Content-Type: text/html A Niche Analysis of the Web, Catalogs and Retail Stores: A Case in Taiwan Cheng Kuo Department of Advertising National Chengchi University Taipei, Taiwan 116 8862-2738-7424 [log in to unmask] Hairong Li Department of Advertising Michigan State University East Lansing, MI [log in to unmask] Vincent Huang Department of Journalism National Chengchi University Taipei, Taiwan, 116 [log in to unmask] Submitted to the Division of Advertising of Association for Education in Journalism and Mass Communication 2000 Annual Conference Phoenix, Aug 9-12, 2000 A Niche Analysis of the Web, Catalogs and Retail Stores: A Case in Taiwan Abstract This study examines channel utilities of the Web in comparison with catalogs and retail stores from the perspective of niche theory. It identified and measured three channel utilities (communication, distribution, and convenience) with a set of 14 questions. Through on-line surveys, information about 909 Taiwanese internet users were collected and analyzed. Scales of channel utilities were assessed using confirmatory factor analysis and then used to examine the niche breadth, niche overlap, and niche superiority of these channels. Findings of the study indicate the Web is superior to catalogs and retail stores on the communication and convenience utilities but inferior to retail stores on the distribution utility Competitive displacement, instead of competitive exclusion, may take place among the three channels in the near future. Implications for marketing and advertising were discussed. Key words: Internet advertising Confirmatory factor analysis Niche analysis A Niche Analysis of the Web, Catalogs and Retail Stores: A Case in Taiwan Electronic commerce has gained increasing popularity among consumers as a convenient shopping channel during the past year not only in the US but also in countries and markets outside the US. The internet user population in Taiwan has increased significantly in the past few years, as the government and semi-official organizations launched " getting on-line" promotional campaigns to attract more people to sign up for internet accounts. A widely circulated magazine has recently survey 1336 people by telephone and reported that among the quarter million Taiwanese people who have signed up for an inter account, about 37% of them (approximately 94 thousand people) are frequent and active internet users. Though Taiwan's E-commerce is still in its early stage in comparison with that of the US, it is a quite promising area. To take full advantage of the Web as a new marketing channel, we need to better understand how consumers respond to the characteristics of the Web in comparison with traditional marketing channels such as retail stores and catalogs. Such a fast growing e-commerce in marketing practice has aroused interest among both scholars and practitioners. They wondered if shopping on the Internet would displace existing retail formats or it would serve as a natural complement to current marketing practices. Peterson, Balasubramanian, and Bronnenberg reviewed the fallacies in electronic shopping visions throughout the 20th century and recommended considering the complexity and heterogeneity of consumer markets in study of the Internet. Burke clearly stated, "Internet will have the greatest impact on marketing communication, a moderate effect on sales transactions, and (with the exception of information goods) a minimal impact on logistics." (p. 352) These comments certainly help us better assess the impact of the Internet and Web as new marketing channel on other marketing channels and practice. This study attempts to further empirical exploration in this direction. Specifically, it will examine consumers' perception of channel utilities of the Web in comparison with catalogs and retail stores. It uses the theory of the niche to inspect the pattern of competition among the three marketing channels and hopefully, to make a valid assessment of the impact of the Web on other two channels and derive useful implications for marketers and advertisers. Literature Review Marketing Channels Marketing activities occur through various channels such as retail stores, catalogs, and online stores. Kotler identified nine functions of marketing channels, including information, promotion, negotiation, ordering, financing, risk taking, physical possession, payment, and the actual transfer of product ownership. Peterson, Balasubramanian and Bronnenberg aggregated marketing functions into three distinctive types of marketing channels: communication channels, transaction channels, and distribution channels. By definition, communication channels enable the flow of various types of information between buyers and sellers. Transaction channels realize ordering and payment activities between buyers and sellers. And distribution channels facilitate the physical exchange of products and services between buyers and sellers. Stewart, Frazier and Martin incorporated marketing functions into two types of channels: communication channels and distribution channels. The latter has a broader definition, referring to "a mechanism through which a product or service can be selected, purchased/ordered, and received by a segment of the firm's customers." (p.190) Channel Utilities Although conceptually distinct, in the context of consumer marketing, these channels frequently overlap, and a given channel may be responsible for multiple functions. The multi-functionality can be best demonstrated on the Internet and Web. For instance, a digital product such as a software program or a research report can be advertised to, paid by, and distributed to the purchaser all through the Web. In this