Fine-Tuning the YouTube Picture

We have taken an in-depth look at Google’s YouTube, as we believe this moaty business has been underappreciated by the market and may be a more important contributor to holding company

YouTube was founded in February 2005 and quickly emerged as a leading site for video posters and viewers. Google spotted the potential exponential growth of YouTube’s users and content library early on and acquired the company in November 2006 for $1.65 billion, a high price for a service that was earning minimal revenue at the time. The deal was a shrewd move in retrospect, in light of the tremendous network effect the business has built over the past decade.

YouTube is the leading over-the-top, or online, content distributor. Based on various reports, the company boasts more than 400 hours of video uploads per day and 5 billion daily video views. YouTube’s latest monthly average user count stands at 1.5 billion, according to its CEO. We estimate that the business generated $12 billion in gross advertising revenue in 2016. We project total YouTube revenue to grow to over $29 billion by 2021, driven by increases in digital video ad demand and subscriber growth.

Key Contributor to Alphabet's Wide Moat We view YouTube as a moaty business with valuable intangible assets associated with user data; it benefits from powerful network effects between viewers and content providers, both user-generated and professional. If YouTube were an independent company, we would probably propose a wide-moat rating for it. We believe that YouTube's parent, Google, holds significant intangible assets related to overall technological expertise in search algorithms and machine learning, as well as access to accumulation of data that is deemed to be valuable to advertisers. In our view, the same can be said of YouTube. Its video search capability, along with its growing content, has kept bringing back viewers. YouTube's massive user base allows the company to collect rich data about content type and topic viewing, video length viewing, time-of-day viewing, and additional user behavior information. With such data and a large user base, YouTube should be able to offer an attractive return on investment for advertisers and build a growing network of ad clients, similar to what we have seen Google's search function accomplish over the years.

In our opinion, YouTube also benefits from a network effect that creates value for users, content creators, and advertisers. With more viewers on the site today, more content creators will look to YouTube for content distribution. In addition, continuing growth of YouTube’s content library drives further viewer growth. Such growth makes YouTube attractive for advertisers. As most online users continue to “Google it” when conducting general online searches, the video content search results are more likely to point them to YouTube. User growth has created other monetizing opportunities, such as subscription, for YouTube.

Content Is King and YouTube Wears the Crown In our view, YouTube's variety of content reinforces the network effect. As YouTube's content library attracts more users who create additional content, upping the value of the platform for current and new users, the network effect flywheel will continue to spin. As the digital revolution and cord-cutting have progressed, many premium content owners have launched their own over-the-top distribution channels, but we think YouTube has become the hub of nearly all types of content. To generate revenue from content, the owners may begin seeing YouTube and its global audience reach as the most effective content distributor.

YouTube is the unquestionable leader in online video sharing and streaming, in terms of viewers and hours of content available in its library. The company has also made available training and production facilities around the world, creating various communities for the users with more popular videos. These offices allow content creators to add a professional touch to their videos, enhancing YouTube’s supply of content.

Most of YouTube’s content is user-generated and short-form. The company also provides premium short-form content in addition to clips of longer content via agreements with a variety of networks. The YouTube platform has helped some of its users monetize their content via advertising, where users take in around 55% and YouTube keeps the rest. Such a take rate is applicable to all content monetized via ads.

YouTube’s popular video types include celebrity (mainly music), how-to, and gaming videos. In our view, the main reason behind the popularity of such content is that they are mostly short-form videos, especially music video, attracting what we think is the short attention span of younger viewers who consume more content on mobile devices. While some how-to videos may be longer, viewers are more likely to sit through the preroll ad to get the immediate answer to their how-to question. Also, those users more likely began their how-to search on Google, which guided them to the company’s video-sharing platform.

