Trends in TV and Internet Use

APPENDIX 1*

Joint Submission of ACTRA, APFTQ, CMPA, DGC and WGC in response to B/TNOC 2011-344 Trends in TV and Internet Use:

The Impact of Internet TV on Canadian Programming

Barry Kiefl, Canadian Media Research Inc. July 2011

* Funding for this study was provided by Ontario Media Development Corporation. Any opinions, findings, conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of Ontario Media Development Corporation or the Government of Ontario. • Introduction Netflix Canada launched in September 2010, at about the The 2010 TVQ Survey involved a sample of some same time as the latest version of Boxee and the second 1,483 Anglophone and Francophone respondents generation of Apple TV were released. The latter provides aged 18-plus. The response rate to the survey ready access to not only iTunes but also Netflix, Youtube, from the initial mail-out was approximately 35%, MLB and NBA streaming video and could presumably offer access to the NHL. These devices plug directly into a relatively high response rate for a mail survey. your TV without the need of a computer. Google TV and To encourage response to the TVQ Survey, the several other Internet TV services will also undoubtedly questionnaire was short and made as simple to make their appearance in Canada soon. The following complete as possible and a small monetary report examines the first available data on Internet TV incentive was included with the questionnaire. All viewers, that is, people who use the internet to view TV data were weighted by age and sex. The 2010 programs or other video. The source for much of the data survey has a margin of error of +/-2.6%, 19 times in this report is CMRI‟s annual TV Trends and Quality out of 20. (TVQ) Survey. The TV Trends and Quality (TVQ) Survey is a national This report also reviews very recent TV viewing survey of Canadians‟ use of and attitudes toward trends as measured by BBM‟s PPM system, television. The primary purpose of the TVQ Survey is to covering the period since the launch of test consumer awareness of and satisfaction with the many Netflix/Apple TV. We also review current TV channels available today via cable TV, DTH or over- comScore internet viewing data and discuss how the-air. The TVQ Survey also measures Canadian use of services such as Netflix can be properly measured and interest in many new technologies such as the internet, in the future. Finally, the report presents an VOD, PVRs, blackberries, HDTV, etc. The 2010 TVQ Survey is the ninth annual survey of its type and since the analysis of BBM data analyzed by program methodology has remained constant, this year‟s results can country of origin and genre and forecasts how be trended with findings from earlier surveys. internet TV may affect the audience share of Canadian programming in future years. The TVQ Survey was conducted in November-December 2010, employing a 17-page mail questionnaire sent to Barry Kiefl former BBM TV survey respondents. CMRI 2 1. Trends in TV Use and Internet Viewing

3 According to the TV Trends Survey, TV viewing has remained stable in the past decade at just under 20 hours per week per person; internet use has grown substantially since we began measuring it in 2004 and surpassed radio listening for the first time in 2010. Internet use doubled from about 5.6 hours per week in 2004 to 11.3 hours in 2010.

Weekly Hours Spent With TV, Radio and the Internet, Anglophones 18-plus, Fall 2002-2010 25.0 TV Radio Internet

19.8 19.9 19.9 20.0 19.4 19.3 18.8 18.8 18.9 17.4

15.0 13.0 12.4 12.7 12.7 10.9 11.011.3

10.0

7.7 7.2 6.6 5.6 5.7 5.0

0.0 2002 2003 2004 2005 2006 2007 2008 2009 2010 4 Source: CMRI (TV Trends Survey) The percentage of Canadians using internet TV is surprisingly high and has more than doubled over the past three years. As of late 2010 some 30.8% of Anglophones indicated they spent 1 hour or more per month watching internet TV or other video on-line. Even more (44.4%) of Francophones claimed to be internet TV viewers. About half of Anglophone internet viewers (15.7% of all Anglos) were heavy viewers (5 hours or more per month), whereas over 30% of all Francophones were heavy internet TV viewers. The analysis from this point on deals with Anglophone data; Francophone results are available on request.

Percentage of People Who Are Internet TV/Video Viewers, Fall 2010, Anglophones and Francophones 18-plus

Heavy Internet TV Viewers

Light Internet TV Viewers

Francophones 31.1% 13.3% Total = 44.4%

Total = 30.8% Anglophones 15.7% 15.1%

5 0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% Source: CMRI (TV Trends Survey) Internet TV viewers were just as likely to subscribe to cable TV or satellite (both over 90%), although heavy internet viewers were slightly more likely to subscribe to cable TV. So at this early point in the evolution of OTT, few cable/DTH subscribers seem to be „cutting the cord‟, at least entirely.

Percentage of Internet TV/Video Viewers Who Subscribe to Cable TV/DTH, Fall 2010, Anglophones 18-plus Cable TV DTH

All Persons 49.9% 45.0% Total = 94.9%

Light Internet TV 49.7% 49.3% Viewers Total = 99.0%

Heavy Internet TV 57.6% 35.4% Viewers Total = 93.0%

0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 100.0%

Source: CMRI (TV Trends Survey) 6 Internet TV viewers are much younger than the general population; approximately one half of both light and heavy internet TV viewers were in the youngest age group, 18-34 years. Younger people are more adaptable to change and likely to switch more of their viewing to internet TV as it becomes more compelling and easier to access.

