Trade Desk Inc (TTD) 2018 Q2 法說會逐字稿

Greetings, and welcome to The Trade Desk's Second Quarter 2018 Earnings Conference Call.

(Operator Instructions) As a reminder, this conference is being recorded.

I would now like to turn the conference over to your host, Chris Toth, Head of Investor Relations.

Thank you, operator.

Hello, and good afternoon.

Welcome to The Trade Desk Second Quarter 2018 Earnings Conference Call.

On the call today from our Hong Kong office is Founder and CEO, Jeff Green; and from our headquarters in Ventura, California, Chief Operating Officer, Rob Perdue; and Chief Financial Officer, Paul Ross.

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, 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 included in our press release and 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 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.

As Chris mentioned, I'm speaking to you from Hong Kong today.

I'm excited to be taking this call from Asia for the first time.

More on that later, but first, on to our results.

I'm pleased to report that Trade Desk had another outstanding quarter in Q2 of 2018.

Let me remind you that programmatic is growing at 21%, while our growth was nearly 2.5x that.

Our revenue was up 54% from a year ago to a record $112 million, which again surpassed even our own expectations.

That 54% year-over-year growth equaled the 54% year-over-year growth we had in Q2 2017.

Q2 also represented the largest increase in incremental revenue dollars we have ever had in a single quarter.

Our business strategy continues to be validated in the marketplace, this quarter more than ever.

We continue to see marketers allocate budgets beyond a few search and social sites that historically got the most incremental advertising dollars.

Our strategy of being the best platform for media-buying and not owning or arbitraging media is more valuable today than it ever was.

In the last 3 months, even more of the top 200 worldwide advertisers signed up on our platform.

In the last 12 months, Ad Age's top 50 worldwide advertisers increased their spend nearly 100% more with us this year than last.

That positions us very well for continued growth, not only for 2018 but in 2019 and beyond.

This quarter, growth came from many areas.

Mobile spend grew nearly 100% year-over-year to account for 45% of spend on our platform, the highest ever.

That is about 4x the industry average for mobile ad spend according to eMarketer.

Data spend again hit another record for the quarter, and spend on cross-device grew by nearly 100%.

Meanwhile, our international business continued its solid growth in both Europe and Asia, up 85% compared with the last year.

Arguably, the most important channel for our company is Connected TV.

Last quarter, we shared the most bullish number that I think we've ever shared as a public company.

Q1 2018 CTV spend increased by over 21x over Q1 2017.

This quarter, I'm excited to report that Connected TV more than doubled from last quarter.

In Q2, nearly everything went right.

We executed well in one of the most dynamic environments we've seen.

We often see a game-changing event or 2 impacting the industry in any given year.

However, in Q2, many significant developments occurred in the industry, all of which, we believe, will yield massive positive implications for our business over the rest of this year, next year and over the long term.

Here are 4 that I want to touch on.

First, Google sharply limited how their DoubleClick ID can be used.

Second, significant merger deals shifted the TV landscape.

And third, GDPR went live in the EU.

And finally, The Trade Desk launched the biggest product in our company's history.

I think it's important that I take a minute to discuss each of these events and how they impact the entire digital advertising marketplace.

First, Google announced in April that they would stop sharing DoubleClick IDs with clients to enable reporting across websites and properties.

This includes YouTube and DBM.

Keep in mind that Google's DoubleClick has about 75% of the global ad serving market share, and this means that they touch and measure about 75% of the ads served on the global independent Internet.

Removing the ID makes comparative reporting go away, so this is a very big deal.

Sharing the ID enables third-party reporting companies to measure Google's performance objectively.

The ID makes it possible for marketers to compare YouTube, Google and DBM performance to the other parts of their media plan.

Taking this away weakens the value proposition of YouTube, Google and DBM.

In my view, Google's decision to remove this ID offering is driven by their increasing need to reduce risk against malicious data enablement, like what we saw Cambridge Analytica do with social data.

The risk is similar for both Google and Facebook.

The risk exists because Google, at the fundamental level of their business, transacts in directly identifiable consumer data.

Google knows so much about billions of consumers because of their core product, their search engine.

Because The Trade Desk does not transact in directly identifiable consumer data and because we don't own a search engine, we can provide a Unified Open ID that enables advertisers to compare every destination on their media plan to every other destination objectively.

Agencies and brands see this, and it is why, in part, we are winning spend.

In our platform, we don't even house email addresses, let alone provide hundreds of millions of people with their own email addresses.

Our data and the data of our third-party partners cannot be directly associated with an individual.

Data in The Trade Desk platform does not include names, phone numbers or social security numbers, for example.

This is the centerpiece of many of the discussions we have with agencies and brands today.

Our value proposition has strengthened because of Google's strategic ID policy change.

The choice for marketers could not be clearer: choose an objective partner with transparent reporting; or choose higher walls where the publisher largely does the measurement.

We have already seen some of the benefit in Q2, and we think this can only lead to more positive outcomes for our business over the long term, which brings us to the second highest-impact development this quarter, the significant merger deals in TV.

To properly understand this, a little context is needed.

The worldwide advertising market is currently at $700 billion and moving towards $1 trillion over the next 10 years.

The biggest part of that market is television, which, according to IDC, is nearly $230 billion this year.

When that TV number is added to web video, social video, mobile video and CTV, video content is approaching half of the growing global advertising pie.

And TV has just started to move to digital.

We are witnessing a generational shift with the global convergence of the Internet and television.

This monumental change is making new forms of distribution possible, in which content owners can reach content consumers directly.

As a result, we are seeing decisive actions from TV and broadcast companies.

We see this as one of the drivers behind the AT&T-Time Warner deal or the Disney and Fox deal, which also includes the control of Hulu.

