Roku, Inc. (ROKU) Earnings Call Transcript & Summary

So I'm Laura Martin, and I'm a senior media analyst at Needham & Company. And I'm here to welcome Dan Jedda, the CFO of Roku. Thank you for being on my stage. I appreciate you being here.

It's my second time. Maybe you'll invite me next year.

Yes, maybe I'll invite you next year be good. So let's talk about -- to sort of level set and say tell us what Roku does and then tell us with your full year outlook, which you've given, was Frndly part of your outlook when you provided it in February because you made this acquisition, maybe you should tell me about that, and if it's in guidance already.

Yes. Happy do that. First of all, thanks for having me. Happy to be here. Love coming to this conference. So it's really a pleasure to be here. So Roku is the largest streaming platform in the U.S. by hours and by broadband penetration. And basically, we are that. We are the operating system on connected TVs, that basically act as the streaming platform. So what we do is we -- or the software that is embedded in the smart TVs, we also have a player that makes any TV that plugs into a port that makes any TV a smart TV with the latest technology. And then we run, of course, the operating system. And we have -- we're #1 in the U.S. I mentioned that we're #1 by a long shot in the U.S. We're #1 in Mexico. We're #1 in Canada. We are growing in Brazil. We're growing in what we call the rest of Latin America. We're expanding in U.K. So we are global. And the way our business monetizes, which is not often well understood, but it is important is we don't monetize the actual sale of the device or the sale of the operating system, the software. We monetize after the sale in all the activities that you do on the platform itself, whether it's through subscriptions, whether it's through content distribution deals or whether it's through advertising where we sell ads across our own app called the Roku Channel. It's an experience. It's not an app, it's an experience or we get -- we often get inventory as part of the content distribution deals from other apps, and we monetize that as well. To your second question, the acquisition of Frndly, we're very excited about it. So we recently purchased Frndly, which you'd call a skinny bundle, if you will, a virtual skinny bundle of about 50 channels, including A&E, including the history channel, including Hallmark. These are very popular channels in a core group of people that watch them. And we purchased them, and we will now -- they'll operate independently, we will now help them drive that business by using our platform to expand their subscriber base and grow over time. And why that's important is we're investing big in our subscription business, which is one of the ways we monetize the platform. Where I wasn't clear, and I do thank you for giving me the opportunity to make it clear, it was on the guidance and what was included, was Frndly included in the guidance? When we gave the guidance in Q1, when I gave guidance in Q1, we guided to $3.95 billion of platform revenue. Frndly was not included in that guidance. Multiple -- we have many initiatives that are ongoing, new product innovations both on the subscription side, on the ad product side. We have new revenue streams, of course, coming, but Frndly was excluded on that. And then when I came and gave the Q1 results and gave the rest of the year guidance, during last month, I said -- I didn't imply, I said that Frndly is now included in the forward-looking guidance. And the reason why that was is simply because we said there is so much macro uncertainty in H2. Many companies, they're pulling guidance. Many companies are not even giving the current quarter's guidance. We did give Q2 guidance. We also gave or we stated our full year guidance. And we simply said we just don't have enough visibility to call it up right now, inclusive of Frndly. And we're going to wait and see. We said we'll give more data as we go through Q2 and into Q3. So the clarity there is we just simply haven't raised our guidance with Frndly until we just have better visibility into the second half of the year, given all the macro. We don't see -- we'll probably talk about it. We don't see a big change in demand in Q2. That's good. We see a change in the way advertisers are buying but not in overall demand. I'm sure that's something we'll get into. But I did want to clarify the Frndly acquisition and how that plays...

That's big. Isn't it tiny?

So first of all, we don't give out the size, but it's not -- I would say that it will be adjusted EBITDA accretive. It will be -- without giving out the size, you guys can do the math on this. It does have hundreds of thousands and getting close to 1 million subscriptions. So it's sizable.

And what's the average price?

It depends on the different tiers. It starts at $6.99 and goes up from there.

Okay. I just wouldn't have thought it was very big compared to your $4 billion of advertising. But I want to stay on this point. One of the things we talked about at breakfast is that you went -- and it's not you, of course, it's Anthony. You went from the high return on capital distribution, which was licensing everything to now we're making our own TVs, which is a lower return on capital. Here you've done it again. You started in streaming, which is where the world is going, and you backward integrated in the thing that's dying, which is linear TV. Please defend?