YouTube has also created its own channels for music and gaming content, in addition to content for kids. While YouTube continues to seek improving relationships with musical artists, it keeps working with Vevo, the leading music video hosting company that is owned by the top three music labels. According to various industry reports, the more than 400 million monthly users who view Vevo content do so on YouTube. According to IFPI and Ipsos, 83% of the male and 81% of the female YouTube users worldwide listened to music on YouTube in 2016, with 93% of them between 16 and 24 years old. YouTube Gaming, launched in 2015, hosts videos made by gamers and is gaining some popularity, although it is going up against Twitch, which Google tried to buy in 2014. Amazon AMZN subsequently acquired Twitch for nearly $1 billion.

Attraction of Higher-Quality Content Various television networks use YouTube to attract viewers to their content, while further monetizing clips of their long-form content via advertising and the live streaming of their programs via licensing fees. Stars like Ellen DeGeneres have launched their own channels on YouTube to further market their content and increase engagement with the audience. The Ellen Show channel on YouTube has nearly 21 million subscribers. Other examples include Jimmy Fallon's The Tonight Show on NBC and Stephen Colbert's The Late Show on CBS. In our view, these traditional network programs are trying not only to widen their audience reach, but also to attract a younger audience, which makes up a big chunk of YouTube users.

With an increasing amount of premium content available on YouTube, a lot of it has also been duplicated and reshared on the platform by other YouTube users, leading to copyright infringement. To alleviate these issues, YouTube brought forth its copyright management system, Content ID, which spots what it believes may be copyrighted content distributed by other YouTube users. When detected, copyright owners can block the video from being viewed, take YouTube’s share (45%) of any ad shown with the content, or later charge YouTube based on the content’s viewership.

To protect and monetize their programs, traditional popular networks such as CBS, ESPN, and HBO Go have their own OTT channels, but those don’t attract a mass audience. YouTube’s access to more than 1.5 billion video watchers around the world and more than 180 million in the United States could help those networks increase their viewers and possibly monetization.

After 12 Years, YouTube Remains Young Given such variety of content, young users (16-34) from all around the world are attracted to YouTube, which makes the platform attractive for advertisers. We expect online video consumption to continue to increase within this age group, which is valued highly by advertisers. Overall online video watching in the U.S. has been increasing as users in the 18-24 age group watch at least three hours of video online daily, up from two hours only two years ago.

In the U.S., an estimated 80% of the 18-24 age range uses YouTube, according to a study by GfK MRI. In the 25-34 age range, about 68% are YouTube viewers. Based on a study by Defy Media, we estimate that about 84% of youths in the 13-24 age range view content on YouTube, compared with 91.5% using other social network platforms such as Facebook FB, Instagram, Snapchat SNAP, and Twitter TWTR. We note that the social network figure is likely to be lower if broken down into each specific platform. In fact, in the same study, when youths in that age range were asked which video source they couldn’t live without, 67% answered YouTube and only 48% said social media.

Various studies indicate that the top 10 countries in which YouTube is widely used are the U.S., Brazil, Russia, Japan, India, the United Kingdom, Germany, France, Mexico, and Turkey. According to Global Media Insight, localized versions of YouTube are available in around 70 countries, and content is shown in 76 languages. In the U.S., YouTube has almost a 79% market share in terms of traffic to multimedia sites.

Ad Dollars Going Toward Digital Video With YouTube's huge global footprint and the variety of online ad format inventory it monetizes on its video content, we view worldwide total digital ad spending, which is estimated to be at around $360 billion by 2021, as the company's total addressable market.

Google has begun to monetize YouTube via subscription models, but we expect the majority of YouTube revenue to continue being generated through online ads. While Google does not provide YouTube advertising revenue data, based on estimates provided by eMarketer, we think YouTube generated $12 billion in gross advertising revenue in 2016 and expect that figure to grow at a 15% compound annual rate (slightly ahead of the market rate for digital video ads) to over $24 billion in 2021. With those estimates, we project that YouTube’s current 6% share of global digital ad spending can grow to 7% by 2021.