Percentage of Internet TV/Video Viewers By Age Group, Fall 2010, Anglophones 18-plus 18-34 35-54 55-plus

All Persons 29.7% 38.6% 31.7%

Light Internet TV 56.9% 26.8% 16.4% Viewers

Heavy Internet TV 47.7% 43.1% 9.3% Viewers

0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 100.0%

Source: CMRI (TV Trends Survey)

7 However, light and heavy internet TV viewers are comprised of roughly equal proportions of males and females and are not much different than the general population in this regard.

Percentage of Internet TV/Video Viewers By Sex, Fall 2010, Anglophones 18-plus

Male Female

All Persons 49.1% 50.9%

Light Internet TV 49.3% 50.7% Viewers

Heavy Internet TV 54.0% 46.0% Viewers

0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 100.0%

Source: CMRI (TV Trends Survey)

8 Interestingly, light and heavy internet TV viewers have much newer TV sets (especially their main TV set) than the general population, another indicator that they are ready for change in the TV viewing environment.

TV Set Ownership and Age of Set, Internet TV/Video Viewers, Fall 2010, Anglophones 18-plus All Persons Heavy Internet TV Users Light Internet TV Users

8.6

How many years have you owned the televisions 7.2 mentioned above? 2nd TV

7.1

7.4

How many years have you owned the televisions 4.7 mentioned above? Main TV

4.7

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 Source: CMRI (TV Trends Survey) 9 Earlier we reviewed trends in TV, internet and radio use and found that internet use has been growing. In fall 2010 we found that internet use is especially high among heavy internet TV viewers (over 23 hours of internet per week or more than double the general population). Light and heavy internet TV viewers are only slightly less likely to watch regular TV*. Interestingly, however, they are lighter radio listeners. Please note that one cannot conclude from survey data taken at one point in time that use of one medium, e.g., the internet, causes less use of another medium, e.g., radio or TV.

Hours Spent Watching TV, Using the Internet and Listening to the Radio, Internet TV/Video Viewers, Fall 2010, Anglophones 18-plus All Persons Heavy Internet TV Users Light Internet TV Users 11.3

Internet 23.0

12.5

11.0

Radio 8.9

7.7

18.9

TV 16.6

16.9

0.0 5.0 10.0 15.0 20.0 25.0 10 *Regular TV refers to traditional conventional and specialty channels. Source: CMRI (TV Trends Survey) Ownership of HD TV sets among light internet TV viewers (53.6%) is much higher than the general population (35.2%) but neither light nor heavy internet TV viewers are much interested in 3D TV, a technology seemingly dead-on-arrival in its present form.

Percentage of Internet TV/Video Viewers Who Own HDTV and Likelihood of Purchasing 3D TV, Fall 2010, Anglophones 18-plus

HDTV is a new type of TV that provides a much clearer picture 35.2% and sound than traditional TV`s. All Persons These TV`s are available at a Light Internet TV Viewers cost of about $1000. How likely 53.6% Heavy Internet TV Viewers is it that your household will be buying an HDTV set during the next year? (% Already Own 32.4% One)

3D TV is a new type of TV that 77.2% provides a three-dimensional picture. These TV`s are available at a cost of about 82.4% $2000. How likely is that your household will be buying a 3D TV set during the next year? (% 71.1% Not at All Likely)

0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0%

Source: CMRI (TV Trends Survey) 11 Ownership of iPods and smartphones is substantially higher among light and heavy internet TV viewers, meaning that mobile internet TV is also an opportunity in future years. Internet TV viewers are also much more likely to text message and surf the internet with their mobile devices.

Percentage of Internet TV/Video Viewers Who Use Other New Media Technology, Fall 2010, Anglophones 18-plus

All Persons

Light Internet TV Viewers

Heavy Internet TV Viewers 14.9% Do you surf the internet with 28.4% your cell phone/smartphone? 31.8%

44.1% Do you text message with your 57.3% cell phone/other device? 63.1%

20.3% Do you personally have a 36.2% smartphone? 30.3%

26.1% Do you personally own iPod? 41.8% 56.7%

0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% Source: CMRI (TV Trends Survey) 12 For the past decade CMRI has been asking Canadians which mass medium is their main source of news. In 2010 the internet hit a new high of 17.5%, but still well back of TV (46.6%) in the general population. But, among both light and heavy internet TV users, the internet surpassed all other media as the main news source, a harbinger of the new media world we are facing .

Main Source of News for Internet TV/Video Viewers, Fall 2010, Anglophones 18-plus Internet Newspaper 17.5% Radio 14.3% Television All Persons 21.6% 46.6%

31.8% 15.7% Light Internet TV Viewers 25.0% 27.4%

38.4% 12.6% Heavy Internet TV Viewers 12.9% 36.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 40.0% 45.0% 50.0%

Source: CMRI (TV Trends Survey) 13 2. Recent BBM and comScore Viewing Data

14 CMRI‟s TV Trends and Quality Survey takes a snapshot of the TV viewing environment in fall each year. BBM tracks TV viewing every day of the year using its PPM audience measurement system, which was launched in September 2009, allowing for comparisons to the year before and after the Netflix/Apple TV launch in Canada. BBM‟s data from the period September 2010 to May 2011 on a month-by-month basis show there was no discernible decline in regular TV viewing after the launch of the OTT services. In the same 8-month period of September to May, the average monthly hours spent viewing regular TV equaled roughly 4.1 billion hours both last year and this year. Note, however, the PPM system measured more stations in the second year of its operation.

