Ad-tech Overview
The chess pieces are being rearranged in ad-tech and I don’t think this quarter is very meaningful in terms of where this will go long-term. These shifts can take a while but our strategy from the beginning has been to stick close to first-party data companies. Magnite is given access to first-party data by representing publishers and the rest of our stocks are firmly cemented in first-party data.
I think it is a false assumption that we have all of the information from 10 days of IDFA changes and iOS 15 being live. The migration towards or away from certain ad platforms can take months, quarters or years. But there is certainly evidence that the shift has begun.
Regarding who will do well long-term, it will be those who work with first party data. There is quite a bit of evidence that this is where the advertising world is headed. Listen to Twilio’s call and you’ll see they are a cloud company banking on this shift.
“Digital growth and digital personalization or how businesses are building their businesses is a tremendous opportunity. And that opportunity is actually accelerated by the [Indiscernible] changes, like on the world of IDFA tags and third-party cookies, and all those things getting change because companies here rely on what's just honestly shenanigans. Like in the changes that have been going on, whether cookies or IDFA tags, these privacy changes are on the right side of history. And so, what Twilio's providing is the antidote to all those changes, which is a personalization and marketing system that starts with first-party data.”like on the world of IDFA tags and third-party cookies, and all those things getting change because companies here rely on what's just honestly shenanigans. Like in the changes that have been going on, whether cookies or IDFA tags, these privacy changes are on the right side of history. And so, what Twilio's providing is the antidote to all those changes, which is a personalization and marketing system that starts with first-party data.”
Here is what Magnite said in their recent call: “It’s been a third-party cookie world with third-party data that has really ruled the ecosystem. And I think publishers are clearly seeing a shift and buyers are starting to acknowledge this is really good information [first-party data]. They have a direct relationship with the consumer.”It’s been a third-party cookie world with third-party data that has really ruled the ecosystem. And I think publishers are clearly seeing a shift and buyers are starting to acknowledge this is really good information [first-party data]. They have a direct relationship with the consumer.” My original thesis on Magnite was centered around first-party data.
Here is what Roku said regarding the shift from third-party data to first-party data in their earnings call, “Yes, as you alluded to and the way you framed up the question, the disruption and the noise around the loss of cookies and device IDs like Apple's IDFA in general is a net benefit to Roku, really for two reasons. First, independent ad tech is very challenged in an environment, where these identifiers are getting more scarce because, they don't have these identifiers. They don't have a direct consumer relationship, whereas Roku does. And so we're always working on our platform with our own first-party data and it's a fundamental advantage for us and ultimately is bringing brands to us.”They don't have a direct consumer relationship, whereas Roku does. And so we're always working on our platform with our own first-party data and it's a fundamental advantage for us and ultimately is bringing brands to us.”
This is what you’ll need to ask yourself – are all of these management teams wrong or is the market wrong? Because clearly the market is penalizing these first-party data companies while rewarding third-party data companies (Snap deeply penalized compared to Facebook which is excessive on third-party data AND Roku penalized vs The Trade Desk which is only third-party data). Granted, the Trade Desk would be more affected by browser cookies being eliminated (this could happen end of 2023) but to have no affects long-term compared to Roku is not the correct outcome.
Hopefully by now, you know we would close a position if we thought the story was materially weaker than we previously estimated. When we close a position, our goal is to move that money into a compounder. However, I also don’t mind standing in front of a train and saying the market is wrong if I think the product isn’t fully understood. How many times will I have to do this with Roku? We can add Q3 2021 to the long string of misunderstood moments in Roku’s history!
We already covered Snap following earnings here, but I will tell you that October 2021 is now branded in my mind as another example of how the market lives in an alternate reality of two extremes. When Snap was experiencing its biggest one day loss, the company was simultaneously launching the most bullish product in the company’s history – Arcadia. This product helps Snap scale AR brand ads beyond its own platform. It’s a critical moment when an advertising company is able to monetize audiences outside of its own feed (or channel), and therefore, it’s bizarre to see a double digit drop during the month of Arcadia’s launch (but will provide a great editorial someday).
Magnite has always been a high flier for us; one where we are counting on the whole being greater than the sum of its two parts. If CTV ads were at saturation like mobile or desktop, then we would not be in Magnite or maybe even Roku. It’s this tailwind that has been able to overcome the current headwinds. Magnite not pictured because a pro-forma was not offered, however, we believe the pro forma growth rate for Magnite is around 25% placing the company between Google and Twitter.
