. Our revenue was up 101% from a year ago to $280 million, significantly surpassing our own expectations.

Our growth was across all channels and speaks to our position as the leading DSP for the open Internet. More of the world's top advertisers and their agencies signed up or expanded their use of our platform, which just continues to validate our business strategy. They are increasingly embracing the opportunities of the open Internet in contrast to the limitations of walled gardens.

Our performance this quarter and year-to-date is led by CTV and premium video. The move from broadcast and cable to digital on-demand content is happening all over the world. While each major media market and nation have different dynamics impacting adoption rates, every major market in the world is heading towards consumption of premium TV and movie content over the Internet. Because of our products, including our new platform, Solimar, our objectivity and market shifts, CTV as a percentage of our business continues to grow very rapidly and is, by far, our fastest-growing channel.

Heading into the pandemic, our CTV growth had been driven by our leading position in the U.S. and Australia. And we continue to enjoy outsized growth in these markets. But now we're starting to see our CTV strategy scale more broadly around the world. For example, our CTV revenue in Europe was up more than tenfold in the second quarter. I'll expand on this in a moment, but I could not be more optimistic about our CTV business.

Overall, we fired on all cylinders in the second quarter, in large part because we realized the value of the investments we have made in our business over the last few years. Just as important, these investments leave us very strongly positioned for growth moving forward. And of course, we continue to invest. Our latest platform launch, Solimar, is the result of more than 2 years of engineering work, and it addresses many of the opportunities in front of agencies and brands today. I'll touch on this in a minute, too.

In order to provide some more color on these results and our optimism for the future, I'd like to focus on 3 key areas. First is our strength in CTV. Even as our overall business doubled over the second quarter last year, our CTV business significantly outpaced that growth, and I'd like to spend a moment on the various factors driving our progress there.

Second, I want to touch on how major advertisers are thinking about the value of the open Internet in contrast to the limitations of walled gardens, especially in terms of how they think about identity, first-party data and performance measurement.

And third, I'd like to focus on international growth. Like last quarter, our international growth outpaced North America, and we are seeing some trend lines that give me great optimism for the years ahead.

So first, CTV. Just to provide some context on our growth in CTV, through just the first half of this year, the number of brands spending more than $1 million in CTV on our platform has already more than doubled year-over-year. And it's not just larger advertisers that are taking advantage of CTV anymore. The number of advertisers spending over $100,000 has also doubled. In total, we have nearly 10,000 CTV advertisers on our platform, up over 50% compared to last year. Large and medium-sized advertisers alike are turning to us as the objective DSP for all digital media, but especially CTV and premium video.

That exponential growth speaks to how rapidly the TV landscape is evolving. We've spoken before about the accelerated consumer shift to digital video, including CTV. And that shows no signs of slowing down. In fact, we reached more households via CTV in the U.S. today than are reachable through linear TV. Today, we reach more than 87 million households. Those trends are now well established.

What is perhaps a little less appreciated is what's happening on the inventory side of TV and how advertiser demand for that inventory is also fueling a shift to CTV. In Q1 and Q2 of 2020, nearly every major advertiser had to pause or rethink their advertising campaigns due to the global pandemic. Some adjusted in weeks, some took months, some are still adjusting. Some companies grew faster because of the pandemic. And of course, some companies are still below their growth and revenue levels prepandemic. Many companies such as CPGs and pharma companies have enjoyed significant growth in that 1 year. They were able to adapt their businesses, pivot their message and appeal to consumers as their lives were upended and changed.

Others such as those in the hotel, cruise and airline industries have been largely treading water because of the various new restrictions we've all been living with. But regardless of where our company is on the growth spectrum, we are seeing the same response today, those companies that enjoyed accelerated growth now need to market effectively to sustain that growth. Those that were struggling and hit the pause button are now playing catch-up, aggressively marketing to make up for lost time.

Advertising and marketing matter more than ever in the formula for business success. The demand for growth, regardless of where our company is on the recovery curve has major implications for advertising. Brands are looking to their CMOs to find value and advertising that can help fuel new growth. The only way to find advertising efficiency in this market is with objective data-driven technology.

And within that context, CTV offers some of the most effective advertising in the history of this space. The combination of moving pictures, sound and data creates effectiveness and value that are unprecedented. We have significant premium CTV inventory at scale via our platform and partnerships as CTV growth moves to AVOD instead of the SVOD models that powered early adoption in the category. Indeed, MoffettNathanson recently reported that the ad-supported video-on-demand market is growing from $4.4 billion in 2020 to about $18 billion as early as 2025. And every major ad-supported platform, whether it's Disney's Hulu, Peacock, Discovery Plus, ViacomCBS' Paramount and Pluto, FOX's [2B] or fuboTV and many others, all are reporting record viewership or ad spend figures, and we see the rapid growth in AVOD in our CTV spend every quarter.

The shift from legacy TV to Connected TV was especially apparent in this year's upfronts, which wrapped up in the second quarter. For the first time in the history of this annual process, every major broadcaster included programmatic packages. And there's a wide range of reasons for that, not least because CTV represents a greater percentage of their revenue than ever before. But perhaps most important, broadcasters recognize that the traditional upfront process is a mismatch. It doesn't work in a digital world where data and personalization are required to succeed. The legacy upfront process is really hard to run in an environment with lots of change and lots of uncertainty.

I believe that this year will mark a turning point in how the process is managed. In today's fragmented TV environment, linear audiences continue to erode, linear supply is shrinking and the prices are rising simply because of the scarcity. This year, broadcasters use that scarcity to their advantage and lock up commitments as the demand for growth intensified. But it is becoming increasingly difficult to predict who will watch, which show for which live sports event, which means linear viewership commitments are harder to make and stand by. Advertisers will then have a rethink and then seek out greater value in a data-driven spot market and a data-driven forward market for digital than the opacity and uncertainty of the legacy upfront market. They'll take budget back. They'll demand a new conversation next year and broadcasters are adapting by making CTV and programmatic a more significant part of the process.

