The FT’s Creative Business section ran a piece today on the radio audience measurement company Rajar. Rajar has come under fire over the validity of its figures and its methodology. Ultimately, any type of measurement is inherently flawed and skewed. It requires people wanting to be measured and that requires a certain type of person that is arguably not representative of the population as a whole. They may well match up on the cosmetic weighting variables (e.g. age, gender, income) but the subtler sociological variables are impossible to track. Rajar is currently based on a panel of radio listeners that make notes in a notebook about their radio listening habits. It is an intrusive method that requires continual action from the individual.
However a fully automated system, such as the wristwatch device used by GfK, isn’t going to get round the fundamental principle that the people you are measuring are not necessarily ‘normal’ listeners. Internet audience measurement is an example of how a fully digitised system still does not address the underlying issues. This approach requires incentivising panellists to install a piece of tracking software into their computers. Who, particularly in these days of RIAA tracking individual file sharers, would want to have a piece of “spyware” monitoring their every move and sending that information to an unknown third party? The argument is even more true of companies (not to mention the principle of having a piece of 3rd party software sitting behind the company firewall communicating freely with the outside world). All of which probably goes a long way to explaining why the data from some Internet audience measurement firms suggests that usage of file sharing networks is declining: perhaps what they are actually measuring is their departing file sharing panellists?