These massive deals have given us a clear indication of where the future of TV is going than we have ever seen in the past.

Netflix used to be the exception, and they have become the rule.

Content owners are showing more commitment than ever to having a direct relationship with consumers.

But as the shift occurs, content distribution is fragmenting.

In the past, distribution was through one cable operator on a TV in a living room.

Today, that line to the consumer is through 5 or 6 different types of devices.

When you add in the number of skinny bundles or virtual MVPDs from companies like Sony, Comcast, Hulu, Sling TV, DIRECTV, fuboTV or many others, there are probably more than 20 different ways to watch ESPN, for example.

This is driving the increase in the number of content owners who are providing their inventory directly to us.

Content owners can eliminate many steps in the distribution channels and monetize their ad inventory more directly and effectively.

Our objectivity from not owning media ourselves makes The Trade Desk one of the most important partners to these TV content owners.

For example, take the most recent World Cup on Fox.

We ran ads across every game in the World Cup, ads that aired via our partnerships with many of the virtual MVPDs, not just a single partner.

But new inventory in CTV isn't just coming from virtual MVPDs.

New inventory is coming from 3 main categories of new distribution.

First, it's coming from new ad-funded channels that didn't exist before the Internet.

These are channels like Hulu and Sony Crackle.

Secondly, new CTV inventory is coming from the new MVPDs we just talked about.

There are a lot more skinny bundles than I ever knew, and some of the big ones, like Sling TV, are doing very well and are great partners.

And third, ads are coming from new channels, sites or players where content owners are going direct to consumers.

For example, over the past year, our relationship with Discovery has grown tremendously.

They are providing premium long-form content on all types of connected devices.

This approach is how new standards in programmatic advertising will emerge.

To that point, we are working with many of the biggest digital publishers in the world, from Baidu to Google, to Alibaba to Spotify, to Pandora, to CBS, to DISH Network and DIRECTV.

We think, long-term, we may be the one company who can partner with everyone in digital media because of our scale and independence, with no conflicts of interest.

As a side note, I want to remind everyone that TV market dynamics are different from other digital channels like social.

No single company dominates the market share in TV, so the industry is unlikely to see a walled garden approach succeed.

The Google and Facebook playbooks for search and social do not seem applicable when no one can or will ever own as much market share in TV as Google has in search or Facebook has in social.

Content owners are going direct to consumers, and that is fragmenting the supply chains and breaking up the aggregation points that cable companies used to be.

Because of all these changes, there has never been more opportunity in media than there is now.

TV distribution is more fragmented than ever as content owners in desperate need of ad revenues increasingly try to go direct to consumers.

Internet TV, especially ad-funded Internet TV, is all up for grabs.

For agencies and advertisers, The Trade Desk is the only way to effectively target across the fragmented distribution channels that characterize the emerging Internet television ecosystem.

Once again, our value proposition is enhanced by this development.

Because we are independent and objective, we can nimbly move where the TV ecosystem moves.

As I mentioned, The Trade Desk had seen great progress in CTV.

After 21x growth in Q1, Q2 was more than double what Q1 was.

Further, our scale on inventory increased by over 7x compared with the year ago, which is perhaps equally exciting.

With more inventory, we have the option to do more targeting and add more value.

This continued growth is perhaps the most exciting thing happening in our company and in our industry.

And that's not just in North America.

The big story in the European advertising marketplace and our third major event of the quarter was the GDPR rollout on May 25.

It was tough on many publishers, but the weekend after the launch date, our engineering and partnerships team put in a huge effort working with publishers and their SSPs who, in some cases, were not ready to ensure the technology was in place to secure the required consents.

But in Q2, we delivered record spend in all 4 of our European offices: the U.K.; Spain; Germany; and France.

GDPR did not diminish spend over the quarter.

In fact, the trust we built with our partners and customers was massive, and we even won additional spend because of GDPR.

And we are seeing similar gains here in Asia.

Our business in China continues to progress.

We are building strategic partnerships with major players like Baidu, Alibaba's Youku and Hong Kong's TVB, which are key to scaling our business in this market.

Though we have strong focus on China as a strategic market, Asia, as a whole, is crucial to our success.

The opportunities here are bigger than we originally thought.

And it's not just with advertisers.

Inventory partners in Asia also see The Trade Desk as an essential strategic partner.

Most publishers are not getting their fair share of spend today relative to the walled garden.

One of the largest publishers in Asia wants to rapidly increase advertising on their content, but they realize they can't fully monetize it on their own because they are not getting enough budget from large global advertisers.

Many other inventory partners in Asia are facing the same problem.

As a result, they are looking to partner with The Trade Desk, a partner they can trust, who is also trusted by global brands.

Our inventory in the region has increased nearly 60% year-over-year.

The secular trend driving all of this is the emergence of the largest middle class in history, and most of this emerging middle class is coming from here in Asia.

It is essential for brands to reach these consumers through the channels they use the most, Connected TV and mobile.

Because we made investments early on in these channels, we are starting to reap the benefits and expect to continue to do so well into the future.

And the last of our significant Q2 events was that The Trade Desk launched the biggest product in our history at the end of June.

Internally, we've likened this release to Apple's first launch of the iPhone.

Our DNA has always been to ship products every week, but with this product, we invested almost 40% of our engineering and resources over the last 2 years to overhaul the entire user experience and make decisioning even easier for our users.

What we call the Next Wave significantly increases our technological lead over our competition.

The result is increased sophistication and ease-of-use for the tens of thousands of people who use our platform every day.

This release was the biggest enhancement ever to our platform, and it's already paying off.

Once again, there's a larger context to understand about why this launch has such an impact.