That conversation will also look at how to evolve the traditional focus on high gross rating points, or GRPs. In the chase for high GRPs, many advertisers are finding that they are underreaching some percentage of their target audience and way overreaching others more than ever. The average may be in their target range, but the actual consumer experience at the edges is highly inefficient.

On the extreme overreach side, brands are actually paying to make consumers dislike them, showing the same ad over and over again to the same person. On the underreach side, showing the ad so rarely that it isn't noticed or even remembered. As a result, more advertisers are demanding that the data-driven agility of CTV become a larger component of the overall TV ad spend. That includes more measurement precision. GRPs become less important in a digital environment when you can have a direct interaction with the viewer and more precise measurement.

For example, we recently worked with Ford to help build a CTV campaign that directly targeted households that were in market for a new car. That's very different from the traditional mass-market linear TV campaign where there's a great deal of waste. It's precision marketing using custom audience tools that identify when a consumer is coming off a lease or buying a car for the very first time. With CTV, Ford can then reach that audience with ads that are relevant to the time and place that the ads are being consumed. This level of marketing precision is simply not possible on linear or broadcast television.

Just as important in this equation is the large and growing CTV footprint. When you combine premium quality data and inventory scale, CTV becomes a very compelling proposition for advertisers and their agencies. In that same work with Ford, we found that 48% of the households reached were incremental to anything addressable through linear. That means if you're limiting yourself to linear, you're missing out on almost half of your potential target market.

It was the same story with large pharma and a medical device company. When compared to parallel linear TV ad campaigns, CTV delivered a 51% incremental reach and a 4x improvement when analyzing cost per household reach. These are not isolated cases. We are seeing many brands shift TV budgets to the data-driven precision of CTV, and I expect this trend to accelerate through the next year's upfronts.

One major global food company is working with us to shift almost 1/4 of its TV budget to CTV by next year so that they can better manage targeting and frequency. A global entertainment brand is doubling down on CTV with us because it allows them to measure the full customer journey from exposure to purchase. And a major telco is enjoying measurable sales lift by shifting a significant portion of its TV budget to decision CTV on our platform.

The evolution of TV advertising is a great barometer for the advertising industry overall. More and more brands want to be able to apply data to optimize their CTV and premium video advertising. They want to be able to measure campaign performance across all channels. They are focused on the quality of the TV supply chain, and they want to ensure that their partners, including The Trade Desk, are delivering more value than they are extracting. We've proved this in the U.S., and we're seeing markets around the world evolve similarly. And I still believe we are merely at the very beginning of the CTV innovation cycle.

I can't put it any better than Jon Halvorson, Vice President of Consumer Experience at the Global Food Giant, Mondelez. Speaking last week with the adage, he said, and I quote, "I've been positively surprised by the advancements that The Trade Desk and other partners in CTV have made versus digital platforms. Their focus on advertisers' needs for advanced targeting, inventory management, and guarantees are setting them apart from YouTube and the rest of the marketplace."

One last point on CTV. As the TV ecosystem gets more crowded and competitive, our strategic focus on objectivity is more valuable than ever. Some platforms that claim to be open and then primarily push their own content will lose favor with other content owners as the market progresses. We have great partnerships with content owners because we bring them objective demand from our customers, and we do not compete with them. We deliberately don't own or favor content.

This also brings me to my second point because CTV is also highly indicative of how advertisers are increasingly embracing the open Internet, both as an alternative to walled gardens and as a part of their digital media plans. And I noticed this in various dimensions.

If you noticed some of the recent deals that we've signed and talked about publicly, some of the world's largest advertisers have signed new contracts with us in recent weeks and months, increasing and extending their commitment to our platform. We are just scratching the surface and demand for our platform is growing across industries around the world. And as these brands work with us, they are making a commitment to the open Internet.

This may be summed up by something that Arun Kumar, Chief Data Officer at IPG said at an ad week event with me a couple of weeks ago. He talked about how brands had relationships with consumers long before the major walled garden platforms came along. And how those brands are working harder than ever to preserve those direct relationships even as platforms try to disintermediate them. He believes there's so much collaborative innovation going on to support the open Internet precisely for this reason.

What he's also getting at there is a point that I think is often misunderstood about the motivations of brand advertisers. Most brands have decades-long relationships with our consumers. These relationships have been nurtured and curated carefully over a wide range of interactions, including long-term loyalty programs. In many cases, consumers volunteer preference information in return for some kind of value.

Given the effort that they put into this over many years, brands have no interest in jeopardizing the trust they have established with their customers. Brands want to provide a relevant and enjoyable advertising experience that respects their relationship with customers. And that experience is best managed and delivered across the open Internet where the advertiser has more insight and flexibility.

In addition to managing their existing customer relationships for the long term, brands also want to find the next generation of customers who share some of the same characteristics as their most loyal ones. And that, in a nutshell, is why their first-party data is so important. Brands know a lot about their most loyal customers, and they want to put some of that data to work to find others like them, or ad impressions that their prospects are most likely to interact with.

But unleashing the potential of that first-party data requires a few things. First, it must respect the privacy of the customer and preserve the trusted relationship between the advertiser and the consumer. Second, it has to be secure, not just from a data protection perspective, but also in terms of brands retaining control of their data. It must be easy. You have to have simple integrations and on-ramps. And then lastly, as they put their first-party data to work, brands have to get symmetric performance data back. They have to know whether it's working and how their prospects are reacting.

Each of these requirements is challenging in a walled garden, none more so than the asymmetrical data relationship. Any data a brand puts into a walled garden is often usable by that platform, too. As media owners, they will put that data to work, to their own benefit as well as the advertisers. And the advertiser won't get the same grain of data back on their campaign performance. They'll get a report card saying that the campaign was successful, which is a bit like grading your own homework. But the results they'd get won't be the same kind of information that the brand can use to continue to refine their campaigns. And that's why quarter after quarter, brands are gravitating to our platform and to the open Internet.

