volume 3
december 2000

Audience research for shortwave broadcasters

by Graham Mytton

What audiences can shortwave broadcasters expect to reach? The answer to that question lies in audience research, argues Graham Mytton — former Head of Audience Research of the BBC — in this essay. International shortwave broadcasting is no longer so unusual in its relationship to audience, feedback, etc. As "normal" radio station environments become more and more fragmented, their audiences become increasingly difficult to measure too. The alternatives that are being developed for the standard surveys of general populations may also prove useful for shortwave radio stations.

Why research matters. All successful businesses need to know how well they are performing. They need information on their customers. How many are there? How much do they consume? What do they like and dislike? What attracts them to the product or service? Why do they choose your product instead of another? In today's world when tastes and behaviour can change very rapidly, businesses need to keep themselves up-to-date with these changes. Companies that fail to have a very good system of market intelligence can very soon go out of business.

Doing this in broadcasting is not at all easy. Radio and TV broadcasting is a rather special kind of business activity. Broadcasters and the people who listen or watch the programmes — the "consumers" — have a rather unusual relationship. It is unusual because when these two encounter each other, when a broadcast reaches a person, nothing really happens. No money changes hands. No physical object is either collected or handed over. A transmitted programme can be heard or viewed by one person or one million. Nothing happens that actually tells you for sure how many customers there were.

It makes broadcasting unlike most other businesses. If you are selling soft drinks, you can tell very quickly how you are doing. You can count the sales you have made. That is true for the mighty Cocoa Cola and the humblest soft drink vendor. Bottles and cans are easy to count. Moreover, money is paid for goods like soft drinks and you can easily add up sales income.

Broadcasting is unusual even in the field of communications. Telephone traffic is easily measured. Newspaper sales are added up and comparisons between different titles can readily be produced. A great deal of day-to-day interaction in communications of all kinds is easily measured. Now with the Internet, one can be almost overwhelmed with the wealth of data on web access and use that can be readily accessed. Millions of transactions and movements on the World Wide Web can be and are being daily computed.

But this kind of consumer information is not so easy to obtain in broadcasting. From the early days various methods were devised to get around this problem. Audience research was developed using the market research techniques of face-to-face or telephone interviews. And in the field of audience research we see the development of two specialised methodologies — personal diaries and electronic meters, designed to measure audience behaviour.

These techniques and methods have become well developed and sophisticated. But they are mainly used to measure audiences for domestic radio and TV in national, regional and local markets. Most large and small broadcasters in this and many other industrialised developed countries have up-to-date and timely data on their audiences. They know their popular programmes. They know their reach, share and cumulative audiences. These data are used as the currency for buying and selling advertising time and space. They are also an essential part of the broadcasting business to aid scheduling and programme making decisions.

What of international radio broadcasting? Are the same research techniques available and appropriate? Can we know, with the same level of detail, who is listening to what and when and how often? Can we know what attracts our listeners and how they make their choices? To some extent we can and we do, although we will probably never have the level or depth of knowledge that domestic broadcasters are accustomed to an enjoy.

What existing research tells us. Audience research in international radio broadcasting is something that I have been involved in for nearly twenty years now. Some of the same techniques are used as in domestic audience research. In my old department at the BBC we commissioned surveys in countries in all parts of the world. From these we were able to say what kinds of people the listeners are, what they like, when they listen, what kinds of radio sets they use, what reception is like and how their behaviour changes over time. However, there were always important differences between the data we produced and the kinds of information that domestic broadcasters are accustomed to. We were never able to say how many people listened to a particular programme. Our picture of the international audience for the BBC was painstakingly built up from data from separate surveys in several different countries, carried out over several years. This was because we never were able to carry out simultaneous surveys of listening in several countries.

In audience research for international radio broadcasters, the major international radio broadcasters like the BBC have a distinct advantage. Their audiences are, in many countries, quite large. They are especially likely to be large where certain conditions prevail. These include the absence of choice in countries where the media are not free from various kinds of government restraint. Another factor can be high levels of political tension when demand for reliable news from outside the country concerned can lead many people to tune to foreign stations on shortwave.

Figure 1: Weekly audiences to international radio stations in Nigeria in any language, July 1998 (n = 3,004)

Let us look at one example (Figure 1). I am taking Nigeria in 1998, before the return to democracy last year and before international broadcasters like the BBC became available via many local radio stations. At this time, in June 1998, all listening to international radio had to be by shortwave. As you can easily see, the top three broadcasters had substantial audiences at that time. The sample of over 3,000 was selected to represent the Nigerian |adult (15+) population of about 60 million. It means that the BBC audience can be estimated to be around 17 million and that of the VOA about 14 million. But you can also see that audiences for other shortwave broadcasters are quite small, and some too small even to show here.

Few shortwave broadcasters gain audiences large enough to measure on normal, general population audience surveys. The BBC, VOA, DW, RFI and some other broadcasters do achieve audiences in some countries that are easily measurable and some of these are very large, at times and in some countries, as large as domestic radio stations. This was the case in Nigeria and is also the case in many other countries, mostly in Africa, Asia, the Middle East and in areas of present conflict and tension, like Albania and the former Yugoslavia.