The emergence and continuing growth and consumption of online digital content has driven growth in digital ad spending at the expense of legacy ad spending types, such as TV, print, radio, and out of home. According to eMarketer, growth in U.S. digital ad spending has easily outpaced that of TV ad spending, and that is expected to continue. In terms of actual dollars, 2016 was the first year that digital ad spending surpassed TV ad spending.

Within digital, video ad spending is continuing to attract more ad dollars. Based on eMarketer estimates, ad dollars in the U.S. allocated to digital video are likely to grow at a 13% compound annual rate during the next five years, which we believe bodes well for YouTube.

YouTube has been generating revenue mainly from inserting ads of various formats along, over, or within the video content. In our view, as YouTube utilizes its user behavior data and overall Google search data, its ad type offerings go along not only with more targeted ad or marketing strategies, but also with initial mass brand exposure ones. Further, ad types can be based on various demographic content data. On the ad production side, YouTube provides support for businesses to create professional-looking ads on their smartphones using the YouTube Director app.

Video ads on YouTube can be categorized as preroll, display, and overlay ads. Preroll video ads (what YouTube refers to as in-stream ads) are inserted and displayed before the main video runs. For YouTube, these include the traditional preroll ads, bumper ads, and TrueView ads. The traditional ones run for a maximum of 30 seconds and cannot be skipped; we believe that these ads represent the highest risk of viewers abandoning the ad’s message and content.

Google’s latest video ad option for advertisers is TrueView prerolls, which can be of any length, with the viewer having the option to skip after 5 seconds. Advertisers do not pay Google or YouTube if the TrueView prerolls are not watched fully or for at least 30 seconds. YouTube’s TrueView offering also includes ad videos not attached to any other content. In these, the entire video or at least 30 seconds of it does not have to be watched; YouTube gets compensated when a user clicks on the ad. While TrueView prerolls are also considered to be ads with the objective to raise awareness, if fully watched by YouTube users, the probability of an overall brand engagement increases, which increases the chances of a transaction, in our view.

YouTube also is now offering bumper ads, which run for only 6 seconds. In our view, these are the most effective top-of-the-funnel ads for companies marketing their brands on YouTube. They get the message to viewers, which is beneficial for advertisers, while minimizing the risk of turning viewers away from watching the full content, which is beneficial for YouTube and content creators.

The company’s display and overlay ad offerings are available only on the desktop platform, mainly because of the smaller screen sizes and lower engagement likelihood on mobile platforms, in our view.

Pricing for preroll, bumper, display, and overlay ads is based on impressions, or how many times they have appeared on or alongside videos watched. Such pricing is measured using cost per thousand impressions or appearances, or CPMs. Pricing on TrueViews is based on how many times the video ads appeared as well as what percentage was viewed fully or for at least 30 seconds. As Google does not provide much data on its CPM or cost per view rates and provides only annual growth or decline rates in cost per click, we estimated YouTube ads’ CPM and applied it in our advertising revenue model.

As of the second quarter, YouTube said it had approximately 1.5 billion monthly logged-in users, up from the 1 billion in 2013. According to various studies, YouTube content is also averaging 5 billion total views per day. When combined with some estimates from eMarketer regarding YouTube’s annual gross ad revenue, we estimate that the business ended 2016 with an overall CPM of $6.58, which translates into $8 average revenue per user.

By applying a similar user compound annual growth rate in 2013-17 to periods through 2021, which also helps increase total content views, combined with a pretty much no-growth assumption in CPM, we estimate YouTube ad revenue to grow from $12 billion in 2016 to over $24 billion in 2021. The main assumption behind our projection is that as a market leader, YouTube will maintain and possibly strengthen its network effect, demonstrated by more users, more content, and higher demand from advertisers. All of this goes hand in hand with what is expected to be a robust and growing digital video ad market.

