Abstract

Cross-cultural marketing and advertising research reveal important influences of culture on the global consumption of the Internet and its World Wide Web. The prevailing concepts of culture in this field of research—drawn mostly from the models of Geert Hofstede and Edward T. Hall—are discussed here with respect to their limitations and utility for the increasing number of web-related contributions from marketing research. As a consequence, additional findings on cultural influences on website design, related structural design criteria, basic conditions, and complementary criteria for culturally appropriate websites are discussed that may impact the future of the Digital Divide.

Introduction

The global spread of modern technology, including information and communication technology (ICT), is commonly regarded both as an indicator of the postmodern era of globalization and as the very precondition for that era of intensive worldwide interactions of people and exchanges of goods, services, information, and capital. Some international marketers have predicted a final convergence of culturally different markets into a "one-world culture" that would facilitate globally standardized marketing activities (Levitt, 1983). This, however, has turned out to be an illusion—too many non-cultural hard factors and cultural soft factors still exist or arise as constraints on international marketing that have to be dealt with continuously, utilizing various strategies of adaptation or localization.

For more than three decades, international and cross-cultural marketing research has focused on the standardization versus adaptation debate, which has resulted in the popular classification of "culture-free" and "culture-bound" products. Thus, for example, non-durable consumer goods like food would be regarded as strongly culture-bound products and therefore as difficult to standardize, while durable high-interest and high-tech or digital products, like ICT, as well as industrial goods, would be regarded as essentially culture-free products and consequently as easy to standardize (Baalbaki & Malhotra, 1993; Meffert & Bolz, 1994). With respect to industrial goods and international technology transfers, however, newer studies reveal that the latter notion needs to be revised. These technical systems, like culture-bound products, are subject to cultural influences to a large degree (Hermeking, 2001).

With respect to digital ICT products, initial research on cultural influences on global Internet usage, interface design, and usability has been undertaken by HCI and localization specialists in recent years. Many of their results appear consistent with a significant number of contributions from international marketing and advertising research, as well as with the author's long-term research on cross-cultural marketing communications and website design.

Cultural Differences in the Consumption of the Internet

From a marketing point of view, research on global Internet usage or consumption has two-fold relevance: First, the Internet as a product of ICT in general is a marketable good and its global or local market potentials are of some profitable interest. Internet consumption may therefore depend on marketing. Second, and more importantly, the Internet is a new channel of communication in addition to traditional media such as newspapers, radio, and TV, through which marketing communication such as advertising or public relations can be spread. Marketing may therefore depend on Internet consumption.

Consumption research is an essential precondition for appropriate product design. It tries to uncover how much, by whom, where, at what time, for what purpose, and according to whose preferences the Internet typically is used, as well as how it is used—if it is used at all.

The Worldwide Disparity

Worldwide Internet consumption data over the last decade show remarkable national differences in the numbers of Internet users. In early 2000, for example, most Internet users still lived in North America (147.5 million), followed by Europe (91.8 million) and the Asia Pacific Basin (75.5 million) according to diverse sources (NUA Internet Surveys—http://nua.ie/surveys/how_many_online/index.html; CIA World Factbook—http://cia.gov/cia/publications/factbook/index.html). In early 2005, the percentage of the population using the Internet ("active web users, at home") was, for example, in the USA, 48%, in Canada and Australia, 46%, in Sweden, 53%, in Germany, 36%, in the UK, 38%, in France, 26%, in Spain, 22%, in Japan, 29%, and in Brazil, for example, only 6% (Nielsen NetRatings—http://www.nielsennetratings.com/news.jsp?section=dat_to). Although those figures change continuously over time (and differences between different sources and their definitions of Internet usage exist—online access quotas, for example, are not valid indicators of real usage), there is a clear continuum of descent from high Internet usage in the (developed, western) North to low Internet usage in the (often less developed, non-western) South.

The diffusion of the Internet from its country of origin, the USA, to other countries of the world depended and continues to depends on several hard factors such as technical infrastructure and income per capita—i.e., economic development. This may be one reason for the Internet's differing popularity in the North and the South—also known as the Digital Divide. But economic development is only one factor. A close look at highly developed countries like France or Japan, for example, with their relatively moderate Internet usage (as defined above), reveals that some cultural soft factors play an even more important role. Specifically, the influence of cultural values on Internet diffusion is reflected in the significant correlation of many countries' Internet consumption data with Hofstede's (1991) country scores along his two cultural dimensions: Individualism and Uncertainty Avoidance. This is demonstrated in Figure 1 with Nielsen's total Internet usage data ("all users") of five referenced countries in 2001.

This general tendency of a decrease in Internet usage from the North to the South, as well as a positive correlation with Individualism and a negative correlation with Uncertainty Avoidance, is also evident in the majority of the highly developed countries of Western Europe. This is demonstrated in part in Figure 2:

Utilizing Hofstede's cultural dimensions, several marketing researchers (de Mooij, 2000; 2004; la Ferle, Edwards & Yutaka, 2002; Müller & Gelbrich, 2004) have demonstrated that cultures with low Uncertainty Avoidance are more open to innovations like the Internet as a new medium of communication; that is, they tend to be early adopters with a high diffusion rate. Empirical research in 11 European countries by Steenkamp, ter Hofstede, & Wedel (1999) revealed that Uncertainty Avoidance, among other dimensions, is a strong cultural influence on "consumer innovativeness" in general. Analyzing consumption data for 56 countries with respect to Hofstede's dimensions and several hard factors, Yeniyurt and Townsend (2003) also found Uncertainty Avoidance, among other dimensions, to be negatively correlated with the adoption of ICT products like the Internet and PCs. This correlation was weakened by high rates of literacy and international trade, but interestingly not by a high economic development (GDP per capita). In addition to cultural influence, the early dominance of the English language on the Internet, due to its origin in the USA, and the proficiency (or lack thereof) in the English language in some of these low Uncertainty Avoidance cultures could also be relevant to understanding their more positive reaction to the Internet. In general, however, proficiency in the English language and its acceptance as a foreign language may also correlate with Uncertainty Avoidance (negative).