I have used data collected by the BBC and VOA to do two things. First I look at listening levels to international shortwave radio broadcasters taken as a whole, and then at levels of ownership of shortwave radio. First, I have taken data from surveys in 57 countries to show the different levels of listening to any international shortwave radio station (Table 1). Most of these surveys are of entire countries, but in a few cases they covered only urban areas.

Table 1: Listening levels to international shortwave radio stations

Note: "Brazil (R)" refers to data from a survey of rural Sao Paulo state. Elsewhere in Rural Brazil, listening levels may be higher. Data from China are from a 10 city study. Listening levels may be higher in other towns and in rural areas.

Generally speaking, levels of listening are low or very low in the more developed countries and higher in the less developed ones. But other factors play a part also. One of the most important is the level of deregulation of radio. Another important one is that there is generally a higher level of shortwave use in countries where shortwave is used for domestic broadcasting. This is the case in nearly all African countries. Note also that listening tends to be higher in former British territories where there continues to be some attachment to BBC listening. Shortwave use in these countries has fallen when and where the BBC has been able to obtain local rebroadcasting facilities.

Access to shortwave in various countries. How much access to shortwave do people have in different countries? It is high where shortwave is a domestic means of radio transmission and low where shortwave is not used domestically. The following estimates are likely to be only approximate.

Table 2: Percentage of radio households with shortwave radios in Africa and the Middle East

Note: One of the main problems in this area is that owners of radio sets often do not know the names of different wavebands, even though they may use them all. Note also that these percentages given here and in the next tables are not of sets but of radio households. In some cases — single set households mainly in the less developed countries — there is no difference.

There are a few things to note in these tables. The first is that shortwave sets are common throughout Africa (Table 2). This is because all national broadcasters in Africa have used shortwave as their main means of national transmission. This now changing and there may now be a decline in shortwave access. This will probably apply to the cheaper sets, while higher priced and more sophisticated equipment will probably continue to have shortwave bands. The same is true of India, although this may now also change as FM spreads. In other parts of Asia, radio itself has gone into decline in many countries. However, even though access to shortwave may not be very high, this does not appear to affect shortwave use as much as might be expected. Access is very low in Vietnam and this probably depresses shortwave use figures. Even so, audiences to foreign radio are not insignificant there. Pakistan data are striking in that even though only a minority of radio households has shortwave capable radio sets, the proportion listening to foreign stations is high.

Table 3: Percentage of radio households with shortwave radios in Europe

There is a surprising range of levels of shortwave access in countries of the former Soviet Union (Table 3). It needs to be remembered also that in some former USSR countries, most people do not have wireless sets, a legacy of the fact that the Soviet authorities encouraged the use of cable (tochka) radio, presumably to block out western radio as far as possible. In Europe, shortwave access is quite high, especially in countries where levels of listening to foreign radio were formerly quite high. But even in Western Europe, many people may have at least one shortwave radio set. In many cases, however, the waveband may not be used. In the Americas, shortwave access is generally low, although in countries like Peru where shortwave has been used for domestic broadcasting, access is higher (Table 4).

Table 4: Percentage of radio households with shortwave radios in the Americas

What audiences can shortwave broadcasters expect to reach? For most shortwave broadcasters, the numbers are scattered and small. This does not necessarily mean that they achieve little or nothing. You may have audiences of less than 0.1%. Sometimes they will be a lot less than this when measured within the context of a national survey representing all adults in a country. But this does not mean an insignificant achievement necessarily.

Shortwave broadcasters are niche broadcasters. Often they are not aiming to reach mass audiences, but people with specific needs, interests and tastes. Just let us suppose that a shortwave international broadcaster reaches listeners in many countries but that in none of these is the audience achieved large enough to measure by the usual general population surveys. If your average reach was, let us say, one person in 10,000 — that is 0.01%. On a global scale that is a lot of people — about 350,000 of them. The problem is that we can never afford the scale and number of surveys that would be necessary to prove it.

So is the situation hopeless? What can we do practically to learn about users of shortwave and listeners to international shortwave radio programmes? What can be done to help the broadcasters understand and know their audiences better?

There is growing interest in niche communication and therefore also in niche research. It is no longer unusual to seek to research small and scattered communities of customers. The problem, as we know, with radio is that no visible or easily measurable transaction takes place. But this is not insurmountable and many people are now working on new ways of creating suitable sampling frames.

International shortwave broadcasting is no longer so unusual in its relationship to audience, feedback, etc. As "normal" radio station environments become more and more fragmented, their audiences become increasingly difficult to measure. Sometimes this is because their audiences are also scattered and small, just like shortwave audiences tend to be. Is there a better way of researching them than standard surveys of general populations? There must be.

