Market Impact and Indicators of Over the Top Television in Canada: 2012

Prepared for:

The Canadian Radio-television and Telecommunications Commission ("CRTC")

Submitted by:

Peter H. Miller, P. Eng., LL.B.

and

Randal Rudniski, CFA RJR Communications Insights Inc.

March 30, 2012

Foreword

This report was commissioned by the Canadian Radio-television and Telecommunications Commission ("CRTC" or "Commission") in February, 2012, to provide an examination of financial, market share, regulatory, technological and other metrics in the competitive balance between over the top television providers (OTT) and regulated Canadian TV providers

The report reflects the research and views of the Authors, and should not be construed as representing any views of the Commission.

Introduction

Barely a day - and certainly not a week - goes by without an announcement or pronouncement on something to do with competitive online or "over the top" television (OTT): Netflix's growth, evolving strategy or demise; Google, Apple, Intel; and consumers doing this that or the other with OTT or TV. "Americans Now Watch More Online Movies Than DVDs"^{Footnote 1} being a recent example. All provide interesting tidbits of information or speculation, but little real understanding.

This is of course hardly surprising. OTT is an incredibly new phenomenon. In Canada, it effectively only came into existence with the launch of Netflix in September, 2010 (barely 18 months ago, at time of writing) - a Television service with no Canadian infrastructure or regulatory obligations that now counts 10% of Canadians as subscribers.

We have become used to these lightening fast developments on the web: Facebook has almost a billion users worldwide, 8 years after launch; YouTube, launched in 2005, now has over 800 million unique users each month and over 4 billion videos viewed a day.

What takes longer to figure out is the impact of these new developments. Do they replace and crush incumbents, are they incremental, or will they slowly but surely replace the old models and old media over time?

As far as OTT and "old" media is concerned, there is certainly no immediate crushing, there is evidence of increased consumption and incrementality, but little consensus on what the future will bring. In fact there is a dearth of both reliable data and the analytical frameworks necessary to make such determinations.

The purpose of this report is to bring some precision to the analytical framework necessary in the Canadian context. We analyze financial, market share, regulatory, technological and other metrics in the competitive balance between OTT and regulated Canadian TV providers (I.e. Canadian broadcasters and BDUs), with a view to identifying:

1. Key indices and metrics which are likely to affect the competitive balance between OTT and regulated Canadian TV providers on a going forward basis, and which therefore might most usefully be tracked by the Commission;
2. The current economic status of OTT in terms of its competitive advantages and disadvantages vis-à-vis regulated Canadian TV providers; and
3. Potential tipping points that would see a material shift in competitive balance, if not viability, along with potential competitive impact scenarios, particularly with respect to Canadian content.

The sections that follow broadly address these three sets of issues.

1. Key Competitive Indicators

In our view, useful indicators in tracking the potential impact of OTT on regulated Canadian TV providers include a combination of technological, consumption and business measures. Within these three broad categories, we see different levels of utility, based on the extent to which indices can be seen as direct or indirect measures of potential OTT impact, how easily they can be quantified, and the availability and reliability of data.

This section provides an overview of the potential indicators that we identified, a discussion of their utility and some interpretation of implications, within the three broad categorizations noted above.

From this analysis, we propose three sets of indices, which we recommend that the CRTC consider tracking to a greater or lesser extent:

1. Core indices. These are a limited set of indices, comprised of primarily direct measures that are, or can be, fairly reliably measured These are indices that we believe the Commission can and should track on a regular basis (some, at least quarterly).
2. Secondary indices. These are a more expansive set of indices, often more indirect in nature, and often with less available or reliable underlying data. While less illustrative of current impact, many of these indices are however potentially predictive in nature, and therefore worthy of tracking to the extent feasible.
3. Qualitative indices. These are indices that are less quantitative then the other two sets, and are therefore hard to assess on the basis of a simple numeric measures. They are, however, for the most part, direct indicators of the potential impact of OTT on regulated Canadian TV providers, and therefore important to track, at least to the extent that public data is available.

We conclude with tables of the proposed indices including guidance on data measurement and availability. For many of the indices, we provide sample data in anecdotal, chart or graphical form throughout the report.

1.1 Summary of Potential Indicators

Our identification of potential indicators started with a list of under 10 measures. Over the course of a few weeks of research, it grew to over 50 -- despite the fact that certain "possible" indicators were rejected at the outset as being too tangential, or too difficult to track, for reasons of complexity or lack of data.

To make our cut of "potential" indicators, a measure had to say something relevant (direct or indirect) about the competitive environment as between OTT and regulated providers, and be at least theoretically measurable. Our list of potential indicators^{Footnote 2}:

A Technology reach and use

B Viewing and Consumption

C Business - Financial

D Business - OTT Competitive environment

Technological

Technological metrics are the early warning system. Some provide a measure of the technological capability of the system to accommodate consumer demand for OTT -- indicators such as the capacity of ISP networks, and the penetration of Smart TVs and OTT-compatible boxes (game consoles, blue ray players etc.). Others provide the most basic indicator of consumer response, from growing OTT penetration, to, potentially, declining BDU subscription levels.

The last few years have demonstrated that the technological barriers to OTT have largely fallen to the wayside. Foremost of these is the very capacity of the Internet to accommodate video consumption of a TV scale and quality. While it is a given that the Internet can not absorb TV levels of traffic today, a gradual but persistent shifting of traffic from BDU to Internet delivery appears to be eminently absorbable^{Footnote 13}. For example, if recent 40% annual capacity growth levels were to continue, in a decade the Internet would be able to accommodate 29 times as much video as it does today^{Footnote 14}.

It is also worth noting that, while cable BDUs expect to migrate to IP based protocols within as little as 2 years, they are likely to maintain analog delivery of basic services for many years after that. Digital or analogue, simultaneous live airings of major TV events, be they sports or the first airings of prime time shows, are far more effectively delivered one-to-many than one-to-one. Indeed, the one thing the Internet cannot accommodate now, or in the foreseeable future, is simultaneous delivery of mega-mass audience events. If only for this reason, regulated TV and BDU subscription will remain vital for sports fans.

Mass live TV events aside, technological impediments to OTT rest with the twin, but relatively easily remediable, requirements of:

1. Broadband penetration; and
2. OTT enabled set-top box (STB) penetration

CRTC 2010 statistics place Canadian broadband (1.5 Mb/s and higher) availability at 96% of Canadian households^{Footnote 15} and penetration at 70%, up from 62% in 2009, and no doubt higher today^{Footnote 16}. Interestingly, according to a report that used 2009 numbers, at 74%, Canada had one of the highest high-speed (128 Kb/s and higher) penetration rates in the G7^{Footnote 17}. (While a high speed Internet connection is sufficient to watch regular online video, a broadband connection is generally acknowledged as the minimum necessary for OTT^{Footnote 18}.)

Thus, not only is Canada advanced compared to its peers, other than a small percentage of rural homes, universal penetration of broadband is foreseeable^{Footnote 19}.

Similarly, Netflix, in particular, has proven that the consumer device side of the equation can be largely solved through the ubiquitous nature of a variety of Internet enabled boxes, whose initial purpose may have had little or nothing to do with watching TV. By tapping into Wiis, Xboxes and blu-ray players, Netflix has allowed its viewers to subscribe without purchasing additional equipment^{Footnote 20}. Inexpensive, easy-to-install, dedicated OTT boxes (AppleTV, Boxee, Roku, GoogleTV) are available for consumers who don't have a pre-existing Internet enabled box, or want a broader Internet TV experience^{Footnote 21}. Moreover, the use of a laptop connected to an HDTV, or an iPad or other tablet as an additional household OTT screen, is increasingly apparent.