Billions of Viewing Hours, By Month, All Households, Persons 2+, All TV Stations, All Canada

Netfilx/Apple TV launch 6.00

Sept-May avg. hours= 4.08 Sept-May avg. hours= 4.15 5.04 5.00 4.79 4.72 4.67 4.65 4.51 4.35 4.42

3.99 3.93 4.00 3.84 3.83 3.86 3.67 3.67 3.71 3.68 3.74 3.52 3.53 3.43

3.00

2.00

1.00

0.00

15 2010/07/25 Jul 10 2009/09/27 2009/10/25 Sep 092009/11/29 Oct 092009/12/27 Nov 092010/01/31 Dec 092010/02/28 Jan 102010/03/28 Feb 102010/04/25 Mar 2010/05/3010 Apr 102010/06/27 May 10 Jun 102010/08/29 2010/09/26 Aug 102010/10/31 Sep 102010/11/28 Oct 102010/12/26 Nov 102011/01/30 Dec 102011/02/27 Jan 112011/03/27 Feb 112011/04/24 Mar 2011/05/2911 Apr 11 May 11 Source: CMRI (BBM) BBM‟s data also show that in the early months of Netflix/Apple TV there was no apparent decline in TV viewing among young adults aged 25-34, that is, those most likely to replace regular TV viewing with OTT services. Their average monthly viewing hours was approximately .50 billion hours before and after the OTT launch.

Billions of Viewing Hours, By Month, All Households, Persons 25-34, All TV Stations, All Canada

0.70 Sept-May avg. hours= .48 Sept-May avg. hours= .47

0.60 0.56 0.56 0.56 0.54 0.54 0.52 0.53 0.50 0.50 0.46 0.44 0.44 0.43 0.43 0.44 0.43 0.43 0.42 0.41 0.41 0.40 0.39 0.40

0.30

0.20

0.10

0.00

2010/07/25 Jul 10 2009/09/27 2009/10/25 Sep 092009/11/29 Oct 092009/12/27 Nov 092010/01/31 Dec 092010/02/28 Jan 102010/03/28 Feb 102010/04/25 Mar 2010/05/3010 Apr 102010/06/27 May 10 Jun 102010/08/29 2010/09/26 Aug 102010/10/31 Sep 102010/11/28 Oct 102010/12/26 Nov 102011/01/30 Dec 102011/02/27 Jan 112011/03/27 Feb 112011/04/24 Mar 2011/05/2911 Apr 11 May 11 Source: CMRI (BBM) 16 Yet, among those households equipped with high speed internet average monthly viewing hours watching regular TV was 2.62 billion hours before the OTT launch and 2.99 billion hours after the launch. The increase is probably associated with an increase in the number of households equipped with high speed internet.

Billions of Viewing Hours, By Month, Households With High Speed Internet, Persons 2+, All TV Stations, All Canada

4.50

Sept-May avg. hours= 2.62 Sept-May avg. hours= 2.99 4.00 3.59 3.43 3.50 3.28 3.01 3.06 2.94 2.98 2.92 3.00 2.91 2.90 2.85 2.71 2.58 2.63 2.54 2.50 2.50 2.36 2.36 2.30 2.25 2.14 2.00

1.50

1.00

0.50

0.00

2010/07/25 Jul 10 2009/09/27 2009/10/25 Sep 092009/11/29 Oct 092009/12/27 Nov 092010/01/31 Dec 092010/02/28 Jan 102010/03/28 Feb 102010/04/25 Mar 2010/05/3010 Apr 102010/06/27 May 10 Jun 102010/08/29 2010/09/26 Aug 102010/10/31 Sep 102010/11/28 Oct 102010/12/26 Nov 102011/01/30 Dec 102011/02/27 Jan 112011/03/27 Feb 112011/04/24 Mar 2011/05/2911 Apr 11 May 11 Source: CMRI (BBM) 17 The PPM system replaced the Nielsen people meter nationwide in September 2009 and the PPM has reported much higher audiences for TV ever since, 21.2 hours per week per person vs. 17.0 hours per week*. That is, according to the PPM system, people of all ages were watching more TV traditional conventional and specialty TV substantially than previously reported.

Increased Viewing by Age Groups, Weeks 1-30, 24 Hours, 2008-09 vs. 2009-10

45.0

Viewing hours 2010 40.0 Viewing hours 2009

35.0

30.0 29.0

25.5 25.6 25.0 21.2 21.8 20.0 19.7 19.4 20.0 18.6 18.9 17.7

Weekly Hours Weekly 17.1 17.0 16.4 16.8 14.9 15.0 12.8 12.4 10.6 10.3 10.0

5.0

0.0 P 2+ P 2 - 6 P 7 - 11 P 12-17 P 18-24 P 25-34 P 35-49 P 50-54 P 55-64 P 65+ Source: CMRI (BBM) Age Group 18 *Data are for the same 72 English TV networks measured by the PPM in 2009-10 compared to 2008-09 people meter data. On a 2+ basis the introduction of the PPM resulted in an increase in viewing to traditional services of some 27% and as high as 67% among young adults aged 18-24. The PPM system currently does not measure all conventional or specialty TV channels. Many U.S. TV stations are not PPM encoded and therefore cannot be measured/reported by PPM. Netflix, Apple TV and similar services are also not currently encoded nor measured by the PPM and BBM has no plans to do so. Increased Viewing % by Age Groups, 24 Hours, Weeks 1-30, 2008-09 vs. 2009-10

90.0

80.0

70.0 65.8 66.7

60.0 57.5

50.0 % 40.0 35.5 34.9 29.6 30.0 27.3

17.8 20.0 15.8 11.5 10.0

0.0 P 2+ P 2 - 6 P 7 - 11 P 12-17 P 18-24 P 25-34 P 35-49 P 50-54 P 55-64 P 65+ Age Group Source: CMRI (BBM)19 comScore, the Canadian internet ratings service, recognized the need to measure internet TV and has been doing so for several years. The most recent comScore data show that Bell Media is the sole Canadian broadcast owner in the top 20 video sites in Canada but its audience is miniscule compared to Google (Youtube). Netflix is not in the top 20 but, if one watches Netflix or other internet TV services via a device other than a computer, comScore currently does not measure that audience. Nor is the computer-based audience measured at the program level; data are released on a monthly basis for content providers, not individual programs.

