Internet diffusion in Japan: cultural considerations.
THE INTERNET currently reaches 426 million consumers worldwide (Pastore, 2001) and is projected to more than double in reach to 1 billion by 2005 (Iconocast.com, 2001). The internet as a medium has experienced unprecedented growth in the United States and is now also increasing around the world. In fact, North America currently accounts for only 43 percent of all internet users, followed by Western Europe (25 percent) and Asia (almost 21 percent). Eastern Europe, Latin America, the Middle East, and Africa account for just under 11 percent of total internet use combined. Although North American consumers are the largest group of online consumers, accounting for 79 percent of total advertising expenditures in 2001 and remaining an estimated healthy 70 percent by 2005 (Halprin, 2000), they are projected to account for only 30 percent of total online users by 2005 (Lawrence, 2000). Therefore, understanding internet adoption across different countries is crucial as companies continue to increase their online marketing efforts with the goal of building consumer relationships and securing new shopping channels online.
The United States currently reports the greatest number of citizens online with 135 million, followed by Japan with 34 million and the United Kingdom with 14 million (Computer Industry Almanac, 2001). However, with respect to internet penetration by country the Computer Industry Almanac reveals a somewhat different picture of internet use. Sweden and Canada have the greatest numbers of internet users as a percentage of their populations with 49.6 percent and 48.7 percent, respectively. The United States falls to third place with 48.4 percent of its population online, while Japan falls to ninth place with only 26.7 percent of its population online and is behind South Korea, Australia, the Netherlands, Taiwan, and the United Kingdom. Similarly, 7,800 internet service providers were estimated to exist in the United States in 2000, compared with only 73 in Japan for the same time period (The World Factbook, 2001). Therefore, we must be careful of making quick assumptions regarding internet user rates based only on the sheer number of users in a country. Even more important for marketers is the need to understand the factors that might account for these differences in internet adoption across countries. This point is particularly true for countries like the United States and Japan, where the two countries are arguably equally developed and economically strong and yet have very different internet penetration rates.
Toward an understanding of differences in adoption rates across countries, two projects were undertaken. The first project sought to gather current information on internet use and users in Japan in an effort to compare data with U.S. numbers. The second project focused on potential cultural factors that may help to explain adoption differences found between the United States and Japan. Toward this goal and to strengthen the generalizability of the results, cultural variables and the number of internet users were examined for the two countries of interest as well as for 48 other countries. Theoretical support is drawn from both the diffusion of innovation literature, as well as from research using Hofstede's dimensions of culture. Implications for marketers extend beyond the two cultures examined, as cultural differences were shown to affect adoption of the internet in markets around the world.
INTERNET ADVERTISING AND USER PROFILES IN THE UNITED STATES AND JAPAN--PROJECT 1
The internet is changing the way advertisers present, sell, and communicate with consumers. Currently a variety of practices are being used to reach consumers. Banner ads and interstitials run akin to more traditional forms of advertising, while websites range from being reminiscent of product brochures to full-fledged showrooms. These "showrooms" provide the ability for two-way communication between marketers and consumers as well as offer a channel for consumers to make online purchases. The internet by its very nature is a global communication channel, with the potential to reach consumers anywhere in the world. For all of these reasons, the internet is a medium that marketers of varied products and services are rushing to embrace with multiple techniques.
Yet, in order for the internet to be an effective medium in the United States or abroad, consumers must have both the desire and ability to access computers with internet capabilities. Although the United States is excelling in terms of penetration rates and e-commerce sales, concern still exists regarding people and countries with differing socioeconomic and cultural backgrounds that either cannot access the internet or who choose to avoid the online experience. The differences in internet penetration and user profiles between the United States and Japan, two similar countries economically, provide evidence of the need for more inquiry into this area.
In order to compare the United States and Japan, it was necessary to obtain a current picture of Japanese internet penetration rates and user profiles. Similar up-to-date data were available from secondary sources for the United States through Mediamark Research (Lake, 2000). The Japanese data selected for comparison was collected by Dentsu, Tokyo using a random sample of 1,000 Japanese consumers to determine the penetration rate of internet use in Japan. A second quota sample using multiple areas in Japan identified 400 internet users using telephone screenings. Each user was then mailed a self-administered questionnaire asking about internet use and demographic information.
The data shown on the right side of Table 1 indicate that about 27 million Japanese adults out of an adult population of 108 million were online as of April 2000. This is in contrast to a year earlier when the Computer Industry Almanac reported only 18 million Japanese online in 1999. Perhaps the gap between internet penetration in Japan and the United States may be quickly narrowing with almost 52 percent of Japanese respondents reporting their first internet experience occurring within the previous year.
In the United States, the diversity found among internet users is rapidly increasing (see Table 1). No longer is the internet in the United States the exclusive playground of young white males that it was in 1996. In fact, online users in the United States are becoming older, more likely to be female, and less educated (Lake, 2000). Internet surfers under 34 years of age now account for only 38 percent of all Americans online compared with 48 percent in 1996. Women, once accounting for 42 percent of the U.S. online population, now represent 50 percent of online adults. College graduates now account for only 38 percent of online users, down from 49 percent of the online population. However, the 42 percent of Americans online that reported incomes over $75,000 fell only marginally to 40 percent in 2000. Online consumers in the United States still tend to have relatively high incomes compared to those not online.