While little in the way of Canadian numbers are available, based on available U.S. data, it is reasonable to assume that more than 50% of Canadians are already equipped with the necessary STB or other terminal equipment to consume OTT on their television sets^{Footnote 22}. Some analysts argue that the rollout of Internet enabled TVs will accelerate OTT further. We see minimal incremental boost given how high this is and the fact that setting up a Wii or AppleTV to get Netflix is no harder than setting up a Smart TV.

Thus, while measures of technological capability remain worth monitoring to some degree, it is increasingly OTT vs. BDU/TV adoption that is the more material technological factor. Moreover, it is the spread between the two that may be most telling.

Today, pay TV penetration stands at 24% of BDU subscribers, the average specialty service at 75% and, of course, OTA TV at 100% of BDU households^{Footnote 23}. Translated into all households^{Footnote 24}, these numbers become 22%, 70% and 99%.

With Netflix now at 10% of households^{Footnote 25}, the spread between it and pay TV is around 12% of homes. Fall 2011 numbers revealed the first concrete evidence of pay TV subscription declines, albeit modest ones^{Footnote 26}. Should Netflix's growth continue to remain strong, the service could well meet or exceed pay TV's levels within the next two years^{Footnote 27}.

The notion that Netflix is having a competitive impact on pay TV should not be a surprise. The potential for it to surpass pay TV in subscription levels and revenue generating capacity introduces an entirely different degree of competitive impact, that will be discussed further in the last section.

The spread between the penetration of OTT and specialty TV, and certainly BDU subscription levels or OTA TV, gives cause for comfort, when the OTT service in question is Netflix. When that OTT service is YouTube, and in particular its forthcoming TV channel version, that comfort level evaporates, particularly for specialty channels with much fewer subscribers and little unique or original programming.

Consumption

Consumption measures clearly provide the most concrete evidence of OTT adoption, but not necessarily, impact. Obviously, an OTT service that replaces the DVD rental outlet or is used for incremental media consumption is of far less consequence to regulated TV providers than one which causes Canadians to drop pay and specialty TV services, or disconnect from BDUs altogether.

Today, evidence of OTT consumption is relatively thin. Current publicly available research tends to be based on survey data that seeks to determine the percentage of Canadians who have watched Internet video or watched a TV program online. In our ever-expanding consumer electronics/media universe^{Footnote 28}, there remains much evidence to suggest that online activities tend to be incremental, if not co-existent to TV viewing^{Footnote 29}. Indeed claims that Canadians watch less television often prove no such thing^{Footnote 30}.

Nielson, BBM and survey data may differ in terms of estimates of TV viewing, but all suggest that, at least up to now, TV viewing has remained relatively steady at 19 to 29 hours per week over the last decade, despite increases in online activity, to between 10 and 11 hours per week^{Footnote 31}.

Whether TV viewing has declined from what it otherwise would have been, substitutable OTT viewing is statistically insignificant, or OTT has truly grown the TV viewing pie is simply not determinable from currently available data.

Moreover, recent survey data on online video consumption (there is even less data specifically on competing OTT consumption) suggest little more than an appetite for OTT, which may well be statistically insignificant compared to TV viewing:

• Media Technology Monitor (MTM) 2010 data which revealed that 50 of Canadians watched Internet video in the previous month and 22% watched a TV program online in a typical week^{Footnote 32}
• CTAM Canada July 2011 data which revealed that 93 per cent of Canadians are aware of Netflix and 48% claimed to have watched video content online in the previous month^{Footnote 33}
• comScore October 2011 data which revealed Canada as #1 in the world in the category of number of online videos watched per viewer per month, at 303 videos per viewer, and #2 in the world in terms of the percentage of internet users who watched online videos during the month at 90.9%. (ahead of the U.S. at 286 and 88%, respectively)^{Footnote 34}

While such numbers help understand the propensity or willingness of Canadians to use the Internet for TV and other Video, they say little about actual consumption, let alone potential impact.

There is some available research that attempts to quantify online video viewing. For example, comScore 4Q 2012 figures put 3 month viewing at roughly 30 hours, almost double that of 4Q 2010^{Footnote 35}. According to MTM, "adopters", viewed 2.6 hours (anglo) and 1.5 hours (franco) of online TV in a "typical week" in 2010^{Footnote 36} -  higher than the 19-20 hours of online video per 3 month period reported by comScore for the engaged 18-24 and 25-34 demographics in 2011.

More recent MTM November 2011 data revealed that Netflix subscribers watched only slightly less regular TV than average – 13 hours per week (plus 5.6 hours watching Netflix), compared with 15.8 hours for non-Netflix subscribers^{Footnote 37}.

Recent Nielsen U.S. figures put overall online video usage at just over 5 hours per month, based on December 2011 data. While YouTube was by far the most watched destination, Netflix viewers had the highest consumption patterns, at just over 10 hours per month, compared to Hulu and YouTube viewers at just over 3 hours^{Footnote 38}. According to Neilsen, by the end of 2011, roughly one‐third of U.S. consumers streamed long‐form content such as a movie or TV show from the Internet through a paid subscription service like Netflix or Hulu‐Plus^{Footnote 39}.

Neilsen's figures are significantly lower than those reported for comScore, which put U.S. Web surfer of online video content at 21.1 hours in October, 2011, up from 16.8 hours in June, 2011^{Footnote 40}. Netflix said its customers watched 2 billion hours of video in Q4 2011^{Footnote 41}, which works out to an average of approximately 1 hour per day per subscriber.

The differences in methodology, and consequential differences in survey results, make it virtually impossible to draw firm conclusions on OTT consumption, other than the fact that it appears to be growing significantly (if not exponentially) but remains, at this time, an order of magnitude lower than that of TV.

Given the lack of timely and, possibly, reliable data, this is certainly an area where the CRTC may wish to coordinate with research companies to get a better understanding of methodology, and possibly even commission its own research. As research conducted under different methodologies cannot be easily compared, the best way of assessing developments is to compare research conducted under the same methodology, survey to survey, and under the most identical conditions possible to avoid seasonal and other variations. As consumer polling vs. set-top box or people metering each have their advantages and disadvantages, doing or tracking at least one of each may also be advisable.

Business

Business measures cover a range of metrics that help to assess the financial viability and competitive advantage or sustainability of OTT vis a vis regulated TV providers.

Some of the suggested metrics are specific financial ones --measures such as revenue and programming budget, which are discussed in detail in the next section of this report. Others are equally specific, but more reflective of the nature of the business  --hours or percentage of original content; delay between broadcast and OTT exhibition. Such indicators speak more to the competitive environment at play, than the financial state of the players.

Finally, other metrics, such as the health of the Canadian rights market, or levels of piracy, may not be directly measurable per se, but need to rely on certain proxies that sometimes only become evident after the fact.

For example, the loss of a significant rights package, be it Canadian pay TV's loss of the Paramount movie package to Netflix in March 2011, or the speculation that Yahoo may outbid Bell and other Canadian regulated TV providers for rights to the 2014 and 2016 Olympics^{Footnote 425}, certainly indicate greater competition between OTT regulated TV providers, but they do not necessarily presage the loss of subscribers to OTT.

That said, an increasing number of well financed foreign OTT players, with an increasing array of substitutes for regulated TV offerings, would be a strong indicator of trouble for Canadian regulated TV providers, depending, that is on the nature of the Canadian industry's competitive response.

