The TAMS panel change, using audited figures supplied by DStv for the DStv penetration on the TAMS panel will be implemented from Monday, 10 December 2012.
SAARF TAMS (Television and Audience Measurement Survey) will use the DStv audited figures for weighting the DStv penetration, PVR penetration within DStv and the bouquet percentages within the DStv universe on the TAMS panel. Audited DStv updated numbers will be included in all future universe updates. Background
When SAARF first started measuring DStv viewership on the TAMS panel about 11 years ago, DStv penetration was growing at a rapid rate. In fact, it was growing so fast that the establishment survey AMPS (All Media and Products Survey) lagged behind with the measurement of the penetration.
In those days, MultiChoice supplied SAARF with audited DStv subscriber data to be used in the weighting.
(The standard audit is based effectively on a customer count, which is not the same metric as households. In order to get to a "proxy" for households a number of filters are applied to the subscriber numbers). Over time, the lag became less and in October 2008, the TAMS Council decided to use AMPS figures rather than DStv audited figures. Current reality
Subsequently DStv has launched a variety of new bouquets and the subsequent growth continued. As part of the TAMS audit task team's investigations a decision has been reached to use DStv audited figures once again alongside the AMPS data, in order to get the most accurate reflection of the current reality.
AGBNielsen have run some test data where it used the current AMPS 2011B 6-month data (data that is currently being used for all industry releases) and superimposed the proportions of tentative DStv audited figures on the data.
From these preliminary results, it looks like an average gain in TVRs of 3-4% on the free-to-air stations and a loss of about 12-16% on DStv commercial channels.
The board task team has reviewed these results and believes they are a more accurate reflection of television viewing. SAARF is working to get the results released to the industry by 10 December 2012. Once more detailed results are available, a more in-depth article will be written about the impact on the data.