Yes. So one, subscriptions are not dying. Subscriptions are growing.

Linear TV.

Subscriptions -- digital subscriptions are growing. Of that, I can assure you. Digital subscriptions are growing and they're growing huge. They're growing through absolute volumes of subscriptions. They're growing through ASP increases as SVOD companies and digital subscription companies continue to raise price. Frndly is growing. It is not going backwards. It's growing. And it's growing well. We would not have bought it if it was not growing, and we think that we can actually grow it much faster. So this is not a situation where we're catching a falling knife, exact opposite. We have tens of millions of subscriptions that we monetize on our platform. Frndly will be an additive of that, and we'll do so under our owned and operated model. We monetize our subscriptions through premium subscriptions through what we call direct-to-consumer subscriptions. And now I would say we also have our owned and operated subscription business, which is Frndly. All of them are growing.

Okay. So I think of it as linear TV, which Wall Street has decided is dying, and I would...

I agree that linear is dying.

But you think of it as a subscription business model, not as linear TV and you're trying to grow the subscription business model.

Correct. It is a digital subscription. And the truth of the matter is, whether it's fast, whether it's live, whether it's preprogrammed, streamers love that. FAST channels are growing incredibly well. That is program TV for all intents and purposes. This is a situation where you have a bundle of channels that streamers love. Hallmark has original content, streamers love that. What they don't want to do is pay $180 in a linear bundle to do that. They are very willing to pay, call it, SVOD level bundle -- SVOD level pricing to have these. And then, of course, it comes with the DVR in addition to the programming, in addition to the original content, in addition to some VOD content, it's holistic. You can't think of it as a linear.

Because I would tell you that Muvi who's the back end for a lot of FAST. His most provocative statement from CES as FAST as dead. And he's the back office for FAST, which made it a very controversial statement. So you don't think FAST is dead.

Not only do not think it's dead, we're seeing exact opposite. We see FAST continue to grow and continue to grow.

Because you're adding new FAST channels, not same-store growth.

We're seeing hours continue to grow. If we are now -- if we are seeing -- I mean, the Roku Channel grew 84% in Q1. Yes, we have original content that does well. Yes, we have fixed content that does well. The vast majority of our hours in the Roku Channel is FAST. It is growing extraordinarily well.

It's not same store.

I also think there's very unique FAST channels out there like the NFL FAST channel, like the NBA FAST channel. It doesn't need to be some 80s shows that you haven't seen in 40 years in FAST that you go -- I mean people do actually to watch that, but it doesn't necessarily need to do that. It can be multiple others. And I guess what I'd say is programming does work depending on the streamer, and we want to give the selection to everybody. If you want to watch a specific FAST channel, you're going to be able to probably be able to find it on the Roku Channel. You want to watch VOD, you're going to be able to see that on the Roku Channel as well.

The issue as FAST as there's 400 of them and discovery is really hard. Even if you'd love to watch the Tour de France, it's really hard to find anything 400...

Hence, the power of the platform. So now you're getting into the unique asset that Roku has that no one else has. So you're right. It's just peer selection. If it is a mile deep and an inch wide, it's just selection. You're right. You have a cold start problem. You have a search and discovery problem. When you own the home screen, when you are the lead in to basically 125 million eyeballs watching, turning on your TV every day. When you are the lead into TV, you can curate the experience and basically help people get to where they want. I'm going to give you an example. The content row at the top of our home screen is incredibly powerful to driving hours and to driving subscriptions. And that is basically an AI-based machine learning algorithm that says -- and we're just getting started on this. We'll get better and better at this. The home screen does exactly that. It helps our streamers increase hours. It helps us monetize the home screen more. And the power of the platform in the case of the Roku Channel, again, a very important point. People say, well, well, the Roku Channel, is it unique? Is it unique? Well, it's unique in many aspects. Yes, we have a huge number of FAST channels. But the ingress into the FAST channel, less than 25% of ingress, which is the entry, by ingress, I mean entry into the Roku Channel comes off the app. 75% and more comes from areas outside the app. Well, where is that? It's on the left NAV. It's on the content row at the top. It's on the live icon that we have, not just the TRC icon. And so it showcases the power of owning the operating system and how you can drive engagement and direct traffic and therefore, ultimately, which is one of our -- which is our top priority to drive monetization of the entire streamer experience.