Top 20 Properties Streaming Videos: April 2011 Total Total Unique Videos Videos per Minutes Minutes Minutes Viewers (000) Viewer per Viewer per Video (MM) (000) Total Internet : Total 22,187 5,495,036 247.7 1,021.7 4.1 22,667 Top 20 Video Properties 1 Google Sites 20,657 2,297,808 111.2 360.8 3.2 7,454 2 FACEBOOK.COM 9,364 72,884 7.8 29.1 3.7 272 3 VEVO 9,172 91,593 10.0 83.0 8.3 761 4 Viacom Digital 6,755 44,288 6.6 37.3 5.7 252 5 Microsoft Sites 5,815 92,232 15.9 36.2 2.3 211 6 SpotXchange Video Ad 4,709 33,785 7.2 4.7 0.7 22 7 Yahoo! Sites 3,464 33,942 9.8 31.6 3.2 109 8 Bell Media 3,265 61,004 18.7 75.9 4.1 248 9 ADAP.TV 3,149 33,976 10.8 7.4 0.7 23 10 Tremor Media Video Network 3,017 32,977 10.9 6.4 0.6 19 11 DAILYMOTION.COM 2,884 32,920 11.4 88.8 7.8 256 12 Crosspoint Media 2,781 31,964 11.5 2.5 0.2 7 13 Break Media Network 2,607 17,189 6.6 37.8 5.7 99 14 Undertone 2,585 56,731 21.9 15.4 0.7 40 15 VIMEO.COM 2,579 9,148 3.5 23.1 6.5 60 16 NBC Universal 2,302 6,329 2.8 5.4 2.0 12 17 AOL, Inc. 2,267 9,633 4.3 22.3 5.2 51 18 Amazon Sites 2,150 9,377 4.4 12.0 2.7 26 19 Joost Media 2,144 8,543 4.0 5.3 1.3 11 20 Turner Digital 2,134 22,179 10.4 28.5 2.7 61 Source: Comscore 20 In the U.S. Nielsen measures TV audiences and also measures online video and is attempting to integrate the online audience data with the broadcast audience data. However, there too if one watches Netflix or other internet TV services via devices other than a computer, Nielsen currently does not measure the audience. However, Netflix in the U.S., despite being a paid service and probably mostly watched via blu-ray players, game consoles and other devices, ranks highly in Nielsen net ratings data, especially in time per viewer, which should be of special concern for regular TV.

Top Online Video Brands by Time per Viewer (February 2011, U.S.) 250K Unique Viewer Minimum MOM % Change in Video Brand Time per Viewer (hh:mm) Time Netflix 9:16 -16.8% Tudou.com 8:08 25.2% Hulu 5:03 -9.6% Cwtv.com 3:32 103.6% Megavideo 2:49 -22.9% StageVU 2:32 -11.2% YouTube 2:14 -6.8% Nickelodeon Family & Parents 2:13 5.9% Justin.tv 1:58 -23.1% Veoh 1:49 -19.8% Source: The Nielsen Company Read as: During February 2011, U.S. video viewers spent an average of 9 hours, 16 minutes watching video content on Netflix using PC/laptops from home and work locations

Source: Comscore 21 In Canada, Netflix is also a paid service* and probably mostly watched via blu-ray players, game consoles and other „unmeasured‟devices, but in the PC/laptop environment measured by comScore Canada, it has seen phenomenal growth since it launched here. Netflix.com has gone from basically zero visitors in September 2010 to over 5 million in May 2011 (see below). Many may be just visiting the web site but, nonetheless, the growth is spectacular, making it one of the most visited media web sites in Canada.

*Note: some Netflix users are on a 1-month free trial subscription. 22 3. OTT Potential Impact on Canadian TV Viewing

23 BBM‟s audience data is coded for program origin and genre through an elaborate system involving all broadcasters and BBM. CMRI has developed a tool for analyzing and tracking this data, which allows the industry to understand the overall performance of Canadian programs in general and by program type. For example, in the years 2005 to 2010 we can see in the chart that the overall audience share of Canadian programs has remained at over 30%. What would happen to the Canadian program share if Netflix, Apple TV and other OTT services were to operate in Canada with little or no Canadian content or regulations to ensure Cancon expenditure and scheduling?

Audience Share of Canadian vs. Foreign Programs, All English TV, September to August, 2004-05 to 2009-10, Persons 2-plus, 24 Hours

100%

90% Foreign 80% Canadian

70% 69.2% 66.8% 68.4% 68.0% 66.1% 66.2% 60%

50%

40%

30%

20% 30.8% 33.2% 31.6% 32.0% 33.9% 33.8% 10%

0% 2005 2006 2007 2008 2009 2010 24 Source: CMRI (BBM) If Netflix, Apple TV and other OTT services were to operate in Canada with little or no Canadian content or regulations to ensure Cancon expenditure and scheduling, CMRI estimates that over the next ten years the audience share of Canadian programming would decline from its current level of about 33% to just under 27%, a decline of some 6 share points. To put that in context , it would mean that Canadians would be spending about 2 billion fewer hours annually watching Canadian content and instead would be spending that time with foreign, mostly U.S. programs (and U.S. content providers).