Walled gardens may provide easy access to scale. You can reach a lot of people very quickly but they don't provide the precision and decisioning that are becoming so vital to today's marketer, and they don't provide much clarity on what content the brand is showing up against, what content they are funding and supporting and how to refine their campaigns into the future.

On July 7, we launched a new version of our platform, Solimar. It is the biggest release in the history of the company and the reception has been fantastic. Again, it beat our own hopes and expectations and has great traction. At the current pace of adoption by the beginning of next year, we expect the majority of the impressions on our platform to be bought using Solimar. A few of you joined us at our launch event in New York and some of you joined the many thousands who watched online. I believe the remarkable interest in these events speaks to how rapidly our industry has evolved over the last 18 months and how Solimar is helping solve for many of the issues and opportunities in front of marketers today.

Perhaps most important, with Solimar, we believe we have created the industry's most advanced measurement marketplace. Not only can advertisers measure against traditional campaign performance metrics, but they can now integrate off-site measurement performance in a way that's only possible on the open Internet. That means they can finally reach that holy grail of connecting their ad spend to actual business goals, whether it's in-store sales or foot traffic into a dealership or demand for tickets. Solimar has an ever-growing roster of third-party measurement data sources, including retailers who are eager to leverage the value of their shopper data so that they can attract more advertising demand.

But you don't get to build that kind of measurement marketplace without simple and secure data onboarding and without the ability of brands and partners to protect their data to make sure it's only being used for the intended purposes. And you don't get to real performance measurement without the ability to enter more precise campaign goals, and Solimar provides for all of that and much more.

We launched Solimar at a time when UID2 is also reaching critical scale in the market. This is important because UID2 allows advertisers and partners to onboard their data in a manner that provides more consumer control and protect their data. In recent months, 3 of the major advertising holding companies: Publicis, Omnicom and IPG have announced their support for UID2. In addition, many of the major independent agencies, such as Horizon are also now leveraging UID2. As well as the world's major agencies, many of the world's leading brands are also starting to use it. We're also working with many of the world's leading tech platforms as they look at use cases for UID2.

The publishing side is also embracing UID2 from a CTV perspective, major CTV industry, tech consortiums, which are owned by the major networks, including Blockgraph and OpenAP are making their identifiers interoperable with UID2. That's in addition to networks such as AMC and fuboTV who are integrating directly and then traditional publishing groups such as Maven, who owns Sports Illustrated and thestreet.com and Newsweek have also embraced UID2.

A couple of weeks ago, Snowflake announced that it would deploy UID2. This one is a little different but very indicative of the scale that UID2 is achieving. Snowflake provides a cloud-based data service that sits on a company's cloud infrastructure and enables data to be managed across clouds. Snowflake acts as a hub for many companies' customer data, their first-party data, and that data can now be activated by using UID2 identifiers.

What's really interesting about all this momentum around UID2 is that it has accelerated since Google announced that it would delay the deprecation of cookies by at least 2 years. You may remember that when Google first announced their intentions, I was somewhat skeptical and that's because the fundamental value exchange of the Internet, free content in exchange for relevant advertising is not going to change.

What will change is how we give consumers more information about that value exchange and how we provide better tools to pay off that exchange in a way that improves the experience for advertisers, publishers and consumers and gives consumers more control. That's what the industry is creating. UID2 is one leading example but more important is the way that the industry is mobilizing to create a better approach to identity. One that reflects the fast-moving cross-channel nature of today's digital advertising landscape.

I cannot put this urgency any better than Joy Robins, Chief Revenue Officer of the Washington Post organization, who spoke at our recent Solimar event in New York City. The Post, as you may know, through its Zeus platform also powers the ad tech stack for more than 100 other publications across the U.S., including many daily newspapers. So Joy's perspective is very insightful. To quote her directly, she said, "We need to make sure we're controlling our destiny. If we do nothing, we give our ad revenue stream to the walled gardens." It's that sentiment across the industry and around the world that's driving so much collaborative innovation in support of the open Internet.

The last area that I want to touch on is our international growth. Once again, our international markets grew faster than the U.S. in the second quarter, and this is particularly encouraging. As this industry speeds towards $1 trillion TAM, about 2/3 of that will be outside the United States, and we are investing to capitalize on that international growth and serve a global advertiser customer base.

Let me give just a couple of examples that put our investments and our growth into perspective. We started planting the seeds of Connected TV in Europe a few years ago, and we are now seeing the green shoots. CTV in Europe is still relatively early in its life cycle compared to the United States. But with the exponential revenue growth I mentioned earlier, it won't stay small for very long.

In Europe, there is a significant consumer shift to streaming platforms even for live sports. European broadcasters have developed their own streaming platforms, which is driving the inventory scale that is so important to advertisers. For example, we're working with Sky, the largest media company in Europe as they make their own content available over the Internet. In fact, our partnership has significantly expanded this year. Sky has a huge presence in the U.K., but they also enjoy strong market share across Europe.

Like so many broadcasters today, Sky is also investing heavily in original content to attract new viewers. They are a dominant force in the European TV landscape and hold significant clout with brands and agencies there. Along with our existing partnership with Channel 4 as well as premium content providers in France, Germany, Spain and Italy, our relationship with Sky gives us access to the majority of CTV ad impressions across the continent.

I want to again underline the significance of our objectivity in our inventory partnerships as we expand around the world. Because we don't own content, we are able to cleanly and clearly partner with the biggest content and broadcast companies around the world.

Let me also spend a moment on APAC and one new market in particular, India. We've only recently opened an office in India earlier this year, but we have already made some incredible progress. It's worth reiterating how large this potential market is. According to research by Global Web Index, Indians are spending an average of 8 hours a day online, most of it on the open Internet, led by OTT content. These dynamics are fueling the rapid expansion of the digital advertising market in India, expected to exceed $7 billion by 2024, up more than tenfold since 2015.