Okay. So you just gave a number that I need to -- I don't understand -- how can 125 million viewers be the right number when we have 90 million homes because there's 2.2 people per home. Are you telling me that there's -- that every home has a Roku device, an Amazon device and somebody uses the LG without any device attached?

No, it's just multiple devices, multiple people at home.

Yes. But since you're counting a single home, why wouldn't it be 180 million people.

Again, we're just -- again, I'm giving you a daily number, first of all, a daily number. Yes. Yes. So it's how many people actually turn on their TV and see it every day. But again, that's bigger than Super Bowl that turn on and watch TV. So you start -- and growing. This is another thing and growing. We track this very closely. The distinct hours per day per viewer are growing in all our countries. Now there's -- in the U.S., it's higher, but you even start to see the catch-up in international countries as they migrate more towards streaming. So all these metrics, the engagement, the streaming households, the hours per day, they're all going in our favor, and that is incredibly powerful when you own the platform.

Does that speak to your confidence around the full year guide despite the macro uncertainty?

So yes, the reason why we feel good about the full year guide despite the macro uncertainty is simply because we have a very diversified revenue stream now than we did, say, in 2022, where a big chunk of our revenue was M&E, media and entertainment. And you know this because you covered Roku a long time, Laura. We really did rely on this one, call it a vertical. It is a vertical for all intents and purposes. And it's a vertical that's owned by a handful of big advertisers in the media world. We are far more diversified then we are back in 2022. If you look at M&E, we don't give it out. But if you look at M&E as a percentage of our overall platform revenue, it's far, far lower than it was in 2022. In addition to that, we've expanded ad product. We've expanded -- it used to be -- this is a big shift to that we've talked about, but it's so important. It used to be that if you wanted to run via a DSP on the Roku platform, you had to come through what was our DSP at the time called OneView, which was a very narrow focused, meaning that you could not come through other DSPs. We changed that about 18 months ago. We opened up to all DSPs. So you have a diversified -- you have ad product diversity, diversity of demand and now diversity of supply, all different now than it was back in 2022 or the prior ad recession. We are well -- we are far more differentiated in all our revenue streams on the platform side of the business. And that gives us confidence that in an ad recession, we'll do fine. If it's deep macro recession, I think -- and everybody is going to be impacted somewhat, but we feel very good of our position because of all this diversification that I just referenced.

Okay. So free cash flow has been where we've really seen big impact from you. I think within the first 12 months that you'd be in the free cash flow improved like $200 million, and you've been really driving -- I think your guidance for '25 is like $350 million of free cash flow. So can you talk about how much of the free cash flow growth is behind us? And how much is front of us? Like what inning are we in here?

Yes. So we guided to $350 million of adjusted EBITDA for FY '25. We said free cash flow, and I said free cash flow is going to be higher than that. Obviously, quarter-to-quarter, you do get some working capital fluctuations. We are extraordinarily tight on free cash flow. It's one of the things I'm most proud of at Roku is how we've adopted this important KOM, it stated KOM, for us, free cash flow, and I will say free cash flow per share. We take dilution very seriously. We watch dilution very closely. And to answer your question very directly, how much is behind versus ahead. I think we're just getting started. I think the -- our ability to grow our platform revenue, which is how we monetize double digits, to grow our OpEx well below double digits and call it, single digits and not high single digits. And our ability to -- we have a lot of tailwinds in working capital and just getting very tight in how we manage our receivables. We're very good at this. We're very good at managing our payables, like all this is a formula for improved free cash flow with outsized growth relative to revenue.

All right. Fantastic. I'm glad to hear that because you've done such a good job with free cash flow since you've been here. I was wondering where we're running now. I'm happy to...

I should also say we're also very CapEx light. One of the beauties of our business is we do not have a lot of capital. The capital really is our people cost, our engineering, our product folks. That is our capital. We don't capitalize it, but that is how we think of allocating capital. On the CapEx side, which, of course, free cash flow is operating cash less CapEx, our CapEx is very light and will continue to be very light.

