Canadian Radio-television and Telecommunications Commission
www.crtc.gc.ca
Common menu bar links
Institutional links
Privacy and Security
Completed Access to Information Requests
Proactive Disclosure
Measuring TV Audiences: Defining'Audience Success'
April 14, 2008
Canadian Media Research Inc.,
Ottawa, Ontario
This report is available electronically at www.crtc.gc.ca
Ce document est également disponible en français
Profile: Barry Kiefl
- Barry is the President of Canadian Media Research Inc., an independent research company established in 2001. Barry was Research Director for the Canadian Broadcasting Corporation (CBC) from 1983 to 2001. While with the CBC he was responsible for all research services for CBC English and French radio and television, including audience research for advertising sales. He has written numerous articles and spoken at many professional research conferences on the issues surrounding audience research and the electronic media. He has conducted research for the Canadian Radio-Television and Telecommunications Commision, including a review of international trends in television and a study of cable and direct-to-home penetration versus over-the-air television viewing. He published a seminal piece on audience research techniques in the 1995 Canadian Journal of Marketing Research. He has been a speaker at the annual conventions of the Canadian Association of Broadcasters, the Canadian Cable Television Association, the Upper Canada Law Society, BBM's Annual Staying Tuned Conferences, the Advertising Research Foundation, the Professional Market Research Society and the Canadian Satellite Users' Association. Barry has been an expert witness in patent cases and has testified before the Copyright Board and other regulatory bodies.
- Barry is an expert in all audience research methodologies, including surveys, market trials, content analyses, field experiments, focus groups and 'diary' and 'people meter' audience measurement systems. He was granted a patent for a Portable People Meter-like television and radio audience measurement system by the U.S. Patent Office in 1995. He has represented Canada on a number of international research projects, including the European Broadcasting Union's "Toward Global Guidelines for Television Audience Research," 1999, which he co-authored. He was Technical Research Director of the Canadian Advertising Research Foundation (CARF) from 1994 to 1999 and authored CARF's "Research Guidelines for the Electronic Media in Canada," March 1997. He has been a member of the BBM Board of Directors and Nielsen's Advisory Board.
The Proposals from Various Parties
Audience Measurement Terminology
Canadian Program Audience Trends
French television: A Special Case
Reconciling Different Proposals
Appendix: French TV Tables and Charts
Executive Summary
- Several broadcasters and other parties have expressed concerns about how 'audience success' is defined by the Canadian Television Fund (CTF). Canwest has proposed that success be defined based on total station viewing rather than on viewing to CTF-funded programs.
- The Association des producteurs de films et de télévision du Québec made the controversial proposal to abolish the audience success component in French television, to be replaced by other criteria.
- A number of other proposals were suggested, including rewarding broadcasters for scheduling programs in prime time and recognizing the value of original, first-run programming.
- At the February 2008 Canadian Radio-Television and Telecommunications Commission (CRTC) hearing there was debate about the audience performance of conventional versus specialty channels; Canadian Media Research Inc.'s (CMRI's) analysis shows that pay and specialty channels have captured a larger share of the total Canadian program audience in recent television seasons within CTF program genres.
- Specialty channels, because they are more numerous and offer numerous repeats of Canadian programs within key program genres, dominate the audience success calculations as presently defined.
- For example, Showcase aired almost 14,000 hours of Canadian drama/comedy programming to generate its high number of viewing hours in the past four years. Space aired almost 12,000 hours. Meanwhile CTV, Global and the Canadian Broadcasting Corporation (CBC) had but a small fraction of the hours aired by specialty channels but much of it was original, first-run programming.
- CMRI recommends that the viewing hours of each network be weighted by a special factor, which incorporates many of the ideas expressed at the CRTC hearing. The Hit Factor is an amalgam of many different elements that contribute to audience success. A network's Hit Factor would be higher if it airs more original programming, a concern of CBC and Global; it would be higher if a network promotes Canadian programming and it will naturally be higher if programs are aired in prime time and prime viewing seasons of the year. It would encourage networks to commission more original programming and to schedule and develop Canadian programming aggressively. Conventional broadcasters who have mass (large) audiences for original programming, such as CTV and Global, would receive a larger share of total viewing hours as a result. In short, it incorporates many of the proposals and ideas made to the CRTC by various parties and reflects true 'audience success' and is a substantial response to Shaw's basic concerns about the CTF.
- CTF should consider calculating audience success based on all Canadian programs within CTF genres, not just CTF-funded programs. This would recognize a broadcaster's total commitment to Canadian programming. CTF should also consider calculating audience success within the cable and direct-to-home universe rather than among all households, which would put all networks on a level playing field. That is, audience success would be determined within the universe that all conventional and pay and specialty channels have at least the opportunity to reach.
- Foreign drama series and movies account for a very substantial audience in French television. Combined, the two drama categories account for about 20% of all French television viewing compared to the 8-9% audience share of Canadian drama. CMRI believes that it may be premature, at least in drama, to jettison audience success as a component in the distribution of CTF funds in French television.
Introduction
Canadian Media Research Inc. (CMRI) was asked by the Canadian Radio-Television and Telecommunications Commission (CRTC) to review the various models and approaches broadcasters and other parties have proposed regarding how the Canadian Television Fund (CTF) measures 'audience success.'
Commissioner Rita Cugini, chair of the February 2008 CRTC hearing on the CTF Task Force Report, stated in her opening remarks at the hearing that "the audience success factor carries the greatest weight in English and significant weight in French and is derived from the total hours tuned that each broadcaster achieves over the course of one broadcast year to programs supported by the CTF." The audience success factor is assigned a value as high as 40% in the determination of the broadcasters' annual CTF funding envelope and is therefore crucial in the administration of the CTF.
