It was a big day for the BRC (Broadcast Research Council of South Africa): The presentation of the BRC TAMS and Establishment Survey/SEM Update.
Clare O’Neil, CEO of the BRC, reminded us that the Establishment Survey (ES) provides the marketing and media industries with a reliable picture of the South African population. Based on a sample of 25 082 respondents, it is representative of 39.5 million adults. The weighting efficiencies are reassuring: 93.8% for households and 84.7% for individuals.
It is the place to go in order to establish the facts about our local adult (15+) market. It gives us single source (IHS based) information about consumer demographics and geographic distribution, as well as their broad media consumption. Usefully there is a battery of common questions for all media e.g. number of days used, time spent and device used, which provide the basis for fair intermedia comparison. There are additional specific media questions as well as lifestyle and product questions.
People unfamiliar with the concept of an establishment survey have battled with the fact that it does not contain branded information. O’Neil was able to reassure those people that the ever responsive BRC has noted and addressed this issue: Working with representatives from the media agency world it has developed a brand questionnaire that goes into field in January 2018. The results will be available in the second half of the year.
The picture of the South African population that O’Neil presented was familiar to those who attended the overview of the six months’ data, and clearly affirmed that South Africans are devotees of broadcast media.
Listenership and viewership
The internet falls into third place behind TV and radio, with its reach of 50% of the population over the past week. 95% of people who have accessed the internet in the past four weeks, do so via their cellphones and 43% use these devices to listen to radio. Indeed, radio listeners avail themselves of a number of devices to keep in touch with their medium. Besides the high incidence of listenership on traditional radio sets/hi-fis and on vehicle radios, 24% of people who have listened in the past month, have done so on a TV set.
With the full 12 months ES data available, the BRC TAM (Television Audience Measurement) panel had also been updated to reflect the new HIS population. Updates to the panel are done twice a year, and will now be done from the ES. The scope of the TAM panel is wider in terms of age, covering everyone in the TV universe over the age of four – some 45 million individuals.
O’Neil showed that the subscriber TV universe now numbers 15.4 million adults while the free to air is 18.4 million. The ES data shows DStv has a 46% penetration of the TV universe, while OVHD has reached the milestone of 1%. Combined Netflix, Showmax and Apple TV reach 0.05%. The ES, with its twice year release, will always show a bit of a lag, so the BRC is tracking the incidence of smart TVs on the panel in order to keep an eye on the growth of internet platforms. A live panel can, of course, provide quicker alerts to such changes.
After the overview of the TV landscape, O’Neil shared the standard ‘vital signs’, which prove that the panel is in good health: It covers 94% of TVs, is polling at 97.1% and has a 93% individual weighting efficiency. There are 2 935 installed meters, and panel expansion continues, with another 200 meters being installed. She demonstrated that the panel is balanced by province and area, and reiterated that, because polling on any day will not collect data from the entire panel, RIM weightings are used to ensure that the reported picture is correct in terms of the overall population.
SEMs and LSMs
Another useful cross check on the reliability of the universe update was the comparison of the previous and new overall TV and individual station average rating data over the broadcast day. The correlations were comfortingly close.
The big news of the day was that the ES SEM (Socio Economic Measures) system has now been applied to the TAM panel. Abandoning diplomacy, O’Neil was forthright in declaring that LSMs were “rubbish”. She told a memorable anecdote that convincingly captured the shortcomings of these measures. As part of a field trip, Robert Ruud, the independent and international auditor of the TAM panel, visited one of the panel homes in the Cape Flats. Coming from Scandinavia, where the Gini Coefficient is enviably low, he was struck by the abject poverty of the home, with its boarded up windows. To him, it seemed like an example of an LSM 2 or LSM 3 household. Back at the Nielsen office, he checked on the home’s classification and found it was an LSM 8. Further scrutiny proved that this was no mistake: The home contained the necessary durables to qualify for the classification. That they were antiquated was no disqualification.
While we are all rightly wary of “sample of one” cases being cited to prove research issues, the LSM have long been debated and criticised. The SEM system offers a considered, relevant differentiating and thoroughly tested way forward, that does not overly rely on durables. The SEM profiles of the station, that were presented looked quite in line with what one might have anticipated. Although the SEMs will not be used as RIM weighting variables, users of the TAMs data should not see the kind of fluctuations that they did in audience data based on LSMs.
However, the LSM measures have not been removed, to accommodate the tracking of buying commitments that are in place. Given the problems with LSMs, it would be sensible for TV planners to get started working with SEMs as soon as possible, even if it is, initially, on a parallel basis to build familiarity.
Telmar system in place
In order to help the industry users become accustomed to the new measures, TNS has developed a set of visual representations which supplement the concise graphs summarising the demographics of each of these groups. These photographs are effective aids to develop a better sense of how each group lives. Nielsen provided another useful tool for users, a TV dashboard with station reach. O’Neil also announced, hot off the press, that Telmar had the system in place so that users can now multi-base the TAM and ES data.
I left the presentation with a sense, that with Herculean effort from the JIC, research and software suppliers, the world of South African audience measurement has been moved significantly forward. It is time for us, the users of the data to commit ourselves to getting the best we can out of it.