Overall, the data reported in Table 1 seem to indicate that the diversification of online users witnessed in the United States is also well underway in Japan. Males (52 percent) are slightly more likely to use the internet than females (48 percent), but there is not a great divide. Similar to the United States, internet use in Japan is more likely to be among the young and the more educated, but again there appears to be diversity among the categories. Interestingly, internet usage rises with income in the United States, but the relationship is not so clear in Japan. According to the data, Japanese internet users are more wealthy than Japanese who are not online. However, the gap between the two groups in Japan is not as large as in the United States, possibly indicating that a broader spectrum of people is online in Japan. High-user indices in Japan can be seen for high-income levels but also across the $30,000 to $40,000 income bracket.
The two countries' users also differ in where the internet is accessed. For example, 50 percent of Japanese users report accessing the internet only at work, 23 percent at home, and 27 percent report accessing the internet in both locations (Media Metrix, 2000). In contrast, U.S. consumers are more likely to access the internet from home (53 percent), followed by work (24 percent), and in both locations (23 percent) (Lake, 2000). So, while internet penetration is growing in both countries, the rate at which the internet is penetrating the Japanese market is slower, and the pattern of adoption by various groups is somewhat different. This observation is by far the most interesting and the reason for examining the factors that may help explain how two countries, arguably similar in terms of industrialization and economic development, have very different diffusion patterns of the internet. In order to examine this issue more closely and better understand the underlying factors, a review of the diffusion of innovation literature is necessary.
Diffusion of the internet
The diffusion of innovation theory explains how adoption takes place within a social system (Rogers, 1983). Diffusion models are often used in the consumer behavior literature to describe the adoption of product innovations within a population. Product adoption has generally been understood to occur as a contagion model, where the number of adopters increasingly interacts with nonadopters in a spreading of the innovation. Finally, a saturation point occurs when there are more adopters than potential adopters and the rate of diffusion begins to decline.
According to Rogers (1983) the adoption rate of an innovation is influenced by: (1) characteristics of the innovation itself; (2) the communication channels used to communicate the benefits of the innovation; (3) the time elapsed since the introduction of the innovation; and (4) the social system in which the innovation is to diffuse. The greater the innovation, the longer it will take to be adopted. The reach of the communication channels available to inform potential adopters will also limit the rate of diffusion. The distribution of adopters is expected to take the form of a normal curve and thus early adoption will be slow but rapidly increase, only to slow down and taper off toward the end of the cycle. The rate of diffusion is also dependent upon the homogeneity of the social system in which the innovation is to diffuse. The more similar potential adopters, the faster diffusion is said to occur.
In an effort to predict adoption rates for the diffusion of innovation, traditional methods have used internal influence models to predict adoption rates, where the number of previous adopters influences the rate of further adoption. However, Rai, Ravichandran, and Samaddar (1998) were able to show that these models are not the best for predicting the diffusion of the internet. The models assume a homogeneous population and ignore external factors that may influence the process such as government involvement and commercial use of the medium. Potential internet adopters are anything but homogeneous, as witnessed by changes in the consistency of users over the internet's brief history. Early users of the internet included scientists sharing information from which the base expanded to include academic researchers and now the many commercial users and everyday consumers. Internationally, potential internet adopters are also very heterogeneous in their needs and uses of the internet. Just as consumers may have similar needs around the globe, how they express their needs can differ from country to country, as can the needs fulfilled by the internet. With respect to the influence of external factors on adoption rates, such as government involvement and regulations, these too differ by country and can further help or hinder the internet diffusion process.
In fact, Lee (1990) found that the rate of adoption of televisions across 70 countries could be predicted by several external factors of a country such as a country's "GNP per capita, literacy rate, number of scientists and engineers, and the proportion of manufacturing and service sectors over the total GNP." The results of his study placed Japan and the United States within the Innovator category, while the remaining 68 countries were dispersed across the other categories of Early Adopters, Early Majority, Late Majority, and Laggards. In other words, in terms of innovation for television adoption, the United States and Japan were classified similarly as Innovators. However, the factors Lee used are still insufficient to help explain the differences that exist between Japan and the United States on rates of adoption of the internet.
For all intents and purposes, Japan and the United States are relatively equal with respect to the four variables used in Lee's (1990) study to determine the innovative nature of a country. According to data compiled from the PRS Group (2000) and the World Factbook (2000), Japan's GDP per capita income ($30,045) was similar to the United States ($31,455) in 1999, as were literacy rates for the two countries (99 percent and 98 percent, respectively) and work force distribution rates between agriculture, industry, and services. Therefore, other factors must be influencing the adoption of the internet beyond economic and demographic indicators.
Culture and diffusion of the internet
Currently, very little literature exists that draws a direct relationship between the cultural aspects of a country and the rate of internet adoption. However, borrowing from the literature on new-product diffusion rates, various cultural dimensions associated with a variety of countries have been found to influence the process. For example, Takada and Jain (1991) found that high- versus low-context cultures can impact the rate of new-product diffusion, with high-context cultures typically experiencing a faster rate of adoption. High-context cultures tend to be more homogeneous and have a greater group orientation. All things being equal, once a product is accepted in a high-context culture such as Japan, it is said to spread rapidly because people in these cultures want to maintain an image that is similar to the group (Inoue, 1996). This phenomenon may be driving what we are just beginning to see in Japanese consumers' entry to online channels of communication.
In another study, Tansuhaj, Gentry, John, Manzer, and Cho (1991) examined how cultural beliefs toward the concept of fate can affect one's willingness to try new products. Asian cultures tend to believe in fate more than western cultures, which could contribute to explaining the differences in internet adoption rates between the United States and Japan. Tansuhaj et al. (1991) found fatalism to be inversely related to people's willingness to try new products.