Here is Pay TV ad spend in the United States alone compared to CTV ad spend of $24 billion. When Pay TV ad budgets stars declining like cable TV subscribers has declined, then we know we are finally in the true market for CTV ads.
Roku Earnings
There is a lot to unpack with Roku. Despite headwinds, it’s technically the strongest ad-tech company in terms of forward growth yet the market is going through a serious (and intense) period of doubt. We will need to discuss Player revenue, what supply shortages mean for smart TVs and dongles and competitors such as Google and Amazon and. The supply shortages affect Roku’s active accounts but Netflix also made it clear during Covid that there was a pull forward in terms of subscribers, which is similar to Roku’s situation.
Even despite these near-term risks, Roku is guiding for 37% growth which is the highest of all ad-tech companies above $20 billion in market cap. Notably, Roku is guiding high (comparatively) on a very large revenue base of $890 million at the midpoint for next quarter. Compare this with The Trade Desk, guiding for 21% on revenue of $339 million.
We will quickly go over the financials before breaking down the main issues that the market is concerned about. Then, we revisit our thesis to see if Roku is still on track. The goal is to always figure out where the market is wrong and being inefficient.
Financials:
Roku’s total revenue grew 50% year-over-year to $680 million. Platform revenue increased 82% to $583 million. Gross profit increased 69% and active accounts increased 23% year-over-year. Sequentially, active accounts were low at 1.3 million adds.
Average revenue per user (ARPU) is at $40, representing an increase of 49%. To grow ARPU from $27.00 to $40.00 in roughly a year is unheard of. Here is how Facebook’s ARPU compares with it taking the company 16 years to reach $40 ARPU while it took Roku four years from the launch of its ad exchange to achieve this ARPU. You’ll see the jump from $27.00 ARPU to $40.00 took a few tries between 2017 and 2020 for Facebook while Roku has nailed this incredible monetization growth in one year.
Clearly, Roku does not have billions of users like Facebook but ARPU is a vital sign as to the strength of an ad platform. It’s also helps to elucidate why Roku management is taking it on the chin with Player revenue (see below) and what they mean by flywheel across Roku’s diversified product line, which is an operating system, an ad exchange that extends to mobile and the web, and a publisher.
The $40 ARPU means that Roku can take losses growing its audience and will be able to make up for those losses over time – and that’s exactly what Roku intends to do.
Player revenue is down 26% year-over-year. There are also losses on the player from $20.2 million in profit in the year-ago quarter to $97.4 million in losses in the current quarter. The company stated the following regarding these losses, “While Roku player unit sales were down year-over-year in Q3 2021 (following the extraordinary demand spike we saw in Q3 2020), unit sales were above pre-COVID Q3 2019 levels. Our player unit costs were impacted by the supply chain disruptions. However, we chose to insulate our consumers from these increased costs to prioritize account growth, resulting in Player gross margin decreasing to -15%. We view this Player gross margin erosion as temporary.”
Regardless of transitory issues with player revenue, the substantial increase in ARPU is helping the margins quite a bit with gross margins of 53.5% up from 47.6% in the year-ago quarter and adjusted EBITDA up 132% from $56.2 million to $130.1 million. The gross margins for next quarter are forecast to be weaker at 43% and adjusted EBITDA will be lower between $65 million to $75 million compared to $113 million in Q4 2020. The company explained that the lower EBITDA is from “investing in headcount, product development, and sales & marketing to drive future growth.”
Roku is guiding for quarterly revenue of $885 million to $900 million next quarter or 37% growth, up from $649.9 million in the year-ago quarter. The company is guiding for gross profit of $385 million, or 26% growth year-over-year. To reiterate, what the market doesn’t like is that margin decrease from giving players away at a loss.
You can read our previous analysis on Roku’s ARPU becoming decoupled from active accounts.
Player Sales:
Player sales is the main reason that Roku got clobbered. I don’t believe it was over active users as they were up 23% year-over-year, which in the face of declining player sales is quite impressive. It’s important to remember that earnings are relative and Q3 of last year saw very strong Covid tailwinds where users were buying hardware and staying home. Apple is the bellwether on this issue of the electronics and consumer hardware boom that is now tapering off.
Roku management emphasized the fact the growth is stronger this quarter than pre-Covid levels. “Meanwhile Roku player unit sales remained above pre-COVID levels and the average selling price decreased 7% year-over-year as we chose to insulate consumers from higher costs.” However, with player unit sales down 26% in the face of tough Covid comps, the market is concerned.
