The Broadcast Research Council of South Africa (BRC) has released the full first year of Radio Audience Measurement (RAM) data. Presentations in both Johannesburg and Cape Town once again reaffirmed the stability of the data, and provided assurance that the survey’s “vital signs” are in good health.
Most media agencies will now start working seriously with the data. (The prevailing decision among these agencies had been to proceed cautiously and allow the RAM survey to run for a full year before adopting it for planning and negotiating purposes). It is, therefore, a particularly apt time to pass on some tips to those strategists, planners and analysts who will be tackling the data for the first time.
Do take a few moments to orientate yourself to the new South African landscape, which RAM presents. If you have been working in the media planning and buying business for some years, you will probably have developed a feel for the numbers. This is going to have to be recalibrated. What is particularly disconcerting is that on one level things do look familiar. The overall population numbers are similar to what you are accustomed to seeing. So too, are some of the profiles of some of the basic demographics e.g. race and province. This is because both the new RAM and the old RAMS use the same source of population data, and these demographics are used to weight the data.
It’s not the same
Don’t be lulled into thinking that everything is the same, though. The new sample frame, which is an accurate reflection of our population, has correctly delivered a picture of a South Africa, which is poorer than we were accustomed to seeing. For example, we were used to seeing roughly a fifth of adults assigned to the R20 000+ household income group; the new data shows us that just under 10% are in that comfortable position.
While we were used to seeing a third of the population falling into the under R5 000 household income group, we now see that 43% are in this precarious position. We may recall that the fortunate LSM 8-10 group constituted nearly a quarter of the population in the previous survey; the new reality suggests that it is less than a fifth. LSM 5-7 accounted for roughly half the population; the new survey shows us it is around 43%.
On the subject of household income, do note that respondents may refuse to answer the income question – either via a flat out refusal or by declaring that they do not know. In previous radio surveys, it was left to the interviewer to assign such reticent respondents to the income group the interviewer thought appropriate.
The merits of this might be debatable, but it did ensure that everyone was neatly classified. We can now see that a quarter of South Africans are not prepared to disclose this information. Fortunately, Kantar TNS has provided a modelled solution to this. If you are using income definitions in your target markets, and therefore want everyone in the population group assigned to an income bracket, make sure you use the modelled summary household income groups.
There are a number of other new demographic descriptors. For example, areas are classified on the basis of Stats SA Census 2011 municipal codes and geography types, into three categories: Metro, Small Urban and Rural. My suggestion is that you go to the BRC website, where you will find a full list of which areas are assigned to which categories. The website also has handy FAQs and an easily digestible Technical Report.
- Do remember that once the new Establishment Survey (ES) is released in a month’s time, this picture of the population will be represented in it, because both surveys are based on a similar sampling frame. This will make it easier for strategists and planners to contextualise radio in the overall South African media mix. The ES will confirm radio’s broad reach and ubiquity.
- Do be aware that the currency data is released in six-month rolling tranches: currently Jan – Jun 2016, April – September 2016, and October to December 2016 are all available for trending. These releases have samples of around 33 000, and should form the basis for planning and buying.
- Do remember that there is also a full Jan – Dec report, which is released annually. With its significantly boosted sample of around 69 000, this is the ideal place for deep dives into the data e.g. into the minutiae of cross device and location usage by individual station. This consolidated annual report also permits community stations which are able to build their samples over the full year. Provided a station has a minimum of 40 respondents, the data is released. So you will find the most robust picture of community stations here.
- Do spend some time examining the performance of individual stations, before plunging into planning as usual. The survey design, sample frame and methodology of BRC RAM are different from what was used previously. The results are different.
- Do spare yourself the inevitable confusion that will result from running a post-campaign analysis on the BRC data, when the campaign was planned on SAARF data.
- Don’t demand compensation for ‘lost listeners’ from broadcasters – unless, of course, you are prepared to pay retroactively for all the newly found listeners that many stations have. Broadcasters, don’t cave into any such misinformed bullying.
- Do be sensible: both buyers and sellers across all media types need to get re-orientated in terms of the new picture of the population and the relative values of both media types and individual media vehicles. The ES will help with the former. The specific media currency surveys provide the information for the latter. The first print currency will only be released later in the year.
- Don’t panic about radio inflation simply because cumulative weekly or daily reach on some stations is down. The BRC RAM data shows a picture of loyal and heavy listenership overall. This means that in many instances, the average quarter hour audiences are higher than we were accustomed to seeing. Media inflation is based on both the rate paid for spots and the audiences delivered by those spots. The irony is that because the overall average quarter hour listenership has risen by almost a quarter, the overall media inflation calculation would be deflationary for the radio sector IF direct survey comparisons could be made.)
- Do remember that the new survey necessitates the “restarting of the clock” in terms of tracking both audience and rate performance. Whilst the necessary audience data accumulates, the simple way to keep an eye on inflation is to simply track rate inflation. Media Inflation watch already uses this approach for TV (due to panel changes), cinema and outdoor.
- Do remember that you need to be careful about setting your reach and frequency objectives. Spend some time familiarising yourself to how the new stations perform against your target markets so that you can set sensible reach objectives.
- Do not be tempted to cut back on frequency simply because it looks higher on the new data. Because the average quarter hour audiences are higher, it is likely to be so. These measurements are not the same as the previous. One needs to straddle the transition. An analogy might be that a set budget buys you a bag of oranges. The weight of that bag differs if you use the measure in grammes or ounces. You need to understand the conversion factor.
Finally, do allow yourself the time to make this conversion. Do enjoy the exploration of our new radio landscape.