Roku Inc (ROKU) 2020 Q3 法說會逐字稿

Ladies and gentlemen, thank you for standing by, and welcome to the Third Quarter 2020 Roku Earnings Conference call. (Operator Instructions) Please be advised that today's conference is being recorded. (Operator Instructions)

I would now like to hand the conference over to your speaker today, Conrad Grodd, Vice President, Investor Relations. Please go ahead, sir.

Thank you, operator. Good afternoon, and welcome to Roku's financial results conference call for the third quarter ended September 30, 2020. I'm joined on the call today with Anthony Wood, Roku's Founder and CEO; Steve Louden, our CFO; and Scott Rosenberg, Senior Vice President, General Manager of our Platform Business, who will be available for Q&A. Full details of our results and additional management commentary are available in our shareholder letter, which can be found on the Investor Relations section of our website at ir.roku.com.

The following discussions, including responses to your questions, reflects management's views as of today, November 5, 2020, only, and we do not undertake any obligations to update or revise this information.

Some of the statements made on today's call are forward-looking and are based on our current expectations, forecasts and assumptions and involve risks and uncertainties. These statements include, but are not limited to, statements regarding the future performance of Roku, including our financial perspective for the fourth quarter and full year 2020; the future of TV and TV advertising globally; the impact of the COVID-19 pandemic on our industry; business and financial results; and the future growth in our business and our industry.

Our actual results may differ materially from those discussed on this call for a variety of reasons. Please refer to today's shareholder letter and the company's periodic filings with the SEC for information about factors which could cause our actual results to differ materially from these forward-looking statements.

You'll find reconciliations of non-GAAP measures to the most comparable measures discussed today in our shareholder letter, which is posted on our Investor Relations website at ir.roku.com, and I encourage you to periodically visit our IR website for important content. Finally, unless otherwise stated, all comparisons on this call will be against our results for the comparable period of 2019.

Now I'd like to hand the call over to Anthony.

Thank you for joining today's call. Q3 was remarkable for our industry and for Roku. More American consumers streamed to their TV than ever before. This included free, ad-supported services as well as live programming and subscription. The Roku Channel benefited from offering all 3.

Major media companies continue to reorganize around streaming. They embrace Roku because of our scale and content marketing capabilities. Partnering with Roku is an efficient way to build and retain large and valuable audiences.

Advertisers reassessed their TV upfront advertising commitments and moved significant portions of their investments to connected TV platforms like Roku. Advertising with Roku gave marketers significant incremental reach over linear TV as well as advanced capabilities to target their advertising and to measure its effectiveness.

Streaming is stoking innovation and giving greater choice, value and control to consumers. Despite the ongoing uncertainty caused by the pandemic, Roku's long-term prospects are strong and our outstanding financial and operational performance in Q3 shows the inherent leverage in our business model.

With that, I'll hand it over to Steve.

Thanks, Anthony. Before we take your questions, I'll walk through operational and financial highlights and discuss our viewpoint looking forward.

We added 2.9 million incremental active accounts in Q3 and ended the quarter with 46 million active accounts, up 43% year-over-year. Sales of player units rapidly accelerated, up 57% year-over-year, while average selling price decreased only 1% year-over-year given less promotional activity due to strong demand resulting in tight inventory levels for certain products. This extraordinary level of player unit growth was driven by a confluence of factors, including significant retailer channel inventory replenishment after strong Q2 player sales, continued robust demand for streaming players in Q3 and a portion of holiday inventory arriving at the end of Q3.

Engagement on the platform continues to grow, with Roku users streaming 14.8 billion hours in the quarter, up 54% year-over-year and streaming hours per active account increasing at a more normalized rate of 9% year-over-year as COVID-related restrictions were lifted during the summer.

Now I'd like to highlight a few financial items. Q3 saw record total revenues of $452 million, up 73% year-over-year, driven by robust growth in both the Platform and Player segments. Platform segment revenue was up 78% year-over-year to $319 million, driven by broad-based strength in both our advertising and content distribution businesses. Roku-monetized video ad impressions reaccelerated to almost 90% year-over-year in Q3 versus roughly 50% year-over-year in Q2 as advertisers leaned into Roku, shifting dollars to streaming as they follow TV viewership and take advantage of increased flexibility given disruptions to the traditional TV upfront process.

The content business benefited from our rapid rate of active account growth as well as strong consumer demand for ad-supported viewing, subscription services and premium movie rentals. Reflecting these factors in our content distribution deal models lead to some significant increases in the estimated lifetime deal values with an outsized portion of that value change accounted for in Q3. As a reminder, revenue recognition for our content distribution agreements can be lumpy quarter to quarter. Player revenue grew 62% year-over-year, the highest growth rate in over 7 years, which was driven by an extraordinary level of player unit growth, as mentioned previously.

Gross profit grew faster than revenue, up 81% year-over-year in Q3 to $215 million. Gross margin of 47.6% increased on both a year-over-year and sequential basis, driven by strong segment margins. Platform gross margin of 61% expanded over 400 basis points versus the first half of 2020 due to margin improvements in both ads and content distribution. Player gross margin of 15% was significantly higher than the same period last year due to fewer promotions as well as lower return rates.

Q3 adjusted EBITDA of $56 million was a record, eclipsing the total adjusted EBITDA in 2019 by more than 50%. This outsized level of adjusted EBITDA was the result of several positive trends combining in the quarter, but also highlights the inherent leverage of the Roku business model as the platform continues to gain scale and monetization increases.