With respect to increasing habituation to and language localization of the Internet, one could argue that it is only a matter of time until all high Uncertainty Avoidance cultures will catch up and show equally high usage rates. However, this does not seem to apply to many of them because some other cultural influences on their Internet adoption, including Individualism and other dimensions identified by Hofstede (Steenkamp et al., 1999; Yeniyurt & Townsend, 2003), seem to play important roles. Differentiating Internet usage in Europe into its culturally most favored purposes, de Mooij (2000; 2004) points out the (negative) correlation of Hofstede's cultural dimensions Power Distance and Uncertainty Avoidance with business purposes, and the (negative) correlation of his cultural dimension Masculinity with education and leisure, for example. Utilizing Hofstede's cultural dimensions and comparing Internet adoption rates in the USA, Japan and other countries, la Ferle et al. (2002) found correlations with Uncertainty Avoidance (negative), Power Distance (negative) and, again, most significantly, with Individualism (positive).

The Role of Cultural Communication Styles

Individualistic cultures' high attraction to the Internet is often attributed to the egalitarian, democratic nature of the Internet (de Mooij, 2000, 2004). The strong influence of Individualism on the consumption of the Internet, which is a relatively impersonal medium of communication, is also connected with the essential and product-specific aspect of cultural communication styles.

Hofstede (1991) refers to Hall's concept of Context and concludes that high Individualism is very often connected to "Low-context" communication, whereas low Individualism (Collectivism) very often is connected to "High-context" communication. According to Hall (1977; Hall & Hall, 1990), members of Low-context cultures have less personal contact with each other, therefore communication must be very detailed and very explicit. A great deal of formal information is communicated in a direct way, often by way of written texts. In contrast, members of High-context cultures have closer and more familiar contacts with each other; a lot of information is already shared among them, and therefore their preferred mode of communication is more informal, indirect, and often based merely on symbols or pictures. No country ratings for Hall's concept of Context exist, but a positioning of some cultures was made by Hall and later adapted by Usunier (1991), on which Figure 3 is mainly based.

A significant body of research on international marketing and advertising has shown some differences in preferred advertising styles and creative strategies between High- and Low-context cultures (de Mooij, 1998; Mueller, 1996, 2004). In general, indirect and transformational advertising messages creating emotions through pictures and entertainment are more favored in High-context cultures like France or Japan, whereas direct and rational advertising messages providing product information above all play a more important role in Low-context cultures such as Germany or many parts of the USA.

There are enormous and persistent disparities in world-wide consumption of traditional media such as newspapers and TV across different countries, which may well be interpreted as a result of culture-specific communication preferences—specifically, High- versus Low-context communication preferences—as TV is mostly based on pictures, whereas print media are more or less based on written text. In addition, most newspapers are read by one individual at a time, whereas TV may be watched and listened to by a collective group. Various hard factors (economic, technical, political) may have additional influence on this pattern of media consumption, but in highly developed countries, at least, they seem to be less significant.

For example, the average consumption rate of daily newspapers among the population in Western Europe between the early 1950s and the late 1990s reveals stable differences with (Low-context) Norway, Finland, Sweden, the UK, and Switzerland showing highest consumption rates (40% - 65%); followed by Denmark, Germany and Austria also showing high consumption rates (30% - 40 %); France, Belgium and Ireland showing only moderate consumption rates (15% - 30%); and finally (High-context) Portugal, Spain, and Italy showing the lowest consumption rates (5% - 15%) (de Mooij, 2004). Newer media data also confirm these differences (Medienlandschaft Westeuropa, 2005).

Of course, exceptions exist to that general tendency of preferences in print media versus TV consumption. Apparent exceptions include some highly developed (High-context) East Asian countries like Japan, Hong Kong, and Korea, where print media are very popular, as well; however, this may relate to their written language, which is more symbolic or picture-like (Usunier, 1991). Such a pattern of both high TV and (relatively) high print media consumption is similar to the co-existence of traditionalism and modernity, of subjugation under and mastery over nature (resulting in "harmony with nature"), and of other ambivalences like multiplex religious adherences, and both centralized and market economies (in China). An explanation for this might be the specific holistic rationale of these "Confucian" (Hofstede, 1991) or "Nipponic" (Galtung, 1981) cultures, often combining and integrating two diverse or contradictory elements into one harmonic whole. This may also be found in the Taoist symbol of Yin and Yang, or, with respect to marketing negotiations in East Asia, in the synergistic, long-term "win-win" approach (Usunier, 1991).

Similar to the general tendency in print media consumption, the different Internet usage rates generally reveal higher usage in Low-context cultures and lower usage in High-context cultures. This reflects the fact that, to date, the nature of the Internet has been more similar to print media, based mainly on a text-heavy, rational, explicit, and informational communication style, which is therefore accepted most of all in Low-context cultures like the culture of origin, the USA. Differences in Internet adoption among these cultures over time, of course, may also depend on their preferences in terms of Uncertainty Avoidance, as discussed above.

Marketing research on frequent Internet users all over the world shows a continued dominance of informational needs over entertainment needs (Anonymous, 2000a). Internet users in the UK, for example, are motivated primarily by information needs and also prefer informational to entertaining TV programs (Hammond, Turner, & Bain, 2000). Internet users in Germany read books and daily newspapers more often than non-users (Anonymous, 2000b). Similarly, in the UK and in the Netherlands, the Internet as an informative medium has been found to be more complementary to the traditional newspaper industry than cannibalistic (Geyskens, Deleersnyder, Gielens, & Dekimpe, 2002). Finally, even studies on the recent increase in Internet usage in Korea reveal that Internet users watch TV less often, but spend more hours reading books than non-users (Rhee & Kim, 2004). Because of their dominant information needs, Internet users worldwide may be regarded as a sub-cultural "information elite." These elites are, in terms of culturally preferred communication styles, generally proportionally larger in Low-context cultures and smaller in High-context cultures (exceptions might be expected in some East Asian countries).

To broaden the basis of Internet usage in the UK, Hammond et al. (2000) conclude that the Internet may need to become more entertainment oriented and more comfortable to use. With respect to Internet consumption worldwide, this means that the communication should be appropriate and acceptable to users from both Low-context and High-context cultures, which has implications for interface and website design.

The plausibility of this very general and preliminary résumé of global Internet consumption, which may be relevant to many millions of individuals as actual or potential Internet users all over the world, depends, however, on the validity of the premises about culture that are implied by the integration of Hofstede's and Hall's work.