Commissioner Cugini effectively and succinctly described the purpose of this report at the hearing: ".the manner in which audience success is defined, measured and rewarded remains a topic of study and debate. We are cognizant of the limitations in the industry's audience measurement systems to serve the CTF's audience success measurement requirements and we continue to seek consensus and solutions to these issues." Thus, CMRI has reviewed the transcripts of the hearing as well as other documents and has prepared this report which compares, evaluates and summarizes the key ideas and proposals regarding the measurement of audience success.
Methodology
The report reviews the submissions of various parties and the transcripts of the hearing dealing with the CTF, which lasted five days, covered a wide variety of topics, and included some far-reaching and controversial ideas about the measurement of audience success. Here is the methodology followed:
- The proposals regarding audience success presented by broadcasters and others were examined and compared in simple language that cuts through the terminology of audience reach, average minute audience, hours of viewing, etc. Easy-to-understand definitions of all the terms used in audience measurement are provided.
- The report also examines audience data from the past four television seasons to understand how various broadcasters and sectors have performed.
- Finally, the report expands on the ideas proposed by various interveners and provides some tools that can be used to craft recommendations regarding audience measurement and success. The report examines one possible approach to improving the measurement of audiences by CTF, which takes into account many of the ideas expressed at the hearing.
The Proposals from Various Parties
- Canwest put forward the most controversial proposal for the English television industry when it suggested that the audience success component be based on total station or all program viewing hours rather than just viewing hours to CTF-funded programs. The concept was to reward a broadcaster for its total audience, not just the audience of CTF-funded programs or Canadian programs, but all programs including news, sports and foreign programming. Canwest's rationale for this position was that a broadcaster that delivered large audiences for its entire schedule would be more likely to attract large audiences for its CTF-funded programs. A number of interveners objected to Canwest's idea. Even fellow private broadcaster CTV questioned this idea on the grounds that it was against the public interest. Canwest used an analysis of total station viewing hours as a rationale for suggesting that the CTF fund be equally divided into pay and specialty and conventional broadcaster components. Canwest's suggestion might have some merit in a less fragmented environment where audience lead-in and lead-out and cross program promotion were more effective. But today, when most viewers have hundreds of channels to choose from and technology that makes switching to other channels and programs quite easy, the idea has less merit; but some mechanism for rewarding large audiences would be very appropriate and this report will explore how to accomplish this. Canwest also made reference to the use of viewing within target demographics rather than of persons aged two and over, and called for audience success to be worth more than at present and stressed the importance of "first-window exhibition."
- The Association des producteurs de films et de télévision du Québec (APFTQ) made the most controversial proposal for the French television industry by suggesting the complete abolition of the audience success component, to be replaced by other criteria. The rationale for this idea was explained by Commissioner Arpin at the hearing: "[translation] We also heard APFTQ representatives suggest that we eliminate the audience success component for the francophone market given that francophones essentially only watch Canadian programs."
- CBC currently receives 37% of CTF funds without a requirement to demonstrate audience success. CBC proposed that it be allowed to compete for the other 63% of CTF funds, pointing to its record of audience success, with the implied suggestion that audiences for only original airings be counted.
- CTV proposed keeping the status quo, stating that audience success is "properly weighted in the current model." CTV acknowledged there is disagreement about how to measure audiences.
- Astral expressed concern about measuring audiences to new video services such as video-on-demand (VOD), personal video recorders, mobile television and video downloading and streaming via the Internet. Astral also pointed to inequities in the distribution of conventional and pay and specialty services, that is, that cable and direct-to-home (DTH) services cannot reach all Canadians and this should be taken into account.
- Corus echoed Astral's concerns about the measurement system taking into account new media and also proposed, among other things, that the audience success factor represent no more than 30% in determining broadcasters' funds. Corus presented an interesting idea that calls for weighting English drama productions by a Momentum Factor which would apply to returning series: programs that had pre-sales and met other criteria would be given special recognition.
- Rogers proposed that audience success account for 80% of CTF funding decisions and that measurement apply to Canadian content in all eligible program genres in prime time (to reward mass audiences). Thus, Rogers' idea stopped short of Canwest's idea regarding total station audiences but Rogers did propose including audiences to all Canadian programs with certain genres.
- The Canadian Association of Broadcasters (CAB) proposed that two measures for audience success be used: average minute audience for conventional television broadcasters and cumulative audience for pay and specialty broadcasters.
- A number of methodological concerns about audience measurement were expressed by various other interveners. For example, the Association for Tele-Education in Canada (ATEC) noted that many of its members do not have adequate audience data from BBM Nielsen because of the small sample size in some regions. TFO pointed out that there are only five Nielsen francophone families in the Ontario sample and added that French television generally should not use audience success to determine the allocation of CTF funds. Télé-Québec expressed concern about a recent decline in the number of Quebec homes in the BBM sample with young children. The Alliance for Children and Television said that with a sample of 4,000 homes BBM Nielsen couldn't properly measure television viewing. These complex sampling and methodological concerns will not be dealt with in this report; they, along with concerns about measuring new ways of consuming video, should be examined separately.
In summary, there were seven underlying themes related to audience measurement that emerged at the hearing or in submissions:
- The audience success measure should apply to a broader range of programs than just CTF-funded programs.
- There should be acknowledgement for original, first-run programs.
- Conventional and pay and specialty channels should use different audience measurements that take variances in their distribution into account (pay and specialty cannot be received in as many homes).
- Broadcasters that seek to attract mass (large) audiences by scheduling programs in prime time should be rewarded.
- French television should abandon the measurement of audience success.
- A factor recognizing momentum should be considered: for example, special recognition for returning series, which obviously would constitute original programming and likely would have had a large audience in a previous season.
- Audiences using VOD, mobile television, etc. should be accounted for in the measurement system.
Audience Measurement Terminology
There are many arcane terms that are used in audience measurement but we need only be concerned with a few of them. The first is average audience or average minute audience.
Average Audience: Average audience is the number of viewers watching a program (or time period) in the average minute of a program. It has been used since the early days of television and was then and remains the measure of most interest to advertisers-it represented the probable audience that would view an advertisement placed within a program. Advertisers in the 1950s usually didn't control where their commercial would air in a show and so they wanted to know the average audience for the whole program.
