Editor’s note: Stephen Phillips is vice president of worldwide services at BAIGlobal Inc., a full-service market research firm headquartered in Tarrytown, N.Y.

While developing countries offer wide opportunities for business growth, they present a special challenge for marketers and researchers because local conditions often defy Western expectations. For example, a manufacturer of automatic paper-feeders recently expanded to Indonesia and found their sales were low. They were about to conduct market research when a local customer pointed out the problem. It wasn’t the product that was at fault, it was the Indonesian labor situation. With wages at pennies a day, it was simply cheaper for businesses to hire full-time employees to feed paper into their photocopiers than to buy an automatic gadget.

Given such unexpected situations, researchers who are only familiar with the developed world need to consider whether they have the time/inclination to study local cultures to ensure research goes smoothly. As an alternate to intensive cultural education, they can hire a research firm that specializes in international work to conduct studies on their behalf. Professional research firms with experience in developing markets can offer the expertise needed to avoid cultural pitfalls and deliver accurate results. Having lived and worked in several Southeast Asian developing countries over the past decade, I offer the following insider’s guide on how to conduct effective market research in this challenging environment.

The "culturally sensitive" game

A novice will often take one of two approaches to conducting research in developing countries. The first is to expect everything to be the same as in the West, a potentially dangerous assumption, as we’ve just seen. The second is to be very concerned about cultural sensitivities and "how things are done here." Surprisingly, for research purposes, by far the best of these approaches is the first, allowing some room for flexibility. This is because local research suppliers can find it fun to take "culturally sensitive" newcomers for a ride:

"We tend not to finish things on time here, it’s a cultural thing."

"We can’t do interviews that long, it’s a cultural thing."

"Often only three or four respondents arrive for focus groups, it’s a cultural thing."

The U.S. manager now working in a culturally unfamiliar environment can get bamboozled by these types of comments, but shouldn’t be fooled. The best business approach is to always expect top-quality research. In fact, in many developing countries, it is often easier to conduct accurate research than in the West because:

  • Populations are generally under-researched, so there is little respondent fatigue. Response rates for door-to-door research often exceed 90 percent.
  • Interviewers are relatively cheap and plentiful, so there are few problems getting large numbers of interviews conducted quickly and inexpensively.
  • Highly structured local cultures or long-standing bureaucracies provide a great information source for identifying specific types of respondents. In fact, it may be easier to draw a good probability sample in some developing countries than in Western nations where privacy is a growing issue.

Start with a solid design

When designing and executing any research project, the key to success is to allow time for solid design and set up. For research in developing countries this means:

Step 1: Gather basic background information on the market.

Find out about population size, geographic spread, telephone penetration, income per capita, life expectancy, the government, urban-rural split. All of this information is easily available through secondary resources, back issues of research magazines or local libraries. With this basic data in hand, a researcher will be able to frame the study accurately and come across in the local country as knowledgeable and interested.

Step 2: Identify your target group.

The best definition of a respondent group is one based on behavior rather than the demographic/socioeconomic measures used in the developed markets. For example, the equivalent in socioeconomic terms of someone with $50,000-$75,000 income in the U.S. might be a person in the developing world who owns a washing machine, stove, and television.

Step 3: Determine the ideal methodology.

Depending on the business issue being investigated, this could be anything from door-to-door interviews to mail surveys to focus groups. However, some of the usual methodologies used in the West, such as mall intercept interviews, may not work in markets where there are few malls.

Step 4: Compare the ideal methodology to local conditions and tailor it as needed.

The lucky researcher will not need to modify the study design at all. However, in most cases some changes will have to be made to accommodate local conditions. For example, in the developing world, mall intercept-type interviews can be conducted instead at bus depots. Methodological fine-tuning often occurs within multi-country studies where comparisons need to be made between markets. The most accurate worldwide view will come if you can keep the study design as standard as possible between Western and developing markets.

Build a local relationship

Once the design is laid out, the best way to conduct a successful study is to select and stay in constant contact with a local partner in the developing country. If possible, go and visit. There is nothing as good as walking around with local interviewers, visiting the areas where customers live and work, to get an understanding of the local research environment. Second best is to call, but be aware that the cost of outgoing calls from developing markets can be much higher than in the U.S. and can be a higher proportion of study costs in markets where labor is extremely cheap. I recall one supplier in India frequently complaining about the cost of faxing daily progress reports. Also, understand that it may take several calls to get through to your local supplier, as connections can be slow in Azerbaijan.

