Online Video: Objects in the Mirror May be Closer than they Appear
SEE LAST PAGE OF THIS REPORT Paul Sagawa / Artur Pylak
FOR IMPORTANT DISCLOSURES 203.901.1633 /.1634
sagawa@ / email@example.com
July 12, 2012
Online Video: Objects in the Mirror May be Closer than they Appear
- Online now represents ~9% of U.S. video viewing, and is growing at a 50%+ pace. That audience is attracting advertising, with YoY streaming ad growth accelerating to more than 100% in recent months and suggesting that online will blow past the 5% share of the total video ad market forecast for it at the start of the year. WebTV leaders have pledged $700M+ toward original content, completing the self-reinforcing cycle by attracting even bigger audiences. Against this, TV audiences have stagnated with a tepid ad revenue rebound from the 2009 drop strengthened by the temporary impact of the Olympics and a well-financed campaign season. However, forecasts still suggest cable spending, already 2% of the consumer budget, will outpace household income, expectations incongruous with larger and faster growing items, like health care and education, and the increasingly viable online alternative. We believe that plans to throttle online competition through aggressive pricing and usage policies, or to extract supernormal pricing for broadband will be thwarted by regulation and/or wireless competition. While it will take years for our scenario to play out, we believe that the end game will be apparent before the end of 2013, with obvious consequences for investors.
- Online video viewership continued its sharp rise in 1H12, driven, in part, by the rapid growth in tablets and connected TVs. Netflix reported streaming more than 1 billion hours of video to its roughly 24 million subscribers in June, up 50% in 6th months. YouTube delivered more than 1.2B hours of video to US households in May and total US Internet video streaming is over 4B hours/month, both growing 50%+ YoY. The installed base of tablets in the US has grown over 50% since year end to more than 51M, while the number of TVs able to connect to the internet has risen 30% to 12M, helping to fuel the ongoing rise in the web video audience.
- The traditional TV audience is stagnant, with TV households and viewing hours both in slight decline for the first time in history. The number of U.S. households using TVs to watch traditional channelized television has dropped for the past two years after 20 years of strong, uninterrupted growth. Also unusually, the number of hours watched per person declined 0.5% YoY for 4Q11, yielding total US TV viewing at roughly 43.4B hours/month. As such, online now accounts for more than 9% of all video viewing in the US, with Netflix accounting for 2%+ on its own.
- Online now tops 5% of video advertising, and growing at a 117% YoY pace. 2013 TV ad growth could disappoint 3.5% analyst expectations, 2014 could be even worse. Online video advertising volume was up 117% YoY in May. At the same time, U.S. television ad spending is forecast up 5.1% for 2012, driven by the Olympics and the Presidential election and consistent with the 5.8% growth in upfront ad commitments made a year ago. However, ad commitments for the new season starting in September were up only 1.6%, reflecting the hit to political spending post November, the weak economy and, perhaps, competition from online players presenting for the first time at the upfronts. Given the online video trajectory, 2013 and 2014 could be painful for TV advertising.
- More than $700M total investment in original programming by web players, adds further fuel to accelerate cycle, reminiscent of ‘90’s push by cable nets. Netflix, YouTube, Hulu, Amazon and Yahoo have each committed tens of millions of dollars toward producing original video content for their sites, with dozens of smaller players like Funny or Die, Machinima, Blip, My Damn Channel, and others already delivering their own material. The strategy echoes the efforts by HBO, The Disney Channel, MTV and others to develop their own original content franchises during the growth era for cable networks. We expect this push by the web nets to be similarly successful.
- 3.4% forecast CAGR for cable fees (2.1% of avg. HH income) will be squeezed by healthcare (5.7%) and education (1.9%), and by churn to web video. Cable rates have risen 6.5%/yr over the last two decades, bringing the average monthly video bill to nearly $90, more than 2% of the average U.S. after-tax household income. Typical industry forecasts suggest 3-5% annual growth in video bills through the end of the decade, in line with the long term trend, but improbably given weak forecast HH income growth and the likely continued 6%+ annual growth of health care and education expenses. Negotiations between networks and distributers are becoming contentious – e.g. AMC v. Dish, Viacom v. DTV. As online content becomes more compelling, higher cable prices or reduced channel offerings risk driving cord cutting.