Logically, analysts and investors know they are not invested in Roku for the player yet the player can weigh on margins. The gross margin in this quarter was strong for Roku at of 53.5% but the guide of 43% is why there was a sell-off (in my opinion). Wall Street has always been worried about Roku’s margins relative to its player revenue.
Roku is a growth machine – comparatively speaking, it’s heads and shoulders above other ad-tech companies in terms of revenue size (roughly $900 million for Q4) with the strongest guide in our universe at 37%. This communicates how management views headwinds or tailwinds – both are an opportunity for a land grab. Here’s one way Roku is seizing the supply shortage: “As mentioned earlier, we chose to insulate our consumers from increased component and logistics costs, resulting in player gross margin decreasing to negative 15% in Q3.” In addition, the company plans to keep dongles stocked so that if smart TVs sell-out or are too expensive for consumers, they can upgrade their current television with a Roku player.
Here's a question from Laura Martin on the call that is important to understand why Roku could come out ahead in light of supply issues: “But if you're going to sell out of those [dongles] anyway because TVs are running out why would you cut price [of the dongles]? Why wouldn't you double price and still sell out and just and still add as many subs, but at a higher price because you've got dongles in stock when all the TVs smart TVs are running out of inventory at the retail level?”
Here was the answer: “So the supply chain — in the case of players we're not — our goal wasn't to not sell out. We are paying more for expedited shipping for — to get chips get in front of the line for chips. So the results of all that is our costs are going up. But we haven't sold out yet. We've just been paying for air shipping and we've been spending money to insulate the retailer and the end customer from pricing issues and supply issues. So far we've been doing that relatively effectively.”
The translation is that they can air ship boxes of dongles and keep them on the shelves because of their small size while TVs sell-out and/or are cost prohibitive for consumers with average of 42% increase in price. “That [TV sales] is down. The market is down 31% year-over-year in part because pricing on U.S. TVs on average is up 42%. And the U.S. TV market is actually down below pre-COVID levels in the corresponding period in 2019.”
When asked why they aren’t doubling the price given the supply constraints (i.e., and appeasing Wall Street on the margins), the answer is that they are actually going to take a hit on the players at about (15%) because they want to keep costs low, which in turn, will grow active accounts. This goes back to the $40 ARPU. Once someone is a Roku user, there are high switching costs and Roku’s advertising flywheel can make up for the loss on hardware.
Competitors:
What management said on the call exactly matches my understanding, which is that Roku has always been competing against Google and Amazon. There is no change to the story here. Here is what an analyst asked: “First, just coming back to TVs for a second. Obviously, there's some new kind of incremental competition in market between Google TCL, Amazon Fire TV branded TVs and kind of what Comcast is doing. It remains to be seen how successful that will be. But I guess the question is that that narrative is there and I'm curious what you guys think about to kind of offset that narrative?”, there's some new kind of incremental competition in market between Google TCL, Amazon Fire TV branded TVs and kind of what Comcast is doing. It remains to be seen how successful that will be. But I guess the question is that that narrative is there and I'm curious what you guys think about to kind of offset that narrative?”
Here is what management said, “But we've been competing very successfully with large companies, all the companies you mentioned since the beginning. And if you look at where we are in terms of that competition, we've gone from no market share in TVs to the number one licensed TV OS brand in the US with about a-third of all TVs sold now running the Roku operating system. We've built an incredibly strong brands around streaming. We've achieved large scale with lots I believe lots of scale growth to continue in front of us. Most of our growth is in front of us.
… So we've been competing very effectively. We take competition very seriously. I don't see any particular dramatic change in the competitive landscape, with all the stuff that's going on. It's just more of the same, and we will continue to compete in market share.”I don't see any particular dramatic change in the competitive landscape, with all the stuff that's going on. It's just more of the same, and we will continue to compete in market share.”
They also stated Amazon Prime was not up for negotiation at this time. “As for the — your Amazon question, we have renewal discussions with hundreds of partners each year. It's normal course of business. Our goal in these discussions is always to reach an agreement that's good for our partner, good for our customers, delivers a great user experience. Despite what you may have read, our Amazon agreement is not up for renewal or in negotiations at this time.”
We had a Member post on recent stats on the Wire from Conviva. Here is what Roku’s lead looks like:
Here is how Roku looks on a Global scale – Roku is green, Amazon is white and Samsung is yellow.
That picture is worth a thousand words as to why we are long Roku. Our thesis here is that Roku is the royal flush in terms of its positioning. You can view our webinar here.