Finally, many researchers have looked to Hofstede's (1984; 1997) five cultural dimensions (Individualism /Collectivism, Masculinity/Femininity, High/Low Uncertainty Avoidance, Strong /Weak Power Distance, and Long-Term Orientation) to search for underlying factors impacting the general diffusion process of new innovations (de Mooij, 1998; Herbig and Miller, 1991; Samiee, 1998; Tellefsen and Takada, 1999). In particular, the amount of individualism versus collectivism and the level of uncertainty avoidance have been found to affect diffusion rates (de Mooij, 1998; Herbig and Miller, 1991). The dimension scores usually range from 0 to 100, with a higher number representing stronger characteristics on one dimension relative to other countries. Table 2 presents scores from Hofstede's work on the cultural dimensions for the United States and Japan. Long-Term Orientation (also known as Confucian Dynamism) was not examined in this study due to the limited number of countries it has been applied to in previous studies.
Each dimension can help contribute to our understanding of innovative behavior and, more specifically, adoption of the internet. In fact, de Mooij (1998) suggests that the most inventive and innovative cultures in the world exhibit a combination of the following four cultural dimensions: individualistic, weak uncertainty avoidance, small power distance, and masculine. Unlike the similarities found on economic indicators between the United States and Japan, these two countries reflect very different dimensions on many of Hofstede's measures.
In light of the current differences between the United States and Japan in internet adoption rates, as well as diverse differences in cultural dimensions and assuming that economic and infrastructure variables do not explain all of the variation, we would expect the following cultural dimensions to also contribute to the process. The following propositions are not meant to imply that factors such as economic variables or technological infrastructure will not impact the internet diffusion process. Structural issues identified by Samiee (1998) such as: PC ownership, computer literacy, network access, language, culture, and the regulatory environment are all likely to influence internet use. We also do not intend to say that each of the cultural dimensions operates independently of one another. However, all things being equal, these dimensions should play a compelling role in the rate of internet adoption across a variety of countries.
The Individualism versus Collectivism dimension refers to how much a culture stresses the importance of the individual over the welfare of the group. The United States is considered to be one of the most individualistic-oriented societies (91) and scores much higher relative to Japan (46), a country reflective of collectivist cultures. According to de Mooij (1998), collectivist cultures put more emphasis on building and maintaining relationships with people within their social structure or in-group than individualistic cultures. Obligation, loyalty, and group harmony come before individual aspirations or goals and may explain why the rate of adopting the internet in Japan has been slower than in the United States.
Herbig and Miller (1991) suggest that the strong need for group cohesion and harmony of the Japanese culture is exactly the reason why Japan is very strong in what are known as process innovations, but weak in radical innovations. Teamwork and an eye for detail allow the Japanese to create "higher productivity, lower costs, and higher quality for an already existing good or service" (Herbig and Miller, 1991). In contrast, the authors suggest that the individual orientation of Americans helps explain their ability to excel in radical innovations or innovations that "require the establishment of new behavioral patterns with no established precedents: computers, photocopying ..." Being independent and having a desire to be unique are qualities conducive to entrepreneurial behavior and embracing new technology.
De Mooij (1998) has suggested that because the Japanese strive for collective harmony, no one wants to stand out by trying something unknown. Thus, until people within one's in-group adopt an innovation, usually through the opinion leaders of the group, the rate of diffusion will be slow. But she goes on to add that once a product is adopted in a collectivist culture, the rate of diffusion is much faster than in individualist cultures. Therefore, we may be correct in stating that one contributing factor to the differences in adoption of computers and the internet between Americans and Japanese stems from the cultural differences on the individualism versus collectivism dimension.
Proposition 1: The more individualistic a society, the faster the rate of internet diffusion.
Uncertainty Avoidance can be defined as "the extent to which the members of a culture feel threatened by uncertain or unknown situations" (Hofstede, 1997). Certainly the introduction of a new product or a new type of technology would create ambiguity and this would be higher for cultures strong on this dimension. The United States has been found to be relatively weak on this dimension (46), while Japan is a culture of strong uncertainty avoidance (92). Many researchers have argued that cultures high on uncertainty avoidance are less likely to be early adopters of any innovation (de Mooij, 1998; Samiee, 1998). According to Tellefsen and Takada (1999), Jan and Maesincee (1995) found that new-product sales were lower for countries high on uncertainty avoidance. The authors also suggest that the influence of mass media to encourage product adoption may be less in high uncertainty avoidance cultures. People in these cultures rely more on interpersonal information sources to guide behavior, and thus rates of diffusion tend to be slower. Again the literature helps support the potential cultural explanation for some of the existing differences found between American and Japanese adoption rates of the internet.
Proposition 2: The higher the uncertainty avoidance that exists within a society, the slower the rate of internet diffusion.
With respect to the Power Distance dimension, the United States demonstrates lower levels of power distance (40) relative to Japan (54). Power distance can be thought of as the types of dependent relationships existing in a culture and how willing members of a society are to accept a hierarchical structure where people expect to be subordinates and/or to have subordinates. According to Herbig and Miller (1991), cultures that exhibit large power distance will be less innovative. The authors suggest that innovation is limited because people in high power distance cultures are encouraged to respect authority, follow directions, and avoid standing out through original thinking. However, this is not to say that one style is better than another. In fact Herbig and Miller (1991) argue that both countries would benefit from learning how to better implement the strengths of the other, in terms of process and radical innovations.
Proposition 3: The larger the power distance existing in a society, the slower the rate of internet diffusion.