Okay. Great. Let's talk about third-party DSPs. One of the things that's been really helpful to the stock is when you opened up your ad inventory Trade Desk and assuming Viant and some of the other DSPs over time to fill, I think we assumed a 50% fill rate of ad units on the Roku Channel. Can you tell about -- talk about your learnings to date, what's been better? What's been worse? And I'm really interested in the relative CPM between what your ad sales force has been selling ever since you got there to the kind of CPMs. We had someone on the stage this morning, not Jeff Green, who was also on this stage, but not him, but somebody else say that the CPM discount is 50% versus a direct sales force. So can you speak to...

Yes. The bulk of our CPMs are not a 50% discount and there's a reason why it's because we have the best data.

So you probably have a floor.

But again, the floor is notwithstanding. We do set floors. But again, we set floors to maximize gross profit and therefore, maximize free cash flow not to maximize gross margin. Like we're very smart about this. So we're going to maximize cash. And gross profit is a proxy for cash when you maximize floors. I can explain that more. But let me get to the point on how to think about DSPs. So as I mentioned, like we're just getting started on DSPs. If you were to use the baseball analogy, what inning are you in? I'd probably say we're in the top of the third inning, quite honestly, in DSPs because where we're going with DSPs, I just mentioned how much supply we have with the Roku Channel, 84% plus the run of network, which is incremental to the Roku Channel. We have a tremendous amount of supply. And quite frankly, it's pretty easy for us to create more supply because that's the power of the home screen. You can create more supply without big investment. And so it's really about demand. Well, how do you bring demand over? Well, you -- of course, you have amazing ad products on the premium side and then you have -- you basically have the ability to buy across the entire CPM curve. Because I don't -- I can't tell you exactly where the CPM curve is going to go. I do believe that there's going to be this premium inventory. And then I think everything else moves to programmatic because of the performant nature of programmatic over time. And we see this shift happening now. Where it ends, I can't -- I don't know I've got my opinions, but I think it's going to be like the sports and the Roku home screens and the original content, that's the ultra high CPM and then everything else is going to be programmatic because of the performant nature of running through programmatic. That being said, if I'm even close to being right on that, and I think I am, the ability to have a performant ad product is what's going to be a big differentiator in the programmatic space. Our first-party data allows us to do that. So as we increase demand, we can integrate deeper and deeper with everybody, with everybody, all the DSPs, we can then use our first-party data to become more performant because -- and I can talk about how we utilize that because I know you talked about how do you monetize your data, I'm sure that's going to be a question. We can talk about that. We can use our first-party data to become more performant. And anything short of the walled-garden approach, which is going to be limiting in nature, like we will be probably -- our goal is to be the most performant inventory as a publisher. And we're starting to see this as we integrate and you can't just integrate with DSPs and with external demand. You have to go deeper and deeper and make sure you set up clean rooms where data can be passed and utilized to make sure you're winning bids in an auction and/or you are performant, if it's just a direct one-to-one buy.

So let's stay with we want to be the most performative platform. How do you beat Amazon?

So again, Amazon right now, you could argue is the wall-garden approach. They're talking about becoming a DSP. There's a lot of talk about DSPs about Amazon being a DSP. Again, outside of the wall-garden approach, again, I would say that, that is the best first-party data to have, but we can do it in multiple ways outside of Amazon. And if Amazon does move into the DSP space, they're going to want to integrate with us because of our massive supply. So we have to wait and see on that piece. But all the other DSPs are going to want to integrate with us. And then we can become performant with our first-party data as we match it up with other first-party data. Let me give you an example, like with Shopify, where we are integrated. We can -- we actually can use our data with other first-party data to become the most performant ads. I'll give you another example. It's a smaller example, but one that's important, like we can partner with measurement companies. We partner with a company called iSpot and you've seen this where not only do we monetize our data because you're saying, "Hey, do you monetize your data? Not only on that, particularly, do we monetize our data. We trade log files. We trade information back and forth. So we can see how performant our ads were. And if they're not performant, we know we can figure out why and make them more performant. So that is a win-win in how to use our first-party data to become to actually help us monetize better than anyone else.

So when I think of performance, I think of a drive to purchase. So I don't understand how high iSpot helps you be performant?