case, the Web serves the functions of both communication and distribution channels. For non-digital products like computer hardware, clothing, and wine, the Web is not able to function as a distribution channel. For consumers, the value of a channel is its utility , which can be conceptualized as a ratio of the benefits that a channel offers over the costs of using the channel. It is often called "net utility" in economics. Because a consumer decision-making process usually starts with searching for information about various brands, a key benefit that communication channels can provide is information. Three kinds of information are especially useful: information on alternative brands, latest product information, and detailed product information. Consumers need to form their consideration sets before making final selections. Thus, a wide selection of alternative brands in a given channel is always preferred. Latest information also is important in that the availability and price of a product is in constant change on the market. Consumers need to be certain about the availability and current price of a product before purchasing it. And detailed information is essential, too, especially when a consumer wants to exam a brand closely before make the final decision. Studies of consumer decision making found that consumers often first narrow their consideration set to a few items and then exam them more carefully . Detail information satisfies this need. Communication channels that are capable to offer these kinds of information are considered to be of high communication utility. It is useful to separate shopping and buying acts because each act involves different considerations . While shoppers are more interested in various product information, buyers are more likely to consider prior purchase inspection, security of payment, and quick access to good they buy. How consumers weight these factors normally affects their choice of channels from which to carry out their purchases. Thus, these factors can be conceptualized as the distribution utility of a channel. This study separates the distribution function from the service function of a channel for two reasons. First, perceived quality of post-purchase service and return policy are important for most durable goods. Consumers often consider these two factors before they decide where to buy a product. Second, the capacity of rendering certain types of service is often a built-in function of a marketing channel. For instance, consumers cannot inspect products personally if they choose to order them from a catalog by telephone. They must go to a store to see the product if they prefer so. The service utility often differentiates one channel from the other. In addition to these potential benefits a channel can provide, costs exist in using a channel. Three kinds of costs associated with using a channel are the access, effort involved, and time consumption. For instance, visiting a retail store may need a car while accessing the Web requires a computer with an Internet connection and some computer knowledge. Use of a channel also may involve different levels of physical and mental effort and consume certain amount of time. These efforts are costs to the consumer using a channel and often affect channel selection. In sum, this study conceptualizes four types of channel utilities: communication, distribution, service, and convenience. A comparison of three marketing channels - the Web, catalogs, and retail stores - is conducted on each of the utilities. The comparison is conducted in light of the theory of the niche. Theory of the Niche The theory of the niche is a theory formulated by bioecologists to describe and explain competition and coexistence of different species in an environment of finite resources (Levins 1968). Dimmick and Rothenbuhler first applied this theory to the study of media competition for advertising revenues. In the study, they reviewed basic concepts and propositions of the theory. One of the concepts is the community, consisting of a set of populations. The community is a system in which populations interact with each other and with their common environment. A population is a set of individual entities that are most like each other than they are like members of other population. For instance, television stations, although each distinct, are more like each other than any member in the radio population. The environment contains finite resources of different dimensions. Each population may compete with others in a dimension of the resources. Thus, the niche of a population "occupies a certain volume within the n-dimensional resource space of the community." In other word, a population's niche is "its position in the resource space denoting its interaction with the environment." A population must have its niche to survive in an environment of finite resources. The niche of a population can be measured in three ways. Niche breadth is a measure of the area of a niche. Specifically, niche breadth is the number and amount of resources used by a population. A population with a narrow niche uses fewer resource dimensions than a population with a broad niche. Niche breadth can be calculated using the formula developed by Dimmick : (1) Niche Breadth Where: u, l = the upper and lower bounds of a scale (e.g., 4 and 1) GO = a gratification obtained rating on a scale N = the number of respondents using a medium n = the first respondent K = the number of scales on a dimension k = the first gratification scale The upper bound of B equals unity and is attained when all respondents rate a medium at the upper bound on all gratification statements on a dimension. The lower bound is zero