To date, Canada's vertically integrated BDU/broadcasters have exhibited a progressive approach to following viewers online, with a strong offering of free catch-up viewing opportunities for virtually all prime time network shows, and many day time and specialty programs  --most recently supplemented by well-designed consumer friendly iPad Apps. There is an argument that such a strategy, while good for the broadcaster, is bad for the BDU  --making cord cutting easier, as a viewer with an OTA antennae and a broadband connection can now receive an excellent selection of the very best (i.e. conventional ) TV shows, live and on-demand, for nothing. Evidently, such an argument has not held sway; be it in the name of experimentation or maintaining the broadcaster's exclusive rights franchise^{Footnote 43}.

Going forward, the number and nature of OTT offerings sometimes seems limitless; every week seems to bring a new speculation or announcements. BBC launched a Canadian equivalent of its iPlayer in Canada in the fall of 2011, a niche service that nevertheless will appeal to anglophiles who can't get enough Britcoms. Zip.ca continues to suggest it will launch an online streaming service, but almost three years after initial announcements, the move appears less certain, especially given recent musings over the unfavourable cost of streaming over DVD delivery^{Footnote 44}. In the U.S., Intel, the chip maker, is reported to be developing an OTT service that it has been pitching to media companies as a "virtual cable operator"^{Footnote 45}.

Of all the recent OTT announcements & speculation, it is YouTube's spending of $100 million^{Footnote 46} to finance and launch roughly one hundred new original content "channels" that may be the most significant^{Footnote 47}. After a 2011 RFP process, YouTube selected a mix of channels in the sports, music, news, education, lifestyle and pop culture genres from both celebrities (e.g. Madonna and Jay-Z) and established, successful internet video brands (e.g. Machinima, and Ted.org), which have been launching since the beginning of 2012^{Footnote 48}.

What is notable about this move is both its scope and its business model.

The channels clearly expand YouTube's current scope from largely amateur, low quality comedic video clips to higher quality, more professional, longer-length programming in a broad array of genres. While often more niche than equivalent specialty TV, many of the channels hit exactly the same genres: Drive, Car and Driver Television, CafeMom (lifestyle), Look TV (fashion), a Pet Channel, a Food Channel, Dance on, another Dance Channel and so on.

Designed to draw YouTube viewers in, increasing their time spent, will they supplement or replace Specialty TV viewing?

YouTube's upfront 100% financing of the channels and their original programming is structured as an advance on advertising revenue, of which the channel provider will receive a 55% share after YouTube has fully recouped. All channels are free to the viewer. Content providers retain copyright, with YouTube's licencing deal being exclusive for 18 months and non-exclusive for a subsequent 18 months.

The issuance of successive RFPs will no doubt depend on the success of this approach. With sufficient return on its initial $100 million investment, YouTube would be able to virtually self-finance its OTT ventures, into more and more areas that compete with regulated TV providers.

And YouTube is not the only established online video provider looking to provide OTT. Yahoo, with its 60 million unique video users in the US, is also looking to deepen its user engagement with an even greater reliance on movie and TV-like content^{Footnote 49}.

1.2 Recommended Indices

The following charts provide our recommended core, secondary and qualitative indices.

We identify the metric, its type or category, the recommended measure, frequency of measurement and potential source of data, as well as some brief commentary, as appropriate^{Footnote 50}.

Notes to charts

* Denotes a metric for which at least some data or charts have been provided elsewhere in this report

• For the purpose of this list we have assumed aggregate #s in the category identified. We recognize that disaggregation into English and French-language markets, individual vs. a representative sample vs. all would be potentially useful in many areas.
• The penetration of basic carriage services in English and French Canada (CBC's RDI and Newsworld; Pelmorex's Weather Network and Metromedia) may provide one of the most comprehensive indicators of potential cord cutting, and could be obtained by the CRTC on a "one stop shopping" basis from the licencees concerned
• Some of this data the CRTC already collects (in annual returns), but not with the frequency or aggregated as recommended; some of it the CRTC does not currently collect.

1.3 CRTC data collection

An important consideration in proposing these metrics is that data be available to, or collectable by, the CRTC. In many instances, the Commission may be able to collect reliable data from publicly available sources^{Footnote 51}. In others, the Commission would have to specifically request it.

We see two types of such instances:

1. Data already provided by licencees in annual returns but for which the Commission might reasonably request more frequent quarterly filings; and
2. Data in the possession of exempt services which the Commission could reasonably request under the terms of the new media exemption order^{Footnote 52}.

In the latter case, we note that the Commission has to date not requested the filing of information from non-Canadian exempt new media broadcasting undertakings (NMBUs), such as foreign OTT services operating in Canada, including Netflix, Google, BBC iPlayer, and Apple^{Footnote 53}.

The Authors believe at this stage in the development of OTT in Canada such apparent hesitancy is no longer warranted. Foreign OTT currently operate in Canada subject to a range of Canadian laws and regulation, including the Criminal Code, Copyright Act, Income Tax Act and Privacy Act. Gathering information, such as that suggested, from foreign OTT would clearly be in furtherance of the objectives of the Broadcasting Act, while at the same time being minimally intrusive.

We also believe that Canadian OTT providers will be less hesitant to provide data if foreign OTT are also required to do so. While it may be acceptable for Canadian OTT alone to provide a limited amount of "core metric" data, the more data the Commission seeks, the more unsupportable unequal treatment of Canadian and foreign OTT would become. Moreover, given recent developments in the OTT market in Canada, and the extent of foreign presence, continuing to collect data only from Canadian or licencee affiliated OTT providers may be of little use, and even potentially misleading. Indeed, contrary to the Commission's 2010 findings (and to distort them) it could be said that:

2. The Competitive Status of Foreign OTT in Canada

This section of the report examines some issues that relate to the competitive status of foreign OTT services in Canada. It is not intended as a comprehensive analysis of competitive strategy. Rather it introduces some empirical data that has bearing on the forces driving industry competition and will demonstrate some measures of how the value chain in the broadcasting system is undergoing change and how the boundaries of the broadcasting system are in flux.

The findings in this section can be summarized as follows:

1. At present, pay TV, VOD and PPV markets appear to be in the same relevant market as certain large-scale foreign OTT services, such as Netflix Canada and iTunes, meaning that they present a reasonable substitute for consumers. Movie-oriented specialty services may also be within the relevant market.
2. Entry barriers are sufficiently low that in these markets, foreign OTT services have assembled larger program libraries than their Canadian counterparts, have made competitive in-roads in acquiring distribution rights for premium content, are available without buy-through requirements and without regulatory obligations.
3. Relative to certain distinct relevant markets, such as pay television, OTT is establishing significant scale in terms of program buying and therefore may influence over time the program rights markets in those sub-sectors. The advantage of incumbency held by the Canadian broadcasters in certain sub-sectors would appear to be shrinking.
4. The large vertically integrated operators appear to have considerably larger scale in terms of Canadian and foreign program spending and therefore may have more advantage related to incumbency.
5. There is some preliminary evidence that subscriber demand in the extended pay television market may be affected as a result of growing OTT subscriber bases.
6. Canadian content carried by Netflix Canada amounts to 9% of its library hours and we estimate that Canadian content expense is roughly 5% of its total programming expense. Accordingly, OTT has become a viable distribution window for Canadian library content, although to the extent to which it may eventually replace demand for licensed services over time, Canadian content spending could be negatively affected.

The figure below shows a schematic of Michael Porter's well-known analytical framework for competitive strategy. The figure identifies the generic internal and external factors that have a bearing on competitive intensity and certain sub-elements that are germane in the current discussion.

The forces include factors such as economies of scale, product differentiation, capital requirements and government policy (such as foreign ownership and licensing requirements), all of which address the barriers to entry of new competition and hence the profitability potential of a market; threat of substitute services, which in the present case would represent alternative platforms that compete for share of consumers viewership and revenues; as well as the bargaining power of buyers and sellers and the degree of internal industry rivalry.