Because they can measure -- they can integrate with advertisers, along with us as a publisher and they can measure causal based lifts, they can measure ROIs, they can -- whatever KPM they're trying to track, they can measure that. So you can do what you're saying is get back to the actual conversion and purchase with other companies. The signals that we give ultimately allows us to be better targeted allows us to actually become more performant over time. And again, if we find that we're not performant for any reason, we'll get the data back. We'll understand why we'll make adjustments accordingly.

Okay. Fair enough. So you do think that -- what you said was premium here is sports, which is unique content, originals, unique content, and then the home screen and the home screen is sold like a billboard. It's almost don't reach practically. So it's a premium fee because of its location. So if you think those are where the pricing power is going to be, everything else you said goes to programmatic. Doesn't that mean we're going to get a downdraft in CPM programmatic sales for less money?

It all depends on the performance nature of the ads. So yes, there'll be some CPM curves that are on the lower end. There will be some on the medium, there'll be some on the high. Whether that's a $6 CPM up to a $25 CPM, all that is based on the performant nature of how the impression performs. That's where we differentiate ourselves. So again, you want to performant nature, you want to go -- you want to use our first-party data to do that. We're going to help you do that, but it's going to be higher CPMs. If you just want massive amounts of impressions and reach at some floor-based pricing, let's say, your a video game app and all you want is to hit a bunch of people for app installs. That's just going to be a lower CPM as an example there. And again, like we can integrate with the largest DSPs, and we do, we can integrate with the smallest DSPs and we do. We can play across every point of the curve of the CPM. And again, the idea being because of our first-party data and how we utilize it, we will become more performant and therefore, be able to charge higher CPMs.

Okay. One of the things that's been a problem in the past with the old Internet, not connected television yet, is that a DSP will take your data and then find audiences that replicate that data at 1/10 of the cost. And so then they undermine your power. So how do you prevent your data from becoming their stocking horse and them making money.

You've got to be very careful and you got to use things like anonymized clean rooms and you got to use signals that can be ultimately changed and/or configured differently. So these are things that we think about as we go into these deals.

Okay. And with the iSpot data, is that a different data than you use? Or do you just give...

That's a different. That's a measurement perspective. That's very, very different. And so that's much -- again, that still has anonymize clean rooms and things like that, but it's a very different agreement than, say, a demand-side platform where, again, your point is a valid one, like we have to be very careful on that. But again, like ultimately -- and I should say like I do want to get your questions on CPM. So ultimately -- like if CPMs do come down, which they may in certain areas, again, that's not a bad thing for us because we have supply. So we...

Those of it undermines the pricing power of your direct sales force.

And again, that's where depending on the nature of how this shakes out over time. On the programmatic side, if you're performant, you're going to -- again, when you go to auction, if it is an auction in this case, and you have more data signals and you're performant, like the demand will come in and perform -- and the auction will work in your favor because of the ability to have these data signals and become more performant. Those are just charge higher CPMs. I'm saying if all CPMs proportionately come down, the ultra premium starts to come down, the whole curve shifts to the right we are not at a disadvantage in that case because we have a tremendous amount of supply to weather any change in CPMs. Not every "publisher" has that capability and think about how expensive it can be for a publisher to create supply. They might have to do more content or they might have to do more sports or they might have to do more marketing or distribution. We don't have that issue in creating supply.

You have this a lot because Google used to link out to a lot of websites and now they're not because they're just answering your questions. Google Search is answering your questions. So a lot of people, supply is dwindling and a lot of these websites that used to get traffic from Google.

Yes, you can see like -- I do sense like when we look at CPMs, we do see these trends down. But again, they're not necessarily as a result of the quality of the impression, it's the result of the supply and the demand. And you do have more demand coming because we've talked about this many times, you do have linear -- the whole linear migration over into digital from advertising. That will also continue only because the eyeballs are all coming over. And it has not been a one-for-one move. Different markets vary. But I've seen reports and the ones that we have, say, 60% of the hours have moved, but yet 30% of the ad dollars have moved. That's going to take care of itself over time and not a long time. I think it's not tomorrow, but it's over, I would say, quarters and years, not 5, 10, 15, 20 years.

