It became clear that, in order to gain the trust of licence payers, the government and advertisers alike, a single agreed system would be needed. Since 1992 listening diaries have been issued and the results analysed under contracts awarded by a company jointly owned by the BBC and commercial stations, Radio Joint Audience Research Limited, commonly known as Rajar. With each sector owning 50 per cent impartiality between BBC and commercial interests is protected, policy decisions requiring the agreement of both parties and representatives of the advertising industry who also sit on the board.
A total of 130,000 respondents are used each year making Rajar the biggest audience research survey in the world outside the USA. It is also one of the most complex consumer surveys. Results are produced for about 340 separate radio services, including 60 BBC stations, but few of them cover exactly the same area as another station (Total Survey Area or TSA) so, allowing for overlaps, data have to be analysed for more than 600 different listening areas across the Great Britain and Northern Ireland. Respondents are each asked to complete a one week diary showing all the stations they listened to, for at least 5 minutes, recorded in quarter hour time blocks. The results are published every quarter for all the stations, although for the smaller stations the figures are based on a rolling sample taken across the previous 12-month period.
Every three months the latest top line results, weekly reach and hours listened per station, are made available to the public free of charge, most conveniently via the Rajar website www.rajar.co.uk, but only participating stations and other subscribers have access to more detailed information on each service. Unfortunately such a huge undertaking does not come cheap and many smaller commercial stations and most community radio services cannot afford to participate (costs typically start at around £7,000 per annum) and are therefore not specifically included in the survey.
It is arguable that, even were it affordable, Rajar would not give community radio stations what they need. It is a blunt instrument. Where a programme, or even a whole radio service, is aimed at a minority group or age-range within the larger community there can be no guarantee that they will make up a statistically reliable proportion of the sample listeners. For example, even if the total sample for a survey is 700, a perfectly respectable sample size in most applications, an individual age band or other demographic cell in the published results will have been derived from a much smaller number of respondents. In one case, which we looked at for the book Managing Radio, of the total respondents in a local radio area only 29 listeners were, for example, women aged 25-34. If we believe our station has a, perfectly acceptable, 25 per cent weekly reach with this age group, we can see that we could only expect seven or eight of these respondents to have heard our programmes at all during a typical week. Rajar’s reported listening to any one show by this sample is plainly going to be greatly influenced by whether these particular eight individuals happen to be at work, asleep or otherwise not available to listen at the relevant time on the particular day or days. These issues are even more pressing where the entire service is aimed at a group who represent some minority within society. Even quite large commercial stations such as Premier Christian Radio, Club Asia, Spectrum Radio and Sunrise Radio have seen audience figures fluctuate according to the number of diaries given to their target audience group, a variable not corrected for in the Rajar sampling and weighting mechanisms.
Rajar is also very slow. It does not produce listening figures for a particular programme on a particular day but rather, for a small station, only an average of the audience for that time slot over the past twelve months. Add to that the time taken for data collection and processing, audience response taking a couple of months to reach a programme manager, and Rajar is not helpful in developing new programming, in sharp contrast to the immediacy of feedback available to internet-based media.
Rajar terminology can be quite confusing and whether or not you choose to use Rajar, knowing what terms stand for can help you understand another station’s results. A useful document of Rajar terminology can be downloaded here.
Continue on to Brian’s thoughts on other research
No Comments, Comment or Ping