Figure 2.1: A Framework for Competitive Strategy

2.1 Relevant Markets

There are currently two main generic types of foreign OTT entrants in Canada including: (1) subscription-based programming services, such as Netflix Canada, which offers access to first pay window feature films, library films, library television series, and a limited amount of unique episodic content; and (2) Transactional online VOD services, such as iTunes, which provide download-to-own and download-to-rent distribution of movies, television series and music. In addition, there are other foreign entrants into the broadcasting sector, such as the device-oriented service, Roku, which has announced plans to enter Canada later in 2012 and which provides online streaming services of non-unique programming, such as Netflix, HBO Go and Epix^{Footnote 55}, through dedicated Roku devices.

As a first step in assessing the relevance of empirical and other indicators of foreign OTT in Canada we must first address the relevant markets for which the data has implication^{Footnote 56}. The broadcasting sector in Canada is not homogeneous, but rather the different components offer different means and models of providing consumer value. In fact, it is likely true that several distinct, or relevant markets, are likely to exist within the broadcast sector.

There is a useful concept under Competition Law that can be used to analyze and interpret the data that exists in the area of foreign over-the-top services in Canada. The concept is called relevant markets. The concept is an essential tool used by Competition Regulators in Canada and globally to analyze competitive matters such as dominance, merger reviews, etc. A relevant market is a distinct market where similar goods or services compete. The market is defined by the geographic location of competition, by the type of product or service itself and by substitutability in terms of demand and supply. Essentially a relevant market is the smallest definable market in which two or more substitute goods or services compete.

While the CRTC does not formally utilize the relevant market concept, the licensing of the Canadian broadcast system in fact corresponds fairly closely to our assessment of relevant markets. The accompanying table outlines four potential relevant markets in the broadcast system and illustrates some specific rights markets where competition from certain foreign OTT services may have an impact. We view these sub-markets as distinct since they have individual properties relating to their distribution windows and pricing models.

For example, Netflix Canada would appear to be most closely related to the pay television segment and so may be considered to be in the same relevant market as pay TV. Both components use a flat-rate subscription-pricing model for unlimited consumption, do not carry advertising, carry exclusive first window pay films and target the same consumers.^{Footnote 57}

iTunes on the other hand may be considered to be in the same relevant market as video-on-demand and pay-per-view. Both utilize a transactional pricing model and offer feature films in advance of the first pay window.

These two foreign OTT services share some characteristics, such as offering a vastly larger amount of programming than domestic services, and utilizing a consumer-pull model, however, their pricing and value models are sufficiently distinct that they may be considered to be in different relevant markets.

The list of relevant markets in the figure below should not be considered to be exhaustive. Other potential relevant markets in the broadcasting sector could include: broadcast distributors; sports-oriented specialty services and news-oriented specialty services.

Figure 2.2: Potential Relevant Markets in the Television Broadcast Sector

In order to understand the impact of foreign OTT on the broadcasting system, it is therefore important to evaluate the impacts on the sub-segments (relevant markets). Over time it is possible, and more likely probable, that OTT services will influence a greater degree of influence over a greater number of sub-segments so that the impact would be viewed as systemic. However, there does not appear to be sufficient data to outline any systemic impact at present, and at any rate the public disclosures of the participating companies would suggest that any systemic impact is still minimal.

In terms of programming markets, Canadian companies have the advantages of incumbency and their historic place in the value chain; while on the other side foreign OTT services have the advantages of leveraging their global scale, absence of regulatory requirements and, in many cases, their lead in the learning curve. These elements exert influence on the dynamics of the industry including such factors as the structure of exhibition windows, access to programming, and Canadian content on existing and new platforms.

Empirical data for many of the foreign OTT entrants in Canada is extremely limited. However, Netflix Canada has provided much data owing to the fact that it is a publicly listed company and has given explicit disclosures pertaining to its Canadian operation. Netflix Canada may also prove instructive because it has provided a financial picture for its Canadian operations and because they have proven there is a large mainstream, mature market opportunity with streaming services.

2.2 Economies of Scale and Cost Structure

The figure below compares the revenues and costs for Netflix Canada, the English-language pay television operators and the two largest video-on-demand undertakings in Canada. As indicated earlier, we consider the pay television market to be most similar to Netflix Canada and accordingly the comparison is likely most relevant for pay television.

The comparison demonstrates some elements of economies of scale and cost structure. It has been widely reported, or perhaps speculated, that Netflix Canada benefits as a result of its global scale, which may translate into a cost advantage versus the regulated operators in Canada.

The figure does bear out that point in terms of G&A and technical expense, which for Netflix as a percentage of revenues is far lower than the average of the Canadian group^{Footnote 58}. For Netflix Canada, its G&A and technical expense totaled $1 million in its first year of operation, or 3% of revenues, compared with the Canadian pay TV and VOD companies at 9% of revenues.

However, the comparison above does not necessarily support a cost advantage in terms of programming expense. In fact, Netflix Canada's programming and marketing expense in dollar terms in its first full year of operation was greater than all of the companies in the comparison, except for The Movie Network and Movie Central. Furthermore, its programming and marketing expense per household and per subscriber approach that of the largest entities in the sample group in its first year of operation. Of course Netflix Canada's program library is far more extensive than the pay television operators in the sample group^{Footnote 59} and so its expense per title is likely to be lower than the pay television operators. However, it is also true that Netflix Canada offers a limited selection of first-window made-for-pay series while on the other hand the pay TV operators offer a substantially greater selection of such titles.

Overall, the comparison at a high level does not illustrate that Netflix Canada has a programming cost advantage versus the Canadian peers that can be attributable to global purchasing power. This may result from the geographic windowing of foreign programming, which is evident in the fact that Netflix Canada does not have the same distribution rights in the first pay window for feature films that it does in the United States. This likely indicates that rights markets have continued to operate geographically and that a global purchase, where big volume purchases result in a lower per-title expense, is not a feature of the marketplace.

However, it is also notable that the total dollar-spend of Netflix Canada surpassed most of the comparison group in its first year of operation while, as we will demonstrate later, it assembled a vastly larger library of content relative to the licensed entities in Canada. This could indicate that rather than leveraging a cost advantage per unit, it is leveraging content relationships to develop the streaming market by out-spending the Canadian comparables.^{Footnote 60} It could also stem from a desire of U.S. studios to protect their Netflix revenues^{Footnote 61}, while for Netflix the incremental costs of a Canadian venture are almost immaterial in the context of its global enterprise. Furthermore, the figures represent Netflix Canada's first year of operation, and so do not accurately represent its state as a mature enterprise, Netflix has indicated that it will double its program spend in Canada during the course of 2012^{Footnote 62}. Accordingly, it would appear that its program spend per household, and potentially per subscriber, will outstrip all of the Canadian companies in the comparison by the end of 2012.

The comparison table highlights the following observations:

1. Netflix Canada's programming and marketing expense in dollar terms in its first four quarters of operation surpassed all of the Canadian entities, except for The Movie Network and Movie Central.
2. Netflix Canada's programming and marketing expense was roughly double the level of the most-recent entrant into the licensed pay television market, Superchannel, which launched service in F08.
3. Netflix Canada's programming and marketing expense per addressable household approached the level of the larger operators in Canada.
4. Netflix Canada's revenue of $0.40 per household was only one-quarter of the average of the large four services in the sample. This is owing to its start-up nature during the time period. We estimate Netflix Canada generated $80 million in revenues for the full year 2011, which would equate to revenues per household of $0.92.
5. Programming and marketing expense per household and per subscriber does not show evidence during this time period of an advantageous cost structure relative to the incumbent services. In fact, there is some evidence that its programming and marketing expense per household and per subscriber will exceed all of the Canadian companies in the comparison in 2012.
6. The comparison appears to show that Netflix Canada is spending considerably more on programming than most of the companies in the comparison. And these expenses are expected to rise further in 2012.
7. Despite its high levels of program expense, Netflix Canada broke even in terms of operating profit in Q3/11, or four quarters following its launch.

