From the early days of radio there were suggestions from various individuals and departments within the BBC that it would be a good idea to find out what people liked to listen to and why. Initially, pressure came from those concerned with education, but persons such as Val Gielgud (Director of Drama) and Charles Siepmann (Director of Talks) were also keen for research in this area to occur. They did not consider letters from listeners that arrived via the Programme Correspondence Section or Radio Times to provide sufficiently scientific evidence of listening habits or people’s feelings about the programmes. John Reith (the Director-General) and others were uneasy about the concept, however, as they feared the introduction of systematic audience research might influence programme planning and result in programme policy focusing on achieving high audience ratings, rather than about broadening knowledge and supplying appropriate content. Adapting the timing of programmes to suit people’s pursuits was already being considered.
Throughout the early and mid-1930s there were increasing attempts to establish some sort of formal or informal study of listening habits, culminating in a paper submitted to the General Advisory Council in January 1936 by Stephen Tallents (Controller, Public Relations). In 1936 the Control Board eventually agreed to establish a small Information Department, under Tallents’ Public Relations Division, to deal with research issues, including Listener Research. On October 1936 R J E Silvey of the London Press Exchange began work as Listener Research Assistant. It was Silvey who established and built up the Listener Research (later Audience Research) department and laid the groundwork for its methods of gathering and analysing data.
Silvey began his research with a series of enquiries into the audiences for various types of programmes. The first of these surveyed listeners of Drama and Features, with 350 respondents from a variety of backgrounds asked to complete questionnaires about each play or feature programme they had heard over the course of four months. Whilst its findings were not ground-breaking, it was the first time there had been an attempt to survey a representative sample of listeners. This ‘panel’ method was considered successful and deemed applicable to other programme departments, including Talks, where Listener Research provided valuable feedback on questions such as the qualities listeners expected to find in a good Talks broadcaster. Less representative studies were also carried out. For example, listeners were asked to send in postcards saying what time they preferred certain programmes, which was of interest to certain departments. As those with no strong opinions were less likely to write in, the information this produced was not considered quite so reliable.
Silvey was in favour of random sampling, first suggesting this in July 1937, even though this was not a method condoned by others in the field. Discussions took place about establishing panels to look at the audience’s reactions to programmes and to create a ‘listening barometer’. The idea was that a sample of homes should be chosen and the reactions of its occupants tested four or five times a year, using questions that were as simple as possible. The barometer was designed to show the relative sizes of audiences for different types of output.
The first experiment in this area was carried out using Light Entertainment programmes, beginning with a broadcast appeal for volunteers to take part in the study. They would be asked to record whether or not they had listened to particular programmes, and how much of them they had heard. The problem with this was that the volunteers were more likely to listen to Light Entertainment programmes, and were also given log sheets listing programmes, meaning they might feel obliged to listen because they had volunteered to take part. A control group was therefore also established, who would be interviewed rather than sent log sheets. Its members were distributed through the regions according to numerical strength of the listening public.
By autumn 1938 there was a great deal known about the listening habits of the BBC’s audience. The first random sample had been taken in January 1938 and there had been detailed enquiries into the popularity of various programmes and types of programmes, such as children’s programmes. The research work itself was, however, still largely experimental and the results took time to be distributed, so the findings did not greatly influence programming decisions. It was around this time that Listener Research became part of the Home Intelligence Department of the Public Relations Division.
With the advent of war in September 1939, Listener Research assumed greater importance. Radio was seen as an important tool for communicating with the public. Management needed vital information, such as the areas of the country where reception was defective, in addition to more general research into changes in listening habits and tastes in programmes during wartime. A special report was also prepared into the numbers listening to propaganda broadcast from Germany, including the infamous broadcasts by Lord Haw Haw (William Joyce). Listener research would also serve during wartime to monitor listener preferences and reassure programme planners that even when there was bad news, listeners were still not averse to hearing light entertainment programmes rather than sombre ones.
