Positioning a cosmetics brand for Muslim women in Indonesia

Editor’s note: Michael Sack is president of Image Engineering, Inc., a Goshen, Ky., research firm.

When it comes to ethnic and cross-cultural research, there are plenty of “experts” who will tell you how difficult it is. They’ll tell you that sample frames are harder to obtain. Respondents are more difficult to reach. Questions must be translated and back-translated to verify that their meaning is clear. Rating scales are subject to systematic, yet undefined, cultural biases. Responses must be interpreted in the proper cultural context.

But the situation isn’t as tough as the experts make it out to be. In reality, it’s probably worse! Because even when all those cultural issues are planned for and taken into account, significant problems can still remain.

1. Middle-class, white Americans will willingly tell you more than any other group. All other cultures and sub-cultures are less trusting, with good reason. For most, revealing true values and motives would give power and influence to those who have real potential to use the information against them.

Ethnic and cross-cultural audiences are particularly distrustful of those outside their community, and of those within their community working on behalf of outsiders. They know commercial enterprises don’t just exist “to design products and services that better meet people’s needs.” They will participate in research projects, but they won’t willingly tell you nearly as much as you would like to know.

2. Even when consumers are willing to reveal themselves, most are unable to communicate what we’d like to know in a form we can understand.

A large portion of consumer motivation exists beneath the conscious level, inaccessible to researchers regardless of their cultural knowledge and sensitivity. In fact, when it comes to subconscious motives, those within a subgroup are often the least able to explain why people of that culture behave as they do. They are too enveloped by the culture to examine it clearly and objectively. (And often, they describe themselves as not the majority.) It’s like asking a fish to describe the nature of water.

This brings us to a critical element of successful cultural-research: practiced ignorance.

A lack of knowledge about language and culture issues can be a real advantage when it comes to learning about people’s values, motives and behavior. The process of practiced ignorance can be used effectively to research any culture, including your own!

Practicing ignorance effectively, however, demands skill. First, it requires data that is not dependent on language and/or cultural norms. Second, it requires providing respondents with the ability to accurately describe culturally bound relationships. Third, it requires a culturally neutral medium for effectively conveying the findings back to people living and working in the culture responsible for the development of marketing applications.

To illustrate each aspect of applied ignorance, I will refer to a consumer-products study focusing on Indonesia. If you know little or nothing about Indonesia, the case study will be perfect for you! If, by chance, you happen to know a lot about Indonesian culture, please stick with me - I’ll need your help later.

The value of not knowing the language

Fluency in a language and familiarity with local idiom are obviously helpful in collecting research data. Nonetheless, complete reliance upon understanding of verbal communication suffers from the assumption that people can and will provide the information you need. Unfortunately, as noted earlier, they usually can’t and often they just won’t.

By the time we reach adulthood, we have spent most of our lives developing and exercising verbal defenses, enabling us to withhold information with a high degree of skill. Sure, we can sound forthcoming, but the bottom line is that revealing personal information is giving power to others. If this is a problem in cultures that value individualism and forthrightness (and it is), think of how it plays in a culture which values and demands consensus and social discretion (like that of Indonesian Muslims).

Finally, words are consciously-used symbols that flow from conscious thoughts. They are a poor tool to reflect our emotional reasoning and our unconscious needs and desires that impact much of our behavior - including our purchase behavior. Emotional reasoning is at its strongest when the topic is the most personal. If this creates problems in researching cultures that value introspection (like the U.S.), imagine how it affects research in cultures that de-emphasize public displays of emotion and recognition of personal identity (like Indonesian Muslims).

Thankfully, words are not the only symbols available to the researcher. Images are powerful symbols that overcome many of the weaknesses of words and effectively complement the information verbally obtainable. Further, literacy issues limit data collection among some groups, but visual literacy is nearly universal. Image-based research has even been used successfully within primitive tribal groups with less than 500-word vocabularies and no written language.

In verbal research there are usually at least 10,000 words that need to be understood and managed to achieve cultural knowledge, but a culture’s visual language contains only about 200 active symbols on any particular topic. The visual language of a target culture on a given topic can easily be learned during the time frame of a normal project, unlike the spoken language.

Visual language is also a powerful collection tool. People rarely express themselves visually. This means their visual defenses are much less practiced than their verbal defenses. Therefore, people tend to be much more transparent when selecting images instead of words.

Also, because a single picture can contain dozens of visual symbols, people provide more information when selecting images. Therefore, patterns emerge with greater clarity (and with smaller samples!).

Moreover, because most visual images carry many symbols, they capture unconscious information in addition to conscious information. This allows access to needs and motives that are unavailable through words.

All these characteristics were both necessary and useful in efforts to do research to determine how to sell cosmetics (which tend to be associated with vanity and Western culture) to Indonesian Muslim women.

Because of the depth of information available through images, the Indonesian study, which was nationwide in scope, required only 150 respondents participating in 12 group interviews. Recruiting and moderating were done by native Indonesians (in the Bahasan language) who had completed a two-day training session in our company’s visual research protocol.

Another advantage of visual research over verbal research is that it is far less demanding upon group moderators. For example, local school teachers are often recruited to facilitate when professionals are unavailable.