The data contained in the table was sourced from Company reports, CRTC databases and has involved some degree of estimation. There are a few points that should be highlighted about the comparison:

1. Netflix Canada's revenues represent its first four quarters of operation. Accordingly, its subscriber base was relatively small and owing to the one-month free promotion, its revenue per subscriber of $4.84 was well below its advertised retail price of $7.99 per month. We estimate that Netflix Canada's revenues for the full year 2011 reached $80 million, which would equate to revenue per subscriber of $6.57 for the full year (including $7.58 in Q4.11).^{Footnote 63}
2. We have estimated that Netflix Canada expended $5-$10 million in marketing expense during the time period. In the comparison table we have removed $7.5 million from its reported programming and marketing expense, which is equivalent to 21% of its revenues over the time period. A run rate of $10 million per year would be equivalent to 13% of its estimated full year 2011 revenues of $80 million, which is consistent with Netflix's marketing expenditure on a consolidated basis.
3. As indicated earlier, Netflix has indicated that it will double its quarterly content spending over the next four quarters, beginning in Q4/11. Accordingly, its spending metrics in the above table are understated relative to their likely result in 2012. The table below shows estimated programming expense per household in 2012 assuming a doubling of Netflix's quarterly content spending by Q3/12 to a level of $145 million for 2012, and based on the historic program spending increases of the Canadian service providers. The figure shows Netflix is likely to spend more on programming in 2012, on a per household basis, than any other comparable Canadian service.

2.3 Scale of Program Buying

There are two generic markets for programming within each relevant market in the broadcasting system, the market for Canadian programming and the market for foreign programming. The following tables show the breakdown of spending for certain relevant markets. It is also understood that foreign programming, particularly U.S. programming is a significant driver of viewing hours and revenues in the broadcasting system. Accordingly it is useful to examine this issue as well.

The tables below show breakdowns of foreign programming expense in the extended pay television market. By extended we have added Netflix Canada to the market as a result of the similarity in its pricing model, consumer target and programming library.

For Netflix Canada, foreign programming expense has been estimated based on its public disclosures and on proprietary research conducted by the Authors. There are three primary assumptions that have been made in estimating its foreign programming expense:

1. We have assumed that it expended $7.5 million in marketing expense in its first 4 quarters of operation, as indicated earlier, and is spending a run rate of $10 million per year in 2011. To the extent to which there is variance to that assumption versus actual spending, its total programming expense may be under- or over-stated.
2. We have estimated that Netflix Canada expended $4.3 million in streaming costs in 2011, which has also been removed from program and marketing expenses in order to derive underlying programming expense. The $4.3 million estimate is based on Netflix Canada's March 2011 disclosure of its streaming throughput in Canada^{Footnote 64}, a $0.02 per GB estimated cost of delivery and assumes one hour of average viewing per day per subscriber, which is consistent with its disclosure in January 2012 pertaining to its U.S. operation^{Footnote 65}.
3. We have reflected Netflix's Q3/11 disclosure that it would double its quarterly run rate of programming expense in Canada beginning in Q4/11. This has been reflected by increasing its quarterly run rate to Q3/12 by which time its quarterly spend has been doubled, and by maintaining its Q4 spending at the same dollar level. This results in total program expense of $145 million in 2012.
4. We have assumed its Canadian programming expense is 5% of its total programming expense. This estimate is based on our calculation of the Canadian content in the Netflix Canada library^{Footnote 66}. As it is expected that its premium library films and first pay window feature films, which are sourced from the U.S., represent a higher cost per unit, it is expected that its Canadian content expense is lower than the 9% content level in terms of library hours.

The first figure shows that in 2010, during which Netflix was in operation for four months, the majority of foreign program spending was incurred by the two main pay services, The Movie Network and Movie Central, with Superchannel also a significant participant.

The second figure shows an estimate for the group in 2012. The assumptions underlying the Netflix results have been outlined above. For the Canadian operators, we have increased their foreign program spending in line with their projected revenue growth.

Figure 2.5: Estimated Foreign Programming Expense - English Language Pay Television - 2010

Source: CRTC Statistical and Financial Summaries; Netflix; Author estimates.

Figure 2.6: Estimated Foreign Programming Expense - English Language Pay Television - 2012

Source: CRTC Statistical and Financial Summaries; Netflix; Author estimates.

The above figures demonstrate that Netflix Canada will be by far the largest buyer of foreign programming in the extended pay market. In fact, its spending should total almost half of the total spending in this market segment in 2012. Over time, it would appear likely that such spending would add inflationary pricing pressures to some segments of the program rights market; and could result in the capture of distribution rights from licensed entities by foreign OTT entrants.

The next figure illustrates a slightly broader market definition, which includes English-language video-on-demand and pay-per-view services. Again in this view, Netflix Canada is poised to become the largest buyer of foreign programming in this extended marketplace.

Figure 2.7: Estimated Foreign Programming Expense - English Language Pay Television, Video-on-demand and Pay-per-View - 2012

Source: CRTC Statistical and Financial Summaries; Netflix; Author estimates.

Lastly in this illustration of scale of program buying we will compare foreign program spending for Netflix versus the vertically integrated enterprises in Canada. The figure illustrates that compared to the largest English-speaking-oriented vertically integrated operators in Canada; Netflix will remain a relatively smaller buyer of foreign programming. The figure below illustrates Canadian and foreign program spending in aggregate for the English-language services of the operators. News and sports services have been excluded in order to make the comparison with Netflix more meaningful. The figures represent our estimate for Netflix Canada in 2012; the aggregate of disclosed spending by the Canadian operators in 2010 (for specialty and video-on-demand) and 2011 (for conventional television); although certain digital specialty services have been excluded owing to a lack of disclosure.

Figure 2.8: Estimated Programming Expense - Vertically Integrated Enterprises

Source: CRTC Statistical and Financial Summaries; CRTC Aggregate Annual Returns; Netflix; Authors estimates.

In terms of Canadian program expense, we have estimated that Netflix Canada expended roughly 5% of its programming expense on Canadian programs, which would equate to total spending in 2011 of $4-$5 million. The figure below illustrates, its expenditure percentage would be below VOD and PPV and well below that of the other sectors. Accordingly, to the degree to which Netflix Canada may ultimately capture subscriber and revenue share from the OTA and specialty sub-segments, spending on Canadian programming could be reduced in the system overall.

These illustrations of foreign program spending reveal a few features of the potential impact of OTT on the system:

1. Rather than leveraging lower unit costs for comparable titles, OTT is leveraging larger program buys with larger program libraries.
2. Relative to certain distinct relevant markets, such as pay television, OTT is establishing significant scale in terms of program buying and therefore may influence over time the program rights markets in those sub-sectors. The advantage of incumbency held by the Canadian broadcasters in certain sub-sectors would appear to be shrinking.
3. Streaming rights for library content may be becoming more valuable than broadcast rights for library content. This may be evident in the Paramount exclusive distribution agreement for first pay window feature films with Netflix Canada and in the significant spend made by Netflix on library films and television.
4. It is likely, although not provable, that Canadian program spending on OTT in percentage terms is well below all other licensed sub-sectors. This would imply that, to the extent to which foreign OTT gains subscriber and revenue share versus pay TV and other sectors, system funding for Canadian content would be reduced.
5. Broadcasters that are in the same relevant market as foreign OTT may be faced with demand-side substitution over time while facing a more difficult environment to gain access to content distribution rights and face higher content costs.
6. The large vertically integrated operators appear to have considerably larger scale in terms of Canadian and foreign program spending and therefore may have more advantage related to incumbency.

