Streaming On-line, On-Demand Video: Over-the-Top, On the Way
November 1, 2010
Streaming On-line, On-Demand Video: Over-the-Top, On the Way
- We believe that it is inevitable that internet delivered video entertainment will eventually replace traditional channelized television. However, there are significant technology, infrastructure, behavioral and industry economic hurdles that must be surmounted before “over-the-top” becomes a widespread reality. While many suppose that these obstacles will protect businesses dependant on the channelized model, we see evidence that the barriers are rapidly deteriorating and that market forces, already in play, will likely drive change faster than expected by most investors
- The most obvious hurdle has been network speeds. While almost all US households have access to the 1Mbps sufficient for broadcast quality TV, only 12% are able to get the ~10Mbps necessary for HDTV and less than 1% have access to the 50Mbps+ needed for 1080p BluRay quality. The FCC’s goal of 100M US households with access to 100Mbps by 2020 is, in part, dependent on capex by cable operators under threat in their core business by Internet video. While these MSOs have obvious incentive NOT to make these investments, we believe that the industry would face severe consumer backlash and onerous regulation were it to take an obstructionist tack
- Furthermore, we believe that the long-term potential of wireless networks for residential broadband is unappreciated. The FCC has targeted 500MHz of spectrum for auction, including 120MHz of broadcast TV frequencies to be cleared near term. While initial 4G deployments top out at ~5Mbps, the technology roadmap targets 100Mbps within 10 years, with fixed-point speeds of up to 1Gbps technically viable. We expect 4-6 4G networks to be built out in the US, and that these carriers will successfully compete for residential broadband market share by the end of the decade
- Content delivery networks (CDNs) now cache most popular content in geographically dispersed data centers around the country, drastically cutting delay (latency) and disruption (errors) by reducing the number of routers that packets must pass between server and user. This highly effective architecture is asset intensive and favors scale operators with greater numbers of well coordinated data centers, most of which gain critical mass from their own popular consumer sites – Google, Amazon, Microsoft, etc
- Until recently, most American TVs were hardwired to a cable provided set-top-box without access to the Internet. This is changing. Many new television models now integrate WiFi and Ethernet connections. 28 million households with game consoles from Sony, Microsoft and Nintendo are also connected with links to video content providers. New BluRay/DVD players and alternative set-top-boxes like Apple TV and Logitech’s Google TV offer a third alternative. Finally, the FCC has proposed that cable MSOs begin to provide cheap termination devices, with security but not navigation or DVR functions, for all new installations and equipment upgrades. Thus, nearly 40 million new households annually would be free to choose internet connected devices
- Tablets are also emerging as a significant alternative for video viewing, both in and out of the home. These devices are inherently connected and are not tied to the channelized entertainment model. As channel owners work feverishly to exclude their content from the assault on the family TV, the tablet is an effective back-door to the living room couch, and a gateway drug for couch potatoes to get hooked on on-line video
- Content availability may be the most contentious issue. Cable TV is a chain of aggregators, with content owners, networks, and system operators each leveraging their highest value programming to sell bundles at the highest possible price. As such, users typically find themselves interested in only a small fraction of the programming for which they are asked to pay and fledgling content producers find it very difficult to elbow in amidst all of the bundling. A benefit of the on-line model is a closer relationship between the content that users actually want and the fees that they pay
- As on-line TV rises, the strong hand of the cable industry will weaken and content owners will be tempted to risk its wrath by going directly to the consumer. A positive cycle has begun by which the growing on-line audience is attracting advertisers, which attracts more and better content, which in turn attracts a larger audience. With much of the most valuable programming – e.g. sports, American Idol, etc. – not “owned” but contracted by networks, content owners will have increasing leverage, driving up costs for the channelized model and increasing the potential for Internet defections. First mover advantage in establishing an on-line brand creates an enormous incentive to break ranks, and makes the current industry attempts to control Internet video inherently unstable
- We believe that Internet meta-aggregators will gain advantage in the next 2-3 years by giving users tools to find and buy content of their liking, and advertisers the ability to accurately target audiences. While it will be tempting for companies in control of compelling content to restrict access and to form joint ventures to gain critical mass, we do not think that closed arrangements can gain sufficient scale. Rather, users will likely distribute their loyalty across multiple providers, requiring navigation and other content management tools provided on an overarching basis. Given the rivalry amongst the major video programming networks and systems operators, it seems likely that this valuable role will fall to third parties, like Google, Apple and Amazon, who can also leverage their expertise in content delivery. We also favor CDNs, like Akamai, Limelight, and Internap, and independent content producers, like 19 Entertainment and Lions Gate. We are concerned about companies with disproportionate exposure to the channelized television model
Don’t Touch That Dial!