The final dimension of the original four, Masculinity versus Femininity, can be defined as the extent of competitiveness and aggression in a culture relative to caring for others and quality of life concerns (de Mooij, 1998). This dimension is also very related to societal expectations of gender roles. According to Hofstede (1997), masculine cultures tend to have a clear distinction between males, who are expected to be assertive with a focus on material possessions, and females, who should be more modest, tender, and driven to improve the quality of life. In contrast, feminine cultures do not have such rigid gender roles, and both sexes are concerned with quality of life, caring for others, and being relatively modest. Although considered a "masculine" culture in relation to some other countries examined by Hofstede, the United States (62) scores considerably lower on this dimension relative to Japan (95). In fact, Japan is the most masculine society of the countries Hofstede examined.
Similar to the proposition for power distance, it is believed that this dimension will have a negative relationship with internet adoption. It is argued that the relatively extreme Japanese need for achievement, success, and "saving face" contributes to slow adoption rates of unknown innovations until it becomes clear that those innovations will be successful. At the point of known success, the Japanese may then step in to embrace and subsequently modify and improve the product or technique (Herbig and Miller, 1991).
Proposition 4: The more masculine a society, the slower the rate of internet diffusion.
TEST OF PROPOSITIONS AND RESULTS--PROJECT 2
In order to examine the proposed relationships, time-series data was gathered on the number of internet users and number of internet hosts by country for the 50 countries examined by Hofstede and others. The data consisted of 1995 and 1998 data which came from the Yearbook of Statistics (2000). The year 1995 was chosen as it was the first full year a commercially available graphical user interface for the web was marketed (Netscape Communications, 1994). The year 1998 was used as it was the most recent data available for all 50 countries. The data were used in two ways: (1) the most recent figures concerning the number of internet users by country were examined, and (2) the growth rate of internet users was used. The growth rate of internet users was calculated as the change in the number of internet users per capita from 1995 (the first year recorded) to 1998 (the most recent available by country from a single source).
A measure of internet hosts was also used to verify the findings concerning internet users. Internet hosts are computers designed to allow for access to the internet and are often used to assess the growth of the internet. Change in the number of internet hosts per capita was calculated for each country as the number of hosts in 1998 minus the number of hosts in 1995. In order to examine each of the propositions, we contrasted Hofstede's dimensions with internet usage in both 1995 and 1998 for each of the 50 countries.
As can be seen in Figure 1, the countries were first ranked by their scores on Hofstede's Individualism/Collectivism dimension. The United States is the most individualistic culture scoring 91, with Guatemala being the most collectivist, scoring 8. Figure 1 shows the relationship between internet usage per capita and the cultural value scores. As is fairly evident, countries ranking high on individualism had a greater rate of internet adoption compared with countries that were less individualistic both in 1995 and 1998. To examine this relationship further, growth in internet usage was calculated for each country as the difference between internet usage in 1998 and 1995. Growth was then correlated with each country's score on the individualism/collectivism dimension and the two were found to be significantly related (r = .63, p < .001).
[FIGURE 1 OMITTED]
Similar analyses were conducted for the remaining three dimensions. Figure 2 presents the relationship between power distance and internet adoption. Malaysia was ranked as having the most rigidity between superiors and subordinates, with a score of 104. Austria was ranked as having the least boundaries with a score of 11. The figure seems to indicate that internet adoption increases as power distance decreases. That is, the more exchange between peoples of different statuses, the greater the level of internet adoption. To examine this finding further, growth in internet usage was again calculated and then correlated with the power distance scores. Again a significant relationship was found, with power distance being negatively correlated with change in internet adoption between 1995 and 1998 (r = -.61, p < .001).
Figure 3 shows us the relationship between uncertainty avoidance and internet adoption. The degree to which a culture tends to avoid uncertainty should be positively related with the risk of adopting a new product. Greece was the most risk averse, scoring 112, whereas Jamaica was the least risk averse, scoring 13. Although the pattern is slightly more obfuscated, the relationship can still be seen. Countries that tend to avoid uncertainty adopted the internet less than did countries that were more inclined to embrace uncertainty. A negative relationship between the growth in internet use and level of cultural uncertainty avoidance was again significant (r = -.40, p < .01).
Finally, the ranking of a culture as either masculine or feminine was examined. As shown in Figure 4, Japan was ranked as the most masculine country with a score of 95 and Sweden the least masculine with a score of 5. Of the four cultural dimensions, it is hardest to see the relationship between masculinity/femininity and that of internet adoption. A correlation of masculinity with internet adoption resulted in a nonsignificant negative relationship (r = -.28, p = .07). Therefore, more research is necessary to better understand the relationship between this dimension and adoption of the internet. Perhaps there is an indirect relationship that is dependent on the configuration of the other dimensions.
[FIGURE 4 OMITTED]
To examine the shared variance associated with the cultural dimensions, stepwise regressions were conducted on the number of current internet users by country and the change in the number of internet users from 1995 to 1998. As can be seen in Table 3, Individualism was the first variable to be entered in the equation. This variable accounted for 43 percent of the variance in explaining the current number of internet users and 38 percent of the variance in the change in the number of internet users from 1995 to 1998, F(1,44) = 34.4, p < .001 and F(1,40) = 26.1, p < .001, respectively. Masculinity/Femininity added significantly to the prediction of both dependent variables, increasing the explained variance to 48 percent for current internet users, F(2,43) = 21.8, p < .001 and 44 percent for the change in the number of internet users, F(2,39) = 17.4, p < .001. Individualism was positively related to both the current number of internet users ([beta] = .68) and the change in the number of internet users ([beta] = .63). Conversely, masculinity was negatively related to both the current number of users ([beta] = -.26) and the change in the number of internet users ([beta] = -.28). Power distance and uncertainty avoidance did not significantly add to the explanatory power of the models for either dependent measure, p > .05.