It's those same market dynamics that informed our premium video and CTV-first approach to India, and this was somewhat atypical for us. In nearly every other market we have led with display. So as we opened in India, the first thing we did was strike important inventory partnerships in CTV. These included a deal with Samsung ads, which gives us access to inventory on Samsung Smart TV devices, reaching 50 million highly coveted viewers. We also struck a partnership with Xiaomi, the world's largest smartphone manufacturer, bigger than even Apple. Xiaomi serves more than 10 billion ad impressions per day to those devices, all of which we have access to.

In addition to device manufacturers, we've also established relationships with the leading content providers such as Disney+ Hotstar, the leading OTT streaming service in India. And these partnerships are already yielding results. Brands such as GSK are working with us in India to drive more data-driven precision in their ad campaigns, particularly with OTT content and seen significant performance improvements.

Our initial progress in India is starting from a small base, but it has been very rapid. And over time, we will build on these relationships and scale our business methodically just as we are doing across Asia and, of course, the rest of the world. All of this progress gives us a great deal of optimism for the quarters and years ahead, and this has just been a snapshot.

In every market where we operate, North America, EMEA and Asia and Australia, we are seeing significant growth. That's because we continue to innovate more quickly and efficiently than others in our industry whether it's the future of television, new approaches to identity or trading platforms that bring advertisers closer to the business results of their work, we are able to invest in the success of our advertising clients.

We are not maximizing profit for the short term. We have discipline around profitability so that we can invest for the future of our business and theirs. And this quarter, more than ever, we can see how our prior investments can yield impressive results. And even as we make these investments, we are still generating EBITDA at rates much higher than nearly all of our high-growth software peers.

In the second half of the year, we expect this approach to show more green shoots as we expand our work in retail. Walmart will launch a new DSP, which integrates Walmart shopper data and is built on our platform. This is a leading example of how we are working with our advertising customers to help unlock the value of retail data estimated at $100 billion to $200 billion market. Each retailer will approach this differently, but we are working with many of them, both here and around the world. But one position they do all share, they are all convinced that the value of their data is best realized on the open Internet, not within the confines of the walled gardens.

Our progress is also increasingly rooted in our ability to drive industry consensus around important issues that build overall trust in our industry. Our strategy will raise all boats. You are seeing that most clearly in the work we are doing around identity, but it's also present in how we approach the supply chain, fraud management, CTV scale and many other areas. We remain convinced that the open Internet is the best platform for our customers to achieve their marketing goals. And there's a large and growing coalition of advertisers, publishers and partners who, not only share that perspective, but who are actively working with us to realize it. All of that contributed to a great second quarter. We have very strong momentum. And with these investments and the hard work of The Trade Desk employees around the world, I expect the wind to stay at our backs. And for that reason, as pleasing as the numbers are for the quarter, I am even more excited about the future growth prospects.

Now I'd like to turn the call over to Blake before moving to Q&A. Blake?

Thank you, Jeff, and good afternoon, everyone. As you have seen in our results, Q2 was a very strong quarter. Revenue of $280 million was up 101% from a year ago. Excluding political spend related to the U.S. elections last year, which represented a low single-digit percentage share of our business in Q2 of 2020, revenue increased around 103% year-over-year.

During the quarter, we benefited from continued improvement in the digital advertising environment from both agencies and brands. Growth was broad-based across all regions, channels and verticals. We saw continued strength in CTV, which again led our growth from a channel perspective. Our year-over-year revenue growth rates benefited from lapping slower growth related to the pandemic during the second quarter of 2020.

With the continued strong top-line performance in Q2, we generated $118 million in adjusted EBITDA or about 42% of revenue. EBITDA continues to benefit from temporarily lower-than-expected operating expenses, partly driven by the virtual environment. This includes items such as travel and live company events that have only just started to slowly resume.

I'm proud of our sustained efforts to consistently generate meaningfully positive EBITDA while continuing to invest in the critical areas of our business that can drive our future growth. From a channel perspective, video, audio and even display all more than doubled in Q2 from a year ago.

Exiting Q2, mobile, currently our largest channel, represented a low 40s percentage share of our business. Video, which includes CTV, represented a high 30 percentage share of our business. Video as a percent of our mix continued to grow very rapidly. The increase in video was driven by CTV, which, by a wide margin, again led our growth during the quarter. And finally, display and audio represented about 15% and 5% of our business, respectively.

Geographically, North America represented 87% of spend, and international represented 13% as international's faster growth year-over-year resulted in a rising share of spend. In APAC, Shanghai and Hong Kong spend growth were both very strong. In terms of our overall growth, Europe led the way, growing well over 100% year-over-year in Q2. All of our offices in Europe: London, Hamburg, Paris and Madrid showed particularly strong growth. As Jeff highlighted, CTV drove our EMEA performance growing over tenfold year-over-year. CTV more than doubled its relative share of spend in Europe. While still small relative to the share of CTV spend produced in North America, we are optimistic about the trends we are seeing during the first half of 2021.

In terms of the verticals that represent at least 1% of our spend, the majority of them at least doubled during the quarter. Those verticals most impacted by COVID showed the most improvement overall, including travel, shopping and automotive. Home & Garden, Personal Finance and Food & Drink were also very strong. We believe that there is still a lot of recovery ahead of us in these segments, and we remain cautiously optimistic as we continue to see signs of improvement.

Operating expenses were $218 million in Q2, up 41% year-over-year. The growth in operating expenses in the quarter was primarily driven by stock-based compensation. Operating expenses, excluding stock-based compensation, grew 32% year-over-year. As we discussed last quarter, the majority of the growth in stock-based compensation expense in the first half of the year was related to the company's employee stock purchase plan. Our growth in operating expenses, excluding stock-based compensation on a year-over-year basis is influenced, not only by lower expense growth in the prior year associated with impacts related to COVID, but also by continued investments in our team, particularly in areas like sales and marketing, technology and development and the team supporting that progress as we continue to scale for longer-term growth.