- Carriers may employ usage limits, expensive surcharges and throttling against online video – we expect regulation and/or competition to blunt these tactics. The cable industry often points to its dominant 83% share of residential broadband as a potent weapon to blunt the threat of online video. Imposing usage limits with expensive surcharges or performance throttling for overages could squelch demand for online video, and allow cable operators to favor their own on-demand services. Price hikes on high usage tiers could boost cable profit from online demand, to the detriment of consumers. Given the cable industry’s poor consumer reputation, we believe these tactics will draw considerable regulatory scrutiny. Moreover, we believe that wireless 4G LTE Advanced services will offer viable competition for residential broadband before the end of the decade.
- This scenario, good for online players and bad for cable, satellite and most networks, will play out over 5-7 years, but should become evident to investors much sooner. The self-reinforcing cycle of growing online viewership, rising advertising dollars, and improving programming will accelerate, and is now big enough – i.e. 10% of video viewing, 5% of video advertising – for its impact on traditional channelized TV to be apparent. Audience ratings and advertising revenues for networks may disappoint, particularly once the Olympics and elections are past. Fee negotiations and shifting ad spend may drive some networks to embrace web video more enthusiastically. Cord cutting may accelerate, even with economic recovery. Any of these indicators of the real threat of online video could reverse the current investor enthusiasm for television.
Exh 1: Monthly TV Viewing Hours, All Demos 3Q2007-4Q2011
The Straight Dope
The online video future is coming faster than you think. Recently, CEO Reed Hastings revealed that Netflix had streamed more than 1 BILLION hours of video to its 24 million subscribers during June, lighting the media blogosphere atwitter. A billion hours is a lot. Netflix averaged roughly 650 million hours a month for its 4Q11 – a billion is 50% growth in just 6 months. Nielsen says 284 million Americans watched 153 hours of television per month during 4Q11, for a rough total of 43.4 billion hours – a billion is more than 2% of all video watched (Exhibit 1). Google’s YouTube is streaming a billion hours a month too, and its total hours of streaming are up more than 10% over the past 6 months and up more than 50% YoY. According to ComScore and adjusting for Netflix, Americans are watching more than 4 billion hours of online video each month, up more than 50% vs. 2011 and accounting for more than 9% of the total video being watched in the country.
Viewers attract advertising. US online video ad sales will top $3B for 2012, up 40% YoY, and constituting roughly 5% of total video advertising. Advertising attracts new programming. Google, Netflix, Hulu, Yahoo, Amazon and AOL have collectively pledged nearly half a billion dollars toward developing new original content, added to the programming already being generated by web channels like Funny or Die, Machinima, Blip and My Damn Channel. New programming will attract more viewers, abetted by the huge growth in tablets and in living room online connections via game consoles, DVD players, alternative TV boxes like AppleTV, and Internet-ready TVs. This is the self-reinforcing cycle driving online video.
The 40-50% annual growth trajectory playing across this cycle is bad news for traditional television. The number of US households accessing channelized TV – television households in Nielsen-speak – hit 115 million in 2010 after 20 years of strong growth and has deteriorated since, a phenomenon that the industry attributes to the weak economy but that may be also related to the growth of online (Exhibit 2). The hours spent watching TV, by the 284 million Americans living in television households, dropped 0.5% YoY in 4Q11, a rare occurrence in a long term record of expanding audiences. Television advertising is projected up 5.1% for CY2012, perhaps reasonable given the coming Olympics and a well-financed political season, but 2013 seems vulnerable after just 1.6% growth in advertising commitments for the season beginning in September.
The problem is rolling over into the negotiations between networks and their cable, satellite and telco distributers. Networks, perhaps fearful of coming pressures on the advertising revenue and certainly accustomed to an unbroken string of fee increases, are demanding fee increases across the board. Distributers, accustomed to jamming rate increases down on their subscribers but, perhaps fearful that consumers may not accommodate much more than the 2% of the household budget they are already paying, are pushing back. Something’s gotta give.