We exited the quarter with a strong cash and liquidity position, raising an incremental $148 million in equity capital via an at-the-market offering, resulting in over $1 billion of cash, cash equivalents, restricted cash and short-term investments. We also have $69 million of available liquidity under our credit facility.

Similar to last quarter, we do not intend to provide formal guidance for Q4 given that the short-term macro environment remains both variable and uncertain. We are closely tracking the potential for COVID-19 or economic-related disruptions as well as the potential impact to historical consumer spending levels or shopping patterns as we enter the holiday season. Instead, we will provide a framework on how we believe the quarter could develop.

Coming now to Q3, we are pleased with the resilience of our business and cautiously optimistic about the holiday season. Barring any significant external risk factors materializing, we estimate that the overall Q4 year-over-year revenue growth will be roughly in line with the last few holiday seasons, which was in the mid-40% range. We expect Platform revenue to account for roughly 2/3 of total revenue.

As I mentioned earlier, Q3 revenue benefited from material deal value increases that we do not anticipate will be replicated in Q4. Also, this quarter, we will be lapping the anniversary of our dataxu acquisition, along with certain content distribution launches.

Like past holiday seasons, we plan to keep Q4 player gross margins close to breakeven, while we anticipate Q4 Platform gross margins to be between the mid-50% and 60%, which is similar to levels seen in Q2 and Q3 this year. We anticipate the sequential growth in operating expenses from Q3 to Q4 to be in line with last year, driven primarily by head count and sales and marketing growth. As a reminder, quarterly operating expense levels from Q1 to Q3 remained relatively consistent in part due to steps we took at the outset of the pandemic to slow the rate of growth of our OpEx.

In summary, we are very pleased with the overall resiliency of our business and with our outstanding Q3 performance. Despite the macro uncertainty, we remain confident in our ability to continue to grow our business into the future and believe the pandemic has only accelerated the long-term trend towards all TV being streamed.

With that, let's turn the call over for questions. Operator?

(Operator Instructions) Our first question comes from Ralph Schackart with William Blair.

Steve, just wanted to get some perspective on the Q4 guide, if I could. Obviously, mid-40% growth is great. But just sort of looking relative to what you posted, 73% growth in Q3. Just curious if there's just added conservatism there. I'm sure there's a big element of the outsized portion of the deal that got recognized in Q3. So just any extra color there. And if you could perhaps give some quantification or some more color on how large the outsized portion of the deal was in Q3 would be great.

Yes. Rob, thanks for the question. Yes, as you mentioned, I mean Q3 was an outstanding quarter. It's a confluence of very positive trends, record player growth, TV sales were great with our partners, and then strong performances on both the advertising and the content distribution side, including calling up on several of our deal models, the lifetime values there.

In terms of Q4, I think the important thing to just focus on is there is a lot of uncertainty and variability that remains out there in the macroenvironment, everything from COVID-related resurgences in the U.S. and around the world to the shape of the holiday season as well as consumer spending levels. So that's -- these are all things we're tracking, which is why we're not giving a formal outlook.

So certainly, you referenced some of our color. I mean we're very happy with the trajectory and the resilience of the business to date. And so we're cautiously optimistic on the holiday season and think it could look similar to the last couple holiday seasons, which were in the mid-40% year-over-year growth range on a revenue basis.

As I mentioned in my remarks, we don't anticipate significant deal value changes in Q4, although that's -- obviously, we do that analysis every quarter.

The other thing I'll note just that's factored into those thoughts is the fact that in Q4 we are lapping the dataxu acquisition, which came on board mid-Q4 as well as launches of some certain new services in Q4. So those are also factors that are -- that were included in our thinking there.

Our next question comes from Laura Martin with Needham.

Great results guys. Scott, let's start with you. I'm very interested in this sentence, the question I get the most is dataxu. Why did they buy it? And why does it matter? And I'm very interested here in these words that say, "focused on new products like incremental reach guarantees." Could you talk about how dataxu is giving you new products that weren't available before with this one new platform?

And then for you, Anthony, you guys say in the letter, 75% player growth year-over-year. Does that imply that our mix of TV ads is now pivoted towards player -- away from the 50-50 mix towards players? And are soundbars having an influence on that extra outperformance or not?

Laura, Scott here. Thanks for the question. I'll talk a little bit about the context around dataxu and then provide some broader color on the newer ad products you referenced. We bought dataxu really to bring the same unique advantages we've had in the media business to a broader platform to a broader scale.

Since buying dataxu we've rebranded it and relaunched it as OneView, the OneView Ad Platform. And we've been layering in a bunch of new features, including, for example, as you referenced, the ability to manage reach and frequency broadly across the marketers. Complete buy, so whether they're buying media from Roku or media from a publisher on Roku or media off Roku entirely, they can manage that and orchestrate that all within OneView and leverage our identity and data in -- at the same time.

The incremental reach product that you're referencing is actually something we took into the upfront, and it was a -- we stepped it up. For several years now, we've been highlighting our ability to measure reach and frequency across linear and OTT, which is a key trigger for marketers to choose to invest into OTT. This year, we stepped it up and offered guarantees in the upfront. We would guarantee that we would deliver true incremental reach over linear to marketers.

So overall, tons of progress on OneView. Significant media spend through OneView occurring in the quarter year-over-year, and we're really just getting going. We've got brands like DraftKings, for example, who is a big sports spender, had to shift budgets out of TV as sports was canceled and delayed, has moved a significant portion of their budget into OTT and is yet now managing most of their spend, their cross-channel spend, through OneView. That's just one example.