Internally, we are talking more and more about the concept of consumer surplus, and by that, we mean giving customers measurable value that exceeds the fees they pay to use our platform.

Let me explain why this is such a focus for us.

Providing our customers with more value than we charge them, for us, is the definition of sustainability.

We could choose to enhance our products and charge more for it, or we could just pass the added value on to our customers and increase their satisfaction with our platform, in other words, increase our consumer surplus.

This is the best explanation for our 95% client retention rate over the last 18 quarters.

In 2018, we have added more value than ever without our customers seeing their platform fees increase.

Within a few days of our recent product release, I met with the CEO for APAC for a large worldwide brand.

The reception was phenomenal.

We talked about objectivity and transparency.

We discussed the new products we launched, and he said, "The Trade Desk has given me so much more with this new product rollout." This major advertiser is now spending more on programmatic and moving more of their brands and campaigns to our platform.

The Next Wave consists of 3 new products: Megagon, a data-focused user experience that enables media buyers to see precisely how their bidding strategies affect their opportunities to win impressions; Planner, an innovative tool that enables media planners to generate a range of campaign scenarios and validate them against data-driven insights; and Koa, the artificial intelligence that drives it all, backed by data from our entire bid stream.

Just a word on Koa and our approach to AI in general, we have developed new AI to make the data-driven parts of advertising more automated.

But we've also made it easier for media buyers to participate in the process too.

We built the best approach where people create hypotheses and machine-test them.

And unlike black box implementations of AI, The Trade Desk has always been transparent about what insights are being generated.

Koa draws a clear line between what the AI found and insights it provides to the user.

The user has the option to incorporate the recommendations or not, and the choice and transparency is always there.

The response to Next Wave has been extraordinary.

We live-streamed announcements that, along with other launch videos, generated nearly 650,000 views globally to date.

Industry press and general business media highlighted the Next Wave with more than 300 stories worldwide.

More importantly, our customers enthusiastically started using the new tools.

In the 1.5 months since launch, we are seeing great traction.

If the current trends continue, the majority of spend on our platform at the end of the year will be on our new product.

Already, more than half of those adopted are Koa-enabled.

Over 2/3 of those using the new product have switched to Koa predictive clearing.

This is an incredible feature that counters publisher moves to first (inaudible) auction, which have caused CPMs to jump approximately 40% over the past couple of years.

Early results for those using predictive clearing indicate CPMs are being reduced by up to 20%.

That's a huge value.

That is amazing consumer surplus.

When we provide our users with new savings of 20%, we're more than earning our platform fees.

That's the consumer surplus we're talking about.

The value exchange has never been better for our customers, and we expect that will help our growth.

The Next Wave is a great example of our investment strategy.

We have proven the value of investing ahead in areas of technology, channels and geographies.

It enables us to grab share in the fast-growing programmatic market that is just over 2% of the entire advertising market.

We've done this consistently.

But it does not stop here.

We will continue to innovate more quickly and efficiently than others in our industry.

This includes areas such as CTV, mobile, global expansion and creating a safer programmatic environment as well in investing in our infrastructure and data centers.

We are not maximizing profit for today.

We are doing the best thing for the growth of our business and profitability over the long term.

But even as we make these investments, we are still generating EBITDA significantly higher than almost all of the high-growth software companies of similar size.

In closing, we had a great second quarter.

The secular tailwind is strong.

And like what occurred in the first 2 quarters of the year, when we have revenue surprises, they tend to be to the upside.

We have incredibly strong momentum, and I expect the wind to stay at our backs.

We are executing well and responding rapidly to new opportunities to create consumer surplus value.

We increased our guidance for the year to $456 million, which is about 48% year-over-year growth.

As great as the numbers are for this quarter, we are most excited about the growth for the year, which gives the best indicator of the strength of our business.

The pieces are all in place for continued success for the rest of this year, for 2019 and over the long term.

Now I'd like to turn the call over to Rob for his comments on our operational performance.

Thanks, Jeff, and good afternoon, everyone.

We had a strong second quarter and a great first half of the year.

Our business continues to deliver terrific results.

Second quarter revenue grew 54% year-over-year.

Mobile spend, again, grew nearly 100% year-over-year and exiting Q2, was at its highest level as a percent of our total spend in the company's history.

Growth on the platform continues across both channels and in regions.

In audio, one of the best values in programmatic today, it grew by just under 200%.

Mobile video grew 156%.

And as Jeff mentioned, Connected TV spend more than doubled from the previous quarter.

We had another exceptional quarter of international growth.

Our offices in Germany, France and Spain all posted over 100% growth in the quarter.

In APAC, Australia led the way, growing over 150% on a year-over-year basis.

It is our goal to always deliver the highest return on investment to the agencies and brands using our platform and create surplus value for them.

We do this by focusing on the 3 core priorities of our business, which are: one, to remain the objective and independent trusted partner to our customers; second, to continue to grow our omnichannel presence; and third, to continue to grow our international footprint.

Our objectivity and the fact that we do not own media differentiates us from our largest competitors.

The choice, transparency and insights that we offer are what creates trust between us and the agencies and brands that use our platform.

They can invest in their own first-party audiences, in CRM segmentation and deeper audience analysis through their own data management platform.

It is important for them to understand and optimize reach and frequency across all their media buys.

Agencies and their brands also want robust measurement to track sales and measure offline sales attribution.

We offer more options for third-party measurement and verification so that they can verify and analyze performance across all the platforms they may use.

We have over 100 data and other value-added services partners in our platform.

And while large competitors stop sharing their IDs and do not offer solutions for offline sales attribution, we provide advertisers a better path to understanding how their marketing spend drives overall revenue growth for their companies.