In December 1939 Silvey introduced a new method of assessing listening figures – a continuous survey that would result in a daily listening barometer. In addition to a posted questionnaire, there were 800 interviews a day, and the results were analysed faster than before. The introduction of the Forces Programme (in addition to the existing Home Service) in 1940 meant additional Listener Research was needed, and audience questionnaires were distributed.
From 1942 to 1945 the Listener Research department was part of the Programme Division, a change from its previous administrative designation. It remained under programming departments until 1958, going to Entertainment division in 1946 and then to the General Division of Sound Broadcasting in 1957.
Post-war Listener Research continued along established lines, but the television service, which had been in abeyance throughout the war years, introduced a new factor into the equation. Initially the Listener Research Department did not feel able to carry out more than ad hoc enquiries in this area, as they were not allocated extra staff to carry out viewer research, and the television service was initially confined to the London area. In 1948 a number of viewers were asked to report their views on six particular programmes, via a Radio Times ‘Vote for Viewers’ campaign, but the results were not useful. The final ‘Vote for Viewers’ in June 1948 invited viewers to apply for a general questionnaire about television, which was designed to give a picture of current attitudes that could be more fully tested at a later date. The 900 responses were not necessarily representative but did provide some indication about people’s attitudes to the programmes on offer. The results showed that viewers were not particularly selective about what they watched and were largely satisfied with the television service.
Later in 1948 a more carefully selected sample of viewers was surveyed, alongside a control group of those who did not own a television set. This survey gave a clearer picture of the viewing habits of various age groups and social classes, and also showed that listening to the radio declined with the presence of a television set. It was also found that people were more likely to concentrate solely on viewing whereas when listening to the radio they might do other things at the same time.
In September 1949 it was finally decided that the Television Service should receive audience information comparable to that provided by Listener Research, although the implementation of this decision was delayed until new transmitters came online, giving greater television coverage. At this time the BBC management still seemed quite adverse to the new medium, and were very concerned that viewers were spending so much time watching television. Once more, they were worried that the audience research results might unduly influence programme makers. A survey was first carried out to find out basic information about the households who had television sets, some information about the types of programmes they watched, and their programming preferences. This showed that items like studio plays were more popular than plays relayed from theatres, and cabaret more popular than revues. More cultural pursuits like opera and ballet divided the audience and it was also found that women preferred outside broadcasts of public and ceremonial events whilst men preferred sporting events.
From 1 June 1950 the Listener Research Department became the Audience Research Department, and a Viewing Panel was established to provide statistical information on television audience numbers and to give their views on the programmes. At the end of 1951 the daily interview survey came to be called the Survey of Listening and Viewing. Surveys looking at variations in viewing between different British regions were also carried out, but concluded that differing tastes in viewing could not be explained by the part of Britain in which people lived.
During the 1950s the influence of television was also seen in the Audience Research Department’s listening figures. According to 1954 figures, evening listening fell by 30% for all three radio services (Home, Light and Third). The audiences also seemed more interested in lighter programmes. Radio talks and discussions (which usually followed the 9pm news) showed big losses in audiences, but news, Saturday Night Theatre and variety shows were less affected. The 6pm news, which had no competition from television, did not suffer at all. Another indicator of the loss of audience to television was the fact that by this time a quarter of licences taken out were for both sound and television, compared to 1% in 1948.
The start of commercial television, in the form of ITV (Independent Television), in 1955 meant that there were extra factors for audience research to assess, and its audience measurement techniques were also questioned due to differences in the figures produced by the BBC and its commercial rival. Whereas the BBC’s calculations of audience size were based on interviews with individuals, asking what people had watched the previous day, the ITV’s Television Audience Measurement (TAM) figures were based on households not individuals, and used meter readings rather than interviews, leading to significant discrepancies between the two sets of figures. The BBC, however, retained the advantage of being able to judge audience reaction as well as audience size. As ITV operated through regional franchises, BBC Audience Research also had to adapt to the need to survey audiences area by area in order to achieve realistic comparisons of audience size in the various regions. These statistics were actually purchased by the ITA (Independent Television Authority), the regional ITV contractors and some of the advertisers, although this practice stopped once ITV spread nationwide.