Average audience is the measure commonly used by CBC, CTV, TVA, etc. to express how many people watched a given program. When Little Mosque on the Prairie debuted in January 2007 with an average audience in excess of two million viewers, CBC was rightfully proud to point to the very high average audience. Similarly, CTV will promote that Corner Gas has an average audience of over one million viewers. In fact, the audience for any show might be larger at the beginning or the end but average audience is a measure of the audience averaged throughout a half hour or hour long program. Even specialty channels, which tend to have much smaller average audiences than the main networks, will sometimes tout the average audience for a program. Average audience can also be expressed as a percentage rating, which is the average audience divided by the total population.
Viewing Hours: The second important metric used by the industry, and employed by CTF when calculating a broadcaster's audience, is viewing hours. Viewing hours is quite simple; it is the average audience multiplied by the duration of a program. Little Mosque on the Prairie had an average audience of about two million viewers for its first episode last year and viewing hours for the show equals 2,000,000 viewers X .5 hours = 1 million viewing hours. Corner Gas, with about 1.5 million viewers, has 1,500,000 viewers X .5 hours equals 750,000 viewing hours. Average audience and viewing hours are closely related; one is a function of the other. Also related to average audience and viewing hours is audience share, which is the expression of viewing hours as a percentage (of all viewing hours in a program's time slot). Note that a program with a low average audience but a long duration can account for as many viewing hours as a short program with a large average audience.
Audience Reach: Audience reach is a slightly more difficult concept and can only properly be calculated with special BBM Nielsen software that counts the number of different people who tuned to a given program or several episodes of a program. Reach is calculated by determining the number of different people who watched a minimum number of minutes (one minute, five minutes, 15 minutes, etc.) of one show or multiple airings of a show. Reach is sometimes called cume audience and in Internet audience research is referred to as unique visitors. If reach is calculated over a large number of program episodes or a long enough time period, say, in the case of Internet websites over the course of a month, it will generate a large number, since more people will have tuned in at least once to a program series with many episodes or over a month.
When CBC reports that 700,000 viewers watched The Border on a given night, they normally are referring to the average audience but CBC could calculate the five-minute reach or the number of different people who watched five minutes or more of The Border that night. This would be the audience reach and it is normally higher than the average audience.
Reach is occasionally confused with adding together the average audiences of several episodes of a program, the so-called aggregate or cumulative audience. For example, if Trailer Park Boys airs on one night with an average audience of 300,000 viewers and on another night with 200,000 viewers, some in the industry will add the average audience numbers together and say that 500,000 people watched Trailer Park Boys. Another example: CBC reports that it has a million viewers for The National each night by adding together the average audience on the main CBC channel and the average audience for the five or six times The National airs on Newsworld each night.
However, this is not a true measure of average audience or audience reach. Some people may watch a program like Trailer Park Boys on both nights it airs and are double-counted in this measure, which serves neither the advertiser nor the broadcaster seeking to know how successful a show was at attracting an audience. In effect, the advertiser and the broadcaster must devote two, three, four or five times more air time to achieve the aggregate audience, which is expensive. Aggregating average audience numbers of multiple episodes generally betrays an effort to boost the apparent success of a show.
The three key metrics, average audience, viewing hours and audience reach, can all be used to determine audience success and we will discuss how to best use them.
Who's on First?
During the February hearing and in the written submissions, numerous claims were made by broadcasters and others about audience success despite the debate about what constitutes success. Whether it was the CBC, Astral, CTV, Canwest, the CAB, Shaw or others, almost every broadcaster asserted in one way or another that they were leaders in audience terms or, in the case of Shaw, that CTF-funded Canadian programming was not generally successful. Ratings data have always been used to promote the success of programs or networks and the data can be sliced and diced to benefit almost all broadcasters. In radio, it is said that ratings data can be 'spun' to make every station the number one station in at least one demographic or one time period.
The problem with most of the claims at the hearing is that they were based on analysis of BBM Nielsen data without revealing precisely how the analyses were done. Following are some examples.
For example, CBC claimed at the hearing that ".nine out of the 10 most popular series in English Canada are on the CBC. We are as well the most efficient user of the fund. We take about 50 percent of the drama dollars in English and generate almost two-thirds of the audiences for Canadian drama in English." CBC's final comments revealed that its analysis was based on 'original' programs, which would exclude the vast majority of tuning. However, as CBC and Canwest suggest, it would seem desirable to reward a broadcaster for audiences to first-run programming.
CAB said at the hearing that: "Private broadcasters command the largest market share in viewing to Canadian television programming services. In 2006/2007 private broadcasters received 83% audience share to drama and long-form documentary programming." CAB furnished more statistics to back up their claim: "[translation] To give you a reference point, in 2006-2007 only a weekly average of 13% of the 88 million viewing hours of Canadian programs were generated by the CBC/Société Radio-Canada."
Astral, meanwhile, filed research findings that echoed the CAB's claims and claimed that specialty channels accounted for the majority of television viewing in CTF genres: "[translation] In the francophone market, specialty channels capture 50% of the total tuning time in these three program categories. In fact, 49% is captured by specialty and pay channels and the other 51% is shared between private and public conventional broadcasters."
In the final comments CBC countered Astral's analysis, stating that:
Finally, contrary to the claims of Astral, conventional broadcasters such as CBC/Radio-Canada are the cornerstone of the Canadian broadcasting system. The analysis provided by Astral during the hearing is flawed: it employs audience data that includes repeat programming. When this data is corrected so that it includes only audiences to original programming, the cornerstone role of conventional broadcasters is revealed: specialties account for 10% or less of all viewing to original Canadian drama/comedy programs in each of the major markets of Montreal, Toronto, and Vancouver.