Income tax was $13.9 million for the quarter, representing a tax rate of about 23%. Adjusted net income for the quarter was $88 million or $0.18 per fully diluted share. Net cash provided by operating activities was $10.4 million in Q2. This was driven by a change in working capital driven by strong sequential growth from Q1 to Q2. I would like to remind you that the timing of cash collections and payments can significantly impact quarterly results.

DSOs exiting in Q2 were 82 days, down 14 days from a year ago. DPOs were 67 days, down 8 days from a year ago. The resulting 15-day gap between DSOs and DPOs is the smallest in the company's history. We exited Q2 with a strong cash and liquidity position. Our balance sheet had $705 million in cash, cash equivalents and short-term investments at the end of the quarter. We have no debt on the balance sheet.

In June, we refinanced a new revolving credit facility and currently have $443 million available under the facility. In addition, we also executed a 10-for-1 stock split.

Turning to our outlook for the third quarter. We estimate Q3 revenue to be at least $282 million, which will represent growth of 30.5% on a year-over-year basis. Excluding U.S. political election spend, which represented a mid-single-digit percent of spend that we benefited from in Q3 2020. Our estimated growth rate in Q3 of this year would be about 38% on a year-over-year basis. We estimate adjusted EBITDA to be approximately $100 million in Q3.

In closing, we're extremely pleased with our strong performance in the quarter. We continue to execute and build on our solid foundation, and I could not be more excited about building on our progress in the second half of the year.

That concludes our prepared remarks. And with that, operator, let's open up the call for questions.

(Operator Instructions) And the first question is coming from Tim Nollen from Macquarie.

Jeff, I'd like to talk about CTV a bit more, if I could. You mentioned your growth in Q2. I think you said significantly outpaced your overall growth. I wonder if you could comment about how that difference looks in the second half and even into next year, CTV versus overall Trade Desk growth.

And also, Q2, you talked about the upfront. Q2 was the quarter that the upfront market occurred. So I'm intrigued by your comments on an eventual decline in the role that the upfront will play for TV or at least certainly for linear TV. My question is, what did The Trade Desk do with advertisers and network groups in the upfront? Or should we think really of your role more as helping expand the scatter market, reaching all those incremental households that are otherwise slipping away from linear? And I guess the scatter market having been so strong going into the upfront, I guess that's what must have supported your growth. So I just want to understand how that continues from here.

Thanks, Tim. So first, on the macro related to CTV. So on a year where nearly everything doubled year-over-year in this quarter-over-quarter basis, CTV, once again, led the way with massive growth in the quarter we just reported. We, of course, expect CTV to continue to drive our growth over the next couple of years and beyond as the leader as it relates to channels in terms of what's happening across everything in digital. CTV is the lead for the open Internet.

I do believe upfronts are driving more the CTV as well. I'll talk about that as I answer your second question in just a second. But one thing I want to highlight about CTV is that unlike any other channel, this is being driven by users. So when programmatic first got started, it was because of a bunch of B2B companies like ourselves wanted to create something more efficient for the Internet.

That's very different than the trend that's happening in CTV, which is users are moving from their cable subscriptions to the Internet because it's better. It's on-demand content is just a better way to consume. They can watch what they want, when they want and that's something that can't be stopped. That's a secular tailwind that is going to benefit us for as far as I can see into the future.

And then lastly, on CTV trends before I talk about the upfront. There is a lot of subscription fatigue, especially in SVOD. So what has happened over the last year or 2, I think it's fairly indicative of what most of us have experienced as individuals, which is we consumed most of our CTV content from Netflix or Amazon 3 or 4 years ago. And now today, we have lots more choices, lots more channels that we're going to, and we're asking questions like, is it on this channel or is it on that channel, a lot more often than we did before when we're logging into our CTV.

And most of those incremental channels that we're adding are AVOD. So there's more and more inventory coming online. So one of the things that we're constantly saying to advertisers is whatever you thought you knew about scale and reach of CTV 6 months ago, well, it's changed dramatically. And that's underlined by the fact that we said 79 million households are reached on CTV, which is more than linear inside of linear or cable television. So CTV has an edge that it didn't have a year ago. So I could go on and on, but I want to answer your question about upfronts.

So this upfront was really interesting in that the CPMs or the cost per unit went up because of scarcity. So despite the fact that you're reaching fewer people, the price went up. And so we think that puts a lot of pressure on the upfront, and it's setting us up for some amount of under delivery.

Most of the forecasts that were given to advertisers, I would say advertisers don't believe. They go into it skeptical and are saying, "We need to be prepared to be flexible." And in fact, one of the large advertisers using our platform said, "we are looking to go to an annual calendar and looking to take as much commitment out of the commitment as possible," essentially maximizing flexibility.

I think that sort of summarizes what's happened. And so as a result, you're seeing people move into the scatter and spot market, and we benefit tremendously from that. So right now, you can -- there's the upfront market, which has historically been a get together or a party, which is a relatively inefficient forward market. And then there's a spot market, and that has more and more moved to be a programmatic-driven market.

I think there is some insight to be gleaned from the ratings that were reported on the Olympics, which are, of course, down 45% over past Olympics. And I don't think that's because NBC did anything wrong. It was just a question of discovery and just the nuance of the pandemic and whatnot. But that does make it so people are asking the question, "What can I expect out of shows in this environment?" And people are looking for more and more flexibility.

And that maybe just brings me to my last point, which is maybe at the very heart of your question, which is what role will we play in the upfronts going forward? One of the biggest opportunities that our company has in its future is to play a central role in a more sophisticated forward market that replaces the upfront, where you can put data to work and it's not an event or a party where everyone gets together, but it's an always on forward market that is data-driven and built on top of the spot market, where people can make commitments and forecasting is better than has ever been done in television because the data is available.