To date, the channelized television industry – cable operators, satellite broadcasters, traditional networks, cable networks, and programming producers alike – have handled the supposed threat of internet based video entertainment without breaking much of a sweat. Total subscribership remains stable, total hours of television viewing per household is up, and advertising is strong – even in the face of persistent recession. While on-line video has also seen strong growth in viewership, time spent watching, and advertising, the lack of any impact on channelized television has led many observers, industry management included, to conclude that the Internet is merely a complement to traditional TV, rather than a substitute, and if it were to become a threat in the future, the industry would be able to react to preserve its prime place in prime time. Investors appear to be buying it, at least to a point, as cable and network stocks have bounced emphatically off of early 2009 lows.
Nonetheless, questions about 2010 cable subscribership declines dog industry executives, who unequivocally deny that competition from on-line video has had any effect whatsoever (Exhibit 1). Rather, it is the weak economy, according to company exit surveys, that has households turning off their cable. While plausible, it is worthwhile to note that alternative entertainment providers, such as Netflix, YouTube, Hulu, and others, report no pause in their growing audience. If on-line video has not yet made a dent in cable industry subscribership and viewership, it is on a trajectory toward a sizeable collision.
Not So Fast!
From one perspective, it is remarkable that on-line video has come as far as it has. Five years ago, network speeds would have been a substantial deterrent. Only 30% of American households subscribed to broadband internet services in 2005, growing to 47% in 2007, and over 55% in 2009 (Exhibit 2). According to Akamai, the average speed on US broadband connections was 4.7Mbps in 2010, up from 3.6Mbps in 2009, and 2.3Mbps the year before. With 1Mbps the minimum speed for a decent streaming video experience, the potential audience for on-line TV has increased several fold in the past few years, with 72% of US Internet subscribers enjoying average speeds of 2Mbps or more (Exhibit 3). On the wireless side, the improvement in connection speeds is also impressive. US 3G penetration grew from 10% of all subscribers in 2006 to 40% in 2009, while average realized speeds increased from roughly 700Kbps to more than 1Mbps, with new 4G networks offering a five-fold increase once available.
Of course, for many viewers HDTV quality is an important differentiator for television to be viewed on the large screen. For this, broadband connections must sustain a minimum 5Mbps speed, a rate currently enjoyed by 30% of US subscribers, as measured by Akamai. Ironically, the key to bringing faster broadband speeds to a broader audience lies in capital spending plans by the very same cable operators threatened by over-the-top video. Today, cable modems represent 80% of all American residential broadband connections. This may partially explain how the US ranks just 16th in the average Internet speeds experienced by its citizens, behind such global technology stalwarts as Latvia and Romania (Exhibit 4). Cable industry executives, who have suggested to the financial community that they could accelerate cash flow by cutting capital spending now that Verizon and AT&T have signaled “no mas” in their fiber and DSL overbuilds, will need a lot of convincing to close the gap. This apparent conflict of interest is all too apparent to the FCC, which stated its goal that 100 million US households have access to internet speeds of at least 100Mbps by the end of the decade and immediately began searching for a stick to make the cable industry hop to it.
One potential stick is regulation. The FCC has threatened to reclassify broadband as a telecommunications service, thus exposing it to price regulation. Such a reclassification would be immediately challenged by the cable industry, both in court and in Congress, where an expected Republican House majority could work to thwart the Obama-appointed FCC majority. However, we believe that many underestimate the low esteem with which cable operators are held by their customers, and the hue and cry that would rise from foot dragging on network upgrades, oppressive pricing or competitive hardball against over-the-top providers. It is notable that in the wake of an enormous financial sector melt down, an on-line tournament to name the “Worst Company in America” run by “The Consumerist” awarded its “Golden Poo” to Comcast, which beat out Bank of America, Ticketmaster and Cash4Gold in the final four, and leaving high profile foes like United Airlines and AIG far behind in the dust. Should the cable industry drag its feet too hard, it would seem likely that its political air cover could disappear.