CBC's analysis was based on the 'sweep week' periods of November 2005 and March 2006, when private broadcasters tend to feature U.S. programming. The CBC analysis was also restricted to three cities. Unfortunately, systematic analysis of original programming versus repeats based on the complete television season is not feasible and so one must be selective about the weeks chosen for analysis.
To shed some light on how various broadcasters and sectors have been performing in audience terms, CMRI has conducted a thorough analysis of audiences for Canadian programs in the past four complete television seasons.
Canadian Program Audience Trends
CMRI maintains historical BBM Nielsen data by program origin and genre in a special aggregate form. CMRI works with the agencies responsible for coding the data, including CTF, and regularly verifies and corrects coding errors. CMRI believes it is the most complete and accurate database for analysis of audiences by origin and program genre.
CMRI's analysis system includes estimates for U.S. networks, both broadcast (CBS, NBC, ABC, etc.) and specialty networks (A&E, TBS, CNN, etc.). BBM Nielsen data include codes for Buffalo television stations and these stations are used as a proxy for the U.S. broadcast networks nationally. Data for 2005-06 to date include estimates for The Movie Network, which is used as a proxy for all pay television in English television. Super Écran has been included since 2006-07. Data for the Toronto stations of CITY, A Channel and CH (now E!) are given weights equivalent to their national audience; these channels have more limited distribution than CBC, CTV or Global but are now available to most cable and DTH subscribers nationally.
Following is an arm's length look at the audience data from the recent past to better understand basic trends in audiences to Canadian programs, especially within key program genres and conventional versus pay and specialty.
The chart below shows the audience share of Canadian programs versus foreign (mostly U.S.) programs in English television over the past four complete, 52-week television seasons. Clearly, as many at the February hearing acknowledged, the challenge to attract an audience for Canadian programs is significant and constant: the viewing share of Canadian programs has varied in a narrow range from 31 to 33% over the past four years.
Digging one level deeper into the origin and genre data for English television, the chart below reveals that indeed pay and specialty channels have captured a larger share of the total Canadian program audience in recent television seasons. Pay and specialty channels capture roughly 50% of viewing to Canadian programs in any given year, followed by private conventional stations at about 30-35% and the CBC (including Newsworld) at 15-20%, depending on the television season. If one were to take these data at face value, it would only be logical to allocate specialty channels a large share of CTF funds.
Please note that these data contain viewing for all programs, original and repeat, and are for the 24-hour day rather than prime time. No systematic data exists that would allow one to examine national historical data of original versus repeat programming, although ideally original programming should be recognized in the CTF audience success component.
In fact, pay and specialty channels capture an even greater majority of the viewing within CTF genres such as Canadian drama series and documentary, as shown below. Most program categories show the same pattern: specialty channels, because they are so numerous and offer multiple repeats of Canadian programs within key program genres, dominate the audience success calculations.
French Television: A Special Case
The viewing share of Canadian programs in French television is the inverse of English television; roughly two thirds of all viewing in the past four television seasons has gone to Canadian programs (see chart below). It is presumably for this reason that many in French television feel that audience success need not be a component of CTF funding decisions.1
It is interesting to note, however, that foreign program viewing in French television centres on specific genres, as shown in the detailed table below.
Foreign drama series and movies account for a very substantial audience in French television. Combined, the two drama categories accounted for about 20% of all French television viewing compared to an 8-9% audience share of Canadian drama over the past four years. CMRI believes that it may be premature, at least in drama, to jettison audience success as a component in the distribution of CTF funds. Despite the predominance of Canadian program viewing, there is still a significant reliance on foreign drama in French television.
| Audience Share of Canadian vs. Foreign Programs, All French TV, September to August, 2003-04 to 2006-07, Persons 2-plus, 24 Hours | |||||
|
YEAR |
|||||
|
2003-04 |
2004-05 |
2005-06 |
2006-07 |
||
| Canadian | News |
17.5% |
18.8% |
18.6% |
16.1% |
| Information |
8.8% |
9.3% |
8.9% |
8.5% |
|
| Documentary |
3.9% |
3.0% |
3.5% |
2.8% |
|
| Sports |
4.7% |
2.3% |
4.8% |
4.5% |
|
| Drama Series |
5.9% |
5.6% |
5.9% |
6.6% |
|
| Movies/MOWs |
2.4% |
2.2% |
1.7% |
1.8% |
|
| Comedy |
3.0% |
3.0% |
2.2% |
3.3% |
|
| Animation |
1.9% |
2.3% |
2.2% |
2.7% |
|
| Reality/General Ent. |
7.0% |
9.0% |
8.7% |
9.1% |
|
| Music, Variety, Quiz |
4.7% |
4.3% |
4.4% |
4.8% |
|
| Light Entertainment |
3.3% |
3.8% |
3.4% |
3.3% |
|
| Other |
5.3% |
4.5% |
4.3% |
3.7% |
|
| Total |
68.5% |
68.1% |
68.6% |
67.2% |
|
| Foreign | News |
0.3% |
0.2% |
0.2% |
0.2% |
| Information |
0.3% |
0.3% |
0.2% |
0.1% |
|
| Documentary |
2.2% |
2.1% |
2.4% |
2.5% |
|
| Sports |
1.0% |
0.9% |
1.1% |
1.3% |
|
| Drama Series |
6.7% |
7.7% |
7.4% |
7.7% |
|
| Movies/MOWs |
15.6% |
13.6% |
12.8% |
12.3% |
|
| Comedy |
0.2% |
0.2% |
0.2% |
0.9% |
|
| Animation |
2.7% |
2.9% |
2.6% |
3.7% |
|
| Reality/General Ent. |
1.1% |
2.3% |
3.0% |
2.5% |
|
| Music, Variety, Quiz |
0.9% |
0.6% |
0.4% |
0.6% |
|
| Light Entertainment |
0.3% |
0.2% |
0.2% |
0.3% |
|
| Other |
0.1% |
0.9% |
0.8% |
0.8% |
|
| Total |
31.5% |
31.9% |
31.4% |
32.8% |
|
| Source: CMRI | |||||
Reconciling Different Proposals
Earlier we defined the three basic terms used in audience research: average audience, viewing hours and audience reach. Audience reach and its surrogate, aggregate or cumulative audience, is a measure that is not useful in this context.