and indicates that respondents rate a medium at the lower bound of all gratification scales. Niche breadth refers to the relationship of a population to the environment. The second way to measure the niche of a population is niche overlap, which refers to the relationship of one population to another in the community. It measures the degree of similarity or competition between two populations, provided the resources are finite. Below is Dimmick's formula of niche overlap. (2) Niche Overlap Where: i, j = medium i and medium j GO = a gratification obtained rating on a scale for i and j N = the number of respondents who use both i and j n = the first respondent Overlap is an inverse measure and low values indicate high similarity between the two media. The lower limit of the overlap measure is zero and indicates that the niches of the media i and j overlap completely, or the niches of the two are identical. The upper limit is u minus l and is reached when respondents rate one medium at the upper bound on all scales and the other medium at the lower bound of all scales. Niche overlap can be considered as an index of the substitutability or complementarity of two media. High overlap indicates that two media are substitutes; they serve the same needs. Lower overlap indicates the two serves different needs and thus a state of complementarity exists between the two media. The third way to measure the niche is niche superiority, which "answers the question of whether one or the other of a pair of media provides greater gratification utility." The formula of niche superiority is as follows: (3) Niche Superiority Superiority S i > j = Superiority S j > i = Where: i, j = medium i and j m i>j = the value of a respondent's rating for those scale items on which i is rated greater than j (the sum of the actual values) m j>i = the value of a respondent's rating for those scale items on which j is rated greater than i (the sum of the actual values) K = the number of scales on a dimension k = the first gratification scale N = the number of respondents who use both i and j n = the first respondent The scores of niche superiority are defined as arithmetic means. As a result, the differences in superiority between two means on a utility may be tested for significance using a t-test for correlated media. If the test yields a significant result, this is interpreted as superiority of one medium over another on that utility. Conversely, the absence of a significant difference indicates that neither medium is superior on that niche dimension. Although these three formulas were designed for measuring media gratifications obtained, it can be used to assess the niche of marketing channels because of the conceptual similarity of media gratifications and channel utilities. The "populations" in this study are the Web, catalogs, and retail stores. Each channel has four niche dimensions as discussed earlier in the study. What these marketing channels are competing is consumers' attention to and use of them. In normal conditions, consumers' attention to and use of marketing channels are in equilibrium because consumers' buying power grows steadily along with a nation's economy. In a given purchase situation, for example, if a consumer buys a book on the Web, it would mean that one fewer book would be sold from a conventional bookstore. Which channel a consumer chooses to use will depend on a number of factors including channel utilities. Research Questions The theory of the niche is both an illustrative framework and a useful analytic tool. Using this theory, the study attempts to answer four research questions: 1. What is the niche breadth of the Web, catalogs, and retail stores on each of the four channel utilities? 2. How much niche overlaps do the Web, catalogs, and retail stores have on each of the four channel utilities? 3. Which channel is superior, and on which channel utility? 4. What are the relationships between these channels? Are they competitive displacement, competitive exclusion, or complementary each other? Methods Data for this study was collected through an online survey which was part of a collaborative project conducted by the research group and a semi-official ISP SeedNET in Taiwan. Unlike most current online surveys that have little control over who will visit the Web site and fill out the questionnaire, the respondents of this study were randomly selected from a pool of internet users to minimize the self-selective bias of the survey respondents. The selected users were sent e-mails three times to invite them to participate in the online survey. Decent prizes and gifts were promised as the incentive to increase rate of the survey. Each selected survey respondent was assigned a unique password to access the web site for the survey, each password could be used only once. That is, the password became invalid after one entry to the designated web site. This device was used to minimize the bias resulting from the fact that some respondents may fill out the questionnaires multiple times for winning the incentive. In a six-week-period, over one thousand respondents participated in the survey. After a careful screening procedure, the researchers were able to obtain 909 clean and valid completed questionnaires. A summary of the demographic composition of the sample is presented in Table 1. Table 1 Summary of Demographic Composition ______________________________________________ Gender (N=909) Male Female 67.9% 32.1% Age Under 21 18.0% 21 to 29 52.5% 30 to 39 21.5% 40 to 49 5.8% 50 to 56 0.6% over 65 0.1% Education Junior Middle School Graduate 4.8% High School Graduate 15.4% Some College & College Degree 70.2% Some Graduate School & With Graduate