2.4 Capital Requirements

Capital requirements have a bearing on the potential for new entry in the industry since the larger the magnitude of investment that is required, the greater the level of risk that is incurred by an entrant. The figure below shows the ongoing capital requirements for three business segments as illustrated by Astral's television business, Rogers' cable business and Netflix on a consolidated basis^{Footnote 67}. There are a few noteworthy observations that can be made: (1) capital intensity can vary significantly by sector; (2) the cable segment is, by far, the most capital intensive of the segments; and (3) the capital investment requirement of Netflix is roughly the same as for Astral's television segment^{Footnote 68}. While both OTT (Netflix) and pay/specialty TV assume capital costs for assembling programming (linear & on demand) and delivering it to the ISP or BDU (via CDNs or fibre feeds/satellite, the remaining delivery costs are deferred to the ISP or  BDU.  In the case of OTT, these costs are notionally passed on to the consumer in the form of bandwidth charges and caps; but depending on a consumer's level of OTT usage, this may not be reflected in a direct incremental cost. For pay/specialty TV, such delivery costs are assumed explicitly by the consumer in the form of a (typical 100%) mark-up on wholesale fees. Not only does this mean a direct and immediate increase in consumer fees, it is unrelated to bandwidth utilization. This pricing model alone is a potentially major competitive disadvantage for pay/specialty TV over OTT, and is also discussed in relation to tipping points in the last section of this report.

Source: Company reports; Author estimates

2.5 Comparative Analysis of Program Offering --- Substitutability of Content; Product Differentiation; Demand for Programming

An examination of the program offerings of OTT services and the regulated broadcasting services yield some useful insights. Netflix Canada offers consumers access to almost 3,000 movies titles and more than 300 television series, comprising over 12,400 total television episodes (see figure below). Its television series are typically at least one year past their original broadcast date and in many cases many years or even decades following their original broadcast release. For feature films, Netflix offers a selection of older library selections as well as first window pay feature films. Recently Netflix has entered the market for original programming and released its first made-for-Netflix series, Lillyhammer, in February 2012. In addition, it has further original series in development^{Footnote 69}.

In terms of feature films, Netflix Canada offers a significantly larger selection of movie titles at any given time than is available on alternative regulated channels, as is outlined in the figure below. We have focused on pay television and movie-oriented specialty services in this comparison, as the programming libraries are most similar and therefore most competitive with one another. The titles for the regulated players in the figure represent those available on the linear channel and online. While any direct comparison between linear services and streaming services is inexact, it may be notable that on an annualized basis the movie-oriented specialty services each offer between 900 and 1,800 titles per year compared with Netflix Canada's 2,957 instant streaming titles.

The number of titles, number of television series, and hours of programming that will be referenced in all cases (except where specifically noted) represent unique and unduplicated titles. The figures are based on the libraries as outlined on the websites of the various services, and accordingly there is some scope that the programming lists as contained on those websites may differ slightly from the actual programming that is available. For example, for Netflix Canada we have observed that a direct title search for programming yields 7% more titles than the list that is populated on its menu.^{Footnote 70} Nonetheless, the sample size is large enough that any under-reporting of titles in the service menus and the similarity in under-reporting by programming segment is not expected to affect the comparison of programming in a meaningful manner.

In terms of the number of its movie titles, the vast majority^{Footnote 71} of Netflix Canada's titles are older library titles. Accordingly, its distribution of these films represents a new means for studios to monetize their libraries. As we demonstrated earlier the scale of its program purchases positions it to become the largest buyer of film in the extended pay market.

OTT services have also made in-roads into the market for the more current distribution windows, such as video-on-demand and the first pay window. For example, Netflix has Canadian first-window pay television distribution rights with two of the top six Hollywood studios that together represented 29% of U.S. box office sales in 2011^{Footnote 72}. These two arrangements provide it with exclusive access to Paramount movies and non-exclusive access to 20th Century Fox movies in the first window pay television market.

The figure below shows a comparison of film offerings for selected Canadian pay and VOD services, as well as OTT services. The figures in the table represent the number of movies that are in distribution in Canada among the top 50 Hollywood movies (in terms of U.S. domestic box office sales) from each 2010 and 2011. Of this sample of 100 movies, 90 are currently in distribution with at least one of the services indicated in the figure below. The percentage figures represent the proportion of those movies that are available by each provider. The total number of movies, and percentage figures, for the total sample group exceeds 90 titles since there is substantial overlap in the availability of titles between certain services. For example, 25 of the 31 films available by Rogers on Demand were available for rent on iTunes. Furthermore all of the eight films in distribution by Superchannel were also available from Netflix Canada.

While Netflix Canada specifically is regarded as having a large selection of library titles and a weak line-up of current movie titles, the table below shows that its movie selection in the first pay television window is fairly substantial and is far greater than Superchannel, although below that of Movie Central.

The figure below shows the licensing arrangements that exist for pay television in Canada. Generally, licensing deals for the first pay window and for made-for-pay television series are arranged on an exclusive basis. The exception is 20th Century Fox, which has a non-exclusive output deal with both Superchannel and Netflix Canada for first pay window films. Library movies are generally not exclusive to a broadcaster. The table below is not intended to be fully comprehensive, although the studios that are identified represent the vast majority of U.S. box office results.

In a much-publicized development, effective March 1, 2012, Netflix U.S. lost the right to air Starz movie titles, including pay window Disney and Sony releases^{Footnote 73}. This has led analysts in the U.S. to suggest that Netflix is now primarily an Internet streaming service for television shows, not feature films. As Netflix Canada's licensing deals appear to be independent of the U.S., and in any event have not included Disney or Sony, no such explicit shifts in emphasis have become evident here.

Figure 2.14: Distribution Rights in the Extended Pay Television Market

Turning to television series the regulated pay television providers offer the vast majority of the currently in-production made-for-pay series and a large selection of library titles of made-for-pay series that have ended their series runs. The table below shows a rough outline of the number of series that are available on pay television in Canada and a breakdown between the numbers of titles that are currently in production versus library titles. The figures in the table should be considered illustrative rather than exact since a relatively small subset of the "made-for-pay titles" have been sourced by the Canadian distributor from cable networks such as TNT and AMC. Of Netflix Canada's totals, one series, Lillyhammer, is unique to it.

The figure reveals that The Movie Network, Movie Central and Superchannel each offer consumers a much greater selection of made-for-pay series than Netflix Canada. Of Netflix's seven titles, only one, Lillyhammer, is unique to it. The other six programs represent back seasons of a limited selection of made-for-pay series, such as Dexter and Californication.

Figure 2.15: Made-of-Pay Series In Distribution

In addition to a larger inventory of films and television series, Netflix also offers consumers access to its library at much lower retail prices and does not have buy-through requirements. The table blow shows a comparison of average retail prices and programming availability.

The degree of Canadian content on Netflix Canada has been the subject of much debate and this issue ties into the ability of foreign OTT to deliver on or affect some aspects of broadcasting policy. An analysis of their library is also important as it reveals the programming that Canadian consumers are purchasing and valuing. The following table shows a summary of the Netflix Canada library and a breakdown of Canadian content between movies and television series. It shows that the service offers 99 Canadian films, representing a Canadian content of 3.3% for movies, and 43 Canadian television series comprising 1,968 total episodes and representing 13.7% Canadian content for television.