We’re On the Air
The other stick is competition. With Verizon having signaled the end of its FiOS fiber optic overbuild after passing under 16 million homes and AT&T U-Verse apparently stalled at just over 10% penetration of the 20 million residences to which it is available, the large majority of US households have no choice for video capable broadband beyond their cable operator. For most neighborhoods, the costs of deploying fiber-based networks to compete with the legacy cable operator are prohibitive relative to the revenue streams that could reasonably be expected to be generated.
The real challenger is wireless. Previous attempts at wireless broadband were based on point-to-point technology that required dish antennae mounted on the sides of houses, line of sight to the transceiver tower, were subject to service deterioration during inclement weather, and had insufficient bandwidth to offer a competitive video product in the pre-on-line-video days. The memory of hype and failure around these business plans colors perceptions of future wireless threats. Indeed, the initial implementation of 4G wireless offers just 5Mbps throughput with poor indoor reception. But that is not the end of it.
The roadmap for 4G takes speed from the initial 5Mbps to better than 50Mbps in 10 years. The technology behind 4G, “multiple input, multiple output” or MIMO, allows reuse of transmission frequencies to send multiple streams to each device multiplying the theoretical capacity of radio spectrum by as many times as the processor on each end can handle. The ITU, which has authority for the 4G standard, has set an expectation that stationary connections should eventually be able to sustain downloads at 1Gbps. With BluRay 1080P programming requiring 40Mbps, it would seem that 4G could handle the future needs of most residences utilizing Internet TV (Exhibit 5).
In addition to the 50 MHz of prime 700 band spectrum auctioned in 2008 for 4G, the FCC plans to make another 500 MHz available over the next decade, 300 MHz of which has been identified and targeted for availability before 2016, including another 120 MHz of 700 band spectrum currently occupied by broadcast television (Exhibit 6). This spectrum is particularly attractive for wireless communications – signals can carry for 6 miles or more from a single cell site and easily penetrate walls. Given the amount of spectrum to be allocated, we believe that new competitors will be drawn into competition and that residential broadband will be a primary target. Foot dragging by cable will not do against competitors able to serve 100 square miles from a single tower with better than 50 Mbps.
No Room For Error (Or Latency)
One major stumbling block for on-line video has been largely neutralized behind the scenes. Streaming video is very sensitive to delays and errors that occur as data hops from router to router across the internet. Each hop introduces latency as packets wait in queue to be processed and sent along, with the potential that congested routers will have to drop packets for retransmission, thus interrupting the stream. Because of this, the difference between a 100 mile trip and a 1000 mile trip could increase the time needed to download a file by more than a factor of 12, a wait of more than 4 hours to download a single film (Exhibit 7).
Content Delivery Networks resolve this by dispersing distributed data centers across the country so that popular content will be located as close as possible to end users. Companies like Google, Amazon and Akamai have invested billions of dollars in building out these networks, and the large majority of streaming media now relies on them to provide a high quality video experience to on-line viewers. We note that the shift to CDN architecture has been extremely rapid. Today 150 networks carry 50% of the world’s internet traffic. In 2007, the same percentage of traffic needed 10,000 networks (Exhibit 8). This shift gives those companies that have invested in the infrastructure and expertise necessary to serve users from a web of distributed data centers a significant advantage over those that have not in addressing the emerging market for on-line video entertainment.
You Can’t Get There From Here
Internet TV didn’t seem like much of a threat to channelized TV as long as it was exiled to the PC. TV networks blithely launched websites to stream their programming on demand as a complement to their daily broadcasts to the living room. TV viewing was trending ever upward, due, in no small part, to the proliferation of TVs in the typical American home. Lap tops with WiFi brought the Internet to the couch, a good thing thought the network operators thinking that they could be used to interact while watching prime time programming. Household TVs were hardwired directly to set-top-boxes provided by cable operators. These boxes offered the main menu for video entertainment consumption, offering navigation and DVR functionality and … NO INTERNET!