Audience Share of Canadian vs. Foreign Programs, All English TV, September to August, 2004-05 to 2019-20, Persons 2-plus, 24 Hours

100% Foreign 90% Canadian

80%

70% 66.8% 66.1% 66.2% 68.2% 67.2% 67.7% 69.2% 68.4% 68.0% 68.2% 69.0% 69.6% 70.5% 71.3% 72.3% 73.1% 60%

50%

40%

30%

20% 33.2% 33.9% 33.8% 33.2% 32.8% 32.3% 30.8% 31.6% 32.0% 31.8% 31.0% 30.4% 29.5% 28.7% 27.7% 26.9% 10%

0% 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Projection:

Projections based on assumed take up and usage levels of internet as determined by the TV Trends Survey Source: CMRI (BBM) 25 2007-2010 and the relative viewing shares of different genres of Canadian program (see appendix). BBM‟s audience data is coded not only for program origin but also genre. In the years 2005 to 2010 news, information and documentary accounted for about one third of Canadian program viewing, sports another 5-6 percentage points, and about the same for dramatic programming, including animation, as well as other entertainment programming. What would happen to the audience for each genre, if internet TV take up and usage grows substantially in the next decade?

Audience Share of Various Types of Canadian Programs, All English TV, September to August, 2004-05 to 2009-10, Persons 2-plus, 24 Hours 40.0%

33.9% 33.8% 35.0% 33.2% 31.6% 32.0% 0.7% 0.8% 30.8% 0.8% 0.7% 2.4% 2.3% 1.9% 0.9% 0.8% 1.1% 1.1% 30.0% 1.4% 2.1% 2.3% 2.1% 2.3% 1.1% 1.1% 3.1% 3.0% 1.5% 2.8% 3.1% 25.0% 2.0% 3.1% 2.4% 2.5% 0.9% 2.7% 2.4% 1.0% 1.0% 2.9% 1.2% 1.2% 1.2% 1.1% 1.0% 1.9% 1.1% 2.0% 1.2% 1.1% 2.1% 20.0% 1.0% 1.8% 2.0% 2.4% 6.5% 6.8% 6.6% 5.0% 15.0% 3.7% 4.9%

1.9% 1.9% 2.1% 2.1% 2.0% 2.1% 10.0% 2.0% 1.6% 1.5% 1.9% 1.8% 1.7%

5.0% 9.3% 9.5% 9.5% 9.2% 9.5% 9.4%

0.0% 2005 2006 2007 2008 2009 2010

News Information Documentary Sports Drama Series Movies/MOWs Comedy Animation Reality/General Ent. Music/Variety/Quiz Light Entertainment Other 26 NHL lockout in 2005 affected sports share Source: CMRI (BBM) If internet TV take up and usage grows substantially in the next decade, virtually all Canadian program genres would lose audience, including sports, since MLB, NBA and NHL games would still be watched but without Canadian broadcast coverage. Sports would lose about one fifth of its audience. Canadian news and information would suffer relatively little because Canadians would continue to want local and national Canadian TV news but Canadian drama series would lose about 35% of its audience, as would Canadian movies/MOWs.

Audience Share of Various Types of Canadian Programs, All English TV, September to August, 2004-05 to 2019-20, Persons 2-plus, 24 Hours 40.0%

35.0% 0.7% 0.8% 0.8% 0.8% 0.8% 0.7% 2.4% 2.3% 0.8% 1.9% 0.9% 2.2% 2.1% 0.7% 0.8% 1.1% 1.1% 2.1% 0.8% 30.0% 1.4% 2.1% 2.3% 1.2% 2.1% 0.8% 2.1% 1.1% 1.1% 2.1% 0.8% 1.1% 1.1% 2.0% 0.8% 2.3% 1.1% 3.1% 3.0% 2.7% 1.0% 1.9% 1.5% 2.8% 2.8% 1.0% 1.9% 0.8% 2.8% 3.1% 2.8% 1.0% 1.8% 0.8% 2.0% 3.1% 2.4% 2.5% 2.7% 2.6% 1.0% 1.8% 25.0% 2.6% 2.6% 2.5% 0.9% 2.4% 2.3% 2.5% 0.9% 0.9% 2.7% 2.4% 1.0% 1.0% 1.0% 2.3% 2.9% 1.2% 1.2% 1.0% 1.0% 2.2% 2.4% 1.2% 1.1% 1.2% 1.2% 1.0% 2.1% 2.3% 1.0% 1.1% 1.0% 0.9% 0.9% 2.0% 1.1% 2.0% 1.1% 2.1% 1.9% 2.0% 1.0% 1.0% 0.9% 1.9% 20.0% 1.2% 1.9% 1.8% 1.8% 0.9% 0.8% 1.7% 1.0% 2.0% 1.7% 0.8% 0.8% 1.8% 1.6% 1.6% 0.8% 0.7% 2.4% 1.5% 0.7% 1.4% 1.3% 6.8% 6.6% 6.5% 6.6% 6.6% 6.4% 4.9% 5.0% 6.3% 6.3% 6.1% 15.0% 3.7% 5.9% 5.8% 5.6% 5.5% 2.0% 1.9% 1.9% 2.1% 2.1% 2.1% 2.0% 2.0% 2.0% 1.9% 1.8% 1.8% 1.7% 1.6% 2.0% 1.6% 1.5% 1.9% 1.8% 1.7% 1.5% 1.5% 10.0% 1.7% 1.7% 1.7% 1.7% 1.7% 1.6% 1.6% 1.6% 1.6% 1.6%