Reach is useful for understanding whether a program or station is connecting with its potential audience, at least occasionally. For example, the cable community channel, which provides a valuable local service, often has a substantial reach over the course of a month or the television season: most people will tune in at least once per year to the channel but it generally has a small average audience and, therefore, accounts for relatively few viewing hours and therefore share of the audience. Almost all channels (or television series with many episodes) have a large audience reach.
For example, CBC television promotes that it reaches about 65% of Canadians per week; this is based on viewers tuning in for at least one minute in a week. Similarly, the documentary series, Canada: A People's History, boasted that it had an audience reach of tens of millions of Canadians. This too was based on viewers watching for one minute in either English or French, on the main CBC channels or specialty channels, across the many months the series aired. Reach may be useful in this sort of public relations exercise but it has little utility for comparing the success of one broadcaster versus another or one program versus another, and, the essence of measuring success is to be able to compare one thing to another.
Average audience, as the CAB pointed out in its final comments, is such a metric; it allows one to compare the audience size of one program versus another. It is the reason that advertisers and broadcasters have relied on this measure for over fifty years. Viewing hours, the product of average audience times the duration of a program, is also a metric that one can use to compare one broadcaster to another. But, viewing hours can distort the relative success of a channel or program series if the duration, the number of program hours, is exceedingly large.
The table below examines the share of all viewing hours to Canadian drama/comedy series (i.e., CRTC codes 71 and 72, funded and unfunded) that various English Canadian networks have generated in each of the past four television seasons. Surprisingly, Showcase and Space lead all networks with about an 18-19% share each the past four years. These networks do this mostly by airing hundreds of repeat episodes of new and older program series that achieve small average audiences but the tonnage, that is, the number of program hours, builds or augments viewing hour totals. CTV was a close third with about 15%. CBC was a distant fourth, followed closely by Global. In total, over the four years, CBC accounted for 6.8% of all viewing hours to Canadian drama/comedy. CBC's poorest year was last year (4.4%), despite the fact that it introduced the hit Little Mosque on the Prairie.
| Share of Viewing Hours of Funded and Non-Funded Canadian Drama/Comedy Series | |||||
| All Persons 2-plus, 2003-04 to 2006-07 | |||||
|
Year |
|||||
|
2003-04 |
2004-05 |
2005-06 |
2006-07 |
Total |
|
|
Viewing Hours % |
Viewing Hours % |
Viewing Hours % |
Viewing Hours % |
Viewing Hours % |
|
| A Channel |
7.4 |
3.7 |
2.9 |
2.8 |
4.2 |
| BRAVO! |
1.0 |
1.1 |
1.8 |
1.8 |
1.4 |
| CBC |
8.1 |
7.1 |
7.4 |
4.4 |
6.8 |
| CH |
3.2 |
4.0 |
2.5 |
2.4 |
3.1 |
| CITY |
2.6 |
2.9 |
2.0 |
2.2 |
2.4 |
| CMT |
0.3 |
0.4 |
0.1 |
0.2 |
0.3 |
| Comedy |
2.4 |
3.3 |
3.9 |
5.7 |
3.8 |
| CTV |
14.7 |
15.5 |
11.0 |
20.0 |
15.3 |
| Family |
2.5 |
3.8 |
2.9 |
7.4 |
4.1 |
| Global |
9.4 |
6.3 |
4.2 |
3.6 |
5.9 |
| History |
0.9 |
1.5 |
3.0 |
2.6 |
2.0 |
| Showcase |
18.9 |
17.9 |
20.0 |
19.1 |
18.9 |
| Space |
17.3 |
21.7 |
20.5 |
14.9 |
18.7 |
| The Movie Network |
0.5 |
0.5 |
0.2 |
||
| Tvtropolis |
0.2 |
2.9 |
4.3 |
1.8 |
|
| Vision |
2.8 |
2.3 |
2.0 |
1.4 |
2.1 |
| W Network |
3.3 |
5.6 |
9.2 |
4.6 |
5.6 |
| YTV |
5.3 |
2.6 |
3.0 |
2.1 |
3.3 |
| Average |
100.0 |
100.0 |
100.0 |
100.0 |
100.0 |
| Source: CMRI | |||||
This analysis shows that viewing hours can be skewed to favour a network with a large number of program hours, even though they may have small average audiences per program hour. Canwest's proposal attempted to counter this statistical phenomenon but many interveners were critical of Canwest's idea of using total station hours. Canwest essentially wanted to reward a broadcaster for delivering large audiences, which the CTF presumably wants to do. What is needed to accomplish this objective is some statistical combination of average audience and viewing hours within CTF genres.
This analysis of viewing hours, which for all intents and purposes replicates the system employed by the CTF, underscores the problem that Canwest addressed in its proposal. Canwest feels that viewing hours based on the CTF genres is weighted grossly in favour of specialty channels, who offer very little original programming. CBC said basically the same thing.