That's something that we're working very closely with some of the biggest names in TV and some of the biggest brands in the world and some of the biggest agencies in the world to make certain that we upgrade the upfronts. But that will make it so that programmatic is no longer primarily in the spot market, but it's also in the forward market.

And the next question is coming from Vasily Karasyov from Cannonball Research.

So Jeff, you had highlighted growth in Connected TV outside of the U.S. Can you talk a little more about the growth you're seeing in Europe, where I think you mentioned over tenfold rates of increases, and would you characterize this growth as an inflection point? Or do you think it's just the low base effect?

And as a follow-up, maybe you can talk about APAC a little more. Especially about the premium video and CTV growth that you highlighted. Maybe you can explain to us in a little more detail the difference between premium video, which I assume is mainly mobile and Connected TV over there.

Absolutely. Thank you. So maybe one of the most bullish numbers that we shared in this report is that CTV growth in EMEA grew 11x year-over-year. That number is really astounding and really surpassed our own expectations. I think one way to characterize what's happened for us in CTV is to say, The Trade Desk was leading in Australia and the United States going into the pandemic. And over the last 18 months or so, everything in the world has changed. Consumption has changed. We had 3, 4, 5 years' worth of change in television happened in that period of time.

And so when I say it's one of the most bullish things that we've reported in this report or in this quarterly report, it's because I do believe those represent green shoots. I do believe it's the start of a change. And where we were leading just in the U.S. and Australia, we're now seeing amazing results in Germany and the U.K. and in Scandinavia and France and Spain.

And we continue to maintain that all of it will eventually be traded programmatically or bought programmatically because that is the data-driven way that is more effective and it's the only way to preserve the amazing state of TV around the world, which is you get premium content, you see fewer and more relevant ads, and that creates a better user experience, which can fund more content. That virtuous cycle can only be perpetuated with a data-driven marketplace. And because that's just the math, that's just the way that it works, it's impossible for that not to take on everywhere around the world, and we're seeing that play out now.

As it relates to APAC. APAC is a little bit different in that CTV is not always the place where premium content starts. That's much more common in the U.S. But because it is such a mobile region where mobile phones are often in the place where content -- even premium content is consumed. You have to think of it as premium video and CTV kind of in 1 bundle, especially in that market.

But I think it's great to look at a case study. There are so many markets that we could report on. But because we just started in India this year, seeing us go into a market as big as ripe as India being the most populated country in the world and then see us create partnerships in, really, a matter of months with Samsung ads, Xiaomi, which has a huge presence in India and then Disney+ Hotstar, to have those 3 content -- partnerships for premium content and to show ads on top of that in the most populated market in the world after only being there for months is indicative of just the momentum as well as how we continue to grab land around the world. So really optimistic about what I'm seeing around the world and convince that our playbook that we've tested and proven in the U.S. and Australia is applicable to many other regions in the world.

And the next question is coming from Shyam Patil from SIG.

As had a couple of questions. The first one, Jeff, can you talk a little bit more about CTV and what you're seeing in Europe? It seems like it could be inflecting and maybe a few years behind the U.S. Would love your perspective on that. And then just a quick follow-up for Blake. I know you're not guiding beyond the third quarter. But when we look at 4Q, is there any color that you could offer just in terms of how to think about seasonality for revenue?

Yes. So I'll just reiterate what I just said about CTV in EMEA. We had 11x growth year-over-year, which is one of the most bullish metrics that we've shared in this report. And we still feel like we're in early stages, but the green shoots that we're seeing, I think, are indicative of more to come. And our partnerships in content around the world, whether that's in London or Germany or Spain or Italy or France, are all indicative of great things to come. Blake, on the second part?

Sure. And thanks for the question. I think, Shyam, I'll just talk a little bit about Q3 and then I'll try to address what I can on your question for the rest of the year in Q4. With Q3, there are so many positive trends and momentum in the Q3, and you've heard Jeff kind of highlight some of those already. CTV's leading the way. The fundamentals of the business are really strong.

The thing on Q3 is excluding political, which I would encourage people to think about as far as how to evaluate the underlying strength of the business on a year-over-year basis. We're showing significant acceleration year-on-year in Q3. I think that -- we're talking about excluding election Q3 of around 38% year-on-year, and we were in the mid-20s for Q3 in the prior year, so really excited about that.

As you think about -- and then moving on into Q4, it does have more difficult comps than we see in Q3. You'll recall last year, and we talked about this a couple of times, political for us in Q4 of last year was the biggest quarter that we had as a company. It was a high single-digit percent of our business. So really meaningful on a year-over-year basis.

Also, if you recall last year, Q4 is driven not just by that high single-digit political spend, but we also had a rebound in digital advertising as well. But I would just say regardless of the comps, there's so much momentum for us and the fundamentals of this business are so strong. I'm really optimistic about it.

And the next question is coming from Youssef Squali from Truist Securities.

Congrats on the solid performance again. Two questions, please. Jeff, maybe going back to Solimar. I know it's very early. I think you've only had it out for maybe about a month now. But just wondering if you can share any early feedback from clients. Any improvement in campaign performance so far relative to next wave, at least within those customers that have used it for a few weeks now? And any impact so far on overall spend? Again, I suspect not, but I just wanted to ask.

And then Blake, maybe the outperformance in the bottom line was particularly impressive. Can you maybe just walk us through the biggest contributors to bottom beat? And what's nonrepeatable as we go into second half of this year and early next year?

Yes. So I'm really glad you've asked this question, and I don't mind just being really open here, which is whenever you release a new -- essentially a new platform, there's always some amount of anxiety. You do as much research as possible. You test the market as much as possible. You listen as much as possible, and then you hope you got it right.

And so we were really excited to ship Solimar, which is the biggest release in the history of the company. We had more engineering hours and people on this release than anything we've ever shipped ever. And to do most of that work during the pandemic, where it's a little bit harder to do all those things like listen, I was anxious to see how it would be received; of course, optimistic, and we're convinced we did the right thing, but you always want to see it in the numbers.