OOPS! Those sneaky game console makers started to connect to the Internet, first for multi-player gaming, but quickly, to use the console as a gateway to video content, like YouTube and Netflix (Exhibit 9). OUCH! Television makers started to include Ethernet and WiFi connectivity built into their sets, bypassing the set-top-box and offering one click access to on-line competition. YOIKS! The same manufacturers are building Internet connections into their new BluRay and DVR players, offering another end-run around the set-top-box. EEK! Big scary Google and Apple announced comprehensive platforms to replace all of the value-added functions of a set-top-box and then some (Exhibit 10). Now traditional networks are worried. Some are trying to block access to their web sites from these new platforms – workable for the moment, but easily surmounted by even slightly ambitious users. OMG! The FCC has proposed a mandate that all new cable hook ups and all equipment upgrades utilize an inexpensive termination device that does not provide navigation, recording or any other advanced function (Exhibit 11). Subscribers would be free to use other devices, such as connected TVs, game consoles, BluRay players or GoogleTV as their primary interface to their cable provider.
Take Two Tablets and Call Me in the Morning
The iPad is another end-run around the cable dominated set-top-box. From conception, the iPad was designed for easy, high quality video streaming. The uber-popular tablet ships with apps that provide one-click shortcuts to iTunes and YouTube, and Netflix, Hulu, ABC.com and others are amongst the most popular user downloads to the platform. After the 6 month head start, a raft of tablets featuring Google’s Android OS are beginning to hit the market, with versions based on Microsoft’s Windows 7, HP’s WebOS and RIMM’s Playbook coming just behind. All of these products will tout streaming video as a killer application. Certainly the electronics industry appears to be gearing for a tablet Christmas.
The willingness of network operators to embrace the tablet will likely drive home the efficacy of TV on the web to a nation of skeptical consumers, who will learn the benefits of video-on-demand, targeted advertising, and interactivity while using their iPads outside AND inside the home. At the very least, the rise of the tablet puts at risk the proliferation of additional TVS within the average American household. Why buy another TV for the kitchen or bathroom, and pay for another subscription and cable box, when you can use your tablet for free? A reversal in the trend toward additional televisions in the home would hit a revenue growth driver for the cable industry and affect measured ratings for broadcast programming. Moreover, a nation learning to love TV on the tablet may be more willing to cut the cord on cable programming earlier than many imagine.
Oh, You’ll Buy It Alright … And You’d Better Like It!
The television business is a hard place for a newcomer. Producers with valuable content and good track records bundle programming together in negotiations with networks in order to launch new franchise properties and nurture existing ones. Most networks are vertically integrated, producing part of their own content, and typically, giving favor to in-house programming. Networks consisting of bundled content are then bundled themselves for sale to distribution. Network owners jam cable operators to take ever larger slates of related networks. For example, ESPN begat ESPN2, followed by ESPNews, ESPN Classic and the Spanish language ESPN Deportes. Thereafter came ESPN-U for college sports, and finally the all on-line ESPN 3. Bundling on top of bundling assures little room for true outsiders despite ample real estate on the typical cable system. As such, the cable line up gets filled with pages of channels that few people watch, but operators cannot drop. Consumers and advertisers are then asked to pony up for the expanding mess, and to date, have had no real option to say no (Exhibit 12).
Barney Versus Godzilla
At first Internet TV was a shiny new toy for television networks. The on-line audience was seen as entirely complementary to the nation of couch potatoes, and programmers tripped over themselves to establish their web presence. NBC-Universal (GE), Fox (News Corp), and ABC (Disney) created Hulu as a joint venture to aggregate programming from their network families for distribution over the Internet through on-demand streaming, while simultaneously maintaining streaming sites dedicated to their own brands. CBS launched TV.Com to aggregate its own content with that of Viacom (MTV, Nickelodeon, etc.), with links to other network sites (Exhibit 13). In a world where streaming video quality was just okay and availability was limited to desktops and laptops, the scenario seemed as threatening as an episode of Barney.
The rapid pace of Internet time has seen Barney evolve into something more like “The Land of the Lost”. On-line video is showing up on living room TVs to siphon off audience from nightly network viewing, and from the perspective of the networks, “It must be stopped!” Thus, Hulu, ABC, CBS and NBC have blocked their content from the fledgling Google TV, despite merely passing users on to these networks’ own sites where they are free to sell ads and subscriptions, as they are from any other browser. While we suspect that these networks have just signed on to police the Internet for intermediaries that can easily mask the identity of Google TV users looking for Hulu, in the long run it is not a viable strategy. While networks may try to strong-arm Google or Apple on their flagship Internet set-top-boxes, they cannot stop all of the ways in which an internet connected TV can get at their content. The ugly history of digital music is an object lesson in the efficacy of trying to hold back the web. These content owners know this, and we suspect that they will join the enemy after wringing some modest face-saving concessions from Google.