CMRI believes that there is a simple approach that would still employ the metric viewing hours within CTF genres but control for the large number of program hours that are largely repeats of old programs.
| Program Hours of Funded and Non-Funded Canadian Drama/Comedy Series | |||||
| 2003-04 to 2006-07 | |||||
|
Year |
|||||
|
2003-04 |
2004-05 |
2005-06 |
2006-07 |
Total |
|
|
Program Hours |
Program Hours |
Program Hours |
Program Hours |
Program Hours |
|
| A Channel |
6,112 |
4,047 |
1,677 |
2,393 |
14,229 |
| BRAVO! |
550 |
568 |
795 |
905 |
2,819 |
| CBC |
451 |
343 |
514 |
273 |
1,581 |
| CH |
913 |
1,581 |
1,432 |
552 |
4,478 |
| CITY |
814 |
746 |
684 |
355 |
2,598 |
| CMT |
72 |
97 |
20 |
62 |
251 |
| Comedy |
758 |
770 |
841 |
1,044 |
3,413 |
| CTV |
246 |
233 |
178 |
656 |
1,313 |
| Family |
336 |
530 |
304 |
583 |
1,752 |
| Global |
441 |
201 |
139 |
152 |
933 |
| History |
118 |
373 |
528 |
320 |
1,339 |
| Showcase |
3,381 |
3,150 |
3,843 |
3,602 |
13,976 |
| Space |
3,065 |
3,145 |
2,736 |
2,732 |
11,678 |
| The Movie Network |
240 |
234 |
475 |
||
| Tvtropolis |
89 |
980 |
1,751 |
2,820 |
|
| Vision |
833 |
953 |
1,021 |
1,024 |
3,831 |
| W Network |
916 |
1,264 |
1,989 |
1,168 |
5,336 |
| YTV |
1,054 |
898 |
380 |
419 |
2,750 |
| Average |
1,254 |
1,117 |
1,017 |
1,012 |
4,198 |
| Source: CMRI | |||||
The table above shows the extent of the problem in the present CTF system. This table mirrors the previous table but contains the number of Canadian drama/comedy program hours that each network has aired over the past four television seasons. Showcase aired almost 14,000 hours of programming in these genres to generate its high number of viewing hours in the past four years. Space aired almost 12,000 hours. A Channel led all networks with over 14,000 hours. Meanwhile CTV, Global and CBC had 900-1,600 hours, a fraction of the hours aired by specialty channels that dominate viewing hours in drama/comedy.
A number of far-reaching proposals and suggestions were made in submissions and at the February hearing dealing with the CTF. APFTQ and others in the French television industry suggested doing away with measures of audience success entirely. However, our analysis of French television viewing data would seem to indicate that there is a strong rationale for continuing to use audience success for French television broadcasters, at least within the drama category.
In English television, there were a number of different suggestions:
- Global suggested that total station audience be the determining factor to determine a broadcaster's audience success, however, this would measure success based on foreign programs, as well as news, sports, etc. In other words, if a network spent enough and selected successful shows from the Hollywood studios, it would potentially skew the CTF results. However, some mechanism for rewarding large audiences within CTF genres would seem in order.
- Astral, Rogers and CAB (indirectly) refined Global's idea by examining or proposing that audiences to a somewhat broader range of programming be counted in calculating the audience component of a broadcaster's envelope: that is, either include all Canadian programs or all Canadian programs within CTF genres. CTF currently calculates the audience for CTF-funded programs and up to 10 non-funded Canadian titles. Expanding the base of programming to include all Canadian content within CTF genres would seem to have merit.
- CAB suggested using average minute audience for broadcasters and cumulative audience for specialty channels. As explained above, cumulative audience is potentially a misleading indicator of audience success and has lead to the problem the CTF currently faces. However, average minute audience is the most useful metric for measuring audience success and should be used by CTF, along with viewing hours.
- Canwest, CBC and by implication, Corus, stressed the need to reward broadcasters for original programming and this is also an important factor to incorporate.
- Rogers and others, indirectly, proposed focusing calculations of audience success on prime time programs, when the largest audiences are available. This is very appropriate but can be accomplished by still measuring audiences outside prime time, for example, when children are more likely to watch television.
- Astral, Corus and others addressed the inequities in the distribution of conventional and pay and specialty services, that is, that cable and DTH services cannot reach all Canadians and CMRI agrees that this should be controlled for in CTF's measurement of audiences.
- ATEC, regional broadcasters, Astral and others had concerns about BBM Nielsen sample sizes and other basic methodological issues such as measuring mobile television and other new ways of consuming video. These issues are mostly related to the people meter and Portable People Meter (PPM) technology employed by BBM Nielsen and were not addressed in this report. Nielsen is aggressively seeking ways to measure audiences to new forms of video, e.g. mobile television and Internet downloading and streaming video. Nielsen's Anytime Anywhere Media Measurement initiative was recently profiled in the New York Times.2 BBM describes its new PPM on its website: "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." Audiences to mobile television and streaming and downloading video on the Internet could potentially be measured with the PPM. CMRI would be pleased to explore these issues with the individual broadcasters or the Commission.
After considering various alternatives dealing with the measurement of audience success, CMRI recommends that the viewing hours of each network be weighted by a special factor, in some ways similar to the idea expressed by Corus but audience based and empirical in nature. The table below shows how the Hit Factor would be calculated.