The cable industry is counting on their domination of residential broadband to carry the day. Tiered pricing and usage caps, with expensive surcharges and performance throttling for exceeding tight limits, could suffocate online video if left unregulated. Comcast has taken it a step further, exempting its own online streaming offerings from its usage caps, drawing the ire of the online crowd and an inquisitive eye from regulators. Given the stunning rancor that American consumers have for their cable companies, we do not believe that the currently benign regulation of cable broadband will remain if the MSOs remain on this tack. Moreover, as we have written, we believe LTE Advanced technology will make wireless broadband competitive with cable modems before the end of the decade.
While the cycle will play out over the next 5-7 years, we believe the end game will be apparent to investors in the next 12-18 months. Invest wisely.
Exh 2: TV Households and Average Number of Televisions
One Billion Hours
Netflix CEO Reed Hastings left a few mouths agape with his casual Facebook congratulations to content licensing head Ted Sarandos and his team on exceeding the 1 billion hour streaming milestone for the first time ever in June. One analyst calculated that 1 billion hours (minus 100 million presumed to be streaming to international markets) made Netflix the number one video network for its 21 million US subscribers, topping all broadcast and cable nets for the month. Of course, YouTube is also delivering well over a billion hours of streaming video per month in the US, albeit to a much larger audience of 153 million unique visitors (Exhibit 3). Collectively, ComScore pegged total US online video streaming hours at roughly 4 billion for May, likely underestimating Netflix’s enormous gains in its number (Exhibit 4). By this accounting online video streaming is up 50% YoY, with Netflix up 50% in just the past 6 months.
Exh 3: YouTube vs. Hulu, Total Minutes Viewed, January 2011-May 2012
Exh 4: Aggregate Internet Video Viewership, Monthly Minutes
According to Nielsen, 284 million Americans watched 43.4 billion person hours of traditional channelized television per month during the fourth quarter of 2011. Note that 4Q is a seasonally strong quarter for television watching, 2Q viewing is typically off 5-7%. Note also that the Nielsen data accounts for multiple members of a household watching the same show, while ComScore and Netflix merely count individual video streams that may be watched by more than one person. Without adjusting for these issues, online still represents 8.4% of total video viewing. After adjusting the Nielsen numbers down 6% for seasonality and assuming that 10% of Internet video streams are seen by a second viewer, online rises to more than 9.5% of the total video audience.
Exh 5: Aggregate Monthly TV / Online Viewing
Nielsen, which is the primary information provider to the television industry, makes estimates of online video audiences as well. For 4Q11, Nielsen estimated that just over 147 million Americans watched an average of 4 hours and 32 minutes of internet video apiece, for a total of roughly 661 million hours per month (Exhibit 5–6). This is quite curious, given that Netflix alone reported streaming approximately that many hours during each month of 4Q11. YouTube reported streaming more than 1.1B hours of video to U.S. viewers in December 2011. Add in Hulu, Yahoo, the network sites, and others, and it becomes apparent that Nielsen’s estimate that online accounts for less than 2% of total video viewing is not credible, yet TV incumbents cling to the figure as evidence that online is a benign threat. We believe that the shift of eyeballs to online programming is a primary factor in the stagnation of the television audience, and that the continued growth in web video viewership is poised to drive an accelerating decline in the traditional TV audience in the future.
Exh 6: ComScore vs. Nielsen, Online Video Unique Viewers, 1Q10-4Q11
Take Two Tablets and Call Me in the Morning
Using the ComScore/Netflix numbers and assuming that online video viewing grows 40% over the next 12 months while television viewing remains stable, online will rise more 300bp as a percentage of the total audience to nearly 13% on a seasonal and multiple viewer adjusted basis. If anything, we believe this estimate is conservative given multichannel television’s tenuous hold on stability and the dramatic spread of video friendly tablets and devices able to connect the living room TV to the Internet.
eMarketer projects the total number of tablets in use in the US to almost double from over 33.7 million at the end of 2011 to more than 69 million at year end 2012, with a further jump to 99 million in 2013 (Exhibit 7). Similarly, the Leichtman Research Group reports that, as of February 2012, 38% of all US households had at least one TV set connected to the Internet, primarily via gaming consoles, but also through connected Blu-Ray players, alternative set-top boxes like AppleTV or Roku, or the TV set itself. This is up 21% YoY and up almost 60% vs. 2010 (Exhibit 8). Moreover, device platforms are evolving to make the consumption of video across multiple devices, including smartphones, seamless and easy for users. This revolution is a huge driver for the growth of the online video audience.
