Survey Research in Japan

SURVEY RESEARCH IN JAPAN

Kazuo Kobayashi, Japan Market Research Bureau, Inc.

ABSTRACT -

This paper gives an overview of marketing research activities in Japan, focusing on size of research firms, customers of research services and data collection methods. The present state of in-home personal interview method in Japan is also reviewed.

INTRODUCTION

This paper was prepared for the purpose of giving a basic understanding of marketing research activities in Japan to non-Japanese research related scholars and practitioners.

It focuses upon the characteristics of survey research organizations and their data collection methods.

Technical aspects of marketing models, data analyses, and implementation of survey data into marketing decision by Japanese marketers are beyond the scope of this paper.

1. SIZE OF M.R. BUSINESS IN JAPAN

The size of marketing research business in Japan was estimated to be approximately US $170 millions in 1983, accounting for 5.8% of worldwide expenditure for marketing research. (Table 1)

There are some 60 research companies conducting survey research and providing data to various customers. As of March 1985, 41 research companies are formal members of the Japan Marketing Research Association (JMRA), which was founded in 1975.

JMRA sets forth essentially the same code of ethical standards for survey research as the Council of American Survey Research Organizations (CASRO), ESOMAR and MRS in the U.K. do.

Each year, JMRA member organizations supply to the Association, on a confidential basis, such data as their annual sales, their office-based staff size, and so forth.

For 1975, the year of the foundation of the Association, 30 members supplied the data, and reported combined annual research sales of Y9.2 billion.

For 1983, 38 members supplied the data and the corresponding figure reached Y25 billion or $105 million, showing a 13.4% compounded annual growth rate for the past eight years. For the same period the compounded annual growth rate for the Gross Domestic Product in nominal terms was 8.0%

It is estimated that survey research sales among JMRA member organizations represents 60-65% of all such business in Japan. Thus, for 1983, the size of marketing research business in Japan as a whole was approximately $170 million. (Table 2)

Compared with the US ($1,275 million, in 1983), survey research business in Japan is only 13% as large. This seems very small if we consider the GNPs and per capita GNPs of the two countries. Japan's GNP level is 35% of that of the US and Japan's per capita GNP is 68% of that of the US.

This low level of research expenditure in Japan can be explained by the following factors:

(1) Japanese consumers are more homogeneous in terms of race, language, income/expenditure distribution and way of life than American consumers, and Japanese consumers are less individualistic than Americans.

Therefore, it is rather easy for Japanese marketers to meet the consumers' demands or desires without heavy spending on survey research.

(2) When Japanese marketers make decisions they tend to give more weight to their experience and intuition than to numbers. Compared to American marketers, Japanese marketers do not have much faith in numbers.

(3) Research organizations in Japan tend to be limited to the function of providing data. Although they maintain a high level of quality in their survey data and use various advanced techniques, they are not skillful enough in implementing the survey data for clients' marketing decisions, and tend to fail to create further research needs.

The situation will change radically in the future. Because consumers are becoming more individualistic and are looking more and more for ways to be different from other people, it is becoming difficult for marketers to predict consumers' needs merely on the basis of their past experience and intuition.

TABLE 1

MARKETING RESEARCH EXPENDITURE: WORLDWIDE AND JAPAN

TABLE 2

JAPANESE MARKETING RESEARCH COMPANIES (1983)

2. SOME CHARACTERISTICS OF RESEARCH FIRMS

No marketing research companies existed before the World War 11. The first research company was established in 1949. A.C. Nielsen's and six other JMRA member firms started their business in the 1950's. (Table 3)

Japanese research organizations are small in size. The average annual sales for JMRA member firms were $2.8 million in 1983. As in other countries, a concentration of sales among larger organizations is found in Japan.