Lawsuit with Google:
Who hasn’t Roku fought (and won against?) – Peacock, HBO Max and Fox have all threatened to remove access before eventually folding. Google is especially in a bad position here as they are asking for search data from Roku customers to be shared with them, which Roku does not do for other apps. They also want preferential treatment in search results. Keep in mind, that YouTube TV has been off Roku platform since April and Google created a workaround for YouTube TV to be accessed through the YouTube app. The percentages shown above help illustrate why Roku fights these apps – they are the top dog in this space.
The chances that Google goes up against a well-informed tech CEO in the court of law on data requests and preferential treatment with search results is very low. Roku can expose Google in ways that a Congress Vs. Zuckerberg was not able to as Congress does not know enough about data collection to handle Zuckerberg. Meanwhile, Roku representatives can easily describe the issues with Google. The discovery process will be enough for Google to fold, in my opinion, so let’s see if this prediction turns out to be true. It will also leave Google wide open to have an example made of the company in terms of Big Tech’s overreach. Google tends to play things safe in this regard as there’s a whole lot of data issues lurking beneath the surface; no reason to wake a monster.
CNBC and others are also publishing favorably for Roku, including viewing a 2019 email that shows Google did ask for preferential treatment: “But a 2019 email from Google to Roku that was viewed by CNBC shows Google did ask for preferential treatment for YouTube in Roku’s search results.” And, two members of Congress are already salivating at the idea of taking Big Tech to court.
Let’s Revisit the Thesis:
Now that we’ve dissected why Player revenue is struggling and why Google is small beans in the Connected TV world (and very unlikely to go to court), let’s revisit why we are here. What management said towards the end of the call sums it up: “I mean, if you think about the big picture, we believe all TV is going to be streamed. That means, there's one billion broadband households around the world. They're going to get all their TV through streaming. So, a pretty small percent of those are actually doing that todayThat means, there's one billion broadband households around the world. They're going to get all their TV through streaming. So, a pretty small percent of those are actually doing that today.”
Here's another way to frame the growth that is in front of Roku: “I think if you just think about the drivers of our ad business because some of your questions were about our ad business, the biggest driver of the ad business is not these kinds of details. It's the fact that if you look at TV time in the US today adults 18 to 49 spend 42% of their TV time streaming. But if you look at the amount of ad spend on streaming versus traditional TV, it's only 22% has moved to streaming. So there's this big gap still and that gap is starting to close, but has a long way to go. That — the rate of that closure because they will catch up eventually and the rate of all viewers moving to streaming those are the biggest drivers of our ad business which is a $60 billion opportunity.”It's the fact that if you look at TV time in the US today adults 18 to 49 spend 42% of their TV time streaming. But if you look at the amount of ad spend on streaming versus traditional TV, it's only 22% has moved to streaming. So there's this big gap still and that gap is starting to close, but has a long way to go. That — the rate of that closure because they will catch up eventually and the rate of all viewers moving to streaming those are the biggest drivers of our ad business which is a $60 billion opportunity.”
Magnite Earnings
Magnite has gone through a streak of acquisitions (2020-2021) following a large merger in the previous year (2019) and also a name change. The company uses pro-forma to indicate growth for the combined company of Magnite and SpotX. Ex-TAC refers to revenue that excludes acquisition costs. For our purposes, pro-forma is what the Street will be judging Magnite on moving forward.
If the acquisition comes together nicely then we can reasonably expect this growth to accelerate (assuming the whole is greater than the sum of its two parts). Usually in tech growth, acquisitions are strategic rather than accretive yet financial analysts are only able to model the accretive growth post-acquisition. Magnite’s strategy is to become the strongest supply-side player globally and to circumvent the need to figure out hardware (like Roku) by going direct to publishers.
SpotX was a strong company coming into the acquisition and it almost resembles more of a merger in that regard. This is why pro-forma growth is low in the first year at 26% in Q3 and CTV revenue is up 51% on a pro-forma basis. Compare this to 290% CTV growth if we look at only Magnite last year. SpotX brings about $30 million per quarter to Magnite and this is primarily CTV revenue.
Losses are widening on Magnite to a total of $24.3 million, up from $10.5 million. The adjusted EBITDA margin of 35% is based off ex-TAC revenue, so the comparison to last year pre-SpotX is not as meaningful. Operating cash flow was $34 million in Q3.
We’ve covered SpringServe in the past here and more about Magnite’s product.