Practical and effective ways of audience research. I think that we can find practical and effective ways of evaluating the impact of and the audience to shortwave broadcasts:

Mail surveys. All broadcasters receive mail. This can be used to generate addresses on a database that can be used over time to learn about those who write, track changes in what they like and listen to, learn about who they are and invite regular feedback and other useful knowledge. The drawbacks are that no quantitative assessment of impact can be made. This is a self-selected sample and may not be representative of the audience as a whole. Those who write are different from those who listen. However, we know from research that responses from listeners can give an accurate picture of reception conditions.

They can also tell us something valuable about an important section of the audience — the committed and keen listeners. What is more, the people who respond to you by writing may be precisely the kinds of people you wish to reach. The very fact that they respond means that they may be what you want to achieve — an interchange or relationship with users, customers, or however you want that relationship to be. In other words, mail surveys, so often frowned on as unrepresentative, may come into their own. Turn the criticism on its head! The letter writers are the ones we want! So any sample of them is by definition a perfect one!

I use the example of people giving out pamphlets on Oxford Circus Underground station in London, advertising English lessons for foreign visitors. You don't need to know how many people read the pamphlet, how many threw it away, how may stuffed it in their pocket and forgot it. You need to know only what the response was. How many customers did you get as a result? That is all that really matters. The same surely can apply in broadcasting, even if we are not selling anything. But most shortwave broadcasters are in the business to get a message across. One way of assessing impact is to measure the response. Effective communication can be designed to elicit response as part of a strategy. Research can be used to capture a lot of useful data about those who respond.

On-air surveys. If you want some more general information about those listening and you wish to stimulate response for research purposes, you can actually ask questions over the air. These can be very good value in terms of the information gleaned. The on-air appeal can attract people to write who normally would not. You can explain the value of and importance attached to feedback. Specific questions can be asked. The BBC Can You Hear Me? exercise was of great value ten years ago in identifying the use of different transmitters at different times and places. Announcements were made at various times of the day and week, and in each language service and to all parts of the world. Listeners were asked simple questions. They were to give their name and address, say where they were listening at that moment, to give the date and time and to say what frequency they were using at that moment.

It gave the BBC an immensely valuable snapshot of transmitter use. It could have been used more for other purposes. It could also have been used to build up a database of useful contacts, but wasn't. But I am reminded of a similar exercise conducted a few years ago for the FEBC by their Burmese programme department. Research inside Burma was then impossible and it is not very easy even now. FEBC wanted to know what kinds of listeners it had in Burma. Were they mostly Christian? Were they mostly young people? What proportion were female? If, as they though, many listeners were young people, what kinds of issues and problems were uppermost to them? They asked these questions on the air and received many thousands of responses, which they were able to use to understand and know their audience better. Research showed that a majority was indeed young and only a minority was Christian. Most were Buddhist. I am convinced that on-air research could yield valuable dividends in shortwave broadcasting.

Special surveys among target populations. Instead of general surveys of whole adult populations, surveys can be specially targeted to those people who the stations wish to reach. If the station or service wishes mainly to reach, for example, the Polish diaspora, there may be ways of obtaining address lists of such people. If the station can define the kinds of people they most want to reach, it is possible that a special sample to find such people may be obtainable and used in a survey. There are some other ideas that may be worth considering. One could conduct or commission audience surveys among target group people only. If you are aiming to reach shortwave hobbyists, just survey them. If you want to reach Poles living abroad, the same applies. If your main objective is to reach the better-educated listener, commission research just among a sample of them.

Internet research. This is growing area for research and one from which shortwave broadcasters may learn. The medium (Internet) is still very limited in its reach and caution should be exercised when considering its usefulness as a medium through which actually to carry out research into anything other than the medium itself. However, a good deal of Internet-based research (as against research into Internet use) is very like the on-air surveys mentioned earlier. Users are asked to respond. As these methods are developed by the fastest growing area in the market research business, there may be lessons to learn for shortwave broadcasting research.

Research through existing shared and other surveys. Lastly it should not be forgotten that much useful research continues to be done among general populations by or for the major international broadcasters. The BBC, the IBB, RFI and other major broadcasters commission many surveys each year. These can serve two functions for the smaller broadcasters. They can provide data on shortwave access, shortwave use, and many other relevant data. They can also be used to measure audiences to stations, although the results for many may be disappointing. Nonetheless, having a general picture of the amount that shortwave is used can be a guide to the strength of the medium.

Note from the author: I should like to acknowledge the financial support of Merlin Communications, which enabled me to come to Washington for this conference, and the active interest and assistance of my old department (Marketing and Communications) in the BBC World Service that made possible the writing of this paper. I am grateful to Merlin Communications and the BBC World Service for assistance and support in providing this paper at the NASB conference. Anyone interested in further information regarding audience research for international radio broadcasting can write to me at gmytton@gn.apc.org.

Dr. Graham Mytton is former Head of Audience Research of the BBC, and now working for Audience Research Training and Consultancy, UK. This article was presented at the 12th Annual Meeting of the National Association of Shortwave Broadcasters (NASB), Friday, May 5, 2000.