We have estimated that Netflix Canada's library has a total of 12,020 hours of programming, of which 1,084 hours can be identified as being Canadian hours.^{Footnote 74} As the exhibit shows, Canadian content for feature film hours (3.1%) indexes much lower than for television series (13.2%). In aggregate, 9.0% of total library hours can be identified as Canadian programming. The figure also illustrates that roughly 60% of library hours are television series and 40% are feature films, which corresponds roughly with the statement of Netflix U.S. that subscriber viewing tends to slightly favor television. In Canada, the television library tends to skew towards family and children's programming with an estimated 75% of total Canadian television hours.

Figure 2.17: Canadian Content on Netflix Canada

It is also notable that the 9% of total hours that can be identified as being Canadian is well below the Canadian content exhibition and CPE requirements for most regulated broadcasting entities in Canada. Nonetheless, the total supply of Canadian content, as measured in hours or titles, is considerable and has been previously demonstrated surpasses the number of titles available on movie-oriented specialty broadcasting services on an annual basis and is comparable with the number of Canadian films titles on The Movie Network and Movie Central.

2.6 Indicators of Demand

As has been widely discussed, consumer demand for foreign OTT services has been increasingly apparent. For example, by the end of 2011, Netflix Canada had assembled a subscriber base of roughly 1.2 million subscribers, 16 months following its launch in September 2010. In its first full quarter of operation (Q4.10), Netflix added more subscribers than accumulated by Super Channel over the three years of its operation. With 1.2 million subscribers, Netflix Canada has achieved a penetration rate of 16% of addressable households with a broadband connection of 1.5 Mbps or higher and roughly 10% of television households in Canada. Its customer base is as large as that of the largest premium pay service in the country, The Movie Network.

To date, the evidence has supported the notion that Netflix Canada has been largely complementary to the regulated services (see figure above). In other words, there has been little evidence to suggest that its growth in Canada has come at the expense of licensed operators. This is also evident in the results reported by the publicly traded companies.

Furthermore, the subscriber penetration trend for the terrestrial BDUs does not at present show an impact that can be attributable to OTT services (see figure below). While terrestrial incumbent BDU basic subscriber penetration has been in decline for some years, this is likely primarily attributable to market share gains by Telus and Bell Canada. The penetration trend line remains intact in the period prior to and after the entry of Netflix and broadening scale of iTunes.

However, there is one indicator that suggests OTT may be influencing the subscriber patterns in the premium pay television market. The figure below shows the attachment rate of pay television subscribers, on a trailing 4 quarters basis. The figure illustrates the growth in pay television subscribers as a percentage of digital BDU subscriber growth. It shows that over the period of the sample, for each 100 digital BDU subscribers added to the industry, 10-15 on average would subscribe to a premium pay television service.

There are two notable exceptions to this average: (1) the steep decline in attachment that occurred in late 2007 and early 2008, which we attribute to the cancellation of the Sopranos; and (2) the sharp decline that has occurred in the most recent two quarters. The reasons for the decline in the current periods are unclear. The question of causality is an open one. It may be that subscriber growth by Netflix Canada has reduced the attachment of premium pay television; or it may be that economic or other factors have influenced the pattern. In either case, ongoing monitoring of this metric may be essential.

Figure 2.20: Attachment Rate of Pay TV to Digital Cable Net Additions - Trailing 12 Months

2.7 Summary of Competitive Status

The analysis provided above demonstrates certain characteristics of the landscape as it stands at present:

1. At present, pay TV, VOD and PPV markets appear to be in the same relevant market as certain large-scale foreign OTT services, such as Netflix Canada and iTunes, meaning that they present a reasonable substitute for consumers. Movie-oriented specialty services may also be within the relevant market.
2. Entry barriers are sufficiently low so that in these markets, foreign OTT services have assembled larger program libraries than their Canadian counterparts, have made competitive in-roads in acquiring distribution rights for premium content, are available without buy-through requirements and without regulatory obligations.
3. The bargaining power of foreign OTT relative to incumbent Canadian services appear to be growing as a result of their significant expenditures on content.

3. Potential tipping or inflection points

The defining of "tipping", "trigger" or "inflection" points can be a useful means of representing material shifts in competitive balance as between OTT providers and regulated players within the Canadian broadcasting system.

While a tipping point can be defined as a singular event, for example when a service loses 5% of its subscriber base^{Footnote 75}, the Authors believe that a series of different tipping points may prove more useful in the current public policy context. Evidence of competitive impact, material or substantial competitive impact, and impact threatening viability are all potential tipping points, but each would indicate to public policy makers distinctly different levels of concern. Accordingly we would propose following these 3 types of tipping points, across at least 4 different parts of the system, as the well as the system as a whole.

Viewing, revenue, ARPU, profitability and many other possible indices discussed earlier could all in theory be used in such a set of tipping points. We believe, however, that subscriber base is the most useful. Among its advantages, it is:

• Simple and easy to understand;
• Accurate. There are no different sampling methodologies to contend with;
• Easy to track. All subscription based services (pay, specialty and BDU) regularly monitor it;
• Flexible in application. It can be applied to a particular service, a genre of service or a class of services;
• Expressed as an absolute number, percentage of BDU households, or percentage of all households (and combined with the flexibility noted above), it can be normalized to isolate cause and effect^{Footnote 76}; and
• Indicative of the potential of a broadcasting business, not its relative strength resulting from the strength of particular programming, or advertiser interest.

The precise determination of tipping point thresholds would require a more involved process than simply accepting the suggestions made above. We note however the following:

• Pay TV, which as discussed in Section 2 appears to be in the same relevant market as current OTT services, may also be the most resilient to competitive impact. In the U.S. there are at least 5 competing regulated pay TV providers, comparable to Canada's two. While the outbidding by OTT of movie packages held by Canadian pay TV providers may well cause the loss of some subscribers, there can be a material reduction in program costs too. That said, that pay TV in Canada should survive direct competition from OTT does not mean that all the players would survive, or that the current regional structure or business model of the original licensees could survive either. The loss of premium pay TV content, such HBO or Showtime original series, would threaten the viability of the Corus and Astral's existing "gold plated service" business model. This is not an unforeseeable scenario. Should OTT services like Netflix achieve comparable penetration levels to pay TV, there would no doubt be some temptation for Time Warner to take HBO Go direct to Canadian subscribers ("retail"), rather than through agreements with Corus and Astral ("wholesale"). This is, however, another area where Vertical Integration could pay dividends; the vertically integrated pay TV provider will have far more means at its disposal to encourage the HBOs and Showtimes to stay on board.
• Category B Specialty channels are largely Ethnic and English-language competitive channels, with 2 million or less subscribers. The low penetration levels of these channels leads to modest revenues and programming budgets, which in turn can lead to a strong reliance on acquired programming and a corresponding lack of distinctive original Canadian content or Canadian "angle". The result, particularly for English-language Cat Bs, is a strong vulnerability both in terms of potential subscriber loss to competing free Google and other OTT substitutes, and a lean, if not fragile, business model that cannot withstand much in the way of material subscriber losses. It is also these channels that are most vulnerable to general cord shaving – be it caused by OTT competition, or simply the abundance of quality on demand content from larger conventional TV sources, now competing for viewer attention.
• Category A Specialty services mostly comprise the 3 rounds of analog specialty channels launched in the '80s and '90s, plus the then Cat 1 digital round in 2000^{Footnote 77}. They have the best (and protected) genres, the largest subscriber base and hence the highest subscription and advertising revenues, and programming budgets and profits. They therefore have both more capacity and ability to respond to OTT competition. They can absorb higher losses, and have the resources to respond competitively with distinctive original content. They nevertheless can be vulnerable, initially to outbidding for exclusive foreign content from OTT, and ultimately from loss to substitutable OTT channels, if such OTT channels are able to provide equal quality, and are not in genres where Cancon is a driver^{Footnote 78}.
• BDU Subscription is dependent on TV viewers finding sufficient value in some aspect of a BDU TV offering to plunk down their hard-earned cash every month. The fact that viewers are prepared to pay for something that they can actually get for free elsewhere is not new -- in fact, it has been a reality for cable since its inception half a century ago (the most popular services being the free OTA channels). In addition to content choice, BDUs have always offered convenience – be it the historic convenience of not maintaining rooftop antennae, or today's convenience of on demand viewing through VOD and the PVR. The more BDUs are able to protect their "best TV choice" franchise, while matching OTT convenience, the less cord-cutting will occur.