Finally, bear in mind that many local suppliers in developing countries are top professionals, so U.S. researchers should try not to come across as the "Western expert." As in any business relationship, local researchers like talking to colleagues and sharing experiences, so use this natural inclination to build rapport and relationships to help get things done.

Create the best questionnaire

Again, the best approach is to write the most accurate questionnaire possible, then check its design and wording against local cultural conditions. Most designs and questions translate well from one language to another. Even complex conjoint designs have been conducted in fairly primitive markets. However, there are certain hazards to be aware of.

  • Scales: Scales are probably the biggest discussion point in any international study, both in their use and in their interpretation. Verbal scales can cause problems as there is little or no difference between "excellent" and "very good" in some languages. Numeric scales are probably better to use, with some written description at each extreme such as "most likely" and "least likely." I personally favor 11-point scales from 0-10 as almost all schoolchildren are marked on this scale from Wuhan to Washington.
  • Sensitive questions: Income is a sensitive question anywhere, but particularly so in countries where a lot of the money changes hands at a "gray" (i.e., bribery) level and where governments are intrusive. Respondents may be afraid of answering or will simply lie. Most countries have some adapted system of determining social class and it is probably best to use this. If you are particularly interested in income then you can ask, but consider supplementing those responses with questions on expenditure and ownership of household items.
  • Lifestyle issues: In politically repressive countries beware of asking "lifestyle" questions that border on the political. Do not simply delete these questions, but work with the local research firm to get as close to the intended meaning as possible without causing offense.
  • Translations: Once the questionnaire has been translated, get it back-translated, i.e., from the local language back into English. This is an important quality control step and helps prevent any sloppiness on the part of local suppliers. Beware of words that mean similar things even in English, such as "masculine" and "macho." There may be no way of differentiating them in the local language. If the back-translation disagrees with that of the original, be careful. The local language may have changed and new words may have developed. The local research firm should help explain any discrepancies.

Fieldwork and analysis

Having selected the ideal respondents, methodology, and questionnaire and tailored them to local conditions, the fieldwork and analysis now needs to be done. Here, standard research rules and quality control procedures apply. If the local supplier has been selected well, you should be dealing with someone who knows research and can speak to you "in your own language" should trouble arise.

One concern that can be infuriating when doing quantitative research is data processing incompatibility. Check what DP system each supplier uses up front and do a dummy data transfer using the first few questionnaires. This simple systems test helps avoid annoying time delays at the end of a study.

The main problem usually experienced during the analysis has to do with responses to scales. Each culture answers scales differently. The Southern Chinese are far less generous than the Filipinos when rating any product or advertising. In addition, different groups within a single country can answer scales differently. Frequent travelers among Thailand’s upper class may respond to scales differently than their lower-class compatriots. In fact, they may respond more like frequent travelers from the U.S. than like any other group in their home culture.

There are several different ways of standardizing such divergent data, such as weighting responses or adjusting the scales up or down for a given market. International research companies often keep databases of results for the purpose of adjusting scales, but the key is to be aware of the problem in the first place when reporting results.

Troubleshooting

Examples of the more obvious research problems that occur in developing countries include:

  • Inappropriate data collection method: In a market with 15 percent telephone penetration, you would not consider a telephone survey of the general population. Door-to-door interviews would be more productive. However, if your audience is upscale consumers or businesspeople, then a telephone survey may still be appropriate.
  • Wrong sample design: Doing a nationally representative survey may be inappropriate. If all the wealth and 80 percent of your client’s sales are in the top three cities, then stick to those locations. For example, trying to do a nationally representative study in China would be very costly and take forever. There are over one billion people to represent and most of them have little or no money. Research must be tailored according to local realities.
  • Unreasonable quotas: If your study requires quotas for each decile up to age 65, but life expectancy in the market is only to age 50, you will have problems. Quotas need to be adjusted to match local demographic profiles.

It can be done

Overall, conducting insightful market research in developing countries is not that difficult to do. With some knowledge of the local culture, a good relationship with a reliable supplier, and some forethought regarding possible pitfalls, it can be done. By using an international market research firm, U.S. managers can shorten the learning curve and get data from a developing market more quickly and effectively. But the ultimate keys to success are an inquiring mind and adherence to the pragmatic rule "make no assumptions."