Exh 7: US Tablet Penetration, 2010-2015
Exh 8: U.S Households with Connected Devices for Viewing OTT Delivered Content
A Word from Our Sponsors
Advertising makes up more than 90% of the revenues for broadcast television networks and more than half of the revenues for the average cable network. In total, advertisers are expected to bestow nearly $70m on multichannel television during 2012, up a robust 5.1% YoY, driven by the quadrennial Olympics and presidential election. Meanwhile, online video advertising spend for 2012 is expected to top $3 B, forecast up 50% YoY by e-marketer, perhaps a conservative estimate given that ComScore reported streaming video ad volume up 117% YoY for the month of May (Exhibit 9-10). Indeed, the online video advertising data shows a sharp acceleration in the hours of ads streamed to on line viewers over the first five months of 2012, suggesting that any 2012 estimate of online video ad spending based solely on the 2011 spending trajectory will be significantly short of reality.
Exh 9: Video Advertising Spend versus Viewership
Television advertising revenues have rebounded somewhat since a sharp fall in 2009. It is important to note that major networks sell much of their primetime inventory during a preseason “upfronts” process that plays out each May. The 2011 upfronts delivered a 5.8% growth in spending commitments for the season just ended, while the 2012 upfronts yielded lackluster 1.6% growth (Exhibit 11). On a calendar year basis, this has translated to forecasts of 2012 television advertising spending growth ranging from 3.5% (Strategy Analytics) to 6.8% (MagnaGlobal), including major boosts from the London Olympics and a fiercely contested and well financed election season. While these events provide significant air cover for traditional TV advertising through November, the tepid 2012 upfronts and the big growth in online video advertising casts a long, dark shadow over 2013.
Exh 10: Monthly Online Video Ads Viewed, January 2011-May 2012
Exh 11: Major Network Upfront Ad Sales, 2000-2012
57 Channels and Nothing’s On
With the passage of the Cable Act of 1984, cable systems were suddenly free to import an unlimited roster of non-local networks to their channel line-ups. This was the genesis of the modern multichannel television service, prompting billions of dollars in infrastructure investment by cable and satellite carriers to expand reach to nearly every household in the country. It also spurred the rise of the cable network – HBO, ESPN, CNN, MTV, Nickelodeon, and The Disney Channel became anchor tenants on the set-top-box grid, greasing the skids for the long list of channels available today. Initially, cable networks largely relied on programming drawn from archives full of reruns and feature films, interspersed with easy to produce news-like presentation shows. With time, a growing audience and support from advertisers, the cable network pioneers began to invest in producing their own original quality programming, beginning with made for TV movies, cartoons and special events, establishing the production-cost friendly reality TV genre, and eventually moving to the expensive scripted comedy and drama series that were once the sole province of the traditional broadcast networks. Today, cable networks command nearly 75% of the multichannel audience and win well over half of the Emmys awarded annually for programming excellence (Exhibit 12).
Exh 12: Network versus Basic Cable Primetime HH Viewership
This history is important background for industry observers that doubt the potential for quality original programming from online video providers. The audience is there – 4 billion hours of monthly streaming, growing 50% YoY and representing roughly 9% of total viewing, puts online video clearly in the mainstream. The advertisers are there – online video ad impressions have more than doubled YoY in recent months, portending that full year 2012 spending will handily beat expectations for $3B and 5% share of the total video ad market. Now the major web video players are following the cable network playbook and beginning to invest in their own original programming.
Exh 13: Online Video Libraries and Original Content
Netflix is launching 6 original series, including the already streaming “Lilyhammer”, a much buzzed about return for the cult classic comedy series “Arrested Development”, and the David Fincher project “House of Cards” starring Kevin Spacey and reputedly carrying a $100 million commitment over two years (Exhibit 13). All in, Netflix has likely committed more than $250 million to these projects. Hulu has announced 10 new series, including a movie discussion show with indie icon Kevin Smith called “Spoilers”, and a quirky travelogue from director Richard Linklater called “Up to Speed”, along with several programs originally produced for British or Canadian broadcast and now exclusive to Hulu. These projects are likely a healthy chunk of Hulu’s $500 million annual programming budget.