5.0% 9.5% 9.5% 9.5% 9.4% 9.3% 9.2% 9.2% 9.1% 9.1% 9.0% 8.7% 8.8% 8.6% 8.5% 8.3% 8.1%

0.0% 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Projection:

News Information Documentary Sports Drama Series Movies/MOWs Comedy Animation Reality/General Ent. Music/Variety/Quiz Light Entertainment Other 27

Source: CMRI (BBM) Canadian movies/MOWs would be severely affected by foreign OTT services. Canadian movies/MOWs currently account for about 1.2% of the total English TV audience (see earlier chart) but, given their premium status, a disproportionate share of this audience is currently delivered by The Movie Network/Movie Central (e.g., 18.2% in 2010). These premium movie services are the first Canadian channels that would likely be negatively affected by OTT services, which rely heavily on movie offerings, mostly from Hollywood.

Audience Share of Canadian Movies/MOWs, Conventional vs. Specialty, All English TV, September to August, 2004-05 to 2009-10, Persons 2-plus, 24 Hours

100% 0.0%2.9% 4.3% 5.7% 3.9% 4.6% 4.8%

90% 16.2% 17.7% 18.2% 20.6% 20.8% 80%

70% 51.9%

60% 39.7% 37.1% 33.3% 38.9% 37.3%

50%

40%

25.8% 30% 25.8% 32.6% 34.4% 29.7% 32.2% 20%

10% 19.5% 12.2% 7.6% 5.7% 9.1% 7.5% 0% 2005 2006 2007 2008 2009 2010 CBC (incl. News Network) Canadian Private TV Canadian Specialty TV Canadian Pay TV Other Source: CMRI (BBM) Pay TV not coded by BBM in 2005. 28 Canadian drama series would also be severely affected by foreign OTT services. Canadian drama series currently account for about a 2% share of the English TV audience but this genre too would lose about a third of its audience in an unfettered OTT environment. Currently Canadian specialty channels, which schedule multiple airings of old and current Canadian drama and account for over 50% of the audience to Canadian drama series (see below), would also be the first to be „shaved‟ by cable/DTH subscribers. In an unregulated environment foreign OTT services would be unlikely to give Canadian drama the same prominence and audiences would fall substantially.

Audience Share of Canadian Drama Series, Conventional vs. Pay/Specialty, All English TV, September to August, 2004-05 to 2009-10, Persons 2-plus, 24 Hours

100% 0.9% 0.3% 1.2% 0.0% 1.4% 0.9% 0.7%2.1% 1.0%1.8% 0.9%1.7%

90%

80%

51.0% 70% 58.0% 57.8% 63.8% 64.5% 72.9% 60%

50%

40%

30% 38.0% 32.4% 33.8% 20% 31.1% 31.0% 19.4% 10% 8.2% 4.2% 6.0% 6.8% 5.9% 0% 2.5% 2005 2006 2007 2008 2009 2010 CBC (incl. News Network) Canadian Private TV Canadian Specialty TV Canadian Pay TV Other 29 Source: CMRI (BBM) 4. Measuring the Audience to OTT Services

30 Diaries, People Meters and PPMs: Measurement Issues

1980s

•For most of our TV history audience measurement in Canada was provided by two companies, Nielsen and BBM. Both companies relied on diaries as the method of data collection prior to 1989. In 1989 a new method was introduced by Nielsen which involved a meter connected to the television set. This meter was called the people meter. Despite debates about methodological issues, the industry agreed that meters accurately measured the station a TV set was tuned to. The president of BBM put it best in a September 1991 speech to the Broadcast Executives Society: “…because people meters are connected to the television they eliminate all of the measurement problems associated with signal proliferation, simultaneous substitution and so forth.‟ The industry did not anticipate internet TV at this point.

•After several failed attempts to get Nielsen and BBM to work together, Nielsen eventually launched its people meter system independently of BBM in the fall of 1989.

•Wherever the people meter was introduced, whether in 1984-85 in the U.K, in 1986-87 in the United States and finally in Canada in 1989, audience researchers were consumed with the methodological issues surrounding people meter systems. Would people push the buttons on the hand-held remote? Would they agree to participate for a year, two years, or longer? If people left the room for several minutes, would they log out? These and many other methodological issues arose and were studied in depth.

1990s Changing Environment

•In North American the explosive growth in the number of cable TV pay and specialty channels, starting with HBO in 1972, created the need to update meter technology. Cable TV subscription rates exceeded 50% in both Canada and the U.S. by 1990, providing subscribers dozens of channels. This meant that the meter had to be capable of accurately identifying not only 3 or 4 local channels but many channels, some from distant markets. Cable companies began providing many different versions of converters or set-top boxes that not only gave the subscriber a remote control for the TV set but also allowed access to higher-number channels, including some 31 that were scrambled, premium channels. Diaries, People Meters and PPMs: Measurement Issues (continued)

•TV sets were changing too, becoming cheaper and increasing in number, such that the average household had more than two sets. Small, portable sets were introduced and they were an added challenge to measure. VCRs were also being introduced at this time. Initially, advertisers worried that the VCR would mean that viewers would skip or fast forward through commercials and demanded that meters be capable of capturing this behaviour. The meter now had to be adaptable to many different household technical TV set/VCR/cable TV combinations and permutations.