Okay. Let's talk about ROIs on hardware related revenues such as home security, owned Roku-branded TVs. And in light of your $5 billion impairment charge in '23 for audio and its subsequent deemphasis, it feels like hardware is a riskier -- lower returns for sure than the old business, but also riskier from a point of view, sometimes it doesn't work.

Yes. So the bulk of our hardware revenue that you see I'd say the bulk on a unit basis is still our players. Our players are amazing. Our BOM cost of our players continue to do very well -- continue to come down, and therefore, we can keep our price low. It's sort of neck and neck with us and Amazon on the player market share. We just launched a new player product, which is pretty cool because, because of the way we've engineered it, you do not have to plug in to an outlet, it actually is plugged in and is operated by the TV itself. Yes, I mean, it's cool because yes, you don't have a core, but it also reduces the BOM cost by a sizable amount more than you think and all of this is a positive for us. So I do not ascribe to hardware is bad ROI. Our players have an incredible high ROI to it when you factor in the monetization and the new streaming households that they bring in relative to the cost per acquisition. First party TVs, again, we're very new in first-party TVs. We've only been doing it just under 2 years on first-party TVs. We're getting better at this. This will take...

Roku brand.

Yes, sorry, the first party [ data ], I should be clear. It's the Roku built and branded TV, not the OEM, third-party TVs. And so that just takes time. And that's strategic. You do want to control your own destiny on the hardware as well as have multiple partners, and we do. And so yes, that is a more expensive cost per acquisition, but it's one that we think is very strategic and one we're going to continue to invest in.

Right. Okay. So returns on capital are lower, but it's strategically important because you need to be able to control your own destiny.

Yes, and they're still positive.

I can't -- sorry, sorry, just thinking about your P&L, the gross margins have gone negative on the hardware business.

Right. But again, remember, our strategy -- sorry, when I say -- when I'm talking about every new streaming household that we bring in has a CAC to it, that gross profit is part of the cost per acquisition. The platform side is how we monetize that.

That was true before.

Yes. So we're growing more households as a result of our first-party TVs and therefore, we grow our ARPU.

Why would first-party TVs grow your households?

Because it adds new streaming households.

And you don't think if you just had players, they would just buy those instead of the TVs?

Well, again, we're additive to it. And you don't know. I mean, the TV market is very competitive. There's a lot of low-priced products out there. There's a lot of differentiated operating systems. What I will say is we're #1 by a lot. We talked about this. Our broadband household is over 50% in the U.S. and growing. It's growing every -- it grew last quarter. It's going to grow this quarter, and it continues to grow. And that is a result of our third-party and first-party devices.

Okay. But the margins are falling on hardware.

Yes. Again, the margin is notwithstanding, they are negative on our devices on the first-party TVs -- that's obvious because our players are slightly positive.

Yes, exactly. Okay. So you've been introducing new products and features offshore Canada, Mexico, U.K., A lot of these are now like first in the country for OTT backbone. In several countries in Central and South America, and successful markets take 3 to 5 years to actually get to the point where you've got enough penetration to start actually monetizing through advertising. So can you talk about why this is a good return on capital for Roku compared to just stick into the U.S., building new products, focusing on the U.S. where you have really nice unit economics.

Yes. Here's what I'd say about how we think about international versus the U.S. The reason why Roku is #1 in the U.S. is because, one, it has an amazing product that streamers love, and two, Anthony was incredibly visionary in getting this, call it, first mover advantage. So you are right that international countries take time, but to grow the scale that we're growing to is relatively inexpensive now to get to the scale. What's the -- what you have to have a little patience for and as a CFO, I don't have a lot of patience, but I'm very thoughtful and very strategic, what you do have to understand is the ad market and even the subscription market, how we monetize subscriptions needs to catch up with the overall streaming just like it's still catching up in the U.S. So let me give you a specific example, like in Mexico. Mexico, we have significant market share. We're #1 with significant -- it's actually getting close to the U.S. in terms of market penetration in Mexico, yes, it's doing very, very well. We've been able to get this market share at relatively low what I'd call cost per acquisition. We haven't yet monetized it because the ad market just isn't in the space yet. It's just taking -- similar to the U.S., a longer time to move the ad dollars and it's just a different ad market than U.S.

Why haven't you gone in Mexico in the first place then, why wasn't...

