Editor's note: Jerry Stafford is chairman of SRG International - Europe; Mark Stapylton is managing director of SRG International, Pleasantville, N.Y.; and Geoff Hutton is managing director of SRG International - Europe.

Traditionally, syndicated or multi-client research services are often simply annual or biannual surveys of awareness, image and usage that are funded by the research agency and the resulting data sold to several participating clients. These services are now as well developed in east and southeast Asia as in western Europe, with buyers being able to purchase off-the-shelf reports covering a plethora of products and services. Our firm, Survey Research Group (SRG), for example, offers multi-client services in such industries as automotive, fast food restaurants, automobiles, hair and skin care products, consumer electronics, domestic appliances and many others. Since each participating client shares in the cost of the research, one of the key benefits of a syndicated service is that generally nationally representative data can be obtained at a fraction of the price of an ad-hoc survey.

This undoubtedly has led to the emergence of these standardized services over the past 20 years as companies operating in Asia tend to be restricted by relatively small research budgets and clearly looking to stretch them further. Obviously, there are disadvantages of this type of survey: All participating clients, including your competitors, will have access to the same information and there is far less scope to dictate the content of the questionnaire to suit one's own particular needs.

One particular type of syndicated service, the omnibus survey, is regularly carried out in 15 countries in Asia by SRG and some of its competitors. The omnibus service offers similar advantages as mentioned above but it does not suffer from the inherent disadvantages. Each client has exclusive access to their own information and total control of the questions asked. On the surface it would appear that the omnibus service could be made very standardized in Asia, even more so than in Europe, where one must deal with a different language in almost every country. One might conclude that a questionnaire in Chinese could be used with little if any modification in a number of countries, such as China, Hong Kong, Taiwan and Malaysia, thus avoiding the problems of timing, cost and inaccuracies generally associated with translations. However, due to vast differences across the region in terms of language, culture and geography, the "standard" omnibus survey becomes far less standard than would first meet the eye.

Great care

Whilst the omnibus is a very cost effective and efficient way of conducting research in Asia, great care must be taken in order to make best use of this service, particularly by companies thinking of entering these markets for the first time. It is an extremely complex region, not just in terms of language and dialects, and one which is difficult to comprehend without firsthand experience.

Countries such as Japan, Korea and Thailand, where there is a single culture and a single language throughout the country, are unique in the region. In both Singapore and Malaysia, for example, omnibus questionnaires are always printed in three languages: English, Mandarin and Malay. In Malaysia, in addition to Mandarin, which is spoken by all Chinese-literate consumers, there are three other commonly spoken dialects which have to be dealt with at the respondent level: Cantonese, Hokkien and Hakka, whereas in Singapore the commonly spoken dialects are Cantonese, Hokkien and Teochew.

Whenever possible it is also best to match ethnicity of the interviewer and the respondent. This issue goes beyond language; it has an impact on rapport between the interviewer and the respondent due to culture and behavioral issues. In both Singapore and Malaysia, it is common for research agencies to have both Chinese and non-Chinese interviewing staffs.

China is somewhat unique in the region in that Mandarin, or Puntonghua, meaning "common language," is the official language of the entire country. However, it is usually necessary to interview respondents in Guangzhou in Cantonese, respondents in Shanghai in Shanghainese, and in a number of other local dialects in cities in other areas. Also, due to the rapid development of the key urban cities in China, a lot of "foreign" Chinese (i.e., people who have moved to the cities from other areas) will not speak the local dialect, and not necessarily Mandarin. Whereas, questionnaires in Chinese are printed using the same Chinese characters, the various dialects are not identical in that slightly different words may be used for the same Chinese character and the sequence of words may vary.

At the extreme is India, where there are 19 major languages and over 200 dialects. India is divided into linguistic states. The country can be described as a mini-Europe, with each state like a separate country within Europe with its own language and cultural peculiarities. A national survey in India is similar to a pan-European survey in terms of cultural and language differences the researcher must be aware of and able to deal with. For an omnibus survey conducted in only the five major metros of Delhi, Bombay, Madras, Calcutta and Bangalore, questionnaires will usually be printed in six different languages; English (for the professional classes), Hindi, Marathi, Bengali, Kannada and Tamil. A survey that even approaches national representation in scope will generally be printed in at least 12 languages. Imagine the difficulties of timely preparation of visual stimuli for the more complex questionnaires that we all know too well.

Inconsistencies

Geography, economic development and infrastructure are all factors that often cause inconsistencies in samples and interview methods across countries in Asia. Samples, omnibus and otherwise, are almost always nationally representative and include all SES classes in Hong Kong and Singapore. Omnibus surveys by telephone are also well developed in both countries.

Based on geography alone, however, national samples in China and Indonesia are almost unthinkable. China is predominantly rural and Indonesia consists of several thousand islands, the majority of inhabitants of which cannot be considered consumers by Western definition. In Thailand, only 21 percent of the population live in urban areas, and two-thirds of the total urban population live in a single city, Bangkok. Many surveys are carried out in Bangkok only, or at most Bangkok and a few Upcountry cities. In these three countries, telephone interviewing is, for the most part, not a consideration.

When conducting omnibus or any other type of research in Asia the following guidelines will serve the international researcher well.

  • Be sensitive to cultural differences, not only between countries but within countries as well.
  • Learn to rely on local researchers, who usually know their countries best, and encourage their input. But ask questions.
  • Focus on the objectives not the methods. Asia is a very diverse region often requiring different methods and approaches to a particular research problem.
  • Allow sufficient time for planning and preparation of surveys, translations, checking and rechecking of translations, printing of questionnaires, preparation of visual stimuli, etc.