•In 1993 DirecTV satellite service was launched in the U.S., soon to be followed by several competitors and Canadian counterparts. People meters had to be able to measure stations/networks that were being delivered on this new digital platform. In addition, terrestrial broadcasters were developing standards for digital TV allowing multiplexing of their channels. The cable TV industry was also experimenting with digital technology and two-way, interactive TV and VOD. Digital TV signalled that old meter technology that was based on reading the frequency that a TV set was tuned to would no longer be sufficient since digital television did not employ frequencies.

•In the early 1990‟s, of course, few people were far-thinking enough to realize that the meter might not be hard-wired to the TV set and that in the future more than one means to identify stations/programs might be required. Although, even early on in the history of the people meter the TV industry in North America was rapidly changing and this affected the ability of meters to accurately identify the channel/program tuned.

32 Diaries, People Meters and PPMs: Measurement Issues (continued)

•The seminal 1999 EBU publication on audience measurement standards, Towards Global Guidelines for Television Audience Measurement, summarized the challenges facing audience measurement companies:

…digital compression will cause a need for audience research companies to make sizeable investments in meter design to ensure that current peoplemeter systems can adapt to digital compression and transmission. The systems will have to be provided with new techniques that will ensure correct channel and program identification, as the direct relationship between measured voltage and channel will not necessarily exist. Indeed, although new techniques such as passive meters may offer the possibility of major advances in audience measurement, it is impor•tant not to lose sight of the fact that peoplemeters, which are today's industry standard, also face serious challenges. Without a major investment in research and development, electronic measurement may not be able to contend with the technical changes brought about by the introduction of digital TV.”[1]

•The radio/TV research standards of the Canadian Advertising Research Foundation, issued as long ago as 1997, also recognized the challenges that the ratings companies would soon be facing:

“The Internet and its successor and perhaps co-medium, interactive television, will not be measured as anything else we have experienced. There is nothing on the shelves of either experience or technique which we can dust off and plug into this new environment. The new media are no longer passive. They invite and demand participation and control - especially control. While there will still be media-based opportunities to accumulate large audiences, the Media will not control the timeframe or continuity of viewing and "large" will be a smaller "large" than we currently know. How do we measure audiences which dynamically create their own "networks" at will?”[2] [1] Towards Global Guidelines for Television Audience Measurement, European Broadcasting Union, 1999. [2] Canadian Advertising Research Foundation, Research Guidelines for the Electronic Media in Canada, March 1997.

33 Diaries, People Meters and PPMs: Measurement Issues (continued)

•In 1992 BBM signed an agreement with an American company, Arbitron, to develop a portable people meter system, the PPM. After years of testing this technology was introduced by BBM in 2002. And in 2004, BBM announced that PPM would be the technology of choice, replacing people meters. By fall 2004, BBM had a working PPM system in place in Quebec and it has been operational ever since. Arbitron has commercialized the PPM not only in Canada but also in the UK, China and other countries.

•The PPM is a portable, passive system that reads encoded, inaudible signals registering channels that are potentially being watched. The encoded signals are embedded in TV broadcasts and respondents carry with them or wear a small device that is activated by the inaudible signals whenever they are near a TV set (or radio). At the end of each day the survey respondent must also remember to place the meters into base stations that recharge the devices and send the collected codes to BBM for tabulation. No longer does a home TV set need to be hard-wired, as with the people meter.

•Following the merger of BBM and Nielsen in 2006, which effectively created a monopoly in audience measurement, PPM was chosen as the only technology to measure national and large market TV audiences. By the fall 2009, the PPM replaced the people meter across Canada.

34 Diaries, People Meters and PPMs: Measurement Issues (continued)

•BBM describes the PPM: “The Portable People Meter, developed by Arbitron, is a pager-sized device that is carried by a representative panel of television viewers. It automatically detects inaudible codes that broadcasters embed in the audio portion of their programming using encoders provided by BBM and Arbitron. At the end of each day, the survey participants place the meters into base stations that recharge the devices and send the collected codes to BBM for tabulation. The Portable People Meter can measure exposure to any electronic media, which has audio that can be encoded – television, cable, and radio, even cinema advertising and in-store media.”

•Some preliminary analysis of PPM data has been done (see earlier section). We know, for example, that the change to the PPM has resulted in increased viewing levels. This PPM „effect‟ is most noticeable once viewers start their day and is consistent from about 7:30am to about 11:00pm weekdays. Weekends follow a similar pattern, suggesting that the effect is not related to out-of-home activity. One thing we know for certain is the definition of the „audience‟ has changed dramatically from being in the viewing room with people meters to being in the household or within earshot of the TV set with the PPM. In the case of people meters respondents were asked to periodically confirm they were present in the TV room, while in the PPM system, no confirmation is requested.

35 Diaries, People Meters and PPMs: Measurement Issues (continued)

Measuring Internet TV Viewing

As early as the late 1990‟s the industry recognized that the internet was a potential source of TV viewing and in 2010 this became a reality with the launch of Netflix/Apple TV in Canada. With regards to OTT services, BBM and comScore, the internet ratings service in Canada, are currently not providing the kind of ratings data that CBC, CTV, TVA, etc.are used to receiving. BBM informed CMRI that it has no plans at this time to measure Netflix or other OTT services.