Excursion: Some Remarks on the Implied Cultural Concepts

A first look into the body of international and cross-cultural marketing research reveals the great popularity of Hofstede's work, implying that the concept of culture represented by his theoretical model is widely accepted in this field—or perhaps, as critics say, "all too blindly followed" (Holden, 2004). The same, albeit to a lesser degree, holds true for Hall's work. Like most theoretical concepts and models, Hofstede's and Hall's works were intended to describe and explain a very complex matter (here: culture) in an understandable and operational, but consequently also simplified, way. The high complexity of culture has led to hundreds of definitions (Müller & Gelbrich, 2004, report more than 240 different definitions). A closer look should be taken at the reasons for the popularity of Hofstede's and Hall's cultural models in marketing research.

Hofstede's work was influenced by the model of basic value orientations (on man himself, man and nature, time, activity, social relationships), which Florence Kluckhohn first published in the 1950s. Kluckhohn´s model is based on the theories of Functionalism, which also had much influence on Hall's work (Rogers & Hart, 1998). According to the theories of Functionalism, culture comprises material and immaterial products as instruments to cope with problems to which all human beings are exposed during their existence. These problems are identical worldwide, but which instrument out of a variety is preferred depends on the specific priorities or values that people in social units or groups, i.e., cultures, collectively share. Consequently, different groups or cultures are characterized by different sets of values that guide their existence.

Although neither Hofstede nor Hall focused on marketing matters explicitly, their implied Functionalistic concept of culture is attractive to that discipline, which arranges for the supply of products and services that have to meet the needs and demands (i.e., problems and values) of respective customer target groups. On the other hand, the missing marketing context behind Hall's and Hofstede's work might impose a basic problem (Holden, 2004). Marketing, however, is an interdisciplinary field that integrates many diverse but relevant theories.

Hall's model of cultural categories, which originally included 10 universal anthropological "Primary Message Systems," focused on the interdependence of culture and communication. His model, of which only four categories (Time, Space, Context, Information Flow) gained larger popularity, has proven to be useful in communications research (Roth & Roth, 2001). Since communication is central to all marketing activities and Hall's research approach was quite broad, his model seems to be useful for marketing matters as well.

Hofstede's model of (originally four, now five) cultural dimensions, which predominantly relate to universal aspects of social relationships (e.g., hierarchy, group orientation, gender roles, trust / risk-taking) resulted from factorial analysis of data of work-related values of international personnel in a multinational company (IBM) in the late 1960s. His sample size of questionnaires (which added up to more than 116,000) is enormously large. The possibly obsolete databases and the generalizations related to people outside IBM are frequent criticisms that Hofstede (1991) continuously counters. De Mooij (2004) points out that several replications of Hofstede's study on different samples have proven that his data are still valid. Universal aspects of social relationships of the kind referenced may also apply well to international marketing, which has to create relationships between the supply and the demand sides, which are culturally different.

A more critical aspect of both Hall's and Hofstede's models is the local fixation of culture. Culture is generally defined by nation, regardless of the relativity of national borders and the ethnic or cultural diversity within them. Some countries still may be characterized as historically developed social, linguistic, and cultural units, but surely not all countries. This simplification, however, makes culture as a complex, influential soft factor more comprehensible and operational for international management and international marketing theories in particular. For example, according to Cateora's classic model of the international marketing environment, which is shown in Figure 4, each country as a national market is characterized by specific conditions or "forces" of several hard factors (e.g., legislation, politics, economy, topographic facts, infrastructure, standards of technology, etc.). National culture can be integrated into this set of "uncontrollables" for each country market on an equally important level in order to react to that market scenario with appropriate marketing strategies and instruments as "controllables" (Cateora & Graham, 1999).

Holden (2004), however, criticizes this concept of national culture and rejects such marketing models as synonymous with an obsolete notion of simply exporting goods and services from one country to another as homogeneous units. Many criticisms also exist with respect to Hofstede's quantified approach, which uses index scores for each cultural dimension in order to demonstrate relational differences between countries. This rigid quantification, which makes the soft factor culture, or elements of it, equal to hard factors, might mask any dynamics of cultural change and individual as well as sub-cultural value differences inside national cultures. The strong influence of the concept of Cognitive Anthropology on Hofstede's model, according to which culture as a central determinant (i.e., "producer") is based on fixed sets of collectively shared and generally unconscious "mental programs" from which all human behavior in a society results, contributes to his quite static and strongly generalizing model (Roth & Roth, 2001).

Such generalizations about culture, however, are common practice and very operational, especially for international marketing. Often marketing deals with aggregated units, such as segmented clusters of thousands or millions of individuals; each cluster is to be addressed specifically as a homogeneous target group. Such clusters are differentiated or positioned with respect to their typical market-related characteristics (e.g., consumer behavior). Similarly, Hofstede's index scores as aggregated "proxies" represent degrees of some cultural values at which they are most frequently shared by individuals in a country, representing the statistical modus (mean) of such values. Different individual values are distributed around this modus following the normal distribution (de Mooij, 2004), which is demonstrated (for Individualism) in Figure 5 in a very simplified way. Each country thus represents a (five-dimensional) cultural cluster, which may be combined with culturally similar countries (i.e., very similar index scores) to create larger cultural clusters; all clusters are positioned along the cultural dimensions. This very pragmatic way of "culture mapping" makes Hofstede's model ideal for mass marketing (de Mooij, 2004).

Because of the high number of indexed countries (originally 40, now more than 80), which, compared to other culture models, seem to be quite precisely positioned, Hofstede's model is most attractive to international marketing. As de Mooij (2004) states, his cultural dimensions provide excellent variables, as they can explain (together with national wealth) more than half of the differences in consumption and consumer behavior. Müller and Gelbrich (2004), however, have found some critical statistical deficiencies, for example redundancies and internal correlations among Hofstede's five dimensions. According to them, his model includes only three independent dimensions (Individualism, Masculinity, Uncertainty Avoidance). Not surprisingly, among these dimensions are the two that various research results have also confirmed to be relevant in Internet consumption, as has been discussed above.