Lexus, the national team, is really leaning into our unique capabilities, adding its on-the-home-screen sponsorable experiences and then using OneView to reach consumers off Roku for that one-two punch of a cross-platform reach. So overall, we're very excited about progress we've made with OneView.

Anthony, do you want to take the second question?

Yes. Laura, so I think the stat you're referring to is player unit sales were up 57% year-over-year, which is a great number for us. In certain international markets, like in Canada and the U.K., players sales actually doubled year-over-year. And then Roku TVs also had an extremely strong quarter. They continue to perform very well.

So we saw -- and then audio products like we just introduced, the Streambar, and we have some other audio products in the market, those are still not -- still a lot smaller than player unit sales but those are growing as well. So I think we're seeing our active account growth engines on all fronts doing really well. And the result was active accounts in the quarter were up 43% year-over-year to 46 million, which is great.

And we also -- I mean we also sell lots of units that don't result in a new active account but they result in more Rokus going into existing Roku households, which strengthens our relationship with those customers as well.

Our next question comes from Michael Nathanson of MoffettNathanson.

I have a couple. Whoever wants them, please do. Can you talk a bit about The Roku Channel this quarter? It looks like streaming hours slowed down a bit. That makes sense, but Roku Channel picked up a ton of share. So what's behind that?

And then I thought it was interesting that you guys announced a deal with Amazon to move Roku Channel across to their platform. What does that do to perhaps your content strategy and your partner's willingness to work with you? And do you see one day maybe moving to more original content as you build that Roku Channel further?

Michael, this is Anthony. Yes, so The Roku Channel is doing extremely well. It's an important asset for us. And reach doubled in The Roku Channel in the quarter. But I think Scott will be best to answer most of your questions.

Yes. It really was an incredible quarter for The Roku Channel. Active account growth now reaching active accounts with about 54 million people, double the pace of an already fast-growing underlying platform in terms of hours and reach, growing faster than any other top 10 channel. Some of this was driven in part by the launch of our EPG, our live channel grid, which we began rolling out in June. It's a great product, a great way to channel-surf.

All this growth really accrues to Roku's benefit but also, of course, our partners who are participating and putting content into The Roku Channel. So partner earnings within The Roku Channel more than doubled in the quarter. So overall really strong performance for The Roku Channel.

As -- in answer to your question about sort of content and off platform, as The Roku Channel grows and generates even bigger and better financial outcomes for our partners, it just attracts more partners, more content depth, more verticals. It opens up yet more possibilities for programming for us as a platform. So it's a real vehicle for us in terms of the value that it creates for our users, for Roku and for content partners participating in it.

Regarding off-platform syndication of The Roku Channel, our primary focus as a company remains on our platform where we have the greatest advantages, the marketing tools, the reach, data, but free, especially AVOD is a powerful recipe. And so where we see an opportunity to take The Roku Channel on to other platforms, we will explore it.

Our next question comes from Justin Patterson with KeyBanc.

I was hoping to get a finer point on the shopper data program with Kroger. How has the uptake been with CPGs? And how many other opportunities like this, could you envision out there down the road?

Scott, do you want to take that?

Yes. Well, of course, CPG as an advertiser category is an enormous category. And especially, I think COVID has taught us just how much consumers are willing to move their shopping online. The Kroger deal has been a very strong deal for us in terms of signaling what we're capable of given our scale and our data. We mentioned in the shareholder letter that Snyder's, for example, saw a 250% return on ad spend. We have a bunch of brands now in and leveraging that program.

And there are lots of opportunities. It is a big market. The key for us as a platform is that first-party customer relationship and our reach into our platform with advertising. So we anticipate lots more to come on that. CPG became our fastest-growing vertical in the quarter, just as an indicator of how much potential there is.

Our next question comes from Jason Helfstein with Oppenheimer.

So Steve, just can you go into a bit more detail, I mean you called out that you had an -- you recognized an outsized portion of the content distribution lifetime value in the quarter. I think second quarter kind of also benefited, maybe you didn't call it out as much in the letter, but I recall, I think there was something in the Q about it. I mean just help us understand kind of what those triggers are, because it feels like this kind of is actually a pattern because it's such a strong part of your business.

And then just with one of you, I mean how do we think about the sizing, if we think out over the next 2 years? I mean you gave some commentary in the letter with DraftKings, but I mean kind of how meaningful could that become over the next 2 years?

Steve, you want to take the first question?

Sure. Jason, thanks for the question. Yes, in terms of the content distribution deal models, and this is a similar quarterly view that we have in our analysis that we do for other types of deals as well. We're in 606 accounting. We are valuing the lifetime of the deal and the content distribution deals are generally a year or 2. And so when we see trends, what we're trying to better understand is whether those trends are sustainable and whether we should reflect them into the future periods. And in those cases, that can increase or decrease the overall level of the deal if they're positive or negative.

In this case, in Q2, we obviously had trends that were in the early days of the pandemic in the lockdown. At that time, we didn't feel comfortable creating -- extrapolating those out too far. But certainly, with more time, what we're seeing is a sustained acceleration in active accounts, as Anthony mentioned, driven by very strong player and TV sales as well as just very strong consumer demand for not only ad-supported viewing but also subscription services. The premium movie rental business that was pretty nascent has increased with some titles -- kind of notable titles there.