| Average Minute Audience of Funded and Non-Funded Canadian Drama/Comedy Series | ||||||
| All Persons 2-plus, 2003-04 to 2006-07 | ||||||
|
Year |
||||||
|
2003-04 |
2004-05 |
2005-06 |
2006-07 |
Total |
2006-07 |
|
|
AMA |
AMA |
AMA |
AMA |
AMA |
Hit Factor* | |
| A Channel |
8 |
7 |
11 |
8 |
8 |
0.1 |
| BRAVO! |
13 |
15 |
15 |
13 |
14 |
0.3 |
| CBC |
125 |
156 |
96 |
105 |
119 |
2.1 |
| CH |
25 |
19 |
12 |
28 |
19 |
0.6 |
| CITY |
22 |
29 |
20 |
41 |
26 |
0.8 |
| CMT |
34 |
28 |
32 |
20 |
28 |
0.4 |
| Comedy |
22 |
33 |
31 |
36 |
31 |
0.7 |
| CTV |
415 |
503 |
411 |
200 |
322 |
3.9 |
| Family |
51 |
54 |
63 |
84 |
65 |
1.7 |
| Global |
147 |
235 |
201 |
155 |
176 |
3.1 |
| History |
51 |
31 |
38 |
54 |
41 |
1.1 |
| Showcase |
39 |
43 |
35 |
35 |
38 |
0.7 |
| Space |
39 |
52 |
50 |
36 |
44 |
0.7 |
| The Movie Network |
15 |
14 |
14 |
0.3 |
||
| Tvtropolis |
19 |
20 |
16 |
17 |
0.3 |
|
| Vision |
23 |
18 |
13 |
9 |
15 |
0.2 |
| W Network |
25 |
33 |
31 |
26 |
29 |
0.5 |
| YTV |
35 |
22 |
53 |
33 |
33 |
0.7 |
| Average |
67 |
76 |
64 |
51 |
58 |
1.0 |
| Source: CMRI | ||||||
| *Ratio of a channel's AMA/Average AMA | ||||||
The Hit Factor is simply a network's average audience divided by the average audience of all networks. In the case of Global the calculation would be 155,000 viewers/51,000 viewers = 3.1 in 2006-07. This factor would then be applied to Global's total viewing hours, that is, Global's hours would be multiplied by the new factor. If a network's factor is less than 1.0, it would not be used to penalize a network, that is, it would only be used if the factor was greater than 1.0. This respects the specialty channel business model of repeating even original series very frequently, which lowers the average audience. Similarly, so as to not over-compensate conventional broadcasters, who air mostly original programs and repeat them usually only once in a television season, the maximum value of the Hit Factor could be set to 2.0. The new factor would reward networks that attract mass or large average minute audiences and increase the total viewing hours of those networks; CTV, Global and CBC would all have a larger share of total viewing hours as a result. Two specialty channels, History and Family Channel, also had positive Hit Factors in the 2006-07 television season.3
CTF currently calculates the total viewing hours and then percentages them and divides the funding according to the percentages. If viewing hours are weighted or multiplied by the Hit Factor, it changes the percentages substantially. For example, in the 2006-07 television season CTV's percentage of viewing hours increases from 20 to 30% and Global increases from 3.6 to 5.4%. Showcase drops from 19.1 to 14.4%, etc., as shown in the table below. Global would see further increases in 2007-08, given the success of some of its new Canadian dramas.
| Viewing Hours of Funded and Non-Funded Drama/Comedy Series | |||||
| All Persons 2-plus, 2003-04 to 2006-07 | |||||
|
Current CTF Calculation |
Calculation Using Hit Factor (Max. 2.0) |
||||
|
2006-07 |
2006-07 |
2006-07 |
2006-07 |
||
|
Viewing Hours |
Viewing Hours % |
Viewing Hours |
Viewing Hours % |
Hit Factor* | |
| A Channel |
18,136 |
2.8 |
18,136 |
2.1 |
0.1 |
| BRAVO! |
11,975 |
1.8 |
11,975 |
1.4 |
0.3 |
| CBC |
28,621 |
4.4 |
57,243 |
6.6 |
2.0 |
| CH |
15,558 |
2.4 |
15,558 |
1.8 |
0.6 |
| CITY |
14,710 |
2.2 |
14,710 |
1.7 |
0.8 |
| CMT |
1,215 |
0.2 |
1,215 |
0.1 |
0.4 |
| Comedy |
37,291 |
5.7 |
37,291 |
4.3 |
0.7 |
| CTV |
131,000 |
20.0 |
261,999 |
30.0 |
2.0 |
| Family |
48,737 |
7.4 |
80,446 |
9.2 |
1.7 |
| Global |
23,487 |
3.6 |
46,974 |
5.4 |
2.0 |
| History |
17,164 |
2.6 |
18,176 |
2.1 |
1.1 |
| Showcase |
125,644 |
19.1 |
125,644 |
14.4 |
0.7 |
| Space |
97,779 |
14.9 |
97,779 |
11.2 |
0.7 |
| The Movie Network |
3,225 |
0.5 |
3,225 |
0.4 |
0.3 |
| Tvtropolis |
28,283 |
4.3 |
28,283 |
3.2 |
0.3 |
| Vision |
9,515 |
1.4 |
9,515 |
1.1 |
0.2 |
| W Network |
30,071 |
4.6 |
30,071 |
3.4 |
0.5 |
| YTV |
14,011 |
2.1 |
14,011 |
1.6 |
0.7 |
| Total |
656,421 |
100.0 |
872,249 |
100.0 |
|
| Source: CMRI | |||||
| *Ratio of a channel's AMA / Average AMA | |||||
The Hit Factor is an amalgam of many different elements that contribute to audience success, all of which were discussed at the February hearing. A network's Hit Factor would be higher if it airs more original programming, a particular concern of CBC and Canwest; it would be higher if a network promotes Canadian programming; it will naturally be higher if programs are aired in prime time and prime viewing seasons of the year, an issue raised by Rogers. It would encourage networks, conventional and specialty, to commission more original programming and to schedule and develop Canadian programming aggressively. Specialty channels that lost some ground in the future might be encouraged to air fewer repeats and to promote and schedule Canadian programs more effectively to increase average audiences. In short, the Hit Factor incorporates many of the proposals and ideas made to the CRTC by various parties and reflects true 'audience success' and is a substantial response to Shaw's basic concerns about the CTF.
CTF should also consider, as did the analysis in this paper, calculating audience success based on all Canadian programs within CTF genres, not just CTF-funded programs. This would recognize the total environment in which Canadian programs generate audience, a concern of Canwest, CAB, Rogers and others. In response to Astral, Corus and others, CTF should also consider calculating viewing hours, the average audience and the Hit Factor within the cable and DTH universe rather than among all households, which would put all networks on a level playing field, that is, audience success would be determined within the household universe that all conventional, pay and specialty channels have at least the opportunity to reach.
My thanks to Kathy Corcoran and Richard Koo at CTF for their patience and insights.