Hall's model is less criticized, since it neither strictly quantifies culture nor explicitly excludes intra-country differences, as is indicated in his emphasis on Anglo-Americans as one specific group in the USA (see Figure 3). His most popular four (qualitative) categories are based on anthropology, ethnography, and real life experiences, so they may well relate to as many different aspects of life as Hofstede's (quantitative) categories do. Because only a few countries are mentioned as ideal examples within his categories and because their positioning along his categories is vague, Hall's model is less operational and therefore less favored by marketing researchers compared to Hofstede's. His model is also more hermeneutical and flexible, however, so it provides a good basis for cultural insights and tendencies, which in each single case should be combined with detailed cross-cultural and ethnographic research.

Referring to globalization, critics of Hofstede's (and Hall's) models argue that the differentiation into national cultures and national markets is losing importance, whereas trans-cultural or global target groups, which were dispersed into different nations but shared a similar set of values and consumer behavior, are becoming more important. De Mooij (1998, 2004) and other marketers favoring Hofstede's model state clearly that real global target groups, e.g., young adults, are a myth, since even with identical consumption data (e.g., ownership of products) their consumption patterns (e.g., buying motives, forms of usage, etc.) generally do differ depending on country and/or culture. For example, with respect to Internet consumption, a study of US-American and South Korean students (Choi & la Ferle, 2004) revealed that both groups use the Internet frequently—i.e., both have an average usage time of nearly two hours per day. The importance of this medium as a source of information among other media differs, however (i.e., magazines rank first in the USA, as compared to TV in Korea). In addition, the Korean students, in contrast to their US counterparts, might well use the Internet primarily for participation in online communities, but less for e-commerce (Park & Jun, 2003).

Critics of Hofstede's concept, who argue for individual or intra-national dynamics or cultural changes, prefer a concept of culture as a fluent construct, a "product" of human beings, which is also created in each (international) marketing relationship in a specific way (Holden, 2004). This dynamic cultural concept—almost the opposite of Hofstede's—is represented by the theories of Symbolic Interactionism and Clifford Geertz (Roth & Roth, 2001). In general, distinct individual and sub-cultural or minority differences within a dominant (national) culture are a result of more conscious processes of building or preserving a distinctive identity. Consequently, Hall's and Hofstede's models, which have been constructed on the idea of unconscious values or "mentalities" as the basis for cross-cultural differences, may well fail with respect to intra-country, multicultural matters (Moosmüller, 2000). Both opponents and proponents of Hall's and Hofstede's concept of culture may be right: The utility of these models depends on a basic differentiation of the research matter into a (conscious, identity-related) multicultural context versus an (unconscious, mentality-related) international context.

Thus if the Internet, for example, is consumed in a country as a result of unconscious cultural communication preferences or as a result of unconscious values of being prepared to accept this new technology, Hofstede's and Hall's models and their cultural premises will be appropriate concepts for describing and explaining the cultural backgrounds. They probably will not work well, however, if several individuals increasingly use the Internet to observe and to imitate a new lifestyle from abroad as a kind of resistance against their dominant culture, or if Internet usage by a part of the population of a country is denied because it is regarded as an attribute of a denied lifestyle of another undesirable part of the population. Therefore, unless international and cross-cultural marketing research deals with intra-country, or multicultural but with inter-country, comparisons, their prevailing cultural models can make sense. They provide initial, if vague indicators about basic cultural factors, which require deeper research including additional cultural criteria and more ethnographic details. This also holds true for the cultural criteria behind the Internet consumption patterns discussed above, which have manifest implications for global website design.

Cultural Influences on Website Design

Since the World Wide Web (WWW) is the most popular usage platform on the Internet, its interface design or website quality in particular is of high relevance for the cultural acceptability and diffusion of the Internet as a medium of communication. By analogy with product design, website design can be described as a specific set of instrumental or technical, economic, social, aesthetic, and symbolic attributes or qualities of a website that contribute to its users' satisfaction, which in turn depends on the users' cultural habits and values (Hermeking, 2000). In sum, a culturally well designed website may be defined as communicating the right information at the right place with the right layout in the right manner and in the right time according to the culture of each of its users.

Basic Website Characteristics

Often the general criteria: (1) site quality, which is also equated with usability, (2) establishment of trust, and (3) creation of positive affect during website use are quoted as the most essential website characteristics (Lengert, 2000). Research on Internet shoppers from 12 countries in North America, Latin America, and Western Europe by Lynch, Kent, & Srinivasan (2001), for example, reveals that these three criteria are of importance everywhere both for Internet purchase intentions and website loyalty; however, the extent of their impact varies significantly across cultures and across different product categories.

Regarding the establishment of impersonal, abstract trust on the Web, Morrison and Firmstone (2000) refer to the general phenomenon of a high trustworthiness of things that are familiar, expected, and conform to habits, in contrast to a low trustworthiness of things that are unfamiliar, strange, and unexpected. This phenomenon can be carried over to the forms of communication via the Web. The more the design of a website conforms to culturally-familiar communication styles and cultural habits, the more trust is established. Trust is another essential argument in support of culture-specific website design, something that also may increase general Internet acceptability.

This aspect is equivalent to the "not-invented-here-syndrome" or "country-of-origin" effect well known in international marketing, which may result in a generally more positive image of products or services from the home country than of those from foreign countries (Knight & Calantone, 2000; Usunier, 1991). Accordingly, even global brands are often marketed and advertised as if they were large local brands, which creates reputation and trust (Morrison & Firmstone, 2000).

Positive affect finally depends on a user's experience both of successful website use, which is subject to culture-specific definitions of website usability, and of a culturally appropriate or expected ratio of emotional versus rational messages communicated. In sum, these very general website characteristics are subject to cultural communication styles, but they also depend strongly on each other. Accordingly, some more operational, culture-related design criteria need to be introduced.

Culture-Related Design Criteria

In promoting people or organizations and their products, services, or ideas, communication via the Web is generally equivalent to marketing communication like advertising or public relations. In addition, it also combines several basic elements of traditional communication via newspaper, radio, TV, telephone, or direct mail in a specific way (Ju-Pak, 1999; Morrison & Firmstone, 2000). Consequently, both the standardization/adaptation discussion and central criteria of cross-cultural marketing and advertising research on culture-specific advertising styles are relevant to website design.