And so in this case, we reassess these material deal models every quarter, but what we had was a portion of those increasing in deal values. And so that's reflected in that -- in the quarterly results because when you reflect an increase in deal value, an outsized portion of that is impacted to the current quarter.

The results can be lumpy with that. And certainly, as a reminder, that is increasing our assumptions for not only the current period but in future periods as well, which is why that can be lumpy on the quarter-to-quarter. So again, I think for us, I would say that, that process of looking at the deal model is definitely a quarterly part of the business, but it can go up or down. And in any given quarter, a lot of them kind of stay the same.

Jason, let me take the second part of your question about OneView. It's a great question. Let me just -- without getting too frothy here, let me just take it in parts to highlight how big an opportunity this is for you, so -- for Roku.

So on the Roku platform, we are fast growing, of course, as an ad sales team. OneView allows us to access the yet additional, the still significant majority of impressions that are not sold by us. They're sold by publishers on our platform. It enables us to enable a marketer to use the same identity and data when buying elsewhere within the Roku ecosystem in the same way that they have been when they're buying media from Roku.

But of course, as big as we are in OTT, there are other -- there's other video OTT media to be accessed outside of Roku. So that's kind of a part 2 in terms of additional opportunity. And then, of course, there's omnichannel, there's desktop and mobile and tons of use cases of marketers who reach a user online and then want to retarget them in OTT, reach a user in OTT and then want to follow through on desktop and mobile. The opportunity here is really very significant for any marketer, and TV and OTT is a centerpiece in their marketing strategy.

The key success factors, in our view, the right to win in our view as a platform and as a DSP really comes down, in our view, especially in this world of cookie challenges, device ID challenges is that first-party customer relationship, a platform at scale, advanced targeting and measurement, unique ad products. These are all the things that we think make OneView strongly appealing to our clients.

Jason, this is Anthony. Let me just add a couple things. I would say from my point of view, the most interesting thing is that even though we have a nice robust ad business, the vast majority of advertising spend is still in linear television, it's not -- it's still not in streaming.

And one of the interesting things in the quarter was the fact that advertisers are increasingly seeing streaming as something they need to start allocating a bigger portion of their budgets towards. So it's still -- there's still a lot of opportunity because, of course, all those ads are going to switch to steaming eventually, and we think they're all going to be programmatic eventually as well through platforms like OneView.

Our next question comes from Thomas Forte with D.A. Davidson.

So I wanted to ask about the maturation of the AVOD market. So you talked about the changes to the upfront. I thought that adding Roku Channel to Amazon Fire TV was a real milestone for the company. And on its earnings call, Amazon made very positive comments about the AVOD market and its own efforts. So I wanted to, in particular, talk about the maturation of AVOD, but 2 things in particular. What do you see as the opportunity for political ads on AVOD? And how did they perform in the quarter? And then what are your current thoughts on your international opportunity for AVOD?

So this is Anthony. I'll just say that AVOD is obviously a huge opportunity. We're very bullish on it. But I think Scott's best equipped to answer those questions.

Yes. I mean, AVOD is vital to the OTT business model. You framed the question as maturation as though we're reaching maturation. I mean I think it's really still an extremely fast-growing and very high-potential market. We see that in the pace of growth of The Roku Channel itself. We highlighted some stats in the shareholder letter.

And I think if you look at new services coming to OTT, a majority of them are actually hybrid business models. They're subscription-based, but they're keeping the price down and they're augmenting their revenue model through advertising. I think also consumers have a somewhat limited budget for the number of services they're going to sign up where they're looking to save money as they leave the pay TV ecosystem. All to say, it just makes ad-supported experiences in the OTT environment that much more important.

You asked about political. We had a strong participation in that category. Our capabilities are a key differentiator for political marketers. But that said, it was a relatively modest contribution to an otherwise really big quarter for us in advertising. It was an important segment, a newer segment for us as a younger ad business. A great segment to participate in, but certainly not anywhere close to a large share of our overall ad business.

And then on international, I can just say a few words on that. So if you just think big-picture what is our international strategy? Well, a big part of it is -- a lot of it is to replicate -- most of it actually is to replicate what we've done in the U.S. internationally. And in the U.S., we started by just focusing on building active accounts and then as we started to reach scale, layering on monetization options. And that's what we're doing internationally as well.

So we're focused primarily in this phase on building scale in the key markets. And then as we start to reach scale, we'll start to layer on monetization options. And that started, but it's still pretty nascent. For example, The Roku Channel is available in the U.K. and it's available in Canada. But that's our overall strategy.

And then I would say, thinking about AVOD, I mean, SVOD is fairly -- pay TV is popular in the United States, but in almost all international markets, AVOD is by far the dominant business model. Free is very important worldwide.

Our next question comes from Vasily Karasyov with Cannonball Research.

So I've been hearing some concerns from investors about the press reports about Walmart selling TVs with Comcast operating system. So it does feel a little bit like the scare from last year with the Xfinity Flex. But I would appreciate your view on how this potentially can or cannot impact you. And then I have one for Steve.

Sure. This is Anthony. So we are the #1 TV OS in the United States and now in Canada. And we've achieved that position despite a very dynamic marketplace with intense competition from large tech companies and large TV companies as well. And there's a few reasons we think we're successful or the most successful in the TV OS business. And one, of course, is just the Roku OS itself, our software. We've built a platform that's purpose-built for television. It's the best software platform for television. It has lots of advantages for the entire ecosystem and for manufacturers, in particular. It allows them to build TVs that cost less. And so price -- it gives them a pricing advantage. So that's the big advantage, and we continue to refine that advantage. But we do -- but so software, of course, is a big part of what we do.