To further examine the impact of the cultural dimensions on internet adoption, internet hosts were also examined. Following the procedures used for internet users, the cultural dimensions were regressed (stepwise) on change in the number of internet hosts from 1995 to 1998. As can be seen in Table 4, Individualism was again the first variable that entered in the equation, accounting for 45 percent of the variance In explaining host growth F (1,45) = 37.0, p < .001. Masculinity/Femininity added significantly to the prediction of host growth increasing the explained variance to 51 percent, F(2,44) = 22.7, p < .001. Similar to the findings concerning the number of internet users, individualism was positively related to the change in the number of hosts ([beta] = .68), while masculinity was negatively related to change In the number of hosts ([beta] = -.24). Power distance and uncertainty avoidance did not significantly add to the explanatory power of the model, p > .05.
The goal of this paper was to establish the importance of understanding cultural influences when viewing the diffusion process of the internet across countries. Knowledge in this area is important for advertisers who are currently using or considering the use of the internet in their marketing strategies. In particular, this information could impact decisions regarding promotional mix choices and message design. If culture influences consumers' abilities and desires for online communication then it becomes crucial for advertisers to understand the factors that might inhibit or enhance the rate and/or pattern of internet adoption.
Data collected on internet penetration and user profiles for the Japanese market demonstrate different rates of internet diffusion and patterns of adoption from those found in the United States. The percentage of online internet users is considerably less in Japan than the level experienced in the United States. And although economic indicators, literacy rates, and the ability to create the necessary infrastructure are often the culprits, Japan and the United States share relatively similar characteristics on these elements. However, the two countries are quite different on cultural dimensions, and these appear to explain some of the variance in internet penetration and patterns of adoption. These differences may further help to explain internet penetration across other countries and across future innovations. Therefore, our understanding of the diffusion of the internet can be applied to (1) understanding or predicting the process in other countries and (2) encouraging the adoption of this new medium (and other similar innovations) by consumers at different stages of the diffusion process through marketing-mix and message-design decisions.
While reviewing the potential managerial implications, a few limitations of the study should be considered. First, it is not our intent to suggest that economic development or other potential factors do not play a significant role in diffusion rates of the internet. Certainly, a country without a strong telecommunications industry or one facing economic turmoil would have other elements inhibiting internet adoption beyond culture. However, the results of the study suggest that culture should also be considered along with the other economic and demographic variables.
Second, Hofstede's dimensions have been criticized for being overarching as well as for being used outside of their original IBM domain. In a review of articles incorporating Hofstede's dimensions, Sondergaard (1994) described three major negative responses. Some researchers believe that the dimensions are simply a product of the time period in which they were created and therefore are no longer relevant. Others question the generalizability of the dimensions to populations outside of IBM employees. A third set of criticisms concerns the survey measures used to capture values. However, Sondergaard (1994) then reviewed 61 replications of Hofstede's work and concluded that the studies actually confirm the existence of the cultural dimensions across populations and time periods. While this evidence supports Hofstede's work, as do the numerous studies across disciplines that treat the dimensions as a means of explaining cultural influences, the use of these cultural variables remain controversial and care should be taken in interpreting the results.
Third, internet user rates and the number of hosts change daily so caution must be taken in making decisions based on the data presented. There are many research firms specializing in internet marketing information; however, little data on Japanese consumers could be found in English language publications. Even though the data on Japan were collected in 2000, it remains interesting information about a country with the second greatest number of internet users worldwide and is useful for comparing adoption rates and patterns.
While these above limitations are important to consider, our intention was to shed light on the pattern of internet use across different countries to ascertain if culture in fact influences adoption. Through this process, we were also able to compare the demographic profile of Japanese and American internet users to ex amine similarities and differences in the types of people that are online in these different countries. Our findings led us to ask why we were seeing the differences found and provided the impetus for examining Hofstede's dimensions as possible indicators of adoption patterns.
It is hoped that the current study provides variables that will be useful in predicting the adoption of the internet in other countries. Further, the cultural dimensions may be useful in predicting the adoption of future technologies. It would be valuable to followup with a study that examines if similar patterns of adoption could be seen within countries across various products and even product categories. In addition, adding economic and infrastructure data along with the cultural dimensions in a single model could be used to determine the relative importance of each in the adoption of innovations.
Predicting diffusion rates and patterns with cultural variables
Examining Hofstede's cultural dimensions increases our understanding of differences in internet adoption between American and Japanese consumers, as well as provides guidelines in predicting internet diffusion rates and patterns in other countries. In particular, the fact that Japan is a collectivist culture, with high uncertainty avoidance and large power distance characteristics, helps to explain the slower adoption process of computers and the internet in Japan. However, other countries sharing similar cultural configurations may also follow a related pattern of internet adoption.
* Collectivist cultures, such as Japan, often prefer to do things as a group and therefore adoption of the internet should be slower in the initial stages than in a more individualistic culture such as the United States.
* Risk averse cultures, such as Japan, should also be slower because more time is necessary to assess potential pitfalls. Fewer people will be motivated to blindly take an unknown risk without some initial guarantees of the limited risk involved.
* Masculine and hierarchically structured cultures, such as Japan, place a lot of emphasis on success and protecting the image of one's title and/or family name (i.e., saving face) and thus may avoid the internet until its usefulness and the process of how it works can be clearly ascertained.