This is demonstrated by an increasing number of contributions from international advertising research. One example is Okazaki's and Alonso's (2003) comparison of 150 websites of Japanese multinational companies in Japan, Spain, and the USA, which were analyzed by analogy with a classification of advertising appeals into "soft sell" appeals (indirect approaches creating emotions and atmosphere by visuals and symbols) versus "hard sell" appeals (direct approaches highlighting product features with explicit information and competitive persuasion). In general, the former advertising appeals prevail in High-context cultures whereas the latter prevail in Low-context cultures (de Mooij, 1998; Mueller, 1996, 2004). Accordingly, the Japanese websites analyzed revealed the highest rate of soft sell appeals, whereas the US websites revealed the highest rate of hard sell appeals and vice versa; the Spanish websites revealed a nearly equal rate of both appeals.

Earlier evidence for the analogy with cross-cultural advertising research is given in Ju-Pak's analysis of 310 commercial websites from the USA, the UK, and South Korea (1999). Her results reveal some remarkable national differences in preferred creative strategies and, in particular, in regard to verbal and visual content appeal. Rational, fact-based appeals (especially visual) are dominant in the USA and the UK, whereas emotional, symbolic appeals (both visual and verbal) are dominant in South Korea. The UK shows the highest rate of text-heavy layout, whereas South Korea shows the highest rate of text-limited, visual layout.

These differences are equivalent to the preferred advertising styles of Low-context cultures (here: UK, USA) and High-context cultures (here: South Korea) that were discussed above. This very general relation between content, layout, and the culturally preferred communication style best suited for advertising and website design is represented in Figure 6.

With respect to the creative strategies chosen, messages of both traditional media (print, TV) and new media (websites) can generally be adapted in varying degrees to either High- or Low-context communication preferences. Because of its visual and acoustic nature, however, to date TV is the most appropriate medium for the transformational or entertaining communication highly preferred in High-context cultures, whereas print media based on written texts are more appropriate for the rational, informative communication preferred in Low-context cultures. To date, the nature of the Web overall seems to be more similar to print media, as has been discussed above; Ju-Pak's (1999) results, however, indicated quite early that culturally adapted creative strategies, along with localized website design (when appropriate), can make this new medium appropriate to the communication preferences of High-context cultures.

Some additional design characteristics analyzed by Ju-Pak (1999) reveal cultural differences that are not unexpected. The length of each web page seems to correspond to copy format, as the UK leads both in text-heavy layout and in shorter pages, whereas South Korea leads both in visual layout and in longer pages. South Korea utilizes much more (High-context type) multimedia presentation (text, sound, and/or video) than the USA and the UK, where presentation more often is based on (Low-context type) text only. Finally, according to Ju-Pak's results, the degree of interactivity according to number of hyperlinks or clickable items is higher in (High-context) South Korea.

As Ju-Pak (1999) points out, the type of products featured differs by country, but the national differences in website design mentioned above are not due to the products featured. Interestingly, the vast majority of products of the South Korean websites analyzed, as well as the majority of products of the other two countries, were computer-related, digital high-tech, or high-interest goods. These products are alleged to be culture-free and easily marketable in a standardized way. With respect to their promotion via the Web, this is clearly not the case. Nevertheless, some differentiation into both brand and the product type represented is necessary in general, since the familiarity with a certain brand or type of product, its complexity or required know-how, and its relevance for consumer (B2C) or business (B2B) interests may influence the demand for explicit information as well as the time budgeted for website use, and may therefore influence the creative strategy for website design as well.

In an attempt to order and to verify these vague cultural tendencies, the author has started a qualitative, long-term research project that has compared many randomly chosen websites of different international companies and brand types in various European countries (Germany, UK, France, Italy, and Spain) as well as the USA and Japan over the last five years. The initial background was a website design consultation project for a multinational European company that has been continuously followed by diverse student research surveys. The structural design criteria already mentioned:

1. content appeal
2. layout
3. length of pages
4. multimedia presentation
5. interactivity

have been supplemented by:

6. structure of content
7. total volume of website
8. degree of navigation support

which have also been proven to be culturally relevant criteria.

Of course, many external conditions exist in addition to the represented brand and product type. These conditions, combined with the cultural values and communication preferences of the website users addressed, may have influence on the appropriateness or perhaps on the designer's decision as to the degree of cultural adaptation of international websites. By analogy with Cateora's marketing environment model (see Figure 4), these influential factors represent country-specific conditions as "uncontrollables" for the design of websites along the introduced (and other) design criteria. This is shown in Figure 7.

Some General Findings

Some of the general findings are that websites of global companies of the New Economy that are based primarily on the Web and e-commerce both tend to be strongly standardized and dominated by rational content appeals, by text-heavy layout presenting small pictures only, by low multimedia presentation, high interactivity, large website volume, and by deeply structured content accompanied by intensive navigation support. This, for example, is shown in the homepages of Amazon in (High-context) France (http://www.amazon.fr) and (Low-context) Germany (http://www.amazon.de) in Figure 8.

In the Old Economy, websites of companies representing industrial goods and primarily addressing business users (B2B) tend to be standardized to a similar degree. This, for example, is demonstrated in Figure 9 with the homepages of ABB UK (http://www.abb.co.uk) and ABB France (http://www.abb.fr).

The general nature of these websites is strongly equivalent to Low-context communication, which according to Hall and Hall (1990) often correlates with a "Monochronic" cultural time orientation. In Monochronic cultures, time is linear, passing ("time is money"), handled precisely, and divided in an orderly fashion. Thus action chains very often are compartmentalized and sequential; one thing is completed after the other. Correspondingly, the websites that carry a great deal of explicit informational content are detailed, compartmentalized, structured in an orderly way, and also supported by many explicit navigation tools such as sitemaps, index registers, and search engines to facilitate quick orientation and time-saving site use. In addition, low multimodality and a text-heavy layout of voluminous websites reduce download times and correspond to a Monochronic time orientation.

Websites that mainly address consumers (B2C) reveal different tendencies. Global brands of durable, high-interest products are generally characterized by only modest standardization, which is focused on universal brand-related colors or graphics and uniform frames or structures of content. With respect to other design criteria, High-context cultures generally exhibit a higher degree of cultural adaptation, often revealing text-limited layouts including more colored backgrounds, larger pictures, and a much higher rate of animated illustrations or moving visuals in particular. Multimodality tends to be somewhat higher in High-context cultures, including jingles or occasional options for downloads of video or radio interviews, for example. Corresponding to the visual layout, navigation support tends to be less intensive in High-context cultures, where symbolic visual elements like pictures or hidden pop-up icons are frequently used as clickable items to lead to further website content; such websites consequently appear to be less detailed and voluminous. This is demonstrated in Figure 10 with the homepages of Sony USA (Low-context - http://www.sony.com) and Sony Japan (High-context; the latter also includes many visuals as pop-up items-http://www.sony.co.jp). (See also Würtz's analysis and examples of websites from Japan, this issue.)