Our objectivity measurement solutions and shared ID have led us to winning new business and an increasing spend from existing customers in the quarter.

The Trade Desk's other value propositions, such as independence, transparency, ease-of-use, more sophisticated technology, world-class client service and a focus on providing surplus value to our customers also helped our pipeline of new business become stronger than ever.

A great example of this shift in spend is from a large global food company that we worked with in Q2.

The Trade Desk had been a regular part of their programmatic business, but we had a smaller piece of their spend relative to our competition.

Our key performance metrics across multiple channels regularly outperform the other platforms they used.

For example, viewability came in consistently 15% to 20% higher than all other platforms while maintaining comparable cost efficiency.

But when third-party measurement and offline attribution ability came into play, it really changed the game for this customer.

The ability to tie online ad exposure to offline sales results allowed the team to deepen our relationships and prove the value of programmatic advertising on our platform.

We became the largest platform for their programmatic business in Q2 and are now on pace for over $15 million in spend this year, which really dwarfs the prior run rate that we have with this customer.

Next, I want to focus on our omnichannel presence.

It's no surprise that nearly all of us use multiple Internet connected devices each day, and one of our fundamental beliefs is that programmatic advertising should reflect this fact.

To be successful in programmatic advertising, marketers need to coordinate their messages across multiple channels.

And in Q2, we continued to see marketers advertise across more channels than ever before, which include mobile, video, Connected TV, audio, native and display.

In Q2, our customers using at least 6 of these channels increased by 156% versus the same quarter a year ago.

We also see agencies and advertisers take full advantage of omnichannel campaigns by incorporating cross-device data.

And with our recent Next Wave product launches, it is easier than ever to layer cross-device data by using our identity alliance product.

This enables the media buyer to leverage cross-device data from the 4 major cross-device vendors that fast increases the size of a target audience for every single impression.

A great example of an advertiser increasing the ROI of their ad dollars using cross-device data is a large international hotel chain that through their agency had a goal to increase their target audience and capture more online booking conversions.

By activating all 4 identity alliance cross-device partners through our platform, the agency was able to drive a 20% increase in new customer bookings.

These are significantly better results for that advertiser.

Every dollar spent on cross-device data easily paid for itself, resulting in an average 31% increase in ROI across the 4 cross-device partners.

Moving on to our third priority.

We are continuing to widen our geographic footprint to serve our customers locally in the markets that are important to them.

International spend growth percentage continues to outpace that of the U.S. significantly, and in the second quarter, that trend continued.

Our international business amounted to 15% of our total spend, which is the highest percentage ever.

The teams in nearly every one of our offices outside the U.S. reached their all-time spend record in Q2.

The adoption of programmatic and the market growth we saw across both Asia and Europe was very strong, but it is the success we had in Europe in this quarter that I want to highlight on this call.

So as GDPR went live, our team worked around the clock to ensure the technology was in place to secure the required consents with both publishers and SSPs.

And our work really paid off this quarter.

As other large platforms cut off external SSPs and publishers, we maintained and strengthened those relationships.

When large advertisers came to us stating that their DSP was unable to access certain publishers or sites that they wanted to access, we were there for them.

This enabled us to sign new MSAs and move more spend over to our platform.

The new business wins included large global brands such as a major food company, a global airline and another large beverage company.

We have seen this momentum continue as we exit Q2, and our prospects in Europe are well positioned over the long term.

Now overall, we feel really good about our second quarter success and our prospects for the remainder of the year and in 2019.

We have made tremendous strides in increasing our customer engagement and retention, hiring and training the next generation of our team and continuing to advance our international strategy.

The advertising industry is still in the early stages of its programmatic transformation, and we see a huge multiyear opportunity in front of us.

Now I'm going to turn the call over to Paul to discuss our financials.

Thanks, Rob, and good afternoon, everyone.

As you've seen in the numbers, the first half of 2018 is off to a record start and we were pleased with our Q2 financial performance and execution.

Revenue increased 54% year-over-year, equal to the 54% year-over-year increase in Q2 of last year.

Adjusted EBITDA increased 46% year-over-year, and GAAP net income was $19 million, all while we continued to invest aggressively in areas critical to our future growth.

Revenue for the second quarter was a record $112 million, which was above our prior expectations and reflected the increased spend by our existing customers and the addition of new customers and advertisers.

For the quarter, approximately 89% of our second quarter gross spend came from existing customers who have been on our platform for over a year.

With the growth of our business, our operating expenses grew to $86 million in Q2 of 2018 from $53 million during the same period in 2017.

This increase was primarily due to our increased investments in our platform operations and increased personnel, primarily in technology and development, as we invest for future growth.

Total other expense net was $1 million, and the income tax expense was nearly $6 million in the quarter for a tax rate of 23%.

GAAP net income for Q2 was $19 million or $0.43 per fully diluted share.

Our adjusted net income was $27 million or $0.60 per fully diluted share compared with adjusted net income of $23 million or $0.52 per share in the comparable period.

GAAP and adjusted net income in Q2 of 2017 reflected a discrete tax benefit of $8.6 million, primarily related to incentive stock options.

Adjusted EBITDA was $37 million, with a corresponding margin of 33% of revenue during Q2 2018.

The increase in adjusted EBITDA dollars reflects the strong growth of our top line, offset by our increasing investments in product, people, global expansion and corporate expenses.

Of the $9 million revenue increase above our prior expectation, approximately 75% of that revenue fell through to adjusted EBITDA.

Net cash provided by operating activities was about $1 million for Q2, and our trailing 12 months of operating cash flow and free cash flow were $56 million and $42 million respectively.

We have 0 debt on our balance sheet, and our cash position exiting the quarter was $142 million.