Round and Round It Goes, Where It Stops, Nobody Knows!
The process by which Internet TV is gaining traction has begun slowly, but it is a self-reinforcing positive cycle which will accelerate and gain strength. Even largely restricted to the PC, on-line video has gained a large audience of regular viewers. The number of videos served each month by YouTube and Hulu has doubled over the past year, with over 15 BILLION videos served per month between the two services (Exhibits 14 and 15). This viral growth in audience, and the inherent advantages of a medium where the individual user can be identified, characterized, targeted and engaged interactively, has subsequently attracted advertisers. US Internet video advertising has grown from essentially nothing to over $1 B in the past 5 years, and is poised to continue high double digit growth in years to come. Advertising, of course, attracts content, and content owners that eschew this increasingly important venue will harm their own brands as others build recognition, scale and experience in the new medium. Of course, new content and easier access via TV sets and tablets will attract more viewer eyeballs, which will draw more advertising dollars, which will, in turn, convince more content owners to join the party (Exhibit 16).
For consumers, this could be an enormous boon, a choice between the scheduled, channelized content that they already enjoy and a vast library of programming available on demand over the internet. Sophisticated search tools would help sort through the panoply of options, while advertising would be increasingly tailored toward individual interests, increasing its usefulness and appeal. Advertisers would also gain from choice and from the increased capability of on-line delivery to target and engage consumers. New programming creators would have an opportunity to build their audiences directly, while owners of the most compelling content would have the option of cutting out layers of middlemen to go directly to consumers. The burden falls on the cable system and network aggregators, who profit from a gatekeeper role that will be superfluous in an all on-line world.
You’re Not the Boss of Me!
It is typically presumed that television networks “own” the programming that they broadcast. While strictly speaking this may be true, it obfuscates an enormous risk for these aggregators. In large part, the most valuable content on traditional and cable networks is not conceived and created by the networks themselves, but rather, purchased under contract from a largely independent base of individual producers and performing talent. This distinction is important. Few programs remain popular and powerful for more than a few years, and those that do typically result in substantial renegotiations for the division of the spoils of success. As on-line entertainment becomes an increasingly viable alternative to channelized TV, the creators of content will gain considerable leverage in these negotiations.
In a digital world, where piracy drastically reduces the shelf life and value of catalog content, syndication will likely cease to be the goldmine that it once was – how much can you charge consumers for Seinfeld reruns that are available for free? New content, and in particular, time-urgent content that is not susceptible to either piracy or DVR commercial skipping, will be increasingly valuable. This is not good for traditional networks. For example, live sports, considered the most valuable programming on television, are NOT owned by the networks that broadcast their games and matches (Exhibit 17). Networks may control their broadcasting rights through contracts, but these contracts, typically less than 10 years in length, expire and inspire frenzied bidding. Already leading sports teams and leagues have begun to develop their own networks – e.g. The NFL Network, YES, NESN, and others – to minimize the amount of advertising revenue that must be shared. Each time these contracts come up for negotiation, the price will go up, particularly if the networks aim to control rights to on-line broadcast as well. Eventually, we believe networks will no longer be able to afford to pay for exclusivity.
Similarly, talented content creators – the J.J. Abrams, Seth MacFarlanes, Simon Fullers, Mark Burnetts, Jerry Bruckheimers, et al. – are essentially free agents. Just because hits like Lost, Family Guy, American Idol, Survivor, and CSI have been financed and marketed under the auspices of network contracts doesn’t mean that the next round of hits won’t find alternative distribution through emerging Internet brands. In this, channelized networks that deliberately set roadblocks for on-line viewing heighten the risk that they will emerge irrelevant on the other end, like so many record labels and newspaper publishers. Some network executives have already broken ranks to step bravely toward the brave new world, insuring that those that cling to a purely defensive posture will take the worst of it if the scenario we envision comes to pass. As such, we believe that the current antagonism toward Google TV, Apple TV and other meta-aggregation initiatives as the whole of the content world acquiesces to the inevitable.