Because over time, this will also happen where you'll have a big ad market move to digital, and we're going to be -- we are #1 and we'll continue to be #1, and we'll be able to monetize that. And our subscription business, which we are now very focused on monetizing, we believe we can monetize that in the international locations as well through having great subscription-based products and initiatives. And I don't mean like owned and operated subscriptions, I mean like our ability to have like a subscription manager where you can manage all your subscriptions on the Roku. So I'm not saying we're doing that, but that's just an example of how we can utilize our OS and help us monetize this. Like we did launch TRC in Mexico. It's doing extraordinarily well. It's growing...

That's an ad product.

Yes. And so we are selling adds. But again, the market will take time to catch up. What I'm saying is the cost to get into these countries when you get in early is not significant.

Yes, because there's no ad market there.

There was no ad market. What was the ad market in digital in the U.S.10 years ago.

The ad market was always huge.

So I'm going to give you an example, I'm going to give you an example, Laura, on how to think about this because I'm going to use my prior employer example. So I talked to Anthony back in -- when was it, 2016, 2017 pre-IPO, Anthony talked to me. A great guy he is. Yes, '16. '17 was IPO. And I said to him, I said, Anthony, I said, how -- I come from Amazon, I see what's going on in some of these devices. I said, how are -- how is Roku as somebody that -- as a CFO who could hopefully maybe be part of back here, how is Roku going to make money? I don't see it. You're just not going to make money on hardware. He says we're not a hardware company. We're not going to be a hardware company. We are -- he called it a services company at the time. And then I noticed that they were hiring "ad salespeople" in the market, okay? I'm going to pivot that to my former employer where I had a similar conversation about a year later because I stayed -- I was with my employer, and I said to them, "Hey, like why don't we go and start looking at ads? And the comment back was, this is all a Netflix play. Everything is Netflix. Ads don't matter. My point on this, meaning that at the time in 2016 and '17 and '18, there was no ad market on the OS. It was all the hours were Netflix at the time. Netflix -- maybe YouTube, but Netflix was an SVOD service at the time. It took a long time for the ad market to materialize on the digital platform. And the opportunity, you do not want to be #4, #5 in this space. Yes. Yes. So now you're starting to see the power of the platform like more companies are trying to get into this space. This is not -- it's not going to be easy. Roku had 14 years to get to this broadband penetration with a big focus on the streamer experience and building an awesome platform, knowing that the monetization was eventually going to come even if it wasn't going to come in '16 and '17 and '18, even in 2022, so much of it was M&E. So much of it was a handful, a handful. I'm talking 10 big companies making up a disproportionate share.

And then the box office closed. We saw the dependence on that.

Well, and you saw a lot of companies that were SVOD-only players to realize, boy, this AVOD business is going to be little bit -- again, it's capitalism at its finest. You see a big bucket of money and you get a lot of people going in. The good news for Roku is we are #1 in this space. We're going to continue to be #1 in this space, which means we get the opportunity to monetize more at scale than anyone else.

Questions from the audience? Anything from the audience. Okay. Great. All right. So I love that. Let's do new revenue streams. Let's talk about data as a revenue stream. We did a little bit. But should you be selling your data to LLMs on an anonymized basis because they're out of data, they need to -- and should they be selling data, I get that you're doing a little deal with iSpot, they're tiny. Should we be selling data more aggressively the way Vizio does? It's not bold or anonymized...

Yes. So here's what we're not going to do is you're not going to say -- you're not going to get me to say because we don't look at it this way. It's like, hey, our ARPU is x amount of dollars in our average revenue per user is data sales. That is not strategic for us. We could do some amount of data sales. LLM is notwithstanding, I mean, that would be like -- that would be -- these LLMs want more conversational threads. If they came to Roku said, "Hey, we'll pay you a lot of money for your data, maybe we'd consider it." I don't know. That's a whole separate thing that the large language models probably aren't interested in. That being said, there's always the possibility to do data sales and where we have strategically used our first-party data in the examples that I gave, which will ultimately lead to higher CPMs. It will ultimately lead to more wins in the auction because of our performant nature. That is how we're going to monetize our data. And do we monetize it in other ways? Yes, we do. We don't talk a lot about it. But again, I want to give you a very specific example of something that could be very big. And again, just follow me through on this, like we have a product called Ads Manager. Are you familiar with Ads Manager?