Magnite did state in normal conditions, they calculate the growth for CTV ads would have been about 40% year-over-year in the upcoming quarter. “From a comp perspective, if you were to remove political from 2020 Q4, our guide – so our guide straight up is about a 23% year-over-year growth in CTV. If you were to remove the political comp, you get to the low 30s in a year-over-year growth scenario. And if we look at the weakness that Michael mentioned in travel, in automotive, we believe that’s impacting us to perhaps $3 million to $4 million in CTV in Q4. And if you factor that in, we’re at about a 40% year-over-year growth rate.”
For Magnite, the company was more insulated because Android revenue helped make up for any loss on iOS revenue. The company estimates that their exposure is in the single digits and they are seeing a high opt-in rate (likely through the publishers). Magnite’s big risk comes at the end of 2023 when Google will end cookies on Chrome. In the meantime, Magnite plans to build solutions for “first-party publisher segments collected in a privacy compliant manner.” It’s something to monitor but not a concern at this time as it’s two years away.
Magnite’s earnings call is one of my favorites to listen to as the management is often asked industry questions and they give very insightful answers about what they think is going on in the ad industry. Here’s an example of where they were asked about Snap and Facebook.
Analyst: “Hi guys. Obviously, Snap and Facebook felt the brunt of Apple’s privacy changes. And I think you kind of alluded to this and talked about it before, but do you believe ad spend shifted out of social media in totality where you don’t play and into the open Internet or CTV as you talked about earlier? And does that – did that benefit results at all, or it sounds like this could be a long-term issue for social media players. So, is this a catalyst for the open Internet as we push forward?”
Answer: “Yes, Nick, this is Michael. Good question. I believe so long-term, yes. But there is a class of advertisers that really became expert at advertising in the mobile ecosystem and relied heavily upon IDFA that they are not going to be able to shift their spend overnight, right. They are just so used to that ecosystem, the attribution measurements. They have grown to trust their models are based upon from a conversion and lifetime value of the acquisition. All those things have to be reworked not unlike an advertiser that’s lived on Nielsen household ratings and linear TV having to get used to more of the measurement in CTV. And so – so, I think there is no question in the open web will be a beneficiary from that. I just think there will be an evolutionary – an evolution period where these marketers will have to treat their models, get comfortable with new methodology, new attribution and continue from there.”
Here's another explanation:
“We read a lot about the advertisers that have stalled their campaigns or stock spend or decreased spend because they are having a really challenging time working with attribution and customer acquisition costs, etcetera. Those guys are extraordinarily lower funnel. They are extremely sophisticated mobile advertisers. To think that they might jump from that world right into the world of CTV is probably a bit of a stretch. And those are the guys that spend $1,000 a day, $5,000 a day on the Facebook, Instagram, etcetera. But then there is a whole other slew of marketers that do social video advertising, that have much larger campaigns, that also take into account brand attribution, that I think are perfect. And our team is set up for that, Jason.”To think that they might jump from that world right into the world of CTV is probably a bit of a stretch. And those are the guys that spend $1,000 a day, $5,000 a day on the Facebook, Instagram, etcetera. But then there is a whole other slew of marketers that do social video advertising, that have much larger campaigns, that also take into account brand attribution, that I think are perfect. And our team is set up for that, Jason.”
Those quotes are actually bullish for Snap as what Magnite is communicating is that savvy mobile marketers need time to rework measurement and attribution but that doesn’t necessarily mean they will jump ship to CTV ads. They did indicate the push for CTV ads could come from smaller advertisers and we saw this with Roku’s Shopify announcement, as well.
You can read more about Magnite’s private market place here.
Addt’l Earnings Reports Write-Ups:
· Snap’s earnings were covered here. You can view our LTBH webinar on Snap here.
· Twilio’s earnings were covered on the forum here. You can view our LTBH webinar on Twilio here.
· Regarding Atomera, this story centers around the announcement of the JDA. Until that happens, there’s not much to update. Knox is watching this one closely on technicals.
· Bradley covered Sunrun Earnings and ZoomInfo on the forum here.
· Royston covered Cloudflare earnings and AMD earnings here. You can view our LTBH webinar on AMD here. The Dark Horse! Which means a competitor that is greatly underestimated.
· Keep an eye on the forum for a Palantir update today/tomorrow and any others that would require a response on what the ER offered. If there’s a beat, we actually de-prioritize these as it means our thesis is playing out and we prefer to spend our time writing analysis on any misses.