Google’s YouTube is taking a broader approach, sprinkling $100 million in seed money across as many as 100 new programming “channels”, including content produced by personalities like Jay-Z, Amy Poehler, Rainn Wilson, and Deepak Chopra, aiming to generate at least 25 hours of new programming every day. YouTube’s approach to content breaks with the 30 minute/60 minute format long established by broadcast television, with short form 5-10 minute videos its bread and butter. Recently, Google stepped up its commitment with an additional $200 million pledged to marketing the new channels, and with a $35 million direct investment in its third most watched content partner, Machinima. Amazon has also taken a “seed money” approach, launching Amazon Studios with a call to independent producers to pitch their ideas for original comedy or children’s series. Chosen projects receive $10,000 to develop their concepts, with the potential for further funding to produce a pilot episode, and, eventually, a commitment to produce the series. Thus far, Amazon has given an initial greenlight to four projects.
Yahoo and AOL have also committed significant funds toward original programming with the aim of bolstering their sagging brands. Along with YouTube, Hulu, Microsoft and music video leader Vevo, Yahoo and AOL participated in April’s first ever “Newfront” event, presenting their online programming to advertisers as alternatives to the television content featured at the concurrent “upfronts”. This is a milestone for the online video business, as it crowds forward to compete with the established broadcast and cable networks.
Obviously, the hit rate on all of these original projects will be considerably below 100%, but it seems likely that some of these series will gain traction with both viewers and advertisers. The exclusive presence of well publicized, popular programming should accelerate the diaspora of the television audience to the Internet, further pressuring traditional networks to consider a more aggressive approach to putting their content on the web. Thus, the self-reinforcing cycle spins on.
You Can’t Have Everything … Where Would You Put it?
The channelized television racket has had a great run, with a cycle of its own. As audiences grew, the owners and producers of content demanded higher fees from networks. Networks then demanded higher fees from cable and satellite distributers, citing their rising content costs. Distributors typically made a token effort to push back, before capitulating and pushing the higher costs down onto their subscribers. As a result, the price of expanded basic video service rose an average of 6.1% per year from 1995 to 2010 according to the FCC (Exhibit 14). Over the same period, the CPI was up at a 2.8% CAGR.
Exh 14: Average Expanded Basic Cable Bill versus CPI, 1995-2010
Given this constant diet of rate hikes and with the industry’s indolent stance on customer service, it’s not surprising that cable MSOs carry a sorry reputation with their customers. In recent years during the March Madness basketball tournament, consumer advocate website the Consumerist has invited its readers to vote for the most hated companies in America. The likes of Comcast and Time Warner Cable make the final four every year, while Comcast won the dubious distinction of worst company in America in 2010. In 2011, 5 of the 19 worst companies in America according to the American Customer Satisfaction Index were Pay TV companies: Time Warner Cable, Comcast, Charter, Cox, and Dish.
Despite their brutal reputation for customer indifference and price gouging, have enjoyed revenue growth and strong profitability, a testament to the lack of competition in the market. The primary alternative to cable, satellite, is available only to households with unobstructed views of the southern sky, requires the installation of an external dish, and suffers from weather related interruptions. Meanwhile, telcos have halted their video investment – if you don’t have them already, Verizon’s FiOS and AT&T’s U-Verse aren’t coming to your neighborhood.
Exh 15: Average Digital Cable Bill as % of HH Income, 2000-12
Exh 16: Forecast Consumer Spend as a Percent of Median Household Income After Taxes
Online video throws a stick in the channelized television cycle. With a viable alternative for video entertainment and with new pressures on household budgets, the trajectory of annual cable price increases is not sustainable. While SNL projects 3.6% growth in Cable ARPUs through the end of the decade, consumers are facing nearly stagnant household income levels. Household income after taxes and CPI are projected to grow at 2.0% and 1.9% CAGRs through 2020 respectively. Expected costs for necessities such as health care and education are expected to grow better than 5% and make up a more sizeable share of consumer spending. Health care spending is on track to make up 8.5% of Household income in 2020 up from 6.7% in 2012 (Exhibit 15–16). Education spend will eat up another 900 basis points of income as it grows. It seems unrealistic to expect consumers to carve out more of their budget to feed the cable industry, particularly with the growing audience for online video.