And that's exactly what we've seen. We've seen amazing traction. We're nearly a month in and the response has been fantastic. It does represent a behavior change, though, for the users, so they have to have relearned something. But those that have taken the time to do it have all given us fantastic feedback that it is a better experience, which is exactly what we expected. And so because it is a better experience, that's why we're so optimistic that the majority of our ads will be bought on this new platform by the beginning of next year. And that represents more efficiency for the user.

I'm especially excited about this going into Q4. Because often what happens is Q4, as all of you know, is the biggest quarter of the year for us, but it's also where you prove yourself because much of the advertising spend is sort of determined by a calendar cycle, where in Q4, you spend the most, you perform the most. And then in Q1, you make a decision about where you're going to spend for the next year. And so if we have a really strong Q4, it sets up our Q1 really nicely for next year. And so I'm optimistic that the performance will set us up for next year.

So I couldn't be happier with the way Solimar has gone so far. It's very early, as you highlighted, and it represents a behavioral change. But I'm convinced that, that's going to happen over time. I'm really excited about what that means for efficacy on our platform.

And then I'll just take the second part of that question on the beat on the bottom line for EBITDA. So more than half of the beat in the quarter on EBITDA was just really driven by the top line outperformance in the business. As many of you know the -- we don't have a lot of variable costs associated with a higher top line. So when we see it come in, it flows almost directly off and down to EBITDA, so you see that.

We did have some expense benefits as well. Some of that's timing but it was broad-based. There wasn't anything, like, that I would major call out, a little bit lower platform ops expense, a little bit of timing on some fixed marketing expense. We had better bad debt than expected a little bit because of just the continued health of receivables, but nothing major that I would call out.

And the next question is coming from Laura Martin from Needham.

So Jeff, you were talking about how the CTV open Internet consortium was integrating into 2.0. And then you said fubo was integrating directly into 2.0. I was wondering how that affects your ability to target whether people are going indirect like in affiliating versus like fubo directly integrating into 2.0.

And then my second one is on the Walmart deal. Are we still on track to be fully integrated with Walmart by 4Q? And as your -- do you get paid on that because you get the 15% platform fee? Is that how you make money on the Walmart deal when CPG spends money on your platform?

You bet. I appreciate the question. So as many may know, one of the things that is really important about preserving the amazing state of CTV is to provide relevant ads on the shows that you're watching when it's ad-funded. And because of the nature of devices in CTV, it requires a bit more collaboration between the content owners as well as those providing ads.

And so what we've needed, and many people talk about this in the context of cookies and browsing, but that's not really what UID was designed for exclusively. It was designed to create a better currency for the entire Internet, especially for Connected TV. So that a content owner can have the same understanding of a user, pass that to those of us representing the advertisers so that then we have a common understanding that can provide relevant advertising as well as make certain that we don't show the same ad 5x in the same commercial break, make certain that we don't show them an ad that's irrelevant to them, and that makes them more effective.

So all the TV companies have different ways of integrating with Unified ID 2.0. Some of them have gone to coalitions like Blockgraph or OpenAP and said, "Hey, you're a consortium and you're a technology consortium. So why don't you do the work for all of us?" And there are others who have said, "Well, working in a consortium can take too long. And we want to make certain that it integrates directly and we have data that we want to put to work directly."

So from our standpoint, we're indifferent, whether a consortium does the work, whether the content owners themselves do the work. Many have put them on parallel paths, which is a commentary on how critically and strategically important it is for them and for us to get the IDs set up so that we can provide those relevant ads. But fubo has moved to do it directly. To me, it's exciting because they could access it in other ways, but they want to make certain that they do it as quickly as possible. We expect that trend to continue just because the future of television is dependent on it.

And then the second...

Just Walmart on...

Yes, yes, Walmart. So yes, Walmart is on track. We continue to have really fantastic discussions with them. As it relates to how we'll be paid, we're paid the same way that we would with any other partnership where we have our standard platform fees, and then they put their data exclusively to work in this version of their DSP, which is, of course, built on top of our platform.

And the next question is coming from Matt Swanson from RBC Capital Markets.

Yes. I've got 2 as well. They're both going to kind of be follow-ups to Laura's a second ago. Now I mean, great traction on UID 2.0 ads. You used the term critical scale. Could you just help us with a little more color on what you think of as far as what is critical scale in terms of -- if cookies went away tomorrow, what do you think of the efficacy of the replacement? And if there's any way to think about innings or percentages of how far are you away from kind of the ideal state, that would be really helpful.

The other -- on the Walmart example is just -- are there ways that you can use this as kind of a POC, proof of concept for other retailers? I'm sure with Amazon doing what they are with their DSP. It's becoming an increasing area of focus for other retailers who maybe don't have the internal tech platform that Amazon does. And it seems like that's a great place for Trade Desk to come in.

Yes. So the term critical scale, you pulled out, I think, a very important phrase that we used in our prepared remarks in the earnings report, which I do think is indicative of the moment that we're in right now as it relates to the way the Internet is going to work.

So -- and then you asked about if cookies were to go away today. So to be clear, Google announced that they don't expect cookies to go away before 2023. And they said -- and even then, we'll just -- we'll see when we get there.

So I've been maintaining from the very beginning that I'm not certain that it's in the best strategic interest for Google to get rid of cookies at all. But that's not what this was about. Creating UID was not about replacing cookies. It was about creating a better Internet. And that doesn't just apply to the browsing Internet, which as everyone knows, that's a small minority of what the Internet is. Of course, your mobile devices and the related apps as well as things like Connected TV, which is some of the most effective advertising, maybe the most effective advertising that's ever been done at scale, which is all dependent on different IDs.