It is important that the TV industry, along with the Commission and the ratings companies work together to find a solution to what will eventually become a major measurement problem. In the U.S. Arbitron and comScore are working on an integrated system that captures regular TV viewing, mobile and internet TV. This has been organized by the Coalition for Innovative Measurement (http://www.cimm-us.org/index.htm). Nielsen in the U.S. is also working on the multi-screen measurement issue.

As Brent Bernie, CEO of comScore Canada, told CMRI :

The CMRI analysis confirms what we have seen in our proprietary work and syndicated reporting. The rapid development and availability of new technologies that allow access to the internet is changing the way that Canadians access content. While television remains the primary means by which Canadians access original television content, a recent Canadian study we were involved with indicated over twenty percent of Canadians view this content both on television and online and seven percent choose to view through the online medium only.

This creates a new measurement challenge. Consumers will use whatever channel makes sense with their lifestyle and needs to pursue and enjoy the content they want. This then demands that cross media measurement is essential to truly understand the behaviour of different segments of the population. As the leading measurement company in the digital space comScore is uniquely qualified to address this challenge. With syndicated measurement services in the mobile area and traditional online sphere, comScore has built the most comprehensive dictionary available to understand the internet infrastructure. Our selection as one of the vendors to work with the CIMM group in the U.S. underlines both our commitment to cross media measurement and our credentials. The capability to do similar work in Canada exists today. What is required is the corporate will and the investment to begin the learning. 36 Appendix: Explanation of OTT Impact on Canadian Program Audience Share

• In the past four years the TVQ Survey has measured the percentage of the population using Internet TV and the weekly hours spent using internet TV per user:

2007 2008 2009 2010 Internet TV Viewers 15.5% 19.2% 25.4% 30.8% Weekly Hours Per Viewer 2.1 2.7 3.8 4.5

• As shown in the table, the percentage of Anglophone internet TV viewers has doubled since 2007; and the time spent using internet TV per user has more than doubled in this period.

• In fall 2010 OTT viewing accounted for under 1% of TV viewing. CMRI assumed that the introduction of OTT services would substantially increase in 2011 and in future years. By 2020, it was assumed that OTT services would account for 15% of all TV viewing and that a small proportion, 10%, would be to Canadian programming. The remaining 90% would accrue to foreign programs, which would result in a substantial decrease in viewing to Canadian programs in the system overall. The rapid growth of internet based such services as Youtube, First Row Sports, Hulu and others, coupled with the new easy to operate devices such as Apple TV led to CMRI‟s high expectations for OTT use. • CMRI used historical shares of viewing accounted for by major genres of programs and the relative performance of Canadian content within each major genre (news, sports, drama series, etc.) to establish a ratio of how OTT would affect each major genre separately (see next page).

• These two factors, assumed OTT growth over the next 10 years and the historical performance of Canadian news, drama, sports, etc., were used to estimate the long term impact of OTT services.

37 Viewing Share of Canadian and Foreign Programs, All English TV, Persons 2+, 24 Hours, 2005 to 2010

2005 2006 2007 2008 2009 2010 Country Canadian Program Genre News 9.3% 9.5% 9.5% 9.2% 9.5% 9.4% Information 2.0% 1.6% 1.5% 1.9% 1.8% 1.7% Documentary 1.9% 1.9% 2.1% 2.1% 2.0% 2.1% Sports 3.7% 6.6% 4.9% 5.0% 6.5% 6.8% Drama Series 2.4% 2.0% 1.8% 2.0% 2.1% 1.9% Movies/MOWs 1.0% 1.2% 1.2% 1.1% 1.2% 1.2% Comedy 1.1% 0.9% 1.0% 1.1% 1.0% 1.0% Animation 2.9% 3.1% 2.7% 2.4% 2.4% 2.5% Reality/General Ent. 2.0% 2.3% 2.8% 3.1% 3.1% 3.0% Music/Variety/Quiz 1.5% 1.4% 1.1% 1.1% 1.1% 1.1% Light Entertainment 2.1% 1.9% 2.1% 2.3% 2.4% 2.3% Other 0.8% 0.8% 0.9% 0.7% 0.7% 0.8% Canadian Total 30.8% 33.2% 31.6% 32.0% 33.9% 33.8% Foreign Program Genre News 3.2% 3.3% 1.5% 1.4% 2.1% 1.7% Information 1.2% 1.0% 1.5% 1.3% 1.5% 1.4% Documentary 4.2% 3.6% 3.4% 3.6% 3.7% 3.6% Sports 3.6% 4.0% 3.3% 3.5% 3.5% 3.4% Drama Series 14.1% 14.3% 15.5% 13.5% 14.2% 14.4% Movies/MOWs 8.9% 8.6% 9.0% 8.7% 8.5% 8.7% Comedy 5.8% 5.6% 6.0% 6.0% 6.1% 6.0% Animation 5.1% 5.0% 4.8% 5.1% 5.0% 5.0% Reality/General Ent. 10.3% 8.9% 10.3% 11.1% 10.8% 10.7% Music/Variety/Quiz 1.2% 1.6% 2.5% 3.0% 2.6% 2.7% Light Entertainment 2.6% 2.4% 2.5% 2.6% 2.4% 2.5% Other 9.1% 8.5% 8.1% 8.2% 7.1% 7.8% Foreign Total 69.2% 66.8% 68.4% 68.0% 66.1% 66.2% Source: CMRI (BBM)