Degree 9.6% _________________________________________________ Two thirds of the respondents were males, 32.1% of them were females. The gender composition was quite similar to that of other national surveys. However, the respondents were younger in ages, over 51 percent of them were under the age of 25. Having a high level of education was another characteristic of the sample, over 70 percent of them were either college graduates or going to college at the time of survey. As for their residential areas, 77 percent of them lived in cities, whereas, only 23 percent resided in less urbanized countryside or rural areas. It may indicate that internet is generally more easily accessible and adopted more widely in more urbanized areas. Fourteen channel attributes were measured for retail stores, catalogs, and the World Wide Web. The 14 channel attributes are listed as follows: 1. information on alternative brands, 2. detailed product information, 3. latest product information, 4. easy access to information 5. effortless to use 6. least time-consuming. 7. wide selection of goods, 8. pre-purchase inspection, 9. security of payment, 10. quick access to goods bought, 11. post-purchase service, 12. easy exchange or return, 13. privacy of personal information 14. purchase in secrecy Each attribute was measured by asking, "Thinking about retail locations (or catalogs, and the World Wide Web) how ideal is it in providing information on alternative brands?" The respondent was expected to indicate his or her perception on a five point rating scale, ranging from 1 (least ideal) to 5 (most ideal). information on alternative brands)?" All other attributes were measured in a similar fashion. Results Confirmatory factor analysis (Arbuckle and Wothke 1999) was first conducted to test the scales of channel utilities. Four dimensions were originally specified in the analysis: communication, distribution, service, and convenience. However, the question items for the service dimension did not converge well, and was dropped from the analysis. The resulting factor loading of each dimension are summarized in Table 2 Goodness of fit index was .944 for the Web, .905 for catalogs, and .936 for retail stores. The model of the Web utilities fit the data best. The reliability (Cronbach's alpha) of each dimensional scale in each model was calculated and presented in Table 2. Because alpha tends to underestimate reliability of congener measures , these scales should be considered acceptably reliable even though some alphas seem relatively low. Table 2 .Confirmatory Factor Loadings of Channel Utilities Web Catalogs Retail Stores Communication Information on alternative brands .706 .622 .688 Detailed product information .664 .700 .760 Latest product information .534 .536 .741 Reliability (Cronbach Alpha) .68 .68 .74 Distribution Prior purchase inspection .706 .617 .430 Security of payment .759 .492 .412 Quick access to goods bought .560 .507 .417 Reliability (Cronbach Alpha) .73 .68 .74 Convenience Easy access .813 .579 .732 Effortless to use .881 .740 .832 Least time consuming .602 .622 .754 Reliability (Cronbach Alpha) .69 .70 .73 Model parameters__2,df.,p_ 265.26,27,.000 450.10,27,.000 274.47,27,.000 Goodness of fit index .944 .905 .936 Niche Breadth The first research question deals with the niche breadth of the three channels. The results of the niche breadth analysis are in Table 3. On the communication utility, the Web has a larger niche breadth than both catalogs and retail stores; the latter two were relatively close in magnitude. On the distribution utility, niche breadth score of retail stores is larger than other two channels. The niche breadth of the Web is the smallest of the three on the distribution utility, this may indicate that the Web is the most under-developed distribution channel of the three. On the convenience utility, however, the niche breadth of the Web is larger than that of retail stores, which in turn, are larger than that of catalogs. Table3.Niche Breadth Channel Communication Distribution Convenience Web .82 .36 .73 Catalogs .64 .45 .56 Retail Stores .62 .92 .67 Note: Niche breadth ranges from 0 to 1,with 1 for maximum niche breadth. In addition, the niche breadth of each channel can be examined by calculating the average of the scores of three utilities for each channel. The average of three utilities is .637 for the Web, .550 for catalogs, and .736 for retail stores. Retail stores is higher than both the Web and catalogs. The web is slightly greater than catalogs, suggesting a promising future for the Web even though it is the newest channel. Niche Overlap Table 4.Niche Overlap Channel Communication Distribution Convenience Web/Catalogs 1.53 1.02 1.58 Web/Retail Stores 1.91 1.96 1.90 Catalogs/Retail Stores 1.21 1.31 1.51 Note: Low values indicate high similarity or overlap on dimensions of the two channels. Table 4 contains the values of niche overlap for each pair of channels, which answer the second research question. On the communication utility, the Web and retail store are more dissimilar given the largest niche overlap. Whereas, catalogs and stores seem to have a greater overlap of the three pairs. On the distribution utility, the niche overlap of the Web and stores is the highest because of their lowest score whereas the niche overlap between the Web and catalogs is the lowest given their highest score. And on the convenience utility, the niche overlap score is the highest between the Web and retail stores, indicating less overlap. Meanwhile, we observed a slightly higher overlap between the Web and catalogs and between catalogs and retail stores. Niche Superiority Table 5.Niche Superiority Communication Distribution