From the above guidelines we may anticipate that countries such as Brazil or Mexico may also be somewhat slower to adopt the internet due to strong collectivist and risk-averse characteristics. Initial aversion to the internet might further suggest similar patterns for the adoption of other technological innovations such as TiVo and Satellite TV and radio. In practical terms, the use of cultural influences to help understand and predict adoption patterns allows advertisers the opportunity to better prepare their strategies and capitalize on the different patterns of adoption.
Communication strategies for encouraging diffusion
Knowledge of cultural influences on the diffusion process can help marketers to enhance internet adoption by adjusting their marketing and advertising strategies. Tellefsen and Takada (1999) demonstrated how the speed and pattern of new-product sales can be influenced by media availability. They found that societies with higher levels of mass media availability spread new-product information faster. The researchers went on to present methods for selecting media more effectively, given that potential adopters rely on different information sources depending on their stage in the diffusion process. Along similar lines, results from the current study suggest that it may be more important for collectivist cultures to encourage personal selling targeted to group opinion leaders in the early stages of an innovation rollout over heavy mass media advertising. Similarly, for risk-averse and masculine cultures, it may be more effective to initially provide free trials or demonstrations of products to reduce monetary risk and concerns of embarrassment.
Promotional mix decisions
The Japanese most frequently use the internet at work, while Americans more frequently use it at home. This simple finding may help to explain why the internet is not diffusing as fast as could be expected in Japan. In the United States, internet users are encouraging their friends and families to get online as a means of communication or a hobby. However, if most Japanese are using the internet at work, the social aspect of the new medium may not be valued to the same degree. Rather, the Japanese may seek the functional benefits of the internet but mostly for business (not pleasure). Without access to the internet for pleasure, content and e-commerce applications will be slower to flourish. There may be less perceived utility in services that people cannot use or do not feel comfortable using at work. Therefore, managers may want to look at innovative ways to draw Japanese consumers online socially, as well as reduce any perceived risks associated with the medium.
McDonald's is undertaking this very technique in Brazil by providing internet access across several restaurants in an effort to get people familiar with and excited about using the internet (Penteado, 2001). Coincidentally, Brazil is considered a highly risk-averse culture with an index of 76 on Hofstede's (1984) uncertainty avoidance dimension. The company's ultimate goal is to be able to create interest and trust in home internet access, where consumers would eventually order McDonald's online. Other carefully targeted PR tactics would also be beneficial.
Breathe Right nasal strips entered the Japanese market very successfully by receiving free airtime on a top celebrity show. The host of the show and guests not only displayed the package and discussed the product on air but also actually wore the product for a substantial period of time during the program (Cateora and Graham, 1999). This example illustrates how managers can identify opinion leaders within a culture and help to speed the rate of adoption.
Knowledge of cultural influences on internet adoption may also be used in message design to help encourage consumers to go online, as well as to enhance advertising messages in general. Several researchers have examined the use or effectiveness of cultural values in advertising appeals across countries (Albers-Miller and Gelb, 1996; Shao, Raymond, and Taylor 1999; Shavitt and Han, 1994; Yin, 1999; Zhang and Gelb, 1996). For example, Zhang and Gelb (1996) found that culturally congruent appeals resulted in more favorable attitudes. Therefore, countries that are collectivist in nature might respond better to internet appeals using "Join together in something good," whereas individualistic cultures may be more likely to respond to "Be among the first." in an effort to reduce risk, consumers from high uncertainty avoidance cultures may be appealed to using "Talk with someone who owns one," while a low uncertainty avoidance culture might appreciate "If you haven't tried it, you don't know what you're missing." However the dimensions are applied, there are plenty of opportunities to increase the effectiveness of targeted communications by combining cultural variation and diffusion knowledge.
Internet adoption rates and patterns are crucial for advertisers to understand with respect to effective message placement and promotional mix choices, If the desired target audience is not online, the internet becomes an unattractive media selection. If advertisers would like to capitalize on the benefits of two-way interaction online, then promotional mix choices and advertising messages could help encourage online consumer behavior through cultural considerations.
The propositions presented in the current paper not only shed light on the cultural similarities and differences between the United States and Japan but also provide guidelines that can be used to understand the diffusion of the internet in other countries around the world. Beyond traditional economic and demographic indicators, cultural dimensions such as individualism, power distance, uncertainty avoidance, and masculinity were shown to influence internet adoption. In fact, individualism accounted for 43 percent of the variance in explaining user adoption. However, more research is necessary to understand the interaction effects of the various dimensions. Knowledge of cultural influences is beneficial for advertisers in predicting adoption rates and patterns of adoption as well as providing guidance for promotional strategies and message design.