Yes.

It's basically the small- and medium-sized businesses who don't want to go through DSP are too small for an agency. They want to do ads. They could never do CTV ads because, again, they're not -- they're too small to hire an agency. It's very local. It has to be very targeted. Now there's products out there that are self-service that within 5 clicks, you literally can be running a CTV -- on connected TV, a video ad for your 5 chain burger shop or whatever, your auto chain of stores that you've got. And one of the areas that, again, like you can say, well, what's the market for that. I don't even know what that market is. It could be huge. It could be all of -- like a big chunk of the performance market could move over because most of these folks do social, they might do CPC bidding. Why wouldn't they come over? And if it's performant, run a video, because video is way better than a lot of these other ads. It's just their performance and you can measure. We now can do that through our Ads Manager product. And the way we help with the monetization is now we'll integrate with measurement companies. We have one that we integrate via API again, via API, where we share the campaign data, the advertiser shares their data so they can do causal based lifts, true causal based lifts to understand how the ad did. These are after like 5 clicks to get it, and you can measure the performance of your ad campaigns and measure the true ROI. That's how we're utilizing our data in that case. No one can do that. Now again, if our ads manager will do well, but there's going to be other companies that do this as well. We'll integrate with all of them because they want our supply because again, we're over half to broadband households. And then on top of it, we talk a lot about AI. This is another area where we're experimenting with internally and partnership-wise. These videos, you have to create a video. AI will create a video just like that high-quality, targeted, how you want it as an advertiser within seconds. You can have a Gen AI created video and be up and running to spend an X amount of budget in your local area with targeted KPMs of whatever you pick in your self-service checkout. It's pretty cool.

One of the things Trade Desk said is they don't want to do creative on their own platform, but they're going to create a marketplace where you can just select which creative and then they'll charge it -- they take a 15% take.

Yes, there's is going to be multiples.

And then a lot of guys on my stage are like, no, no, we have created built-in that part of the self-serve as you start here, the next step just download all your existing creative from stills like from TikTok, whatever, and then their Gen AI -- platform makes the creative and then the segment and then you send it out. But so you can either do a captive or you can just put a marketplace.

Yes. Yes, you can do it. We could create the product and we are experimenting the product ourselves. We could partner with somebody. There's so many companies that can do this. It will become a commodity. This will become a commodity. And that's what we want. Like that's the beauty of it. We want it to be commoditized. So again, you literally because of the nature of our supply and our performant nature, we'll integrate with everybody in this new SMB space, just like the DSPs. It's no different than the DSPs on the demand-side platforms. We're going to do it in now this SMB market, and we'll be agnostic.

I think Wall Street believes that SMBs will be the primary driver of connected television revenue growth over the next 3 years.

I would love it. I agree. I'm super excited about our "Ads Manager" product. And again, even if ours isn't the best, and I think it will be, I want somebody else. It's fine if somebody else does it because they want to integrate with the market leader in CTV, that's us. And so I've often -- I've challenged people and say, "Hey, what's the size of this market?" I haven't been able to figure it out yet because the market...

Really, over time.

Well, the market hasn't even formed yet, like it's just getting started. And how much this shifts -- like linear to digital, how much this shifts from the performance social or any other budgets that they're spending money on over into this is to be determined, but I think it's very exciting.

That was one of my learnings from the CTV panel. I just asked a simple question like who do you think your primary competitors? One guy says YouTube, which makes sense to me. Another guy says TikTok because they're taking social dollars into his platform for CTV because it is performative than CTV. Another -- like they had all -- it was 4 guys on a panel. No one thought they competed with the same person. I'm like what we can't define the competitive set, but the takeaway was actually a different company can compete with different ones depending on what you think you do better as a CTV option to what was before which is super interesting and flexibility of CTV.

Yes. It's interesting. And I think one of the things we've been out talking a lot about because I do think it's misunderstood is the actual power of the OS itself. And again, you see it now by more companies realize just they did with AVOD, you had the companies say -- you had companies, including my former employer said, we'll never have ads. And again -- and that exactly -- and that all changed. The same is like OS, people are not starting to see the power of the OS. The good news for Roku is we're #1 and growing.

Okay. Thank you, Dan.

Okay. Thanks.