The Broadcast Research Council is about to release its first Radio Audience Measurement data. Already, the picture – particularly around race – is VERY different…
Earlier this month, I wrote enthusiastically about my first glimpse of the new RAM (Radio Audience Measurement) data, shared by the Broadcast Research Council of South Africa (BRC) and TNS with the Advertising Media Forum. Excited by this brand new approach to monitoring radio listening habits, I decided to visit Clare O’Neil, CEO of the BRC, to ask her to tell me more about what we should expect to see from the new study.
With her customary passion, O’Neil started by pointing out that the objective of this new approach was nothing less than the transformation of the audience measurement system. The word ‘transformation’ deserves some reflection, if we are to understand the implications of this new research. The Collins English Dictionary defines it as, “a change or alteration, especially a radical one”. Indeed, radical difference is what the industry must prepare itself for, as all comparisons with the data of the past will be irrelevant. Of course, in our South Africa, ‘transformation’ has greater implications: it is about the meaningful change we need to make to successfully build our country and its people.
The first step in delivering appropriately transformed data was developing a new sampling frame. The sampling frame lists and describes the units from which the sample for a survey will be drawn. As such, it is the absolute bedrock of trustworthy and reliable research. Working with population data supplied by specialist demographers IHS, TNS developed a new sample frame for this survey. Stratification by province and area type ensures that all racial groups are represented.
O’Neil proudly explained to me that 22 years after the end of apartheid, the BRC RAM sample frame truly reflects the South African population.
Of course, while we want to be future facing, we do have to cast a look backwards if we are to understand the sort of changes we will find in the new data. The fact is that the previous South African Audience Research Foundation (SAARF) RAMS sample looked very different and did not correspond particularly closely with the realities of our country.
There are arguments for disproportionate sampling – often most compellingly based on cost. For example, rural interviews are more expensive and arduous to undertake, so a decision might be taken to under-sample in rural areas. After all, we have become accustomed to hearing there is always the panacea of applying weightings to realigns the distorted picture. However, when weighting efficiency is low, the error range in the data is significantly increased.
Poor weighting efficiency is familiar territory for O’Neil, and an issue which she determinedly solved. When the BRC took over the task of restoring the Television Audience Measurement (TAM) panel to health, the audit showed weighting efficiency had fallen to an abysmal 39%. Working with Nielsen, and the BRC research committee, O’Neil oversaw the overhauling of the panel, which now runs well ahead of the global gold standard of 70%. There were numerous other remedies applied to improve the TAM panel, but weighting was one of the significant benchmarks the international auditor, Robert Ruud, examined before stating in his 2015 report that South Africa now has a “high quality well-functioning TAM panel”.
In addition to the new BRC sample frame, probability proportionate to size (PPS) sampling is applied to ensure a correct and random selection of starting points per area type and province. These changes will give the media industry a new and more realistic view of the structure of the South African population. Looking at the familiar metric of LSMs, we should expect to see a smaller proportion of the population falling into the upper end and a wider spread across the middle sectors.
These changes will be the equivalent of the media industry having the ‘scales removed from our eyes’. With our new sight, we should expect the world to look different. It was too soon for O’Neil to share a preview of individual station audiences as TNS is carrying out a reassuringly thorough validation process before revealing the figures. However, to me, logic would suggest that this revised and real population structure will produce solid listenerships for stations with mid-LSM audiences. Conversely, stations appealing to the higher end of the LSM spectrum should see ‘smaller’ audiences.
O’Neil also reminded me that another transformation objective of the new RAM was to ensure that the dynamism of radio listening could be captured and analysed for the first time. The new RAM diary has been designed to explicitly capture cross-device and cross-location listenership. Respondents are not merely reminded that it does not matter how they listened; they now have to actually fill in where and on what they were listening to the radio.
Before sharing the results of the cross-device and location questions, O’Neil stressed that we were only looking at the first wave of the study comprising data from some 7 500 households, collected between 7 and 31 January.