TABLE 1 Online Users in the United States versus Japan All U.S. U.S. Online Adults Users Index Total (millions) 199.4 90.5 100 Sex Male 48 50 104 Female 52 50 97 Age * 18-24 (15-19) 13 17 129 25-34(20-29) 20 23 117 35-44(30-39) 23 26 119 45-54(40-49) 17 21 121 55-64(50-59) 11 9 77 65 and older (60+) 16 4 24 Median 42.7 years 38.9 years Household Income $150k or more (93k+) 4 8 186 $100,000 to $149,999 9 15 175 (75k-93k) $75,000 to $99,999 11 17 152 (56k-75k) $50,000 to $74,999 21 26 127 (37k-56k) $30,000 to $49,999 23 21 90 (<37k) $20,000 to $29,000 13 7 54 <$20,000 19 6 32 Median $45,156 $65,466 Education Post graduate 7 13 178 Bachelor's degree 15 25 164 Attended college 27 35 131 High school grad 33 22 68 Did not graduate H.S. 18 5 26 Occupation Professional, manager 20 36 180 Technical, clerical, sales 19 26 140 Craft, precision production 7 6 87 Retired or self-employed 35 18 51 Other 19 14 74 Race White 83 86 103 Black 12 9 73 Asian 3 3 118 Other 2 2 102 Marital Status Single (and Divorced) 24 28 116 Married 57 61 108 Divorced, other 19 11 84 All Japanese Japanese Adults Online Users Index Total (millions) 108.5 27.0 100 Sex Male 48 58 121 Female 52 42 81 Age * 18-24 (15-19) 7 12 147 25-34(20-29) 17 25 147 35-44(30-39) 16 21 131 45-54(40-49) 16 19 119 55-64(50-59) 18 16 89 65 and older (60+) 28 8 29 Median Household Income $150k or more (93k+) 22 27 123 $100,000 to $149,999 (75k-93k) 15 14 93 $75,000 to $99,999 (56k-75k) 22 22 100 $50,000 to $74,999 (37k-56k) 14 22 157 $30,000 to $49,999 (<37k) 21 15 71 $20,000 to $29,000 <$20,000 Median $70,082 $75,945 Education Post graduate 22 36 164 Bachelor's degree 17 20 118 Attended college 6 10 167 High school grad 48 28 58 Did not graduate H.S. 6 6 100 Occupation Professional, manager 10 12 120 Technical, clerical, sales 31 34 110 Craft, precision production 12 16 133 Retired or self-employed 18 12 67 Other 8 3 38 Race White -- -- -- Black -- -- -- Asian -- -- -- Other -- -- -- Marital Status Single (and Divorced) 37 47 127 Married 63 54 86 Divorced, other -- -- -- * Age ranges in parenthesis are those for Japan. Dentsu, Inc. Proprietary Study 2000 Mediamark Research, Spring 2000 (Lake, 2000) TABLE 2 Hofstede's Five Cultural Dimension Scores Cultural Dimension United States Japan Individualism 91 46 Power Distance 40 54 Uncertainty Avoidance 46 92 Masculinity/Femininity 62 95 Long/Short-Term Orientation * 29 80 * The four original dimensions are based on the results from 50 countries and 3 regions, while the fifth dimension is based on later results from 23 countries. TABLE 3 ANOVA Summary Table: Cultural Dimensions on the Current Number of Internet Users and Change in the Number of Internet Users 1995 to 1998 Model SS df MS F [R.sup.2] [beta] Current Internet Users 1 Individualism 162577.3 1 162577.3 34.4 .43 Residual 207956.5 44 4726.3 Total 370533.8 45 2 Individualism and .68 Masculinity 186703.5 2 93351.7 21.8 .48 -.26 Residual 183830.3 43 4275.1 Total 3370533.8 45 Change in Internet Users 1 Individualism 104943.3 1 104943.3 26.1 .38 Residual 160757.0 40 4018.9 Total 265700.3 41 2 Individualism and .63 Masculinity 125174.7 2 62587.3 17.4 .44 -.28 Residual 140525.6 39 3603.2 Total 265700.3 41 TABLE 4 ANOVA Summary Table: Cultural Dimensions on the Change in the Number of Hosts 1995 to 1998 Model SS df MS F [R.sup.2] [beta] 1 Individualism 6420.9 1 6420.9 37.0 .45 Residual 7801.3 45 173.4 Total 14222.2 46 2 Individualism and .68 Masculinity 7228.8 2 3614.4 22.7 .51 -.24 Residual 6993.5 44 158.9 Total 14222.2 46
ADVERTISING AGE. "Ad Age Dataplace." Advertising Age, August, 2000. Online: http://adage.com/dataplace/topmarkets.
ALBERS-MILLER, N. D., and B. D. GELB. "Business Advertising Appeals as a Mirror of Cultural Dimensions: A Study of Eleven Countries." Journal of Advertising 25, 4 (1996): 57-70.
CATEORA, P., and J. L. GRAHAM. International Marketing (video cassette). Chicago, IL: Irwin McGraw-Hill, 1999.
COMPUTER INDUSTRY ALMANAC, 2001. Online: http://www.c-i-a.com/200103iu.htm.
DE MOOIJ, M. Global Marketing and Advertising: Understanding Cultural Paradoxes. Thousand Oaks, CA: Sage Publications, 1998.
HALPRIN, D. "Web Advertising Grows Outside the US." December 14, 2000. Online: http:// www.emarketer.com/analysis/eadvertising/ 20001212_web_ad_nonus.html.
HERBIG, P. A., and J. C. MILLER. "The Effect of Culture Upon Innovativeness: A Comparison of United States and Japan Sourcing Capabilities." Journal of International Consumer Marketing 3, 3 (1991): 7-53.
HOFSTEDE, G. Culture's Consequences: International Differences in Work-Related Values. Beverly Hills, CA: Sage, 1984.
--. Cultures and Organizations: Software of the Mind. New York. NY, McGraw-Hill International, 1997.
ICONOCAST, 2001. "Internet at a Glance 2001." Online: http://www.iconocast.com.
INOUE, O. "Advertising in Japan: Changing Times for an Economic Giant." In Advertising in Asia: Communication, Culture and Consumption, K. T. Frith, ed. Iowa State University Press, 1996.
LAKE, D. "Access Up, Divide Shrinks." The Standard, June 19, 2000. Online: http:// www.thestandard.com/research/metrics/ display/0,2799,16072,00.html
LAWRENCE, S. "The Net World in Numbers." The Standard, February 7, 2000. Online: http://www.thestandard.com/research/ metrics/display/.