Appendix:
French Television Tables and Charts
| Viewing Hours of Funded and Non-Funded Drama/Comedy Series | |||||
| All Persons 2-plus, 2003-04 to 2006-07 | |||||
|
Year |
|||||
|
2003-04 |
2004-05 |
2005-06 |
2006-07 |
Total |
|
|
Viewing Hours % |
Viewing Hours % |
Viewing Hours % |
Viewing Hours % |
Viewing Hours % |
|
| Historia |
0.7 |
0.6 |
0.3 |
0.5 |
0.5 |
| Radio-Canada |
37.2 |
38.0 |
32.1 |
29.7 |
34.0 |
| Séries+ |
15.9 |
14.2 |
15.3 |
11.0 |
13.9 |
| Super Écran |
0.0 |
0.0 |
0.0 |
0.1 |
0.0 |
| Télé-Québec |
10.1 |
8.9 |
7.6 |
8.5 |
8.8 |
| TQS |
4.1 |
6.0 |
4.8 |
4.9 |
4.9 |
| TVA |
26.1 |
24.9 |
31.4 |
26.8 |
27.3 |
| VRAK |
0.0 |
0.0 |
0.0 |
12.2 |
3.6 |
| Ztélé |
6.0 |
7.4 |
8.5 |
6.3 |
7.0 |
| Total |
100.0 |
100.0 |
100.0 |
100.0 |
100.0 |
|
|||||
| Program Hours of Funded and Non-Funded Canadian Drama/Comedy Series | |||||
| 2003-04 to 2006-07 | |||||
|
Year |
|||||
|
2003-04 |
2004-05 |
2005-06 |
2006-07 |
Total |
|
|
Program Hours |
Program Hours |
Program Hours |
Program Hours |
Program Hours |
|
| Historia |
497.0 |
615.0 |
277.0 |
353.0 |
1,742.0 |
| Radio-Canada |
734.2 |
479.8 |
538.9 |
692.2 |
2,445.0 |
| Séries+ |
3,244.0 |
2,810.0 |
3,269.0 |
3,331.0 |
12,654.0 |
| Super Écran |
75.3 |
75.3 |
|||
| Télé-Québec |
714.5 |
656.7 |
614.8 |
667.8 |
2,653.9 |
| TQS |
68.0 |
138.5 |
83.0 |
130.6 |
420.1 |
| TVA |
237.5 |
159.0 |
198.0 |
197.0 |
791.5 |
| VRAK |
1,554.5 |
1,554.5 |
|||
| Ztélé |
2,626.0 |
2,923.0 |
2,951.5 |
2,759.0 |
11,259.5 |
| Total |
8,121.2 |
7,782.0 |
7,932.2 |
9,760.3 |
33,595.7 |
|
Source: CMRI |
|||||
| Average Minute Audience of Funded and Non-Funded Drama/Comedy Series | ||||||
| All Persons 2-plus, 2003-04 to 2006-07 | ||||||
|
Year |
||||||
|
2003-04 |
2004-05 |
2005-06 |
2006-07 |
Total |
2006-07 |
|
|
AMA |
AMA |
AMA |
AMA |
AMA |
Hit Factor* | |
| Historia |
8 |
5 |
5 |
10 |
7 |
0.1 |
| Radio-Canada |
311 |
445 |
346 |
313 |
346 |
1.6 |
| Séries+ |
30 |
28 |
27 |
24 |
27 |
0.1 |
| Super Écran |
14 |
14 |
0.1 |
|||
| Télé-Québec |
87 |
76 |
72 |
92 |
82 |
0.5 |
| TQS |
368 |
242 |
339 |
276 |
292 |
1.4 |
| TVA |
673 |
880 |
921 |
995 |
857 |
5.0 |
| VRAK |
57 |
57 |
0.3 |
|||
| Ztélé |
14 |
14 |
17 |
17 |
15 |
0.1 |
| Average |
213 |
242 |
247 |
200 |
189 |
1.0 |
|
Source: CMRI |
||||||
|
*Ratio of a channel's AMA/Average AMA |
||||||
| Viewing Hours of Funded and Non-Funded Drama/Comedy Series | |||||
| All Persons 2-plus, 2003-04 to 2006-07 | |||||
|
Current CTF Calculation |
Calculation Using Hit Factor |
||||
|
2006- |
2006-07 |
2006 |
2006-07 |
2006-07 | |
|
Viewing Hours |
Viewing Hours % |
Viewing Hours |
Viewing Hours % |
Hit Factor* | |
| Historia |
3,631 |
0.5 |
3,631 |
0.2 |
0.1 |
| Radio-Canada |
216,890 |
29.7 |
340,140 |
20.7 |
1.6 |
| Séries+ |
80,364 |
11.0 |
80,364 |
4.9 |
0.1 |
| Super Écran |
1,055 |
0.1 |
1,055 |
0.1 |
0.1 |
| Télé-Québec |
61,765 |
8.5 |
61,765 |
3.8 |
0.5 |
| TQS |
35,982 |
4.9 |
49,615 |
3.0 |
1.4 |
| TVA |
195,967 |
26.8 |
975,634 |
59.2 |
5.0 |
| VRAK |
88,896 |
12.2 |
88,896 |
5.4 |
0.3 |
| Ztélé |
45,640 |
6.3 |
45,640 |
2.8 |
0.1 |
| Total |
730,190 |
100.0 |
1,646,739 |
100.0 |
1.0 |
|
Source: CMRI |
|||||
|
*Ratio of a channel's AMA/Average AMA |
|||||
Notes:
1In the appendix, the reader will find a series of other tables and charts related to trends in French TV audiences between 2003-04 and 2006-07.
Louise Story, "Nielsen Looks Beyond TV, and Hits Roadblocks", New York Times, February 28, 2008 http://www.nytimes.com/2008/02/26/business/media/26nielsen.html?_r=2&pagewanted=1&sq=a2%20m2&st=cse&scp=1&oref=slogin
CMRI is also exploring other ways to calculate the average audience to reflect the different scheduling strategies of specialty channels and conventional networks.
| Date Modified: 2008-06-05 |