Our DSOs at the end of Q2 were 101 days, an increase of 6 days from the same period a year ago.

DPOs for Q2 were 82 days, an increase of 9 days from the same period a year ago.

We are continually striving to close the gap between DSOs and DPOs, which benefits our cash flow from operations.

For Q3 of 2018, we are expecting revenue of $116 million and adjusted EBITDA of $33 million.

And for the full year 2018, we now expect revenue to be at least $456 million, which approximates to 48% growth year-over-year and corresponding adjusted EBITDA to be $140 million or nearly 31% of revenue.

With that, I will hand it back to Jeff in Hong Kong for any final comments and of course, Q&A.

Jeff?

Thanks, Paul.

In closing, let me reiterate that while we are excited about The Trade Desk's current performance, we see even more potential for the future.

As the worldwide advertising market grows to $1 trillion, we believe it will move to programmatic.

Programmatic is the fastest-growing segment of advertising, and The Trade Desk is growing faster than anyone in programmatic.

When we see surprises, they typically are to the upside.

There is a generational shift happening with the convergence of the Internet and TV globally.

Massive markets like China are just starting to adopt programmatic, and I believe it is highly probable that the programmatic industry in years ahead will see accelerating growth.

We see the opportunity, and now is the time to invest in growing market share and revenue.

We believe The Trade Desk is well positioned to realize this growth for the rest of the year, next year and beyond.

That concludes the prepared remarks for this afternoon, and now the operator will open it up for questions.

(Operator Instructions) Our first question comes from the line of Tim Nollen from Macquarie.

I have a broad question, which is kind of multifaceted, and the main gist is on your CTV business.

If you could help us understand where the ad inventory is coming from.

And what I'm wondering is if you could elaborate a bit on the discussion of the virtual bundles that have come to market.

You did not talk about devices so much, Roku devices and so on.

I'm sure that contributes as well, and they have inventory that they can make available too.

If you could just sort of help us understand how that works.

And then in conjunction with -- you mentioned some of the operators offering the bundles, and you even mentioned a network group like Discovery.

So if you could maybe just explain a bit better for us, please, where all this inventory is coming from and how it kind of aligns.

And also, you seem to be in a position where you've got great volume growth and also great price growth.

And I suppose that happens when a market is developing rapidly, but it's a very nice position to be.

And if you could maybe explain a bit of that as well, please.

So that's kind of all about the CTV inventory.

And then if I could ask another question, which is about the ad agencies that you really very much choose to rely on for the majority of your spending, if you could explain why that is so important to you and why you also have relationships directly with advertisers.

And I guess behind that is, could you work with advertisers that wanted to bring programmatic buying in-house?

If so, why, or if not, why not?

Great.

Well, first of all, thanks for the question, Tim.

Lots to cover there.

So first, in terms of where the inventory comes from, there's 3 main places where it has been coming from.

The first is from the TV stations or channels that are new since the Internet has arrived.

So if you're on your Smart TV menu or on your Amazon Fire or Apple TV or Roku device, it's pretty much all of them after Netflix and Amazon.

So the Hulus of the world, the Sony Crackle, all of those that have ads on them below that, and more and more of them have premium content and there's more and more consumption.

The second is -- are those skinny bundles.

And when I look at how many have popped up, I'm blown away by how many of them are -- how many of them there are.

There's a lot of people out there.

And I suspect most of the people on this call are the ones that can afford to pay for cable.

And so like me, you're unaware of how many of them are out there, but there are dozens and dozens of them, and they're all dependent on ads.

And then the third, and honestly, this is the one that I'm most excited about, it is the Fox and Discoveries of the world.

So we've definitely had a lot of conversations with all of those so that we can go direct as much as possible and get access to that inventory.

The devices, I think there's an open question about the long term on how much inventory can or will be tolerated from the devices.

I think most of the inventory is going to come from people who own the content, and that's why I'm so excited about the Discoveries and Foxes of the world.

So there's only a small amount of inventory available from the devices, but that may change and sort of an open question as to where that will come from.

But it's a small contributor to us today.

I know there are lots of parts to the question.

Let me just touch on the ad agency question, and we'll -- we need to go to the next one.

Our agency relationships are really critical to us because they historically have been the very best at servicing accounts.

And when I say the very best, I mean, what they really offer is scale and a global approach.

It doesn't mean that there aren't ways for them to improve, and most of the big ones have had some trouble transitioning to this new digital world.

But I think, often, people underestimate how much gravity they have when they employ -- the big ones employ nearly 100,000 people, or WPP employs more than 100,000.

And so for us, that represents a huge army of people to service accounts and leverage our platform.

There's no way that we have the operating leverage in our business without partners like that.

And at a time when things are more complicated, we think brands need to rely on those even more.

That said, we are signing more and more deals direct with the brand, but that's not because we're trying to cut the agency out, it's that we're trying to provide the brand assurance or reassurance that their data is safe and that we have contractual obligations with them to make certain that we're protecting and using their data the way that they want to.

And when we do that, it gives us the liberty to put that to work, and we're able to do a better job for them.

Our next question comes from the line of Shyam Patil from SIG.

Jeff, I just had a quick one.

You talked about the changes that Google and Facebook had made regarding their policies around data sharing and usage.

Just wondering if you could talk a little bit more about what kind of opportunities all of that has opened up for The Trade Desk and how you see those opportunities materializing over the course of this year and next year.

First of all, thanks for the compliment and for the question.

So I'm really glad that you asked this question because I think what Google has done is it has really opened a door for us.

It's a huge opportunity.

And I'm not sure that we've ever explained well enough why this has created such an opportunity.

So there is essentially a Google ID that is assigned to each user.

