How Audience Measurement Approaches Construct Different Audience Imaginaries

The final presenter in this ECREA 2016 session is Jakob Bjur, whose interest is in the media measurement of media work. There is now plenty of work on audience measurement systems, and also a growing wave of criticism of these systems: such systems are viewed as capturing audience labour, but with very one-dimensional metrics that generate measurement currencies that are very far removed from actual audiencing practices.

These systems have generated market-wide conventions for benchmarking media performance and trading audiences between media organisations and other stakeholders (such as advertisers) – which also means that more complex audiences that cannot easily be reduced to simple numbers may not be catered for.

There are similarities here to studies of scientific laboratories: here, scientific theories are being produced by an assemblage of technologies and processes including a variety of inscription devices that generate standardised measurement outputs, which in turn become the source data for scientific publications. The audience 'ratings machine' works in much the same way: it generates the data, the graphs, and the machine then recedes into the background and is taken for granted, without being challenged any further. Audience measurement creates numbers, but the meaning of these numbers is being created elsewhere.

Audience measurements, then, could be used in a very different fashion, could be read in very different ways. This would serve to unpack the black box of audience measurements, and remove the taken-for-grantedness of such measurements in order to open them up for renegotiation. When this happens, one image of the audience will be replaced by another.

To measure a media consumption act, the audience's engagement with the content space within a specific consumption situation needs to be analysed. This could be done at different sample scales, using self-assessment of technical measurement, and using active (audience-involvement) or passive (observatory) measurement. Changes in such measurement could generate a new visibility for different media consumption acts – this could move from occasional snapshots to continuous minute-by-minute measurements.

In radio, this leads to the identification of a new type of listening, and a new type of listener: apparently stable and slowly changing listener bases are revealed as fast and fluctuating; this may give the impression of listener attention spans that are shorter than they used to be.

Changes in measurement also lead to different programme design: for instance, the move in Sweden from telephone-based consumption surveys to automated station identification by the measurement devices means that it no longer is important that audiences recognise the stations' call signs, so station jingles disappear from the radio altogether. In other words, broadcast content is redesigned to be reliably recognisable by the measurement boxes, rather than by actual humans.

All of this appears to point to one crucial realisation, Jakob suggests: what we measure today, and how, affects what is left to measure tomorrow.