The top 12 JMRA member organizations account for 67% of the total member sales. (Table 4)

As in most countries, marketing research activities in Japan are geographically concentrated in the nation's major markets. 35 out of 41 member organizations of the JMRA have their main office in Tokyo. (Table 3)

TABLE 3

YEAR OF FOUNDATION AND LOCATION OF JAPANESE RESEARCH FIRMS

As the Tokyo metropolitan area accounts for 23% of the total population and Japan is a relatively homogeneous country, most marketers are satisfied with qualitative and quantitative data from this area. in many cases, marketers will only use data from the two main city areas, Tokyo and Osaka, or possibly, the third largest city, Nagoya, as well. The demand and supply of marketing research services for the nation as a whole is relatively small, although several research organizations have national field force network.

3. SOURCE OF REVENUE

Who are the customers of survey research in Japan?

The largest customers are advertising agencies, which account for 22.7% of total research expenditures in 1983. Advertising agencies frequently purchase survey research on behalf of their clients, and in such cases ultimate purchasers are not advertising agencies but the advertisers.

The second largest customers are foods manufacturers (19.2%), followed by Textiles and Chemicals (13.4%). After these come, the central or local governments (7.3%), electric appliances and machinery (7.1%), mass media (6.4%), automobiles and ship building (5.4%) and service industries (5.0%).

Overseas customers account for only 1.2%, however, this does not necessary mean that multinationals are inactive in commissioning survey research in Japan. Major multi-nationals have already established their own subsidiary or affiliated companies in Japan and they are more enthusiastic in conducting research than their Japanese competitors .

4. DATA COLLECTION METHODS

In-home personal interviews are still a major method of data collection in Japan, although as described later environment for this research method in Japan is becoming tougher. In 1983, 47% of total research sales, or 62% of Ad-hoc surveys sales came from this method.

Mail surveys and telephone interviewing are not popular yet in Japan. Mail surveys account for 4%, and telephone surveys accounts for 3% of the total research sales. Phone ownership is high -over 90% of homes, but no WATS-line type facilities are available.

TABLE 4

JAPANESE RESEARCH FIRMS BY SALES

Central Location Test (CLT) accounts for 7% of the total sales. This method is used for producttesting and advertising copy or CF testing. Unlike in US, however, very few research organizations maintain halls at constant places.

Qualitative research also accounts for 7% of the total sales. In particular, focus group discussions have become popular in recent years. Japanese consumers were once thought to be less talkative and less dynamic than people elsewhere. Nowadays, although they are still rather polite in expressing their views, they are easily brought into a meaningful discussion. A typical focus group session lasts around two hours among 6-8 participants. Several organizations have facilities with one-way mirror and video-tape recording equipment.

Panel surveys account for 14% of sales, according to JMRA data. This percentage is definitely understated, however, because the sales of two giant companies, A.C. Nielsen and Video Research, are not included. Most syndicated research services common in many countries are also available in Japan.

Omnibus survey services are not so pop ular in this country, but several organizations provide such services to marketers in major cities as well as to the nation as a whole.

The expense of doing marketing research is not cheap. Costs are perhaps close to those in the U.S. or continental Europe, ahead of the U.K. (Table 5)

TABLE 5

RESEARCH SALES BY SURVEY METHOD (1983)

5. DATA COLLECTION THROUGH IN-HOME PERSONAL INTERVIEWS

Let's focus on In-home Personal Interviews - the major method of data collection in Japan - in some detail.

There are several features characterising of in-home personal interviews in Japan:

(1) Respondents are selected on a probability basis using names and addresses from the Government's Local Inhabitants' Register as sampling frame. Use of this sampling frame makes probability sampling both practicable and economical. Interviewers are not necessary involved in the sampling operation.

(2) Advance letters are sent to respondents in most cases.

(3) A briefing session for interviewers is held. Interviewers are usually not employees as such of the research organizations, but are paid per interview or by the hour, with a retainer as well in some cases.

(4) Like interviewers elsewhere, they go out in all weather, almost always 1-iy public transport (The problem of parking makes it hard to use a car). This makes it difficult if the survey requires them to carry bulky or heavy items.