LEE, C. "Determinants of National Innovativeness and International Market Segmentation." International Marketing Review 7, 5 (1990): 39 49.
MEDIA METRIX. "Japanese Internet and Digital Media Measurement Results Revealed by Media Metrix Japan." (2000). Online: http:// www.mediametrix.com/press/releases/ 20000601.jsp.
NETSCAPE COMMUNICATIONS. "University of Illinois and Netscape Communications Reach Agreement." (1994). Online: http://home. netscape.com/newsref/pr/newsrelease9.html.
PASTORE, M. "The World's Online Populations." (2001). Online: http://cyberatlas. internet.com/big_picture/geographicks/ article/o,1323,5911_151151,00.html.
PENTEADO, C. "McDonald's Offers Web Free Access in Brazilian Restaurants." September 3, 2001. Online: http://www.adageglobal.com.
PRS GROUP. "U.S. and Japan Country Profiles." The PRS Group, August 2000. Online: http: www.prsgroup.com.
RAI, A., T. RAVICHANDRAN, and S. SAMADDAR. "How to Anticipate the Internet's Global Diffusion." Communications of the ACM 41, 10 (1998): 97-106.
ROGERS, E. The Diffusion of Innovations. New York, NY: Free Press, 1983.
SAMIEE, S. "The Internet and International Marketing: Is There a Fit?" Journal of interactive Marketing 12, 4 (1998): 5-21.
SHAO, A. T., M. A. RAYMOND, and C. TAYLOR. "Shifting Advertising Appeals in Taiwan." Journal of Advertising Research 39, 6 (1999): 61-69.
SHAVITT, S., and S. HAN. "Persuasion and Culture: Advertising Appeals in Individualistic and Collectivist Societies." Journal of Experimental Social Psychology 30, 4 (1994): 326.
SONDERGAARD, M. "Hofstede's Consequences: A Study of Reviews, Citations and Replications." Organization Studies 15, 3 (1994): 447-56.
TAKADA, H., and D. JAIN. "Cross-National Analysis of Diffusion of Consumer Durable Goods in Pacific Rim Countries." Journal of Marketing 55, 2 (1991): 48-54.
TANSUHAJ, P., J. W. GENTRY, J. JOHN, L. L. MANZER, and B. J. CHO. "A Cross-National Examination of Innovation Resistance." International Marketing Review 8, 3 (1991): 7-20.
TELLEFSEN, T., and H. TAKADA. "The Relationship Between Mass Media Availability and the Multicountry Diffusion of Consumer Products." Journal of International Marketing 7, 1 (1999): 77-96.
THE WORLD FACTBOOK. 2000. Online: http:// www.cia.gov/cia/publications/factbook/index.html
YEARBOOK OF STATISTICS. "Telecommunication Services (Chronological Time Series 1989-98)." International Telecommunication Union: Geneva, 2000.
YIN, J. "International Advertising Strategies in China: A Worldwide Survey of Foreign Advertisers." Journal of Advertising Research 39, 6 (1999): 25-35.
ZHANG, Y., and B. D. GELB. "Matching Advertising Appeals to Culture: The Influence of Products' Use Conditions." Journal of Advertising 25, 3 (1996): 29-46.
CARRIE LA FERLE is an assistant professor in the department of advertising at Michigan State University. She received her Ph.D. in advertising from the University of Texas at Austin; an M.A. in advertising from Michigan State University; and a B.A. in sociology from the University of Western Ontario. Her research interests include cross cultural/ international consumer behavior and consumer role socialization as well as each area in relation to the internet. Dr. La Ferle's work has been published in the Journal of Advertising Research, the International Journal of Advertising, the Journal of Current Issues and Research in Advertising, and Gazette: The International Journal For Communication Studies, among others. Prior to her career in academia, Dr. La Ferle obtained industry-related experience in Canada, the United States, and Japan.
STEVEN M. EDWARDS is an assistant professor in the department of advertising at Michigan State University. He received his Ph.D. at the University of Texas at Austin (advertising), and M.A. and B.A. from California State University San Bernardino (psychology and marketing). Dr. Edward's work explores persuasive communication in new media environments. He is currently examining forced-exposure situations online and consumer avoidance of advertising. Other areas of interest include human perception and cognition in 3-D virtual environments, specifically related to e-commerce applications. Recent journal articles have been published in the Journal of Advertising Research, the Journal of Advertising, the Journal of Current Issues and Research in Advertising, and the Creativity Research Journal. His research has been supported by the Media Research Club of Chicago and the American Academy of Advertising.
YUTAKA MIZUNO is a professor of marketing in the Faculty of Design, Management and Technology at Kyoto Institute of Technology in Japan. Before he joined the Institute, he had worked in industry for more than 20 years at Dentsu Inc. in Tokyo, where he was in charge of Research and Development for advertising and lifestyle studies. He has written more than 60 papers and articles that have been published in Advertising Science (Japan Academy of Advertising), the Journal of Marketing and Distribution (The Japan Society of Marketing and Distribution), Japan Marketing Journal (Japan Marketing Association), and Bulletin of Nikkei Advertising Research Institute, among others.
CARRIE LA FERLE
Michigan State University
STEVEN M. EDWARDS
Michigan State University
Kyoto Institute of Technology
|Printer friendly Cite/link Email Feedback|
|Author:||La Ferle, Carrie; Edwards, Steven M.; Mizuno, Yutaka|
|Publication:||Journal of Advertising Research|
|Date:||Mar 1, 2002|
|Previous Article:||Reaching the male consumer by way of daytime TV soap operas.|
|Next Article:||Recall of brand placements in computer/video games.|