The other thing I would say is we have just world-class TV engineers. We have a great talent base. Those engineers come to work every day with the goal of making streaming better for our customers. We're very focused and we have a great engineering team.

And then the Roku -- I would say, the Roku brand itself is a very strong brand. And for customers, it stands for streaming and great value. And so for those reasons and others, we've been very successful competing in the TV space. It's a very competitive market. It's been competitive for a long time. But we're the #1 streaming OS, and I believe we're going to maintain that position.

In terms of Walmart, I'll just say a few words. I mean Walmart is a big retailer, a very strong partner of Roku's. We have a great relationship with them. They sell millions of Roku players a year. They sell millions of Roku TVs for various Roku OEMs, including TCL, Hisense, RCA, Philips, JVC. We build -- we helped them build onn-branded, which is their house brand, Roku TVs -- smart TVs. And that's a business that's been growing extremely well for them. So it's a great partnership, and it's a long-standing partnership, and we've put a lot of work into making sure that it stays strong.

All right. Steve, I have a question about OneView revenue recognition. I think on the last quarter call you said that you recognized at least the majority of it on a gross basis and that had an impact on the margin in the quarter. And then you said that going forward, you will be recognizing an increasing proportion of it on a net basis. So I was wondering if there was a change in impact from Q2 to Q3 from that on the margin and the Platform segment.

Yes. Vasily, yes, thanks for the question on that. Yes, you're right. I mean what I said was that the historical dataxu business there are effectively 2 parts, the kind of typical DSP business, which was net revenue treatment; and then a historical dataxu managed service business, which was on a gross revenue basis, and that as part of -- as Scott mentioned, rebranding and sort of integrating the operations to the OneView, we were shifting people over to new service agreements, and those were kind of net revenue treatment.

So that is an ongoing process. That wasn't a significant factor in the quarter. The increase -- or the expansion of the platform gross margin, both on a sequential and a year-over-year basis, was largely a result of just strong margins in both the advertising and content distribution business. So that trend will continue over time as we continue to shift more of the existing client base over to the new service agreements.

Our next question comes from Steven Cahall with Wells Fargo.

Just wanted to dive into the ad revenue a little bit. How do we think about the funnel for your ad buys between self-service, upfront deals and maybe buys through agencies that might be using their own DSPs? And I was wondering if at some point, you might look at restricting some of your high-value inventory to dataxu.

And then could you give us some color as to how much your ad revenue is more like these brand marketing campaigns or upfront campaigns versus how much is more ad hoc reach extension? And how do you kind of convert somebody from a reach extension customer to something more strategic?

Scott, you want to take that?

Yes, sure. Well, I guess, Steve, in answer to your question, we're here to sell how marketers want to buy. It's why we've got a media business and a DSP. It's why we also support third-party DSPs who want to be buying on the platform. And it's really being able to participate in all those different ways of buying that allows us to keep expanding the segments that we serve in the ad market.

We are strongest in all of those segments when a marketer comes to us and wants to take advantage of our deep first-party customer relationship, our identity, our data on users, our ability to optimize in a way that third-party ad tech cannot. And we can employ that value proposition, whether they're buying media from us, which, for example, many TV marketers do. They are IO, guaranteed Nielsen demo-driven often when they first come to the platform, all the way through to programmatic buys transacted through OneView.

Our long-term view is that most TV advertising, this will take years to play out, is transacted programmatically. And referencing back to Laura Martin's question, that's why we made the move in acquiring dataxu, incorporating and relaunching OneView is so that we could enable marketers to leverage our unique assets or scale regardless of how they're buying.

So I hope that answers your question. I think maybe Steve wants to take the second part of your question on the -- more color on the revenue front.

Yes. It's Steve. In terms of revenue recognition, that's largely on the ad business related to when we run the ads. There can be some differences in the models themselves. So there can be a bit of timing, but it's fairly straightforward.

Was there a particular question within that?

Yes. I'm sorry. Maybe I tossed it to you, Steve, and I think -- and Steve Cahall, maybe we need clarification. But I think your question was, how do they come to the platform? Is it -- are they starting from a reach extension strategy or some other case?

I would say, what -- for TV marketers, their primary reason to come to Roku is the loss of reach within the linear pay TV ecosystem. There are just whole classes of consumers that are simply no longer reachable through linear television ad buys. About half of TV time for adults 18 to 34 has moved to streaming. Those users are just not reachable anymore. So that's typically what brings a marketer into the Roku fold is the desire to keep reaching them on the big screen.

That said, with OneView, we have the ability, of course, to help them run an omnichannel campaign to reach a user first on the television and then retarget them off-device. And we've also seen significant growth in the performance marketing, D2C segment of marketers who have social budgets, they are optimizing to a bottom-funnel outcome, a site visit, an add to cart, a product purchase, a shopper data outcome. And those budgets come to us really from a different budget line at the client, and they're taking advantage of our deep data assets and ability to optimize those outcomes. So I wouldn't call those TV dollars as much as performance marketing dollars. It's a significant growth segment for us.

Our next question comes from Jason Bazinet with Citi.