Convenience Web_Catalogs Catalogs_Web Web_Catalogs _2.40__0.24_ _1.42__0.35_ _2.40__0.34_ t_27.076___ t_14.794___ t_21.700___ Web_Stores Stores_Web Web_Stores _2.85__0.48_ _6.80__0.06_ _2.05__1.23_ t_22.353___ t_67.769___ t_6.073___ Catalogs_Stores Stores_Catalogs Stores_Catalogs _1.22__0.98_ _5.71__0.03_ _1.93__0.68_ t_2.579_ t_70.324___ t_11.898___ Note:Numbers in parentheses are superiority values for the channel above the number. ____p_.001,___p_.01, __p_.05 The third research question is answered by the scores of the niche superiority of each channel in Table 5. On the communication utility, the score of the Web is higher than catalogs and retail stores, indicating niche superiority. Meanwhile, catalogs is slightly superior to retail stores in this regard. On the distribution utility, retail stores is by far the most superior channel than the other two as indicated by the high scores. When comparing catalogs and the Web, the former one is slightly superior to the latter one in this regard. On the convenience utility, the Web is leading both catalogs and retail stores; retail stores, however, is considered better than catalogs. Summary and Conclusions Interesting findings have been uncovered from the analysis, and are summarized as follows: (1) The niche breadth of the Web is the greatest in both communication and convenience utilities but the lowest in distribution utility. Retail stores are the highest in distribution utility. The niche breadth of catalogs is slightly higher than the Web in distribution utility but is the smallest one of the three channels in convenience utility. The niche breadth of catalogs indicates its less favorable position because of their lack of leading scores on any of the three utilities. (2) The Web and retail stores have the least overlap on all three utilities, as indicated by the high scores. Meanwhile, high similarities exist between the Web and catalogs on distribution dimensions, high overlap between catalogs and retail stores in communication utility. Niche overlap means substitutability or complementary of two channels. A pair of channels with low scores on niche overlap implies the high possibility for each other to substitute. A pair of channels with high scores on niche overlap means the high potential for each other to complement. (3) The Web was found to be superior to the other two channels on communication and convenience utilities. Whereas, retail stores is far more superior to the other two on distribution utility. In comparison with the Web and retail stores, catalogs seem to have less niche superiority on all three utilities. None of the three channels is superior to any other channel on all three utilities. In other words, no total appropriation of the other channel's niche is found among the three channels. Thus, it is less likely that any competitive exclusion will occur and that any less competitive channel is doomed to extinction. The Web, though the youngest in its developmental stage, is already considered superior to both catalogs and retail stores on the communication and convenience utilities, but it is still inferior to the other two channels on the distribution utility. This pattern of niche superiority of the Web foresees a result of competitive displacement. In other words, alteration may occur in the niche of a competitor to lower the level of competition. It is likely that any competitive exclusion will occur and that any less competitive channel is doomed to extinction. As a result, the Web is though less likely to replace retail stores, at least in the near future, but some partial appropriation of niche space penetration is possible. Nevertheless, the Web has become a threatening competitor against catalogs in satisfying consumer's needs for product information (communication) and for convenience shopping. Implications for Marketing and Advertising The findings of the study have a number of implications for marketers and advertisers. The study has revealed the weakness of the Web on the distribution utility. Lack of certain aspects of those utilities, such as experiencing a product before purchase or immediate access to certain products, has negatively affected the total utilities of the Web as a marketing channel. Web marketers need to overcome these limitations. Recently, some e-commerce Web sites have established virtual reality stores, which give the consumers a feel like shopping in a real, physical store. It represents the direction of next generation of e-commerce. For retail stores, an increase in communication and convenience utilities is a real challenge. Opening online storefronts to compete head-on with other virtual stores is one solution but more importantly, retail stores need improve themselves in this regard. As Burke wisely pointed out, retail stores need to offer "greater shopping convenience, more current and complete product information, custom-tailored products and services, and potentially lower prices." (p. 355) It is interesting that on the distribution utility, catalogs are perceived to be superior to the Web and both have high niche overlap as non-store shopping channels. However, these findings do not indicate the likelihood that catalogs will threaten the growth of the Web as a marketing channel. Instead, an interpretation could be that because catalogs have been in use for many years, consumers feel comfortable with their distribution utility. Compared with the Web, catalogs are weak on the communication and convenience utilities. It will be a good strategy for catalogers to establish their presence on the Web to take its advantage while maintaining their current superiority on the distribution and service utilities. References