And it makes it so that, that user can be identified when you go to every website or every destination.

So you used to be able to look at, oh, here's this user on Spotify, here's this user on YouTube, here's this user on whatever destination or device that they're on.

And then you could compare the performance of all of those, and it made it so that people could use the data and the insight to do attribution modeling and figure out who deserves credit for actually moving the needle on any given sale and basically do analysis of where am I spending my money and what is working.

Because they're afraid of that data -- this is my view of why they're making this policy change.

They're afraid of that data being used with SERPs data, supersensitive data or with PII so that some bad-intending data company could get insights that they shouldn't.

That's actually the reason why, from inception, we've never played in directly identifiable information with consumers.

So because we don't have a search engine, because we don't -- we've never used social security numbers, we don't even use email address.

We made it possible that we can activate the data and most importantly, provide reporting that makes it possible to compare Spotify to The New York Times, to when the time comes, YouTube and every other destination on the Internet, whether that's TV or display or a mobile app.

So if you're a marketer and you feel more pressure than ever to make your marketing dollar work and if you have to do more analysis than ever to defend where you spend your money, if you just lost the ability to compare things on a relative basis because of conservatism related to this ID, you're now looking for an alternative.

And so I would argue that the very best pitch that Google had sort of against us when they were selling DBM, and to be clear, I don't compete with all of Google, just a small division of it called DBM, but their very best pitch was that we have a unified stack that's built around an ad server that can compare everything else.

And even though you may argue that we're not objective when we steer dollars to Google.com or YouTube.com, we open that up so that you can measure on a relative basis.

They've made that harder than ever.

So I think their objectivity pitch is more of an obstacle than it's ever been, and that opens up the opportunity for us to give more transparent and more comparable reporting, which I think is only now beginning to pay off.

And we'll see what happens in the future.

But it signifies a lack of strategic commitment to what I think is a low priority at Google, which is to buy the rest of the Internet because I think most of their money comes from properties they own, and that's where they should focus.

And to me, it seems like that's exactly what they're doing.

So I think it's a game-changer, but it's technologically nuanced.

Our next question comes from the line of Youssef Squali from SunTrust Robinson Humphrey.

Jeff, I guess just following on that last question, I'm just trying to understand why is it necessarily kind of risky for Google to do that, which caused it to effectively limit how DoubleClick IDs are used and not present a potential risk to you guys over time?

Maybe if you can just kind of help us understand that, that will be great.

And then in terms of the product launch, how should we be thinking about it as a potential driver of extra spend?

How -- what's the rationale there?

How would that happen?

You bet.

So the -- as it relates to the first part of the question, the biggest reason, in my view, why there's risk to them sharing an ID like that is because it can be combined with sensitive personal information.

To me, that's where everything goes wrong in using data on the Internet is when you get to the sensitive information.

And if you think about how sensitive some of the information is that you share with Google, and I mean, in the search engine, right, like any time you have a medical problem, you do what I do, you go ask Google.

It's just anything that is sensitive, like the details of your divorce, your worries about your kids, like anything, you're going to type into Google.

And Facebook, I think especially because of the Cambridge Analytica challenges, what was happening is just a whole bunch of insights about users who were able to be stripped away so that then data that you actually gave to Facebook, personal data, was put to work.

So the reason why this doesn't pose a risk for us long-term, at least in the same way that it affects them, it doesn't mean that we don't have to be sensitive, of course, we do.

We have to be sensitive to consumers' data.

We have to be super careful.

But for us, part of being super careful meant we never ever ask for personal, like sensitive personal information, and we don't get directly identifiable consumer information, like name, social security number, medical conditions.

We don't store that.

We don't use that.

We don't ask for it.

We don't have a search engine.

And we don't have a social networking site where we ask all these personal questions.

We have no way to identify a user personally.

It's only in an anonymized way and without any of that sort of sensitive data that comes through a search engine.

So by that inherent difference, there's not an ability to activate it, and it makes it easier for us to be open in sharing with our partners an ID so that they can activate their data.

And of course, by working with the biggest brands in the world, we all have similar sensitivities and create this sort of coop where we work together.

So it works because of the type of data that we play in.

What was the second part of your question again?

Yes.

The second part was around the product launch and how should we be thinking about it as a potential driver of spend or extra spend as you go forward.

Yes.

So of course, the biggest launch in our company's history is I would summarize it all as an effort to make the media-buying process easier, but at the same time, more sophisticated.

And normally, you have to choose, either you make it easy and you lose power, meaning the power of the actual technology, and you trade power and sophistication for easy, and we just figured out a way to give the user both.

So it is more powerful.

It is more effective.

It's even more transparent.

But it is easier to use.

And we figured out ways to reduce the number of clicks, and we've done a better job of automating the things that can be automated and still giving the user the ability to inject themselves wherever and whenever they want without a lot of steps.

So we just made media-buying better.

And then by making a planner, we made it possible for them to, in a data-driven way, do better allocation.

So I don't think most people realize how bad the planning process in most of advertising is, like how you decide how much money to put in Connected TV or how much you put in mobile.

Like it historically has not been very data-driven, it's been mostly guessing.

So by not only making the media-buying process data-driven, we've also gone a little bit upstream to make the planning process data-driven, which is just going to make buying better and it's going to create better results for everybody as well as eliminate a lot of the back-and-forth that often happens in the media-buying process, which is this part works really well, I want to put more dollars to it, but it's just not possible, there's not more inventory, there's not a way to get more of that one thing.

So there's a lot of budget shuffling.

So we got rid of huge amounts of time wastes in our space, but we're just slowly getting our current customers moved over.

So the vast majority of them are still on our old platform.

So we're slowly moving them over.