The buy-side is very focused on the pandemic sort of causing a big shift in streaming and digital ad dollars. And you guys referenced in your letter some of the benefits you got as the linear TV ecosystem got gummed up with COVID. My question is, how do you think this plays out once the world goes back to normal? Do you think prudent investors should be sort of moderating their growth expectations sometimes next year? Or do you think the secular trends and all the steps you're taking will just allow you to power right through that without a deceleration?

This is Anthony. From my point of view, I think what we're seeing is a trend towards streaming that's been building for years now. And I think there's a lot of things that are causing it to accelerate and to continue to be a big trend, whether it's better customer value, more options, the ability to save money. There's just lots of content. The content that's available on streaming has continued to build. And at this point there's really nothing that you can't get by streaming. There's -- in fact, there's more content streaming than there is on traditional pay TV now.

So I think the pandemic certainly helped accelerate that trend and maybe tipped over some things a little bit, but it's a trend that was happening anyway and will continue to happen for some time because everyone doesn't stream yet. So I think we've seen strong growth. And it may moderate a little bit, but I think we're going to continue to see strong growth.

I don't know. Scott, did you want to add anything to that?

Yes. We're not going back to the way it was, to be clear. I mean I think COVID triggered a lasting durable change in how CMOs and marketers are thinking about their TV ad spend. We -- in Q3, we saw a 17% drop in linear viewing. Roku was up 54%. 92% of Roku cord-cutters are very satisfied with their decision to cut the cord and aren't planning to go back.

So I really think this is a one-way transfer function. We don't go back to the older spending patterns because the audience isn't there. Marketers need to follow the audience into OTT, and they stay because of the enhanced capabilities. So I think that's really what all this transition is helping teach marketers is that there's a better way, there's a more robust toolset in OTT. And we see it in our own stats, nearly 100% retention among the advertisers who spend over $1 million. These budgets are not flowing back to television, traditional TV.

I mean another proof point of that, I think, is just media companies, how they're increasingly reorganizing around streaming.

Our next question comes from Jeffrey Rand with Deutsche Bank.

Congrats on a good quarter. How are you guys thinking about advertisers' willingness to spend as the pandemic continues? And are you seeing certain industries still holding back on advertising spend?

Scott, you want to take that?

Yes. I mean, certainly, there are some segments that are still slow to return. The ones you might expect, travel, quick-serve restaurants, things like that. But we saw real strength in most other segments. I mentioned CPG growing as quickly as it did. So I think that the big news here is just the shift, the realization that COVID has helped accelerate the need to reinvest in OTT. And of course, we're still early. Proportionate to viewership, we're nowhere close to capturing the kinds of budgets that are spent in television. But the transition is very clear and the ultimate cause of the acceleration of spending in OTT.

And our next question comes from Mark Zgutowicz with Rosenblatt Securities.

Maybe a 30,000-foot and a 5,000-foot question. On the higher end, where is the measurement discussion trending in your discussions with linear buyers today? Is there any tangible measurement solutions that are on the horizon that you see more linear buyers talking about and essentially that's loosening -- would loosen the broader bottleneck? Obviously, you're seeing nice flow in linear today, but the broader bottleneck I'm talking about.

And then a little more micro. To what extent did political or its crowding-out effect on nonpolitical spend, particularly in October, have on your pricing dynamics? And where is that trending now?

Scott, that's another one.

Yes. Let me take the measurement question in parts. So one of the reasons we invested and partnered with Nielsen years ago to enable demographic measurement in our platform is so we could provide marketers with a common currency across linear and OTT. So -- and we've continued to add to that capability the ability to measure reach and demographic holistically across linear and OTT. We launched that capability years ago.

We followed that with optimization using our data to target specifically cord cutters and people who are harder to reach in linear. This year, we followed it up with a guarantee that we would not charge marketers if we reached somebody who had already been reached in linear. So I think we've been the market leader in providing TV advertisers with a common currency and frankly, a bridge from linear television into OTT.

I think the really exciting stuff happens after marketers start to invest in OTT. OTT is a native digital platform that enables measurement up and down the funnel. It is an OpenIAB VAST-based platform. And so this enables, on any given ad, multiple simultaneous kinds of measurement. We're not headed towards a world where there's a single currency. Sometimes when people talk about measurement challenges in OTT, I think it's because they're harkening back to a day when there was one currency. We're not headed towards a world where there's one currency. There will be many. And OTT will be all the more stronger for it because marketers of all shapes and sizes will be able to pick and choose the currency that they care most about, which best fits their tactic.

And then you asked a question about political ad spend. As I mentioned earlier, it was a meaningful and a relatively new category for us. We weren't particularly active during the last congressional or presidential cycle. And so it's an area where we've put good energy into. And I think our platform offers powerful capabilities and was very attractive to political advertisers.

That said, it wasn't a huge portion of our overall business and so didn't have a substantial effect on, like, overall pricing or other dynamics.

And our next question comes from Mark Mahaney with RBC Capital.

Sorry, I came in a little late. You may have already talked about it, so you can be brief. International, just give us an update on where you are in terms of rolling out both player and platform revenue, developing both player and platform revenue in international markets.

Mark, this is Anthony. I'd be happy to talk about international. We did touch on it briefly, but just to summarize. Streaming obviously is a large international opportunity and phenomena and we're making good progress. And I'll just give you some examples. We introduced players in Brazil, for example, following up from our TV launch several months ago in Brazil, which is off to a great start. We introduced the Streambar, which is a new 2-in-1 product that I'm super excited about. The Streambar basically is a streaming player that also makes the sound better on your TV. And we launched that product in the U.S., but also in U.K., Canada and Mexico.