Two waves of data
The first data, comprising the first two waves, will be released on 25 August. The field work for the full study, based on a 30 000 sample, will only be completed in December this year. The industry will, therefore, only have the complete data set in the first quarter of 2017.
With this caveat in mind, it seems that the first results of cross-device listening look eminently sensible: 75% of the seven-day audience claim to listen on a radio, while 40% listen on a cellphone – testament to the wide distribution of affordable FM-enabled hand-sets. In fact, some 8% now listen exclusively on a cellphone. 28% claim to listen on a vehicle radio and 9% to listen on the television. The results of the location questions also seemed understandable. The bulk of listeners (87%) still tune in at home, but over a third (35%) listen ‘on the go’. 12% listen at either their place of work or at university
The detailed device and location prompts in the diary should help respondents to more accurately recall and record their listening experiences. This, in turn, could lead us to see more listening than we we were accustomed to recording in the past. Indeed, O Neil pointed out that this first tranche of data had shown that people were listening for longer periods than had been recorded in the past: roughly half the radio audience are ‘heavy listeners’.
This heavier and longer listenership recorded is also expressed in a different daily curve from what we are accustomed to seeing. As I mentioned in my previous column, the daily curve is now a “sleeping lion” shape, as Jennie Beck – director of media at TNS Global and Kantar Media, memorably described it. Overall quarter hour listenership is higher, especially in the mornings, and is more evenly spread across the day. Reassuringly, this shape of listenership is similar to that seen internationally.
With mobile listenership allowing people to make their personalised choices, rather than having to accommodate someone else’s listening preference, it struck me we could see listeners engaging with their chosen stations for longer periods.
O’Neil also showed me the results of overall listenership by province. This data showed uniformly healthy listenership across all nine provinces. There were, however, some interesting differences which pointed to the sensitivity of the study. A good example was the shape of the daily curve in Limpopo, which showed a pronounced evening peak – understandable in a more rural and poorer province. Limpopo was also the province with the highest proportion of listeners tuned in on their mobile devices, again a phenomenon that seems intuitively correct.
After looking at this data, O’Neil and I talked about the likely impact of the of the new study. This early data suggests to me that the new RAM should occasion the reassessment of the value of the medium of radio. The deeply loyal and habitual nature of the relationship that listeners have with the medium will become clearer. The higher average quarter hour ratings for the medium underscore this. Better knowledge of the listener’s commitment to the medium will help planners appreciate the value of the radio audience better.
These metrics should strengthen radio’s place in the overall media mix, as media planners and clients understand the depth of this relationship.
Radio will benefit
It is my impression that radio will benefit hugely from being in the limelight. Recent chats I had with a few media planners led me to believe that the medium receives less attention than it deserves, perhaps because there has been so little new to be said about it in recent years.
Currently the default radio plan is drive-time dominated, as that is when planners and clients will hear their spots. The new radio curve shows that radio is relevant through the day. Broadcasters will have the ammunition to more effectively sell their inventory. Having higher quarter hour listenerships, and longer listening hours means that the broadcasters have more contact opportunities to offer advertisers.
With better understanding of the way in which people are listening, there will be new ways to for them to package inventory based on device and location listening habits. The development of “on the go” packages, perhaps with an outdoor component, seems an obvious example.
Certainly, the broad picture looks very positive for the medium. The challenge is that we are walking into completely uncharted territory. Few of us in the media industry have had any experience of dealing with a completely new currency. Some of us remember transitioning from TV diaries to the first TAM measurement, but the media landscape was simpler then. The challenge for us now is to adjust to a new and more truthful picture of our environment, without constantly comparing it with what we thought we knew to be true.
Britta Reid is an independent media consultant.
Want to continue this conversation on The Media Online platforms? Comment on Twitter @MediaTMO or on our Facebook page. Send us your suggestions, comments, contributions or tip-offs via e-mail to firstname.lastname@example.org.