Facebook More of a Threat Than Twitter While YouTube remains the market leader, other social networking companies, such as Facebook and Twitter, are also trying to grab a piece of the attractive and growing video ad pie. While we think Facebook, with its strong network effect and base of more than 2 billion users, can make headway competing with YouTube, we have strong doubts about the ability of no-moat Twitter to slice out a substantial piece of the pie.

To increase interaction between its users, and possibly attract more celebrities, Facebook launched Facebook Live in early 2016. It launched Facebook Watch, which has more premium and original content, in 2017. In Facebook Live, the company is currently testing the Ad Breaks option, which allows the broadcaster to run 15-second ads while broadcasting live. It is available only to live video providers that reach a minimum number of viewers watching simultaneously. In Facebook Watch, the company is testing midroll ad placements; for the time being, it is not providing any preroll options.

Facebook has been successful attracting premium content from Major League Baseball, A&E, the NBA, and other content providers. It is also funding some original content while attempting to attract more by offering a 55% of gross ad revenue take rate to content creators. While advertisers are still waiting for the eyeball count of Facebook Live and Facebook Watch to increase before purchasing more ad inventory, there is a strong likelihood that this video content will attract ad dollars.

However, we think more ad spending on Facebook Watch may slightly cannibalize ad buys of other formats on Facebook’s News Feed. In addition, ad inventory on Facebook has already reached the saturation point and may turn away users. Further, while Facebook does not provide Instagram’s revenue figures, we do not expect ad revenue from that app to offset deceleration expected in Facebook ad revenue growth. This is mainly because of the lack of enough inventory, as Facebook does not want to turn away Instagram users who can quickly jump on Snapchat. We also think it will take some time for advertisers to be lured away from the already high-ROI-generating Facebook platform to the yet-to-be-proven Instagram.

Twitter acquired live-streaming startup Periscope in 2015. At the time, Periscope had 10 million users, and we would be surprised if that user base has grown significantly since then. Twitter recently started providing preroll ads on Periscope, but with such a small user base, we do not expect ad dollars to significantly shift toward Periscope. Twitter has already integrated many Periscope features into the Twitter platform, which is likely to lower demand for Periscope ad inventory.

A higher percentage of Twitter users access the platform for real-time news than do users of YouTube and Facebook. For this reason, while seeking premium content, Twitter has pursued live events and news-based content more aggressively. Twitter users’ interactions will surround content coming from the NBA, MLB, PGA, Bloomberg, BuzzFeed, Viacom, and many more. While Twitter’s monthly user count of 328 million is far behind YouTube’s and Facebook’s, we think Twitter will attract more mass-marketing campaigns, as most of its content is live event programming. While this bodes well for Twitter, it won’t help the company gain much share going up against YouTube.

Pay Up for No Ads and Better Content While YouTube has successfully monetized its users, it has also understood that some users may be willing to pay for an ad-free YouTube. The company also probably realized that its very large subscriber acquisition funnel will minimize acquisition costs, if any. For these reasons, in 2015, YouTube launched its first subscription service, YouTube Red, which provides all content on YouTube without any ads for $9.99 per month. In early 2017, YouTube launched YouTube TV, basically a no-cord TV-replacement service. YouTube's objective is to attract young millennial cord-cutters to linear TV programming, which remains popular today.

We expect revenue from YouTube’s subscription models to grow at a 66% compound annual rate through 2021, adding $672 million and $4 billion to YouTube’s overall top line in 2017 and 2021, respectively. We also think the subscription models will somewhat lessen Google’s dependency on search ads, resulting in a slight diversification of total revenue. We estimate YouTube subscription revenue to represent around 2% of total Alphabet revenue by 2021, compared with our estimate of below 1% in 2017. We estimate that total YouTube revenue will represent 15% of total Alphabet revenue by 2021 versus about 13% today.