3.1 The Impact of OTT on Canadian Content Spending

In the face of apparently inevitable competitive impact, many argue that the only answer is deregulation of incumbents. As for the impact on Cancon, the view often expressed is that OTT presents new opportunities for Canadian producers, as foreign OTT needs Cancon to compete in the Canadian market. Netflix itself has highlighted its Canadian content offering.

Netflix does indeed present an interesting test case on the impact of OTT on Canadian content. This is particularly so given that Netflix would appear to present a "best case" non-regulated foreign OTT scenario, as:

1. A "first in" OTT provider that does not wish to provoke a negative public policy reaction and hence inadvertently encourage regulation;
2. A program purchaser that, to date, appears to have respected Canadian territorial rights and not attempted "North American" buys; and
3. A new entrant that does not want to be crushed by incumbents, and would hence rather be seen as "complementary" to cable.

Based on our analysis in Section 2, we conclude that:

1. Canadian content expenditure by Netflix Canada is roughly 5% of its total programming expense. Accordingly, any market share gains by it at the expense of pay TV, OTA, VOD and specialty would reduce funding for Canadian content in the system. We estimate that a loss of a pay TV subscriber to Netflix Canada would reduce Canadian content expenditure by roughly $2.05 per subscriber per month, and would by-pass requirements on programs of national interest and incentives for Canadian drama.
2. Video-on-demand presents similar substitution risk to Canadian content spending. We estimate that $18.20 per month in VOD transaction fees per subscriber is equivalent to a $9 wholesale fee for a pay TV subscriber in terms of its funding towards Canadian content. At present VOD revenues per subscriber is in the range of $3 to $4 per month.
3. Licensed services oriented to library films typically garner much lower wholesale and retail fees relative to premium pay services^{Footnote 79}. Accordingly, to the extent to which future premium content distribution rights are purchased by foreign OTT, there is risk of substantial reductions in retail prices for pay TV services for which consumers perceive quality is impaired. This would further strain content funding within the system.
4. On the other hand, to the extent to which foreign OTT services stimulate subscriber and revenue growth, funding for Canadian content could be increased. For example, we estimate Netflix Canada expends roughly $5 million annually ($0.35 per subscriber per month) on Canadian content, although this appears to be composed entirely of library titles.
5. The table below shows estimated spending on Canadian content on a per subscriber basis per month. To the degree to which switching or substitution between services occurs spending on Canadian content could be affected.

3.2 Threat to the Broadcasting System

The foregoing speaks to threats to different segments of the broadcasting system – the players and all important Canadian programming, in particular. The ultimate question is the extent to which threats to these different segments could lead to the demise of the system itself.

Without diminishing the impact on the players concerned, it is fair to say that a material or even significant impact in pay TV or Cat B specialty should not have devastating impact on the system itself. For example, Cat B specialty services were never granted the privileges or expectations of their Cat A predecessors. That they might find themselves in an even more competitive environment, that no longer favours their business model, would not be inconsistent with their initial licensing environment.

Moreover, as noted above, the very shift to on demand itself (whether it be regulated VOD and PVR viewing, or online and OTT) risks reductions in Cancon contribution. OTT only exacerbates that anticipatable reduction. Making on-demand viewing an extension of linear broadcasting, and capturing revenues in a group contribution approach, is clearly the best means of minimizing such negative impact^{Footnote 80}. As the CRTC recognized in its group licensing policy, exhibition requirements will become less effective in an on-demand world, but expenditure requirements need not.

At the other book-end, regional and national sports services represent the one core BDU programming franchise that can not be readily replicated by OTT. TSN and Sportsnet are both in close to 9.2 million households, and RDS in 3.2 million^{Footnote 81}. This places the sports channels in virtually every relevant language BDU household. They also have the highest reach and highest hours tuned of any specialty service, and already command the highest subscription fees. OTT has neither the technical capability nor financial resources to offer equivalent services. And once Canadians subscribe to a BDU for sports, using it for the vast majority of their TV needs only makes sense.

The good news, then, is that it is implausible to see OTT as a threat to the existence or viability of the Canadian broadcasting system in the foreseeable future. The argument made by some that new IP based, on demand technologies and platforms, require a slow managed phase out of our current regulatory system^{Footnote 82} are, at best, woefully premature.

What is true, however, is that Canadian broadcasting is now once again^{Footnote 83} competing in an "open system" with potential direct and material bypass from foreign services. OTT competition means that a broadcast license is about as far from being a guarantee of growth and success^{Footnote 84} as it has ever been. It also means that the health and ability of certain players in the system to contribute and remain competitive in the face of OTT becomes a vital issue. For both the players concerned and the system itself, the outcome will spell the difference between a vibrant ecosystem that continues to contribute in meaningful ways to the economic, social and cultural capital of Canada and an old medium whose contributions and relevancy are in ongoing decline.

Probing, information gathering, tracking and scenario planning will provide the factual basis necessary to ensure that our regulatory environment remains as appropriate as possible. The rest will be up to the ingenuity, responsiveness and strength of our public and private broadcasters and distributers. This is a challenge. But Canadian broadcasting history suggests we should be up to it.

Appendix A: About the Authors

Peter Miller, P. Eng., LL.B.

Peter Miller is a lawyer and engineer with over 20 years of broadcast and telecommunications industry experience, in both private practice and senior executive positions. Since 2005, he has acted as an advisor to select clients in both the public and private sectors, specializing in business and policy development, as well as regulatory counsel.

Peter's legal and consulting practice is largely focussed on the media sector, but is wide ranging in terms of the types of clients and nature of assignments. Assignments vary from regulatory and public policy advice and analysis, to strategic planning and economic impact exercises.

Peter's professional background includes telecommunications network design, acting as a Parliamentary Assistant to Members of Parliament, private practice in communications law, and senior broadcast executive positions.

Peter is a frequent industry commentator who has been actively involved in numerous industry boards and committees. Peter is a past chair of the CAB Specialty & Pay Services Board, past treasurer of Canadian Digital Television and a current member of the Board of Volunteer Toronto.

For more information on Peter, please contact him at info@petermiller.ca

Randal Rudniski, CFA

Randal Rudniskiis the Managing Director of RJR Communications Insights Inc., which is an independent advisory firm that studies the financial effects of investment, competition and regulation in the broadcast and telecommunications industries.

He is a research professional with over 20 years experience in the broadcasting and telecommunications sectors. Randal has specialized in the areas of strategy, analysis and valuation of the Canadian and global telecommunications, broadcast distribution and media sectors.

Prior to forming RJR Communications Insights, Randal was the senior analyst at Credit Suisse Securities Canada, led the Canadian telecom and media research effort and contributed to the global telecom and media teams. Randal was regularly ranked as a top analyst in the Canadian media sector.

Prior to that, Randal was the wireless and media analyst at Scotia Capital Markets, with coverage including incumbent wireless, emerging markets wireless and new entrant wireless operations.

Date modified:

2012-12-14