Websites representing global brands of non-durable, low-interest products reveal an even higher degree of cultural adaptation. Customer-related (B2C) sites tend to be more open for adaptations to communication preferences of High-context cultures, which according to Hall and Hall (1990) are often characterized by a "Polychronic" time orientation. In Polychronic cultures, time is often regarded as circular and repeating, subject to social relationships or needs and therefore handled in a flexible, imprecise way. Action chains are structured in a less detailed way and interrupted more often since many things are done simultaneously. Polychronic appropriate websites include lots of entertaining visuals, animated illustrations, and even real multimedia elements. These elements certainly increase download time, but this does not matter so much. Explicit navigation support is rare since neither a strictly ordered route through the less detailed site structure nor very quick orientation is necessary. Implicit symbolic cues, however, support an intuitive navigation, which enhances both preferred entertainment appeals and positive affect in website use. This, for example, is demonstrated in Figure 11 with the homepages of Coca-Cola USA (Monochronic: http://www.coca-cola.com/usa) in contrast to Coca-Cola Italy (Polychronic; the latter also includes sound effects - http://www.coca-cola.it).

The higher degree of cultural adaptation of websites for non-durable compared to durable products is also confirmed by Okazaki (2005), who analyzed 206 homepages of United States companies in four European countries. The analysis of non-durable product websites is the best starting point to demonstrate cultural adaptation on the Web. Würtz's analysis (2004, this issue) of McDonald's websites in different High- and Low-context cultures confirms the findings on non-durable products, in particular with respect to the high rate of animations and moving visuals and the limited transparency and navigation support in High-context Polychronic cultures.

In contrast to such cultural adaptation, local websites of global consumer brands occasionally seem to be more standardized worldwide by utilizing the typical communication style of their brands´ country of origin. Worldwide websites of a French food brand, for example, would tend to be somewhat more dominated by (French) High-context style, whereas worldwide websites of a German car brand would tend to represent more (German) Low-context style. The latter could be seen in Figure 12 with the visual-limited homepage of (German) Mercedes-Benz in Italy (http://www.mercedes-benz.it) and, conversely, with the visual-heavy (Italian) Lancia homepage in Germany (http://www.Lancia.de).

The "country-of-origin" effect is difficult to achieve successfully. It works only in those cases where the images of the product and its country of origin are positively related in the targeted countries. When successful, this kind of national representation through website design can create adequate trust; however, in all other cases, culture-specific adaptations are more appropriate or even necessary in order to reduce trust-diminishing, "not-invented-here" feelings.

International marketing and advertising research has introduced some additional criteria that are also relevant for culture-specific website design. For example, background colors are generally subject to different culturally-specific aesthetic and symbolic conceptions and preferences (Jacobs, Worthley, & Ghymn, 1991; Madden, Hewett, & Roth, 2000; Usunier, 1991), and generally have higher importance in High-context cultures. These culture-specific symbolic and aesthetic concepts also relate to visual and pictorial presentations, which tend to reveal the highest degree of adaptation. Even strongly standardized websites tend to be different in their visual presentations of people, products, artifacts, nature, etc., since many visual scenes and images, like in advertising, are implicit representations of specific cultural values (de Mooij, 1998; Mueller, 1996, 2004; Tharp, 2001). The visual contents of websites can therefore be regarded as cultural characteristics or metaphors, which, similar to cross-cultural advertising research, often are analyzed by utilizing Hofstede's cultural dimensions (Marcus, Baumgartner, & Chen, 2003; Singh & Baack, 2004; Würtz, 2004).

Of course, verbal headlines or slogans presented in websites differ in quite the same respect. Other linguistic aspects like the overall tone of address or the use of certain words, like "we" (Leonardi, 2002), could provide for additional cues to cultural adaptation, but to date such analysis in international marketing research is rare. Finally, according to Galtung (1981), even logical styles and forms of rational expression differ across some cultures. This, for example, may be revealed by a more frequent use of tables presenting facts and figures in US or UK websites representing "Saxonic" (fact oriented, inductive) cultures, whereas German websites representing "Teutonic" (theory oriented, deductive) cultures often include detailed historical or theoretical verbal explanations. Galtung's hermeneutical concept of culturally different thinking patterns or intellectual styles, which has some analytical potential, was first introduced to international marketing by Usunier (1991) and later taken over by de Mooij (1998).

In summary, many of the websites compared tended to be both standardized and dominated by a general Low-context communication style—regardless of the preferred communication styles in the related countries. This may be surprising with regard to many local website designers and native speakers as translators who, as members of their specific culture, quite automatically might provide for cultural adaptation of the various design elements mentioned above. An exception to this general finding, however, is consumer brand websites (non-durable products in particular), which tend to be more culturally adapted to the relevant degree of High-context communication in some related countries. One explanation for this exception might be that a large majority of consumers addressed worldwide do not, in contrast to businesspeople, belong to the "information elite" mentioned above. Therefore, they may be less open to being addressed in culturally inappropriate Low-context styles. In addition, non-durable consumer products are generally characterized by low involvement with respect to buying decisions (e.g., on low price); therefore less explicit and rational information may need to be provided. As a preliminary result, the differing degree of website adaptation with respect to the represented product or business category is shown in Figure 13. According to this, websites for industrial goods and New Economy products in general show the strongest standardization tendencies, as if they were still to be regarded as "culture-free" products.

To increase website acceptability and Internet consumption worldwide, a higher degree of website adaptation in general seems to be necessary. The operational cultural design criteria and elements that have been described above provide a basis or starting point from which a higher degree of culturally appropriate website design can be approached. Since the technical conditions of hardware, software, and infrastructure (including data transfer) continuously become more and more efficient, there are fewer and fewer obstacles to offering more entertaining and transformational content appeals, including many multimedia presentations, to meet the preferences of High-context cultures beyond their small "information elites." With respect to High-context cultures' media consumption preferences, the appropriate website design may be simply summarized as "less print, more TV style."