YouTube Red provides access to premium original series programming and movies, and subscribers have the option to save videos to view them offline. This feature can help YouTube gather additional user behavior data as it keeps the users on the YouTube platform and Google’s overall ecosystem even when not connected to the Internet. This option is not available on regular YouTube as it would disable the real-time ad insertion and tracking for YouTube, which could have an impact on revenue.

While premium original programming could increase the value of YouTube’s content library, we do not expect it to pose any threat to other players such as Netflix NFLX. In fact, various studies show that for now, many viewers subscribe to or consume content on more than one OTT platform. According to eMarketer, over 96% of OTT users in the U.S. use YouTube while 64% and nearly 41% also subscribe to Netflix and Amazon, respectively.

In the long run, growth in YouTube Red’s original programming audience may have a network effect as a larger audience can attract more premium content creators to YouTube, expanding its premium content library, which in turn can attract more subscribers, as we have seen with Netflix.

Various news sites have published that YouTube Red had approximately 1.5 million subscribers in late 2016, about 12 months after it was launched--a mere 0.1% conversion rate from YouTube’s user base. Such low conversion rate is partially due to the service being available in only five countries. We forecast only a slightly higher conversion rate partially offset with some churn, resulting in nearly 4 million subscribers at the end of 2021, a 14% five-year compound annual growth rate. We do not expect a change in the YouTube Red monthly rate, which results in revenue growing at the same rate as subscribers. YouTube Red is estimated to generate $252 million in revenue this year, growing to $429 million in 2021.

Connecting Millennials With Traditional TV Content YouTube TV is basically a no-cord TV-replacement service offered in the U.S. While such a service may attract subscribers, we believe it puts YouTube in a crowded space with players such as Hulu Live, Dish's DISH Sling TV, AT&T's T DirecTV Now, Sony's SNE PlayStation Vue, and possibly Apple AAPL. However, with the latest trend of more YouTube digital videos being watched on TV screens, we believe attracting YouTube users to the premium content offered mostly by cable operators is more doable now. In addition, this can lure away advertisers from cable platforms, which are seeing slight decline in eyeballs. We estimate YouTube TV to generate $474 million in revenue this year based on our 1 million-subscriber assumption. We assume 9 million subscribers and revenue of more than $4 billion by 2021.

YouTube TV can tap into YouTube’s large U.S. user base (approximately 180 million monthly users in 2016) to acquire subscribers, which we view as beneficial for a latecomer. In addition, the number of people watching YouTube digital content on TV nearly doubled in 2016 and is likely to grow at a similar rate in 2017. Such a growing living room audience is attractive not only for ad buyers, but also for networks that sell content to YouTube TV.

In our view, YouTube TV’s potentially younger audience will probably attract advertisers. And YouTube TV may increase viewership of YouTube Red premium original programming as the YouTube TV subscribers have access to that content for free.

YouTube TV provides live linear programming, including news and sports, similar to programming offered by cable operators and other OTT pay-TV providers. YouTube linear TV content can be watched on any device, and if recorded, it can be stored in the cloud with practically no storage size limitation. Right now, YouTube TV carries the four major broadcast networks (ABC, CBS, FOX, and NBC), plus 46 cable and sports networks such as ESPN.

YouTube TV is offered only in the U.S. for now and covers 49 of the top 50 designated market areas, which represent nearly 68% of TV households in the country. It is likely to continue to expand its reach as it has gradually since the launch of the service in April 2017.

We expect YouTube TV to generate revenue from subscription fees and sales of some local ad spots. We model $474 million total YouTube TV revenue for 2017, growing to over $4 billion in 2021. Out of YouTube’s 180 million users in the U.S., we assumed 1 million are likely to subscribe to YouTube TV in 2017. Going forward, we assume a 2.5% average conversion rate, resulting in 9 million subscribers in 2021. Although Google remains the digital ad behemoth and YouTube is the leader in video digital ad formats, we assume that the company will need to prove itself to networks and advertisers before a bigger chunk of TV ad dollars is allocated to the YouTube TV platform. For this reason, we applied a lower ad average revenue per user in our model for YouTube TV than what we have seen historically from Comcast CMCSA. We use the same rationale to support our 10% operating margin assumption for YouTube TV in 2017. While we do not expect a decline in content fees, we do foresee an increase in demand for YouTube TV ad inventory, which could drive revenue growth and margin expansion.

