Editor's note: Barton Lee and Alexander Wong are principals at East Marketing Research Institute, Guangzhou, China. This article was prepared with the assistance of David F. Tatterson, marketing consultant, international business development, Amoco Oil Company, and Joseph M. Kamen, an independent marketing and research consultant specializing in industrial and commercial research in international markets.

Marketing is most important when customers - a term that includes buyers, decision makers, influencers and users - have a choice. For those who grew up in an age where competition for the customer's time, effort, and money was limited, and a planned economy accepted as a natural phenomenon, the very concept of choice is difficult to understand.

Less than 10 years ago, China started to accept, if not welcome, foreign investment and influence - but mainly on producing for export. Inevitably, the growth of the export-oriented companies increased consumer income and raised aspirations and expectations. Fantasies became reality. Consumers started to think about product-choice and brand-choice mixes and purchase criteria.

The need for marketing research arose to deal with two types of demand stimulation: primary demand stimulation, demand for a product category or type such as computers, dishwashing liquids or disposable diapers; and selective demand stimulation, demand for a specific vendor or brand within a product category.

Existing Chinese institutions were inadequate to fulfill the information needs. A few organizations attempted to "go commercial," but they generally lacked professional marketing research backgrounds. They tended to equate marketing research with public opinion polling, which itself was not highly developed. Several were originally non-profit organizations that couldn't or wouldn't make the leap into the commercial world. Examples: China Social Survey Institute in Beijing and Soft-Science Company, established by the Guangzhou Social Science Academy in 1985.

China's first professional marketing research service, Guangzhou Marketing Research (GMR) was founded in 1988. Many followed. South China Marketing Research (SCMR) was started in 1990. In 1993, Far-East and SRH launched what is now called Survey Research Group (SRG) China, the first joint venture.

Procter & Gamble (P&G) supported the start and development of China's marketing research industry. P&G provided, and continues to provide, training and funding of research projects to such respected marketing research firms as Guangzhou Marketing Research Company and East Marketing Research Institute (EMR) in Guangzhou and All-China Marketing Research Company (ACMR) in Beijing.

The marketing research industry experienced a growth spurt since 1993. Now, approximately 100 marketing research companies exist. However, the pioneering companies remain busy and are concentrated in three cities: Guangzhou, Beijing, and Shanghai.

Types of marketing research organizations

Chinese marketing research organizations are classified into three categories:

(a) State-owned organizations that are outgrowths of government science research departments, statistical bureaus, and universities.

(b) Joint ventures between foreign and Chinese companies. Examples are Survey Research Group and Gallup's branches in Guangzhou, Beijing and Shanghai.

(c) Indigenous marketing research organizations such as EMR.

Joint venture firms tend to be better capitalized, can cover larger geographic areas with consistent methodologies, and have access to superior hardware and software.

Indigenous marketing research organizations are usually better able to conduct business-to-business research, customize research, continuously exercise quality controls, and adapt the research to local and regional economic, social and cultural environments. They also have lower fees.

As the volume of business grows, these marketing research organizations tend to build their own networks of branches in other cities. No more than 10 are able to cover the larger cities throughout China.
Individual organizations have become more interested in considering merging or forming alliances in order to meet clients' demands for broader coverage, particularly for large scale projects.

Clients

Marketing research organizations that have only a local presence tend to have one or two main clients that account for most of the business volume and many smaller clients. These research organizations often do subcontracting work for joint ventures or China branches of transnational - particularly European and American - marketing research companies. The next largest client group are manufacturing and service corporatios that plan to invest in China.

Chinese-owned corporations, especially state corporations, rarely use marketing research, largely because they are not comfortable with the consistency and systematic approaches that are characteristic of marketing research. Indeed, Chinese companies don't highly regard marketing research. They believe that they already have knowledge of the market, see marketing research as something that slows them down and conflicts with the intuitive approaches that they believe have served them well. They feel that structured research simply doesn't and can't take into account highly important - but hidden - personal relationships and highly complex political and social networking.

However, these attitudes are likely to shift as distribution strategy shifts more to "pull" from "push."
Procter and Gamble is not only the largest single client for many marketing research services, it has set the standards of excellence and introduced methodologies and analytic frameworks that have been universally accepted. Perhaps more important has been its training and support of marketing research organizations. These beneficiaries have become the professional pillars within China. We predict that Chinese corporations, witnessing the values of the research that P&G pioneered, will become much more receptive to the more standardized approaches and quantification algorithms.

Respondents

In the late 1980s, when marketing research first emerged in China, people generally did not hesitate to be interviewed. Highly curious and welcoming fresh experiences, they cooperated with interviewers. The refusal rate was very low.

In recent years, however, the fear of crime has grown, particularly in such large commercial and industrial cities as Guangzhou, Beijing, Shanghai, and Chengdu. Citizens are suspicious of strangers coming to their doors. Many have installed security systems - including alarms and expensive locks - that make it more difficult for interviewers to gain access. Combined with the drop in curiosity and growth in boredom, this factor has led to an increase in refusal rates.