These general findings, of course, represent only preliminary tendencies. One of the main reasons for this is the high complexity and contingency of influences on website design beyond cultural values and communication styles (see Figure 6). In addition, research on randomly chosen samples out of millions of websites can hardly provide representative, precise results. (Okazaki & Alonso, 2003, report that over 36 million domains for commercial websites had been created by 2001.) Moreover, the design of websites is subject to continuous change over time because of the highly dynamic nature of the medium, which, for example, transforms strongly standardized websites into strongly localized ones and vice versa. This phenomenon of change, however, might be interpreted as an evolutionary stage of the Internet as a new medium of communication that is striving for acceptance in many different cultures and that is following the principle of trial and error in its interface design.

Conclusion

In spite of their ostensible universality and important role as engines of globalization, the Internet and its Web neither eliminate cultural differences nor are they culture-free products. Cross-cultural marketing and advertising research provides some useful findings that can demonstrate the broad influence of culture on this new medium.

Though global Internet diffusion is still increasing, there are enormous and persistent disparities in worldwide Internet consumption. This phenomenon is similar to the persistently differing consumption of traditional media (like newspapers and TV) across different countries, which can be interpreted as a result of culture-specific communication preferences. Various hard factors (political, economical or technical) also have influence on media consumption, including the Internet, but in highly developed countries, they generally are less significant. In contrast, Hofstede's cultural dimensions of Uncertainty Avoidance and Individualism, and the culturally different communication styles described by Hall's concept of Low- and High-context, represent important cultural influences on Internet consumption.

Both Hofstede's and Hall's cultural categories have been integrated into the contributions from cross-cultural marketing and advertising research to a large degree. The reasons for their popularity and the limitations and utilities of their implied concepts of culture are worth considering. The basis of Functionalism and the focus on nations make Hall's and Hofstede's models comfortable and operational for international marketing research. Hofstede's pragmatic quantitative positionings of countries as national cultures seem to be the most favored, but also the most static, generalizing, and simplifying approach. Hall's model is less operational but more flexible. Both models have high relevance only for international contexts of unconscious (national-) cultural differences; they provide a useful first basis for research into cultural backgrounds.

The significance of these cultural categories—including High- and Low-context communication in particular—has also been demonstrated by many cross-cultural comparisons of preferred advertising styles and creative strategies. Cross-cultural advertising messages reveal remarkable similarities with cross-cultural website communication. Consequently, some culturally relevant website design criteria can be transferred or derived from the standardization/adaptation discussion of international marketing and advertising research. Following that, eight culture-related, structural website design criteria and some basic external conditions for their implementation were introduced, and additional design aspects were discussed.

Comparisons of websites of various global companies and brands in different countries according to these structural design criteria show a frequent lack of adaptation and appropriateness to specific cultural communication styles. Although an increasing number of websites reveal some cultural adaptation on a moderate degree, too many websites are still characterized by a dominant Low-context style (e.g., rational, text-heavy, deeply structured contents), which is preferred worldwide by only relatively few "information elites." These websites are strongly standardized and globally dispersed, regardless of the prevailing High-context communication preferences (e.g., for transformational, visual-heavy, less structured contents) in many target countries.

Accordingly, with respect to the disparities in Internet consumption worldwide, a country's low rate of Internet consumption could also be considered as a general indicator of an overly low quota of culturally appropriate websites. Like the worldwide disparity in print media consumption, a significant Digital Divide will be normal unless the design of websites—and perhaps of the Web in general—completely meets the communication preferences of many High-context cultures. These may be summed up in a very simple way: "less Print, more TV style." Many global consumer brand (B2C) websites, however, reveal a higher degree of adaptation to existing High-context communication preferences, which might be considered as the beginning of a necessary further cultural adaptation of the Web.

These findings seem to verify the cultural relevance of the website design criteria introduced; thus they may be taken as an operational basis for more intensive cultural adaptations of the Web. Since technical conditions are becoming increasingly favorable to such adaptations, this could make the Internet a truly world-wide medium in the future. However, the present discussion is based on a small, probably not truly representative sample of websites (out of many millions), so its conclusions should be regarded as preliminary. This highly complex subject matter richly deserves further investigation.

References

Anonymous. (2000a). Weltweite Suche nach den Vielsurfern. Media & Marketing, 11, 94-96.

Anonymous. (2000b). Internet-Nutzer sind aktiver. Marketing Journal, 5, 267.

Baalbaki, I. B., & Malhotra, N. K. (1993). Marketing management bases for international market segmentation: An alternate look at the standardization/customization debate. International Marketing Review, 10 (1), 19-42.

Cateora, P. R., & Graham, J. (1999). International Marketing, 10th ed. Boston: Irwin/McGraw-Hill.

Choi, S. M., & la Ferle, C. (2004). Convergence across American and Korean young adults: Socialisation variables indicate the verdict is still out. International Journal of Advertising, 23 (4), 479-506.

la Ferle, C., Edwards, S. M., & Yutaka, M. (2002). Internet diffusion in Japan: Cultural considerations. Journal of Advertising Research, 42 (2), 65-79.

Galtung, J. (1981). Structure, culture and intellectual style: An essay comparing saxonic, teutonic, gallic and nipponic approaches. Social Science Information, 20 (6), 817-856.

Geyskens, I., Deleersnyder, B., Gielens, K., & Dekimpe, M. G. (2002). How cannibalistic is the Internet? A study of the newspaper industry in the UK and The Netherlands. International Journal of Research in Marketing, 19 (4), 337-348.

Hall, E. T. (1977). Beyond Culture. New York: Anchor-Doubleday.

Hall, E. T., & Hall, M.R. (1990). Understanding Cultural Differences: Germans, French and Americans. Yarmouth: Intercultural Press.

Hammond, K., Turner, P., & Bain, M. (2000). Internet users versus non-users: Drivers in Internet uptake. International Journal of Advertising, 19 (5), 665-680.

Hermeking, M. (2000). The cultural influence on international product development. Global Player, International Management and Business Culture, 7 (5), 8-13.

Hermeking, M. (2001). Kulturen und Technik: Techniktransfer als Arbeitsfeld der Interkulturellen Kommunikation, Beispiele aus der arabischen, russischen und lateinamerikanischen Region. Münster: Waxmann.