We expect to have the vast majority of them moved over by the end of the year.

So the adoption is fantastic, but it's slow and steady and there's lots of education that comes with it.

So I mentioned in the prepared remarks that it's already paying off.

It is.

We're winning business because of it.

But in terms of its actual impact, I think we're going to be seeing that for years to come, and there's tons of upside left.

So we're just getting started on its impact.

And then, of course, once we move everybody over, we'll actually be developing faster on the new platform.

And one of the benefits that came with just the change as well, to the user was mostly an overhaul in the front end.

There's definitely some overhaul that took place on the back end that makes it easier for us to iterate in the future.

So we will be shipping product even faster than before.

There's never been a product release that we've shipped that had a bigger impact on our business than this one, and we've only begun to realize its potential.

And we think that there's always been distance between us and our competitors, whether they're big companies or small companies, because we've been laser-focused and had a clear vision of what media-buying should look like.

But the distance between us and our competitors has widened in the last 2 months more than ever in our company's history because of the product release that we just put out.

Our next question comes from the line of Brian Wieser from Pivotal Research.

I was curious about your thoughts on whether or not Customer Data Platform, CDP, is -- the product category that's evolving, is something that you think you need to push into, firstly.

And then separately, on your working capital, I'm curious if how those may be skewed by new customers, large ones that you may be working with or maybe working with independently of the agencies, just curious if there's any variation in working capital trends that you expect to see from your evolving customer base.

Or if we'll get anomalies, relatively negative numbers, in big growth quarters just randomly.

You bet.

So I'll take the CDP question, and then I'll have Paul take the working capital question, and then if Rob wants to add any color, so feel free.

On the CDP, so there's lots of different acronyms that our space creates for basically ways for companies to store data.

And sometimes, I think the acronyms actually confuse people, whether it's a DMP, a data management platform, consumer data platform, Customer Data Platform, like all of those are ways for companies to store data and then put it to work.

In broad strokes, I would just say, whether we're talking about CRM data or customer data, loyalty data, every form of data that is useful for an advertiser, over time, they're going to find ways to activate it.

And especially -- like on one hand, you need to respect consumers' privacies.

On the other hand, you need to use the data, which I think effectively is listening to consumers.

When they share something with you, they expect you to use it so that you don't bombard them with messages or "I already bought that product." Like they expect you to listen.

So you have to do both.

And they are going to have to activate it, and we're going to have to plug in to DMPs and CDPs and basically, any way that they want to store and activate their data.

I will say that, over time, I think the number of moving parts is going to reduce.

And so we think we have to make it easy for people to store data, in part, because we're the place where they actually activate it.

We're the place where you actually use the data.

So if you put the data somewhere else and then they ultimately have to plug it into us so that in real-time, we can use it to make better decisions on behalf of the brand or the agency, then more and more, the name of the game is developing quick pipes to get that data into our platform.

We're happy when other companies do it.

We're also happy when they work directly with us.

To us, we care more about data competencies than how many hops there are.

And sometimes, when there's lots of hops, it reduces efficiencies.

So ultimately, we're going to activate all the things that are respectful to consumer privacy and create the better experience and efficacy for brands.

On the working capital side, Paul?

Yes.

Thanks, Jeff.

As far as working capital goes, there really aren't any new trends.

From a working capital management standpoint, we tend to focus on the difference between the DSOs and the DPOs, and those have been improving over time.

If you look at our balance sheet, we had over $75 million in net cash less than 2 years ago at the time that we went public.

And we have $142 million now, and so we're expecting cash to continue to increase in a channel similar to that slope up and to the right for the foreseeable future.

Really, nothing unusual with working capital.

And the only thing I'd add maybe for color.

You talked about if we're more working directly with, say, a brand, is there a difference in time lag.

And while large brands have their own cash conversion cycle, remember that in the agency model, the brand pays the agency, and then the agency has a holding period and then they pay us.

So I'd actually argue that in most cases that we've seen to date, we've been sort of net 0 or even slightly positive because we've taken that link out of the chain, if you will, in terms of getting the cash back.

Our next question comes from the line of Brian Schwartz from Oppenheimer.

Jeff, I want to bring you back to the business results here.

I would assume that you get very little contribution in the quarterly revenue mix from the new customers and the new agency partners that you signed up during the quarter, but this quarter was such a good one, Jeff.

I mean, you grew at the same rate as last year.

And then if I look back into the first half of this year, the revenue and the profit growth has been accelerating here in the first half versus the second half.

So it sort of begs the question, if you got any windfall or any meaningful contribution in the second quarter revenue mix from any of these new customers or agency partners that you signed during the quarter.

And then if the answer is no, which I assume it probably is, then what is it?

What do you attribute this sort of acceleration to the business in the first half of the year?

You bet.

Thanks, Brian.

Appreciate the question.

So we did get a slightly higher percentage of our revenue coming from new customers, which is actually super exciting.

But when I say slightly higher, it's ballpark, 90% from existing customers and 10% from new customers.

But keep in mind that our customers tend to be platform users that use our platform forever and they maintain relationships with lots of advertisers.

And so where most of that growth is coming from is them getting slightly more budget from their advertisers that have already been spending with us and then, in some cases, bringing on new brands, which, in most cases, are included in that 90%.

Whereas they're bringing on a new brand, but because our customer is the same, we're getting more spend.

But in my view, seeing Connected TV after a 21x growth last quarter, quarter-over-quarter, seeing that double again and just seeing all the move into programmatic, like the secular tailwinds are the very best explanation of where all this is coming from because this is not a case of like new media versus old media.

This is really about data-driven media-buying versus testing.

That to me is the best explanation of the change.

Rob, I don't know if you have something to add.