Ok, but We’ll Get’em on the Internet Service
The universal Plan B for cable operators, should their video service take a real hit from over-the-top web video, is to use their domination of residential broadband to fight back and to recoup lost revenue. Nearly 83% of U.S. residential internet connections supporting download speeds of at least 10Mbps are provided by cable operators. 70% of U.S. households have no alternative to cable modem service for broadband fast enough to support high quality streaming video (Exhibit 17). This is the cable ace in the hole.
Already, cable operators have begun to implement tiered pricing and usage limits with expensive rates and/or service throttling for overages, penalizing heavy viewers of online streaming video. Comcast has gone so far as to exempt its own online streaming services (which must be bundled with a traditional subscription) from the usage limits, bringing cries of foul from the likes of Netflix and scrutiny from the DoJ. With their dominance of residential broadband, cable operators can move to squelch online competition, or to profit from it with much higher pricing on heavy users.
Here, the cable industry is walking a tightrope. The Republican controlled House of Representatives fended off an attempt by the FCC to impose common carrier status for cable modem service, with the agency, instead, settling for weak net neutrality rules that prohibited selective service blocking or performance throttling. Of course, this battle played out largely away from the view of consumers and even if it leaves leeway for aggressive moves to squelch online video or to extract supernormal profits, taking this tack leaves the industry open for a fierce backlash. As noted in the previous section, consumer distaste for their cable providers is just one step short of open revulsion. Political clout with Congress may not offer an effective counterweight to the outrage of constituents over blatantly consumer hostile policies by MSOs.
Exh 17: Residential Fixed Connections by Technology
Moreover, Plan B also presumes that there is no effective competition for cable modem service. Verizon and AT&T have largely capitulated to economic reality in suspending their investment in their hugely expensive broadband initiatives – FiOS and U-Verse respectively. However, wireless broadband is an intriguing candidate. Current 4G LTE technology yields just over 100 Mbps in useful aggregated capacity in each cell site – enough to deliver up to 30Mbps in lightly loaded cells, but subject to significant performance deterioration under heavy traffic. As such, 4G carriers have been cautious with pricing, leaving the prospect of competition for residential customers unrealistic. LTE Advanced, an updated version of the 4G standard that has recently been ratified, will change this. LTE Advanced can increase the capacity of a single cell by more than a factor of 5 – even more if additional spectrum can be employed – and support speeds of at least 100Mbps. This standard will be available for commercial deployment by 2014, and given the potential of new spectrum auctions by mid-decade, it is likely that wireless will compete on par with cable modems well before the end of the decade.
Eyeballs bring advertising, which brings content, which attracts more eyeballs (Exhibit 18). We believe that the end game is clear and inevitable: online video will win and channelized television will lose, in a battle that will take years to play out. While we expect diehards to continue to cast their skeptical eyes until the bitter end, we believe the average investor will join our conclusion sooner rather than later.
Exh 18: Content’s self reinforcing virtuous cycle
The online video cycle has already advanced to where streaming accounts for some 9% of U.S. video viewership and more than 5% of U.S. video advertising. While traditional TV network and distribution stocks have been resilient behind a rebounding overall advertising market, the prospect of Olympic and election related ad spending in the next few months, and generally strong near term cash flows, the online video cycle is big enough and spinning fast enough to seriously threaten industry economics by 2013. The threat will show in shrinking viewership, declining advertising spending and new resistance to price hikes at each step of the traditional television industry chain.
Our perspective suggests a cautious outlook for cable, satellite and network stocks despite the near term advertising recovery and strong industry cash flows. We expect news flow concerning weak TV advertising, the shrinking audience, price resistance by consumers, and/or regulatory scrutiny, will intensify in 2013 and stands to scuttle the bull case for investors. We are particularly concerned for cable MSOs and for network owners that do not aggressively pursue their own online opportunities. In contrast, we believe the potential for video streaming to deliver real revenue growth is underappreciated. To this end, we prefer video businesses supported by advertising, and in particular, look to Google’s YouTube to lead the way (Exhibit 19).
Exh 19: Winners and Losers