Each interview group completed 10 picture-sort tasks on average, answering questions by selecting pictures from a set of more than 10,000 magazine images collected from around the world. The tasks were crafted to elicit perceptions, values and motivations regarding skin care products. The resulting hybrid (verbal and visual) data set included more than 3,000 images (with many times more symbols) and a written description of each that form the data points used in analysis.

The value of not knowing the culture

In verbal research, cultural expertise is often used to bridge the gap between a statement and its underlying meaning or motivation. In visual research, the process is reversed. Visual research analysis begins by identifying consumers’ values, motivations and needs. Then it works backward to better understand and/or contradict respondents’ stated explanations. With the process flowing in reverse, the filter of cultural familiarity simply gets in the way, obscuring the identification of visual patterns and biasing their identification. The visual data analyst can best accomplish his or her task with a blank slate of absolute cultural ignorance!


Chart 1 1

Without cultural predisposition, the analyst is better able to see symbolic patterns emerging from the data. As shown in Chart 1, patterns can be classified into six categories: color, shape, function, relationship, context and the degree to which the symbol is animate. These symbols comprise the shared visual language of a culture or ethnic group. The analyst first applies known rules of symbolic communication and then searches for unique cultural expression of functional, relational and context associations in the group’s visual dialect. Specifically, the analyst examines how respondents used that language to describe themselves and their relationships with products and services, and their preferences. This enables understanding of deep-seated values, motivations and needs - all without benefit of prior cultural knowledge.

Having identified these base patterns, the analyst can now benefit from an informed local marketing and cultural perspective. Invariably, the cultural expert will confirm the findings as consistent with their knowledge about the culture. In addition, the research results frequently enable the cultural expert to make new connections and see previously hidden dimensions of culture -- with important implications for marketing.

For example, in the Indonesian cosmetics study, nearly 80 percent of the images selected in answer to the question “How would you most like to smell?” featured people, usually children. (By contrast, in the West, the most frequent image response would be flowers.) The interpretation was that the people being studied are comfortable with natural human odor as well as their general appearance. Further, it was seen that youthful health was associated with a strong human “fragrance” related to odor but totally consistent with a habit of smelling a child’s head as a gesture of affection (Fig. 1).


Figure 1

Additional images selected to represent the concepts of “enhancing beauty” and “repairs skin” were strongly associated with healthy eating and drinking (Fig. 2). This result, and other responses, demonstrated that Indonesian women view beauty as primarily internal and spiritual. In addition, in its most positive light, external beauty is seen as a reflection of good internal, spiritual health and nurturing -- that is, blessing. (By the way, in keeping with the spirit of the process, those of you familiar with Indonesian culture are welcome to write to me to confirm, clarify or expand on this description.)


Figure 2

Applying visual research: leaving ignorance behind

Once visual data patterns have been correctly gathered and interpreted, it is time to leave ignorance behind. Developing culturally appropriate marketing initiatives requires a thorough understanding of appropriate cultural values, messages and communication channels. When this knowledge is combined with the insight provided through visual research, marketing applications become fresh, engaging and effective. Without the perspective of visual research, targeted marketing applications are much more likely to settle for being safe and appropriate, and based on “traditional wisdom.”


Chart 2

Applying this new form of data involves four familiar steps (see Chart 2). First, use research to describe key consumer relationships, including consumers’ relationships with product/service benefits, with the product/service category, and with individual brands. Second, identify the most important relevant consumer need that relates to the category. Third, determine why existing consumer/product relationships have been insufficient to meet this need, i.e., define the obstacles to success and the reason for failure to dominate a market. Finally, identify the positive associations that can help convert negative or incomplete consumer relationships into positive, complete relationships, define “success” in terms of desired function, relationship and value. When this process is complete, the targeted associations should direct all aspects of the marketing mix.

Difficult sell

In Indonesia, the key relationships identified underscored why cosmetics were a difficult sell. Consumers set aside personal preferences in order to avoid the penalties associated with offending the public morality. Functionally, their issues and needs did not focus upon appearance and hygiene, and they placed great value on the traditional and the natural.

In terms of the most important relevant need, Indonesian women desired not to be seen as outwardly attractive but to be seen by others as inwardly beautiful, natural and spiritual. They also desired to see themselves in this same light. Meanwhile, they viewed cosmetics as immoral, unnatural and artificial. Cosmetics by nature emphasized the external over the internal and spiritual.

The brand commissioning the study was the category founder in Indonesia and had previously sustained a 75+ percent market share for many years. The company might have been tempted to view itself as an expert on Indonesia, except for the fact that the brand’s share had recently and rapidly eroded to less than 50 percent.

In response to the image-based research, skin care offerings began to be promoted for their ability to build strength and protect health in consistency with religion and tradition. The message created permission for Indonesian women to enjoy external beauty as a secondary reflection of their inner wellness.

To add substance to this message, the company created added ingredients to existing products as well as a new product offering - a natural drink product designed to nourish and protect the skin. This “skin care to drink” reinforced the brand’s primary goal of promoting internal health and beauty. As a result, the brand regained a 70 share. This success, which occurred in 1994, influenced a trend in “healthy cosmetics” that extended across dozens of countries and multiple continents.

The final word...er...picture

Many “experts” say that cultural research and understanding directly depend upon an insider’s view. My experience suggests, however, that starting with a clean canvas (in combination with local input) can often yield a fresher, clearer portrait of your target audience. So don’t be afraid to plead ignorance - and be willing to practice it.