From a large scale study, we calculated these refusal rates, which include both refusing to open the door and refusing to be interviewed:

Guangzhou, 32.4%

Beijing, 21.8%

Shanghai, 10.0%

Chengdu, 4.8%

Wuhan, 21.3%

Successful methods for increasing cooperation rates include employing only those interviewers who can speak the local dialect, avoiding using male interviews, and strengthening training and supervision.

Sample design

China's population is about 1.2 billion of which 350 million are in urban areas (622 cities and scores of thousands of towns). Of the cities, 32 have populations of at least 1 million, 42 between 500,000 and 1 million, and the remaining 548 less than 500,000.

Almost all marketing research has been concentrated in the 74 cities having populations of at least 500,000. These cities are classified into two types:

  •  National trade centers such as Beijing, Shanghai, Guangzhou, and Chengdu. Almost all national surveys include these four cities.
  • Secondary cities such as Wuhan, Chongqing, Kunming, Fuzhou, Xiamen, Hangzhou, Jinan, Shengyang, Dalian, Xi'an, and Tianjin.

Sample sizes range from 200 to 400 per city, with a mode of 300.

Few marketing research surveys are conducted in smaller cities. Two exceptions:

1. Qualitative and small scale quantitative research in developed areas in such provinces as Guangdong, Fujian and Zhejiang, but almost never in inland areas.

2. Business-to-business projects. For example, EMR has recently completed a petroleum products project in 20 cities in Guangdong and Fujian provinces.

Almost no small towns or rural areas are included in marketing research projects. An exception is qualitative surveys occasionally conducted in suburban areas of the larger cities. These small towns and areas generally have primitive economies and are not attractive to manufacturers and service providers. Large distances between residences, low educational and literacy levels, and poor transportation are barriers.

Sample selection

Simple random sampling is not feasible because no well defined sampling frames exist for including all in the relevant population. Address lists are not readily available, and even if they were, the high rate of moving render them invalid.

The most popular sampling method is multi-stage random sampling: city area, street, neighborhood committee, household, person. For example, Guangzhou consists of eight geographic areas of which one is unsuitable because of its small, widely dispersed population and high mobility. The second level are streets within one or more areas. Every street has between 10 and 30 neighborhood committees, which comprise the third level.

Then, research organization employees are trained to survey and map the area. They walk around the perimeter of the neighborhood committee and measure its area. They start from a randomly chosen household, and following a tightly defined procedure, mark every nth household.

Later, interviewers will attempt to interview those in these marked households, choosing the individual within the household based on both random methods and quota fulfillment requirements.

For central location studies, "convenience" samples are most often used. The client usually specifies the selection criteria. Passersby are stopped, screened for eligibility and quota, and recruited. Recruiting usually involves making an appointment - and gaining a commitment - for the respondent to appear at a central location.

Quantitative-qualitative mix

Quantitative research accounts for about 90 percent of research revenues. The qualitative research includes both focus group and individual depth interviews. On a relative basis, advertising agencies usually favor groups, but marketers demand one-on-one interviews.

Survey media

Face-to-face interviews - both door-to-door and central location - constitute the most common media. Mail surveys suffer from a very low response rate, and telephone surveys have their own difficulties. Some newspapers print questionnaires and ask readers to complete and return them. Typically, response rates are very low.

Specialized services

SRG and CCTV conduct studies that focus on retailing purchase patterns and media behavior. The number of cities they cover is increasing.

No successful general household panels exist. EMR established a 300-base panel mainly on laundry products. Participants periodically received samples of these products and kept diaries of their purchases. Once or twice a month, they were interviewed. High costs and technical difficulties caused this panel to be disbanded after one year. Another attempt to establish and maintain a panel also failed.

Time and costs

Generally, a door-to-door quantitative survey of 300-500 people in one city takes between two and three weeks, about half devoted to field work and half to coding and analysis. Including more cities increases the time by three to five days. Qualitative studies can usually be conducted within own week. The most time-consuming activities are screening and recruiting.

As to costs, consider a typical quantitative 30-minute door-to-door survey of over 500 participants. Foreign and joint ventures will charge between Y200-300 RMB (or approximately US$24-$36) per completed interview. The local companies' rates are between Y80-150 RMB (or approximately US$9.80-$18) per completed interview. Costs per interview are higher for smaller samples.

The charge for a focus group interview ranges between Y3000-6000 RMB (US$360-$720) which includes recruiting, meeting room and interpreter, but not a moderator.

Analytic methods

Quanvert, a statistical program from Quantime Corporation, is probably the most commonly used software, but most clients are not interested in complex analyses, including multivariate analyses, perceptual mapping, and the like.

Emerging trends

We predict that more observational studies will be used: observing people shopping and making choices, noting difficulties and hesitations, watching how they use products, and the like. These studies trigger ideas for new products, more powerful merchandising methods, and enhanced user friendliness.

As in the United States, marketing research has a strong heritage from the behavioral sciences. Very few researchers have had graduate or undergraduate work in business. We predict that in the future, more marketing research will be strategy-oriented rather than fulfill a more limited specific data ga thering and analysis function.