And so in certain markets like in the U.K., player sales doubled in the quarter, which was good progress. Canada, player sales doubled. And Roku TV, it's been #1 for a while in the U.S. in terms of operating systems, smart TV operating system share. It's now #1 in Canada. So good progress there.

Then I would say the strengths that have made us successful in the United States are working as we use them in other markets as well. If you look at our quarterly results, 43% year-over-year active account growth, which was great. I mean, I thought that was incredible given the size of the base that we have to go on. But that active account growth came domestically but also there were strong international contributions to that account growth as well.

Our priority is, like I said, is to focus on active accounts first, build scale and then start layering on monetization. So really that is still -- in most markets scale is still our primary focus. We are starting to -- we've got small ad teams in some markets, testing the market. We've launched The Roku Channel in a few markets like Canada and U.K., but the monetization really will follow, and we're focusing on scale first.

Our next question comes from Richard Greenfield with LightShed.

I got 2. One is really a big-picture for Anthony. You've seen this wave coming for many ways -- many years and have been predicting it. The challenge right now for big media companies, a lot of your partners, is that sports is not only one of the key parts of that whole ecosystem but it's also one of the biggest drivers of advertising. And I think a lot of these big media companies are struggling to figure out an economic model for sports to shift to streaming.

I guess do you have any advice? I mean there's so much money and so many ad dollars tied to sports, and it's sort of the last thing holding up the entire multichannel bundle. How does sports -- how do they get sports into the streaming world? Should they bundle it into existing channels? Should they have true D2C sports streaming? Like how would you advise them as they think about sports? And then I've got a follow-up on The Roku Channel.

Sure. Well, I mean, I think -- well, first of all, you can get sports in streaming. I mean it's still part of the bundle, like you said, you have to sign up for a virtual MVPD usually. I mean there's complicated deals, multiyear, 10-year deals in sports that I'm not really an expert in. But if I could wave my hand, I would -- sure, I think the best thing for sports would be for them to start selling SVOD services like for regional sports networks as an SVOD service or the NFL as an SVOD service. I mean that's what -- that's, I think, the way consumers want to sign up and consume sports in a streaming world. That's how they sign up and see other streaming products on à la carte basis.

Yes. Rich, I'll just add to Anthony's point. Great question, and I don't think we have a short answer for it, but just a couple of stats that you might find interesting. January to September, so sort of bridging sports going away and coming back, heavy sports fans on the Roku platform dropped their linear viewing 26% and grew their streaming 17%. So even as sports came back, the viewership didn't come back.

Among heavy NFL viewers, we saw a 29% drop in linear and a 16% increase in streaming. So the disappearance of sports during COVID kind of accelerated the pressure on that vertical in the streaming. And the viewership, and I think ultimately, the dollars are not going to come back in the same way long term.

And that's exactly what we're seeing in terms of the impact on ratings, and that's why I feel like they have to figure it out. So I appreciate that answer.

The other one, Scott, you've talked before about the fact that roughly 3 people per -- when we look at The Roku Channel numbers you released in terms of viewers, roughly 3 people per household is sort of the metrics so when you try to think about it relative to your active account base. So if I look at 54 million people who watch Roku Channel or reach Roku Channel, that's like 18 million households or roughly 40% almost of your base of Roku active accounts. What do you need to do to get that base up from 40% to 75% or 80%, let alone 100%? Like, is it purely the content? Is it awareness of The Roku Channel? Like how do you bridge that gap?

I think Scott can add. But...

Go ahead, Anthony.

I was just going to say like that's -- getting to a big number like 100% penetration is obviously our goal, and that's what we've been working on from the beginning. I mean Roku Channel started out tiny, started out with 1,000 pieces of content, very small number of users. There's this virtuous cycle where we promote it, we onboard more content, we get more users, we get more advertisers, then we get more content, we do more promotion, we get bigger titles.

And so that's going -- we're just going through that virtuous cycle, and it's getting bigger and bigger as a result. It's the fastest-growing top-10 app on the Roku platform right now. So it's -- I'm confident we can get -- we can keep growing The Roku Channel. It's still pretty small compared to where it can be.

And when you say top 10, Anthony, is that minutes watched or times used? What's the best way of thinking about that metric?

Reach and hours.

Yes, both.

Yes, yes.

Our next question comes from Ben Swinburne with Morgan Stanley.

I want to ask Scott about social budgets and whether you think you're taking share there. Some of your peers or competitors this earnings cycle have talked about a shift in social spend onto other platforms that are viewed as more brand-safe. And it would seem like that could be or is and could be a huge opportunity for Roku. So I'm just wondering if that benefited the quarter, something that you're focused on even further kind of differentiating your platform versus some of those massive digital budgets out there.

And then I wanted to ask Steve. Just on the guidance for Q4, if I'm doing the math right, I think you're guiding to platform revenue growth in the mid-50s or so. Is that deceleration just because you don't have the sort of 606 impact that you had in 3Q? Just looking for any color there.

Ben, I'll take the first part of your question. Yes, I mean I do fundamentally think that we are able to compete for social budget. We don't always know which pocket the marketer is drawing the money from. But we have heard clearly that both because of exogenous factors, pricing pressure, what's going on in the social platforms as well as like maybe, more importantly, the fact that we have the capabilities to optimize to the outcomes that these marketers are focused on. I mean TV has just never been able to compete there. And so that's why our performance marketing segment, while still early, is growing much faster than our overall ad business.