In 1958 the Audience Research department was once more placed under the Administration umbrella, moving away from the programming area to the Secretary’s Division, where it remained until 1969, when it moved to the Information Services Division (Information Division from 1977) of the Public Affairs area. By the 1960s Audience Research had approximately one hundred staff divided into six sections. The Fieldwork section dealt with the needs of the 1,300 field interviewers. The Statistics section prepared the daily log sheets for the interviewers, processed the data received and prepared the daily barometers, charts and other statistics. Audience Research was one of the earliest departments within the BBC to have a computer and they went through about one and a half million computer punch cards each year. The Analysis section drafted the questionnaires, tabulated the answers received, assessed the written comment and prepared the final reports, dealing with about 3,000 different broadcasts. Projects and Developments undertook more detailed, long-term enquiries into audiences. The Information Desk was responsible for the distribution of reports and dealing with enquiries, whilst the Registry and Records managed the three quarter million documents that were accrued each year in addition to recruitment and staffing.
Amongst topics dealt with by Audience Research in the 1960s were the advent of BBC-2 and colour television and also the phenomenon of pirate radio stations. A large scale enquiry was carried out about listening to the most famous of these – Radio Caroline.
In 1981 the work of measuring audience size for all television channels was taken over by the Broadcast Audience Research Board (BARB), a non-profit making company funded by the participating broadcasters and the Institute of Practitioners in Advertising. Around the same time the BBC Audience Research Department became the Broadcasting Research Department, remaining under the Information Division until 1982. It resided here briefly, still under the Public Affairs umbrella, as part of the BBC Data and Broadcasting Research Division. In 1984 it became its own department once again, directly under the Head of Public Affairs, and later the Policy and Planning Unit (from 1988). 1992 saw the compiling of radio audience figures devolved to Radio Joint Audience Research (RAJAR), a company jointly owned by the Commercial Radio Companies Association and the BBC. Despite no longer being directly involved in measuring audiences, the department continued to produce research relating to audience habits and attitudes to programmes and events. By 1995 staffing levels had fallen to less than sixty, as the fieldwork and analysis functions were no longer necessary with the advent of BARB and RAJAR.
In 1997 a major reorganisation took place and, following a Policy and Planning Unit review, the department’s duties were divided. There was to be no centralised audience research department, although any such research undertaken would still be centrally funded. Researchers based within the various programming related areas of the BBC would in future commission research as required and distribute the results themselves. A hub of the Broadcasting Research Department remained, as part of the Policy and Planning Unit (under Corporate Strategy), to act as a central point of contact as to where material was held, and urgent queries were answered by the Information Desk, which was transferred to the Broadcast division. In 1999 the remaining part of the Broadcasting Research Department again became Audience Research, and the other researchers, whilst still working within programming departments on a day to day basis, returned administratively to the centralised department. In the year 2000 this department was moved to the Marketing Division. Audience and consumer research has remained a core function of BBC research and marketing.
Currently the Marketing and Audiences team is situated under the Chief Customer Officer Group and continues to undertake and develop research techniques aimed at finding out more about the BBC’s audiences. In addition to the continued use of BARB and RAJAR reports, the development of technology in the form of home recording equipment and on demand services has meant that there remains a need for the creation of new metrics to better understand audiences and programme impact. In 2010 this was in the form of Live Plus 7, a metric that increased the time period over which viewing figures were collected in order to allow for the new ways in which people were consuming content over different platforms. Much like the early Audience Research department, the Marketing and Audiences team continue to provide BBC staff with access to recent audience information and adapt to evolving broadcasting opportunities and challenges.
