Melinda Chapman | Student Id: 90426817
Most of us have heard of the TV rating boxes (people meters) that some people have in their homes to capture and report who in their household is watching what and when. It’s a long shot, but you may even belong to one of the 6000 households across Australia providing TV programmers and advertisers with the much anticipated ratings figures that reflect the viewing choices of Australia and which largely determine where programming and advertising dollars are spent. Interestingly, if you do belong to one of these households, you won’t tell us because you are not allowed to. A key participation criteria is anonymity. As outlined in a recent news.com.au article by Debbie Chipp (Feb, 2016), households participating are asked not to disclose their identity so they cannot be unduly influenced or targeted by those interested in achieving a particular rating result. Ensuring market research data is valid is integral to it being relied upon for the purposes of, in this case, spending billions of advertising dollars each year.
Can I opt to be involved?
Unlike other types of market research where you can effectively choose to respond to a survey or participate in an interview, ‘self selection’ and thus self selection bias is mostly ruled out in the world of Television Audience Measurement (TAM). If you call up to get a meter installed, the answer will be a definite no. However once your household has been selected and the system installed, you effectively have the ultimate choice of ‘logging in’ to have your viewing included in the research or not. But as with all research, it is impossible to fully rule out research limitations when humans are involved.
‘So how does OzTam, the organisation which undertakes this predominantly quantitative research, select the households and is the data reliable? According to OzTam CEO, Doug Peiffer (as quoted in the aforementioned article), selection factors which include “demographics, census data, and other information collected by a wide-reaching establishment survey, make the final measure both acceptable and statistically representative…… it’s one of the largest samples — in terms of panel size relative to overall population — in the world.”
The extent to which the TV industry regards the TV ratings data as a reliable and valid representation of Australian TV views is demonstrated by the ratings being the key currency of measurement affecting where billions of TV advertising dollars are spent and influencing which programs are screened when. However, with the recent monumental flop of Channel 9’s Reno Rumble, against Channel 7’s My Kitchen Rules (and particularly following Reno Rumble’s lacklustre ratings the previous season), some in ‘TV world’ clearly still rely on other means to determine what gets screened and when.
A bit like scanner data
In the same way scanner data captures what you are actually buying (Coke) as opposed to what you may say you buy to indicate a ‘healthier you’ (Water), people meters aim to accurately capture what people are actually viewing (The Bachelor) as distinct to what they may say they watch if asked in a survey or questionnaire (Four Corners). To this extent, it largely rules out social desirability bias where TV shows (eg news and documentaries) that may appear more ‘socially acceptable’ or ‘socially desirable’ may be over-reported in surveys and vice versa
Traditional TV audiences dwindling
Traditional broadcast TV audiences are dropping as the market becomes even more complex and fragmented with ‘catch -up’ TV and subscription video on demand services such as Netflix. Just 15 years ago, OzTam’s job was simply to measure viewing across 5 analogue free to air channels. This has expanded to “more than one hundred digital free-to-air and subscription channels 24/7/365 – both live and time-shifted up to 28 days from original broadcast – across dozens of demographics.” (OzTam) Earlier this year OzTam released a new tool, Video Player Management to report on the content playing on connected devices (tablets, smartphones etc).
Recently quoted in the Sydney Morning Herald (18/04/2016) Craig Johnson, Nielsen’s head of audience measurement, Asia Pacific, regards any concern about dwindling TV audiences as ‘device centric’ “Programmers and advertisers have never cared whether their video content or ads are watched on a 22-inch screen or a 122-inch screen, so long as someone is watching”.
But are the 6000 households actually watching TV?
Current Nielson research states 3 in 4 Australians browse the internet while watching TV. So are the viewers that the TV industry rely so heavily upon for their decision-making even engaged in what they are watching or is it simply background noise?
Considering my own patterns of media viewing, an obvious limitation of all this data is its lack of ability to capture the ‘level of viewer engagement’. Without viewer engagement, and in an age where we are not only surfing TV channels but also flicking between devices and a myriad of online content, relying on ratings to determine exactly where to spend advertising dollars becomes questionable. To assess a viewer’s emotional engagement would be a significant leap into qualitative research requiring more resources. However a starting point, as noted by Paul Whybrow in AdNews last month would be to assess metrics such as a viewer’s social media activity around the show, who they connect with about the show, their buying habits pre- and post-watching and what they do afterwards.