And I think, by the way, that's one of the most interesting things about OTT advertising is it isn't just going to be a branding media or performance media, it's going to be both all at the same time. There really isn't an analog in any other media out there. Linear TV is going to remain a top-funnel branding media. Social media is very much a bottom funnel. OTT has the opportunity to be both, and it just is going to make for a very rich ecosystem as well as the opportunity to tap into very, very large budgets, not just TV, but also digital.

And Scott, just to follow-up quickly on performance. How are you doing attribution? Is that something that you're able to tie back to mobile cross-platform to really connect the -- close the loop for advertisers in performance?

Yes, it's a mix of both first-party and third-party. We have a wide array of third-party attribution partners. We have our own pixel. And so marketers can pick their preferred solution in OneView. We can automatically optimize to pixel placed on an advertiser's website according to the CPA, a performance target that a marketer provides to us.

And our last question comes from.

Ben...

There was a part 2 on that question, right?

Yes. Sorry, operator, there's a second part to Ben's question there that I want to hit. Yes, Ben, so just on Q4, just for the good of the order we didn't provide any formal guidance. But in terms of some of the color we gave in terms of how we're thinking about how the quarter could look barring in any of these significant external factors that could impact the short term. What I would say about the platform revenue is, certainly, we talked about how the content distribution deal revenue recognition can be lumpy quarter to quarter. The other factors that we talked about that are platform-specific was specifically lapping the dataxu acquisition, which came into the results kind of mid-Q4 of last year as well as we're lapping some launches of new services from Q4 of last year.

And our last question comes from Ruplu Bhattacharya with Bank of America.

Congrats on the strong quarter. Just 2 quick ones, if I can squeeze them in. Just one on The Roku Channel again. So you've done a great job increasing the reach of the channel. Steve, if you can just talk about how the ad load is trending on the channel. Are you happy with the ad load as it is today? Is there room for growth? And can you address the CPM? Do you think you can maintain the CPM? Some investors think that over time that has to go down, but just your thoughts on that.

And the second question is just on international expansion. You've done a great job in the U.S. You've got great distribution relationships. When we look at internationally, that's a more fragmented market. So are you having to hire more people on the ground who have these relationships? And then in that vein, how should we think about OpEx going forward?

So I think the first question was actually for Scott on TRC and ad loads and CPM.

Yes. I'll take that first question. We're quite happy with our recipe of half the load of linear television. We think it strikes the right balance between a great consumer experience while also providing strong monetization opportunities. So we're not planning to tinker with that. I think there's lots of upside potential in terms of how valuable we make that inventory as well as opportunities beyond 15- and 30-second spots. We've done some work in the market with pause-ads and other ad units. So I think it's a great consumer experience as delivered today.

Regarding your question about CPMs, OTT is just a better, higher performance product. And so I think it does reasonably command higher CPMs relative to, say, linear television.

That said, I don't think -- going back to the prior question, I don't think there is long-term one-single CPM to rule them all. In long run, in OTT, lots of different marketers, top to bottom of the funnel come into the ecosystem and compete in a common auction for ad inventory. And just like you see in social media platforms, you don't end up with a singular CPM across all users and all content. That's one of the things I think is most interesting long term about the OTT media.

There was a second part to the question. I think Anthony would take it.

Yes, I can take the international question. So first of all, I'll just reiterate that we're making good progress internationally. And yes, it is a fragmented -- more fragmented market because -- and I would say in big part because Roku was not in those countries yet. And so as we've entered countries and brought the full portfolio of our assets, we're gaining market share and consolidating a lot of the market around Roku.

So for example, Canada, I mean, a small country, but it was the first one we went into after the U.S., and now we're #1 for TVs, and our player market share is very high. We have an ad sales team there. So it's starting to look a lot like the U.S. for us.

But our -- in terms of investment, international is one of our key strategic investment areas that we view. Obviously, it takes people to do it well. The way we focus is we pick our focus countries, which we keep expanding. And then we make sure that we have local content. So we hire local people to do local content deals. We make sure we have local retailer relationships, getting into local channels, local distribution channels. We bring Roku TV to market, which requires a lot of localization but also, in many cases, specific features that are regulated or required by the government, so we add those. We bring our players to market. We bring our audio for TV products to market. So we bring the whole portfolio of products and offer just a very complete and compelling solution. And that's -- and then our market share starts to grow. So that's what we're doing. It's working well.

And at this time, I'd like to turn the call back over to Mr. Anthony Wood for any closing comments.

Thank you, operator. I'd like to close by reiterating how pleased we are with this quarter's results and wrap up with 2 concluding points.

First, reaching 46 million active accounts at 43% year-over-year growth rate shows Roku's momentum. It reveals Roku's scale relative to traditional pay TV providers who have continued to lose subscribers during the quarter. The pandemic is accelerating the shift away from traditional pay TV. I'm confident that the shift to streaming will be permanent for the vast majority of consumers.

Second, the strength of our financial results illustrates the leverage of our business model. The leverage of our business model may deliver over longer term as we continue to execute our strategic plan.

I hope you and your families stay safe and enjoy the upcoming holidays. Thanks again for joining the call, and happy streaming.

Ladies and gentlemen, this concludes today's conference call. Thank you for your participation. You may now disconnect. Everyone, have a wonderful day.