No Man is an Island
The Internet is a big place. Early in the evolution of the World Wide Web, there was philosophical show down between one approach in which service providers would curate the Internet to provide an organized experience in a so-called “walled garden” vs. search engines that would give users tools to find what they wanted amidst the chaos. Played out in public, the battled ended quickly and decisively in favor of the search engine.
As video content migrates to the web, various players are preparing to fight a similar philosophical battle. Cable operators have fired a salvo with TV Everywhere, a service by which subscribers to cable service can watch the same channels streamed to them over the Internet, as long as they can be authenticated as paying customers for cable service. While it sounds interesting in concept, cable operators must negotiate with channel owners to include their content in the service – many have balked or demanded extra compensation, as these players have aspirations for their own web brands. As such, the economics are uncertain and the uptake has been unimpressive. Ultimately, we expect TV Everywhere to be just one of many sources for the same content, albeit with added restrictions for consumer use, a less-than stellar brand reputation and a high price tag to boot given the requirement to maintain a wired cable subscription.
Content owners are working to build their own web brands. Hulu, the joint venture between NBC, ABC and Fox, has successfully built a base of millions of free and paid subscribers. However, American viewership is now spread far beyond the big four broadcast networks, and Hulu may not aggregate enough content to become an appropriate “home page” for Internet TV consumers, particularly if the actual content producers are emboldened to control the fate of their own programming on-line. CBS has countered with TV.com hosting its own programming and linking to content on other network sites. Netflix has an impressive set of movie studio deals and a non-trivial collection of TV content as a part of its subscription deal (Exhibit 18). Meanwhile, a world of smaller network and independent content is up for grabs.
Finding Needles in Haystacks
We believe that households will find need for tools to help navigate the mass of video content coming to the web, and that advertisers will value companies that can help them target their messages with the greatest accuracy and effectiveness. We also believe that no one aggregator will be able to control enough content to become a dominant navigator on that basis alone. Rather, it will come down to who will be able to help me find content in the most effective way and then serve it to me with the highest quality. Search will involve informed recommendations, intelligent search based on incomplete queries, and even image based search. It will have to mine content from all available sources, rather than just those under direct control. Service quality will depend on large, well managed networks of distributed data centers. We term the companies that will perform these functions meta-aggregators, and believe that there will be room for more than one player to thrive.
This description sounds a lot like Google, which has been preparing itself for this role before most of the cable industry realized that it would be necessary. It also sounds a bit like Apple, Amazon and to a somewhat lesser extent, Microsoft and Yahoo. It does not sound like Comcast, Time Warner Cable or Cable Vision. For that matter, it doesn’t really sound like Hulu or Netflix either. If anything, the defensiveness of the cable industry and the indecision of the major network owners increase the likelihood that the ultimate online TV landscape will be defined from outside the traditional industry boundaries (Exhibit 19).
In this context, we see the Internet/Technology crowd rolling over the traditional media landscape. The timing is uncertain – stall tactics like TV Everywhere and blocking access from Google TV forestall the inevitable, while content contracts run their course on their way to renegotiation. We would not predict a tsunami of cord cutting over the next 5 years. In fact, we would expect a modest trickle building to a small, but significant stream. However, more important will be the shift of eyeballs and the further implications. Nightly audiences spend more time on-line, advertisers defect, and content owners renegotiate and experiment with their own on-line projects. From one perspective this smells like a positive reinforcing cycle in its early stages, but from the opposite vantage point, it could be a death spiral.
We see the potential meta-navigators – e.g. Google, Apple, Amazon, Microsoft, Yahoo, etc. – as advantaged in this coming vortex. CDNs will also play a critical role – e.g. Akamai, Google, Amazon, Microsoft, Limelight, Internap, etc.. Anyone who benefits from the rise of video on the web – e.g. Cisco, EMC, F5, Citrix, etc. – should also gain advantage. Traditional distributers – e.g. Comcast, Time Warner Cable, Cablevision, DirecTV, Dish Network, etc. – probably have the most to lose. Traditional set-top-box makers – e.g. Cisco, Motorola, etc. – will have to transition to the new model or face significant declines in that line of business. The fate of content owners – Disney, News Corp, CBS, NBC-Universal, Viacom, Time Warner, etc. – will depend on their assets and execution.