(5) Japanese social conventions make it necessary to do most interviews standing at the door, or just inside the entrance of the home. So research users have to accept limitations in the length of questionnaires (usually within a 30-minute maximum),

(6) Another unusual point, again dictated by social custom, is the need to give a small present to every respondent. A typical gift might cost one to two US dollars, for example a pair of panty hose, a small towel, a handkerchief set, or a mechanical pencil. The present also has to be carefully wrapped.

To illustrate the present state of the in-home personal interview method, I will outline a survey conducted by Japan Market Research Bureau (JMRB) in June 1983 on males and females aged 20 to 59. This survey was used as a case study at the 'Marketing Research Seminar', held by the Japan Union of Scientists and Engineers.

Despite the fact that respondents are generally cooperative, response rates are fairly low.

Reasons include:

1) The increasing variety of lifestyles, and the lower number of people staying at home. in particular, the majority of married women today hold some type of job (as explained later).

(2) An increasing number of refusals or noncooperation due to an anti-industry attitude or the desire to protect privacy. As shown in the table on next page, the current refusal rate is around 15%, while in 1960's the rate was less than 10%.

(3) Changes in dwelling types (more apartment buildings) making approach difficult.

(4) Increased mobility (people moving away may not immediately transfer their residence registration).

The yardstick now for response rates is 50% for adult males and 65% for adult females. For the case study I mentioned previously, the relevant figures were:

TABLE

The low response rates may startle some readers, but reaching even those levels required a great deal of work on the part of the interviewers.

Call-backs by an interviewer at different times are one way of reducing the non-response rate. Table 6 shows the number of repeat visits that were necessary to complete interviews in the case study survey - an average of 2.7 visits for men and 2.3 for women. Table 7 shows that interviews were completed on weekends - Saturdays and Sundays - in 61% of the cases for men, and 49% of the cases for women.

According to our survey, 52% of married women held some type of job. Right now, women spending their time solely in the household are a minority among housewives; this has far-reaching implications in social terms, and affects marketing and information gathering too. It has become nearly impossible to interview many housewives at home on weekdays during the day. With both husband and wife working, such households have two purses, which means greater buying power.

TABLE 6

RESPONDENTS BY NO. OF VISITS BY INTERVIEWERS

TABLE 7

RESPONDENTS BY DAY OF WEEK INTERVIEWS COMPLETED

SUMMARY AND CONCLUSION

(1) The size of marketing research business in Japan is relatively small compared with the U.S. or European countries, particularly in consideration of Japan's high GNP.

This could be explained by characteristics of Japan consumers, marketers and survey research organizations. However, this situation will significantly change in the future.

(2) Geographically, research activities are con centrated in the Tokyo metropolitan area.

(3) Advertising agencies, food manufacturers and textiles/chemicals are major sources of research revenue. Research customers from overseas are very few, while well-known multi-nationals are active through their subsidiaries or affiliated companies in Japan.

(4) In-home personal interviews are still a major method of data collection. Mail and telephone surveys are not popular yet.

The central location test is fairly popular, but very few organizations have facilities at fixed places.

Focus group discussions are becoming popular.

(5) Most syndicated research services such as TV rating, Store audits, Consumer panels, etc. are available.

(6) The costs of research are close to those of the U.S. or continental Europe, ahead of U.K.

(7) The response rate for in-home interviews is 50% for adult males and 657 for adults females, even after several call-backs.

Such low response rate is inevitable because of (a) high not-at-home rate of respondents (more than half of married women hold some type of job), (b) increasing numbers of refusals due to an anti-industry attitude or desire to protect privacy, (c) increasing in a apartment building making approach difficult, etc.

(8) Overall, the demand for marketing research data can be expected to increase, while the environment for collecting survey data will become even more difficult than they are today. it is the survey researchers task to solve this problem through innovative ways.