Valuing YouTube in Context of Alphabet We estimate that YouTube Red and YouTube TV will add $726 million in revenue to YouTube advertising revenue of $14 billion in 2017, resulting in total YouTube revenue of nearly $15 billion. We expect continuing growth in video ads, along with contributions from YouTube's subscription offerings, to drive 20% YouTube top-line compound annual growth through 2021 to $29 billion. We see margin expansion in YouTube, as there will probably be operating leverage in the coming years, driven by top-line growth, which may include higher take rates on YouTube TV ads, and declining sales and marketing expenses, given the company's large user base.

Our valuation of YouTube is $102 billion based on a sum of the parts, which includes 6 times YouTube’s 2018 ad revenue plus 2.5 times YouTube Red and YouTube TV revenue. The sales multiple applied to ad revenue, which values the YouTube ad business at $99 billion, is based on where we currently value Facebook and Twitter. We used the average of Morningstar’s sales multiples representing fair value estimates of Comcast and Dish Network to arrive at the nearly $3 billion combined valuation of YouTube Red and YouTube TV. While we think YouTube TV has a high upside in terms of revenue and margin expansion through our five-year explicit forecast period, we have applied the lower Comcast and Dish Network multiples to YouTube TV and YouTube Red revenue as we prefer to initially wait and track YouTube TV’s success in attracting subscribers during its first two years. The valuation multiple used for YouTube TV and YouTube Red is also supported by where we estimate Hulu has been valued. In 2016, Time Warner’s TWX 10% purchase of Hulu for $583 million valued Hulu at approximately $5.8 billion overall, nearly 3 times what some have estimated to have been Hulu’s revenue.

We estimate the operating margin will expand to 25% in 2021 from 10% in 2017. For YouTube advertising, we model a 45% gross margin through 2021, given the company’s 55% payout. We assume a similar figure for YouTube Red as we continue to think that most of that content will be user-generated, which will have minimal impact on cost of revenue. We think content costs for YouTube TV will be approximately 65% of subscriber revenue in 2017, higher than what cable operators such as Comcast pay. While we assume YouTube TV will keep paying more for its content, we think the gap will narrow and expect this cost to decline to 55% of revenue by 2021. In our view, YouTube’s network effect is likely to lower the company’s user and subscriber acquisition costs. For this reason, we assume operating leverage driven by declining marketing expenses as a percentage of revenue, helping to widen operating margins.

With our updated estimates of YouTube revenue, we now forecast total Alphabet revenue of $126 billion in 2018, up $1.2 billion from our prior estimate. Our forecast for 16% five-year compound annual revenue growth for Alphabet comes on the back of not only continuing growth in digital advertising, but also the company’s growing share in the cloud market. We believe Google will continue to gain traction in the cloud market, and when combined with Google Play and sales of Google’s hardware products, we see Google’s other revenue growing 40% to $14 billion in 2017. For 2018, we expect 27% growth in other revenue, totaling $18 billion.

Regarding Alphabet’s other bets, we assume that most of this revenue is generated through sales of Nest products, in addition to some from Access/Google Fiber subscribers. Our total other bets revenue estimates for 2017 and 2018 are $1.1 billion and $1.5 billion, representing 31% and 39% year-over-year growth rates, respectively.

We expect the combination of strong revenue growth and a slight deceleration in operating expense growth to create operating leverage for Alphabet starting in 2018. We look for YouTube’s margin improvement and its revenue as a higher percentage of Alphabet’s total to drive some of the margin expansion for Alphabet. We model a 24.5% 2017 operating margin, which is likely to widen to a 26% average the next five years.