Sorry, the only thing I would add, Brian, and you're right on, like what I would say is the wins or sort of the acceleration or matching of the prior year growth is as much about the big wins we've had with major brands in Q3 and Q4 of last year.

Jeff has often said on calls like this, where the bigger the brand, the longer it takes them to sort of really ramp.

And so we had a significant amount of meaningful wins in Q3 and Q4 last year where they were testing us and then came onto the platform in a more meaningful way in Q1 of this year and then even more in Q2.

And so we've had a similar trend of wins, I would say, in Q2.

And you're right, those wins in Q2 didn't really contribute in a meaningful way to our Q2 this year, but I would say that gives us confidence as we head into next year.

The size of the wins that we've had, that are just starting with us now testing, really is a harbinger for what we think 2019 could be.

Our next question comes from the line of Peter Stabler from Wells Fargo.

All on data.

First of all, I'm wondering if -- I think last quarter, you gave us the growth rate of the revenues associated with data.

I think you offered us a partial view this year.

I'm wondering if you could give us the total growth number.

And then secondly, if we understand it correctly, Facebook's new third-party data policy is essentially forcing marketers to contract directly with third-party data providers.

So does Facebook training the market this way present any element of risk to you guys as more marketers elect to directly contract with data suppliers, pull that data into their DMPs, do their customer segmentation essentially off of Facebook and then bring that data back in.

Are they training the market in a way that could be negative in the future for your data brokering business?

You bet.

So I'll just respond briefly on the data side.

I don't have the total data number in front of me, but I can say that data spent on our platform in Q2 was the highest that it's ever been.

And then our cross-device, which is, in part, because of the partnerships with every cross-device vendor in the marketplace as well as the acquisition that we made last year, was up over 100% in Q2.

So that one, I think, is especially relevant and the bellwether.

I think the bigger question that you asked about is Facebook training the market to make sort of third-party companies critical to our future in terms of them sort of housing the data, activating the data.

And it doesn't change really anything for us, it doesn't change our leverage or our opportunity, and let me explain why.

So the place where you can make money in data as well as where you have the most amount of impact is if you have the data at the moment that you make decisions about what to buy and what not to buy.

So the closer you get to that decision, the more power you have with that data, in part, because that's where the data is really potent.

Oh, here are the groups of users that are 55x more likely to buy my product than everybody else.

Now I'm going to adjust what I bid.

I'm going to adjust what media I buy.

I'm going to let other opportunities fall by the wayside because that's the one that I really want.

So even if other companies are storing the data, and this relates to the last question about data, where I'm fine if there's lots of data companies out there that are activating or helping big brands store and get that data into the media-buying process.

We're using a tiny fraction of what we will be at end-state.

So the faster that we can get the relevant data, while, of course, respecting consumers' privacy in the process, the better.

And those third-party companies that Facebook is pushing now are super hungry.

They tend to be more agile and more cooperative than sort of the big new media companies who are trying to do as much of the process themselves up until recently.

So overall, I think that trend is very good for us because ultimately, the power sits at the moment of consumption.

Mr. Green, just to confirm, did you want to continue with the Q&A?

Yes, let's do one more.

Our next question comes from the line of Brian Fitzgerald from Jefferies.

We'll try to make it quick.

Maybe at a higher level, Jeff, could you talk about the competitive landscape now that AT&T has bought AppNexus?

Is it easier to win new clients, given that AppNexus is now inside of AT&T?

And then how do you think about the relationship between advertisers and agencies and then publishers and distributors?

How does that change going forward?

Do the sources of future inventory change at all?

You bet.

So first, let me say I love the fact that AT&T bought AppNexus, and I love that they bought Time Warner even more.

I love the AT&T strategy.

I think it's brilliant.

And I mean that more for what it does for TV and what it does for the landscape than just AT&T specifically.

Like I think the tug-of-war over Fox and the need for Disney to own Hulu, for instance, is directly the result of the AT&T acquisitions in Time Warner and AppNexus.

I do think that AppNexus, the majority of their business over the last couple of years has been on the supply side.

So they've become a better partner to us because where they were both buy-side and sell-side 4, 5 years ago and they were half competitors to us and half partner, they've become mostly partner over the last couple of years as they focused more and more on the sell-side.

That's even more the case inside of AT&T.

So I know a lot of people on Wall Street have asked the question like when is TV happening?

When is it going to move to connected?

And when is digital going to drive choices?

Well, AT&T essentially spent $2 billion saying that time is now.

And as they're racing to get 5G out this year, their need for programmatic ad solution is now.

So it sort of answers that question with it is happening right now.

We definitely have one less competitor in that move and that you start to focus more and more on the sell-side.

They give the marketplace, something I've been begging the market for, which is more players doing TV and video on the sell-side well.

So that's super exciting that the market gets a little more focus from one of the best engineering teams in programmatic at AppNexus.

So it's really good for the marketplace.

It also just signifies people who own content, developing distribution.

And whether that's Time Warner being acquired by AT&T or whether that's AppNexus developing the pipes for Time Warner and AT&T, the relationship between content and distribution is getting closer.

That, for the most part, means more fragmentation.

And that, for the most part, puts way more pressure on companies like us to aggregate it all and make it possible for advertisers to control frequency and make informed data-driven decisions, so the changes to the landscape are really, really bullish for our future.

So there's been little M&A over the last 10 years that excites me more than that one.

Ladies and gentlemen, we have reached the end of the question-and-answer session, and I would like to turn the call back to management for closing remarks.

Thanks for everyone joining, and this now concludes the call.

Thank you.

Thank you, everybody.

This concludes tonight's conference, you may disconnect your lines at this time.

Thank you for your participation.