But when you get to a place where the currency is so widely accepted, it becomes critical to doing business. So think of it like Visa or MasterCard, a currency that if you do not accept that currency in your local store, you are going to struggle to get business. You might be able to say, "Oh, I won't accept American Express," or, "I might not accept Discovery." But there are some currencies like either cash or Visa or Mastercard that you have to accept in order to just keep the doors open.

That's where I believe UID is at, that inflection point, where it's become so interoperable with other currencies, that's making certain that you're interoperable, if you're something new or if you're a TV company, you will be operating at a disadvantage if you are not interoperable with UID2. So when I say it's a critical scale, that's what I mean, that inflection point where you can't afford not to be a part of it.

As it relates to the -- to retail proof of concept. It's always great to partner first with the biggest retailer in the world. And so -- and naturally, that creates a bit of a case study, and it lays down the gauntlet for everybody else. That, of course, has been done.

And it would be strategically suboptimal, to say it nicely, for us to not be working on bringing other retailers and their data into the platform. Especially because of what I talked about in the prepared remarks, where we're trying to help the open Internet be an amazing contrast to walled gardens. And that means especially in measurement and the way that results are created and measured.

And by making everything comparable to each other, you make it so the open Internet is a much better place to spend the lion's share of your dollars than in walled gardens, where everything is opaque and you're totally dependent on them to tell you how you did. So in that environment, it becomes really important, especially for companies that sell their products in stores to have insights about how that performed. And they -- of course, they want to see how it performed at Walmart. But there's a bunch of other retailers or a bunch of other stores that they also want to see where it transacted.

And there's, of course, a bunch of other data about transactions that would help them discover what is actually working. And that's what's really critical for the open Internet is it's at a moment where we can actually show what's working and what's not, which is critical given it's so easy to make ads on the Internet. There are ads everywhere. So which ones are working is a question that has to be answered better than ever before because of what's riding on it, which is economic growth and recovery.

And the next question is coming from Brian Fitzgerald from Wells Fargo.

I want to go back to Solimar and you talked about some new capabilities specifically around onboarding and the measurement marketplace. Just wondering if you could discuss the revenue opportunities and the mechanics around those as well. If the measurement marketplace, is that going to be a take rate model and will onboarding have a revenue component? And if so, is that going to be more volume than spend-oriented? Just some more on the mechanics there.

Yes. So both of those, making it easier to onboard your first-party data, as well as making it easier to measure success are not about making more money for us directly. I really like the Amazon metaphor of spinning the flywheel faster and there are some activities that you do in your business to make it so that your core activity goes better and that the value exchange is even more obvious so that your customers spend more, do more with you.

So in both cases, onboarding first-party data and measurement marketplace, my idea was that they're both free. So there's not an incremental take rate for those exact features. But if you get more spend on the platform, you make more money. And if you produce better results, you get more spend and you make more money. And so I'm absolutely after making more on this but not on those features directly.

By giving them away, we make efficacy on the platform better. We make the open Internet better, and we get more of that spend. And that's the ideal. There will be cases in the measurement marketplace where people do pay, and that's good. It's kind of like paid apps on your iPhone or your Android versus free apps. Both are important and having some amount of paid apps so that they're really high quality and create the most competitive marketplace possible are important.

So some of them will be paid, but I think the default may be similar to -- on your phone, should be for free because that creates more usage and creates a better experience.

Thanks, Brian. And Paul, we have time for 1 more question.

Certainly. The final question will be coming from Justin Patterson from KeyBanc.

Great. Two, if I can. First, Jeff, could you expand on the Snowflake UID integration? It sounds like something that could broaden your reach from large advertisers in the mix. That's question one. And then question two, [as it relates to] Solimar. you talked about efforts to simplify the supply chain. Could you talk about what that entails and just how that creates new opportunities ahead?

You bet. So I really appreciate the question because I think Snowflake adopting UID2 is one of the biggest headlines that has happened for UID2 to date. And not enough has been said about it. I don't think most people understand why this is so big.

So first, let me just remind everyone that UID2 is no longer a Trade Desk product. We did most of the early developing, but it's owned by the community. We work together with the open Internet. It's open-source at this point. So it is way bigger than us. And some of the proof of that is adoption from companies like Snowflake.

So if you don't know Snowflake, what they do is they make it really easy for you to manage your data. So in the same way that Wix made it really easy for companies to build websites, snowflake makes it really easy for companies to put their data to work. Now they are typically much bigger companies than those that Wix works with. But that's part of what makes it so exciting is that if they're trying to put data to work, they, of course, need to be leveraging the currencies that have scale. And so that makes it possible for that data to be more used and protected. So that they can do the right thing by consumers, whether they're on the content side or whether they're on the advertiser side.

So when a company who is so focused on making data actionable and listening to customers so that you can make certain that you're doing what they're asking for. Snowflake using this just underscores just how much critical mass UID to already have.

And then as it relates to Solimar and its impacts on the supply chain. So what we are spending a tremendous amount of time doing is trying to shed light on the supply chain. And what we're constantly asking is, are people in the supply chain adding more value than they extract? That's what we expect our own company to do, and that's what we expect everyone else in the supply chain to do.

And that's not for us to decide on our own but we do want to provide transparency so that the market can decide, especially because we represent so much demand. And so we are working night and day to make certain that we're shining the light of transparency on the supply chain, so that anyone in the supply chain that is extracting more value or charging more than the value they create, that they'd be removed from the ecosystem by the market, not by us, so that the supply chain becomes more efficient.

And the reason why this is so critical is because the open market has to compete with walled gardens and one great, in some ways, advantage, walled garden have is that they have a shorter supply chain because they control it all. So we just need the market to do its thing, and in order for it to do that, it needs transparency.

Thanks, Justin, and thanks for everyone for joining the call today. Paul, I'll leave it with you to close it out.

Thank you, ladies and gentlemen. This does conclude today's conference call. You may disconnect your phone lines at this time. Have a wonderful day. Thank you for your participation.