Hofstede, G. (1991). Cultures and Organizations: Software of the Mind. New York: McGraw-Hill.

Holden, N. (2004). Why marketers need a new concept of culture for the global knowledge economy. International Marketing Review, 21 (6), 563-572.

Jacobs, L., Keown, C., Worthley, R., & Ghymn, K. I. (1991). Cross-cultural colour comparisons: Global marketers beware! International Marketing Review, 8 (3), 21-30.

Ju-Pak, K. H. (1999). Content dimensions of web advertising: A cross-cultural comparison. International Journal of Advertising, 18 (2), 207-232.

Knight, G. A., & Calantone, R. J. (2000). A flexible model of consumer country-of-origin perceptions: A cross-cultural investigation. International Marketing Review, 17 (2), 127-145.

Lengert, J. (2000). E-Business: A key-business? Global Player, International Management and Business Culture, 7 (5), 4-7.

Leonardi, P. M. (2002). Cultural variability in web interface design: Communicating U.S. Hispanic cultural values on the internet. In F. Sudweeks & C. Ess (Eds.), Proceedings of the Third International Conference on Cultural Attitudes Towards Technology and Communication (pp. 297-315). Murdoch, Western Australia: Murdoch University, Australia,

Levitt, T. (1983). The globalization of markets. Harvard Business Review, 61 (3), 92-102.

Lynch, P. D., Kent, R. J., & Srinivasan, S. S. (2001). The global internet shopper: Evidence from shopping tasks in twelve countries. Journal of Advertising Research, 41 (3), 15-24.

Madden, T. J., Hewett, K., & Roth, M. S. (2000). Managing images in different cultures: A cross-national study of colour meanings and preferences. Journal of International Marketing, 8 (4), 90-107.

Marcus, A., Baumgartner, V. J., & Chen, E. (2003). User-Interface design vs. culture. In V. Evers, K. Röse et al. (Eds.), Designing for Global Markets 5, Proceedings of the Fifth IWIPS Workshop (pp. 67-78). Kaiserlautern: Product & Systems Int.

Medienlandschaft Westeuropa. (2005). W&V Compact: Fakten und Zahlen aus Marketing, Werbung, Medien, 2, 6-22.

Meffert, H., & Bolz, J. (1994). Internationales Marketing Management, 2. völlig überarbeitete Auflage. Stuttgart: Kohlhammer Verlag.

de Mooij, M. K. (1998). Global Marketing and Advertising: Understanding Cultural Paradoxes. Thousand Oaks, CA: Sage.

de Mooij, M. K. (2000). The future is predictable for international marketers: Converging incomes lead to diverging consumer behaviour. International Marketing Review, 17 (2), 103-113.

de Mooij, M. K. (2004). Consumer Behavior and Culture: Consequences for Global Marketing and Advertising. Thousand Oaks, CA: Sage.

Moosmüller, A. (2000). Die Schwierigkeit mit dem Kulturbegriff in der Interkulturellen Kommunikation. In R. Alsheimer, A. Moosmüller and K. Roth (Eds.), Lokale Kulturen in einer Globalisierenden Welt (pp. 15-31). Münster: Waxmann.

Morrison, D. E., & Firmstone, J. (2000). The social function of trust and implications for e-commerce. International Journal of Advertising, 19 (5), 599-623.

Mueller, B. (1996). International Advertising: Communicating Across Cultures. Belmont, CA: Wadsworth.

Mueller, B. (2004). Dynamics of International Advertising: Theoretical and Practical Perspectives. New York: Peter Lang.

Müller, S., & Gelbrich, K. (2004). Interkulturelles Marketing. München: Vahlen.

Okazaki, S. (2005). Searching the web for global brands: How American brands standardise their web sites in Europe. European Journal of Marketing, 39 (1), 87-109.

Okazaki, S., & Alonso, J. (2003). Right messages for the right site: On-line creative strategies by Japanese multinational corporations. Journal of Marketing Communications, 9 (4), 221-239.

Park, C., & Jun, J. K. (2003). A cross-cultural comparison of internet buying behaviour: Effects of internet usage, perceived risks, and innovativeness. International Marketing Review, 20 (5), 534-553.

Rhee, K. Y., & Kim, W. B. (2004). The adoption and use of the internet in South Korea. Journal of Computer Mediated Communication, 9 (4), http://jcmc.indiana.edu/vol9/issue4/rhee.html

Rogers, E., & Hart, B. (1998). Edward T. Hall and the origins of the field of intercultural communication. Paper presented at the National Communication Association, International and Intercultural Communication Division, New York, November 21-24, 1998.

Roth, J., & Roth, K. (2001). Interkulturelle Kommunikation. In R. W. Brednich (Ed.), Grundriss der Volkskunde: Einführung in die Forschungsfelder der Europäischen Ethnologie (pp. 273-305). Berlin: Dietrich Reimer Verlag.

Singh, N., & Baack, D.W. (2004). Web site adaptation: A cross-cultural comparison of U.S. and Mexican web sites. Journal of Computer Mediated Communication, 9 (4), http://jcmc.indiana.edu/vol9/issue4/singh_baack.html

Steenkamp, J. B. E., ter Hofstede, F., & Wedel, M. (1999). A cross-national investigation into the individual and national cultural antecedents of consumer innovativeness. Journal of Marketing, 63 (2), 55-69.

Tharp, M. C. (2001). Marketing and Consumer Identity in Multicultural America. Thousand Oaks, CA: Sage.

Usunier, J. C. (1991). International Marketing: A Cultural Approach. New York: Prentice Hall.

Würtz, E. (2004). Intercultural communication on websites: An analysis of visual communication of high- and low-context cultures. In F. Sudweeks & C. Ess (Eds.), Proceedings of the Fourth International Conference on Cultural Attitudes towards Technology and Communication (pp. 109-122). Murdoch, Western Australia: Murdoch University.

Yeniyurt, S., & Townsend, J. D. (2003). Does culture explain acceptance of new products in a country? An empirical investigation. International Marketing Review, 20 (4), 377-396.

About the Author

Marc Hermeking teaches and researches as a member of the faculty of the Institut für Interkulturelle Kommunikation, with a specific focus on intercultural marketing (Seminar für Interkulturelles Marketing), Ludwig-Maximilians-Universtaet Muenchen.
Address: Oettingenstr. 67, 80538 Muenchen, Germany