ARTICLE IN PRESS

Journal of Retailing and Consumer Services 11 (2004) 1929

Goods retailers and service providers: comparative analysis of web site marketing communications
Charles D. Bodkina,*, Monica Perryb
Department of Marketing, The Belk College of Business Administration, University of North Carolina at Charlotte, 9201 University City Blvd., Charlotte, NC 28223, USA b Department of Marketing, College of Business and Economics, California State University, Fullerton, P.O. Box 6850, Fullerton, CA 92834, USA
a

Abstract Retailers may engage in a range of marketing communications on their web sites. Our content analysis of web sites for 152 Fortune 500 retailers showed that more protable retailers were more likely to use company specic, shareholder, web specic and customer service elements. We also found differences in the use of advertising, shareholder, company specic and web specic elements among goods, nancial and services retailers. r 2003 Elsevier Science Ltd. All rights reserved.
Keywords: Marketing; Web sites; Communications

1. Introduction Rapidly increasing household Internet penetration underscores the potential for using corporate web sites to communicate with new and existing customers. Via their web sites, retailers of goods and services may engage in a range of marketing communications, such as advertising, sales promotion, public relations and direct marketing. Retailers can communicate directly with potential customers on-line (and vice versa). Since consumers perceive advertisers with web addresses as more customer-oriented, responsive and sophisticated (Maddox et al., 1997), retailers may be convinced of the need for a web site, but unclear about how to utilize the web site in terms of marketing communications (Budman, 1998). The opportunities to use web sites as a new marketing communication medium increase as more and more consumers access the Internet. At the heart of these opportunities are advances in new technologies that have provided retailers competitive advantages. As new technologies were incorporated into web site designs retailers were quick to adapt their on-line marketing communication strategies to meet the needs of the new market. But as these new technologies
*Corresponding author. Tel.: +1-704-687-4394; fax: +1-704-6876463. E-mail addresses: cbodkin@email.uncc.edu (C.D. Bodkin), monica perry@hotmail.com (M. Perry).

become diffused across the retailing industry what can we expect? Currently, we observe some differences in how service, nancial, and goods retailers utilize marketing communications in the physical marketplace, but do differences also persist across retail afliation in the on-line environment? If so, what do these differences indicate about the development of norms in the on-line environment? The specic research questions we address in this study are: (1) Are differences in web site marketing strategies associated with rm characteristics such as net income? and (2) Are differences in web site marketing strategies associated with industry afliation (goods retailers versus nancial institutions versus service providers)? While we will hypothesize web site marketing strategy differences based on industry afliation, we do so by extending the concepts of rm size, industry norms and organizational inertia in the physical marketplace to the Internet channel.

2. Literature review and conceptual denitions We incorporate the relevant literature addressing offline marketing communications, rm size, industry norms and organizational inertia with the emerging literature on web sites. The traditional marketing communication mix includes: (1) advertising, (2) sales

0969-6989/03/$ - see front matter r 2003 Elsevier Science Ltd. All rights reserved. PII: S 0 9 6 9 - 6 9 8 9 ( 0 2 ) 0 0 0 5 8 - 9

ARTICLE IN PRESS
20 C.D. Bodkin, M. Perry / Journal of Retailing and Consumer Services 11 (2004) 1929

promotions, (3) public relations and (4) direct marketing (Bennett, 1995). Advertising plays the same role on-line as it does in the physical marketplace, that is to create awareness, communicate benets, promote trial and urge customers into action (Gardener and Trivedi, 1998; van Waterschoot and Van den Bottes, 1992). As in the marketplace, advertising can communicate company information as well as information about the retailer mix such as products, prices, and/or location (for our purposes we separated typical advertising messages, such as logos and tag lines, from corporate messages that focus on specic company information). While advertising is an important strategy for retailers, so is sales promotion. Retailers make frequent use of a variety of types of sales promotions in the physical marketplace and sales promotions have been given an increasing portion of the total communications budget (Gardener and Trivedi, 1998). The major sales promotion tools relevant to retailer web sites include refunds/rebates, premium and specialty offers, sampling, on-line demonstrations, and contests and sweepstakes. The third promotional tool, public relations, includes news (e.g., press releases), speeches, educational programs, public-service activities, and lobbying or causerelated information. The nal type of marketing communication is direct marketing. Direct marketing takes two forms, selling on-line and customer service. Customer service involves auxiliary services, such as checking order/product availability, purchases/account activity, billing and returns, product use, delivery or problems (Innis and LaLonde, 1994). Whether on-line or off-line, customer service is often the most important inuence on customers evaluation of a seller (Perrault and Russ, 1976; Sheth, 1973). 2.1. Unique aspects of web sites To fully leverage a web site a retailer must understand and be able to use the unique characteristics of web sites as a communication channel. The highly interactive nature of web sites means that web site users have the potential to be more active acquirers of information (Ghose and Dou, 1998). Thus we can categorize corporate web sites based on the use of web specic elements, such as on-line privacy policy, email, customizing the information acquisition process, and screen specialization and multimedia capabilities such as visuals.

income retailers. Subsequently, we develop hypotheses comparing goods, nancial, and services retailers. Comparisons based on various aspects of rm size have for the most part established the existence of signicant differences in how rms of different sizes react to changes and opportunities in the environment. Although not central to their hypothesis, Asthana and Mishra (2001) found that rms that announced their earnings results were larger (based on capitalization value) than those that did not announce earnings. Presumably, investors seek and expect company nancial information from larger rms more so than from smaller rms. By extension, we would expect more protable rms to be more likely than less protable rms to highlight or disclose their positive nancial results via their web sites. Similarly, more protable rms may be more motivated to provide on-line company specic and public relations information. Research conducted for the Council For Public Relations Firms and Thomas Harris/Impulse Research (2002) validated a correlation between spending on traditional public relations (e.g., media and investor relations, annual reports) and a companys Fortune 500 ranking. The increased motivation to provide on-line information may result from the enhanced ability to invest in extensive public relations activities, such as support of various causes and other general company activities. Being more active at the corporate level and in public relations activities, these retailers would have more to communicate on their web sites than less active retailers. Thus: H1a: High income retailers use company specic information more frequently on their web sites than low income retailers. H1b: High income retailers use shareholder information more frequently on their web sites than low income retailers. H1c: High income retailers use public relations information more frequently on their web sites than low income retailers. While more protable rms have a much greater incentive (motivation) to provide company and nancial information they also have greater nancial resources (ability) to spend on particular marketing strategies. More protable rms would have greater wherewithal to consistently compete on a price basis. Competing on the basis of price is difcult for rms lacking deep pockets relative to the competition. One example can be found in the discount retailing industry. In summer 2001 K-mart Corporation initiated their BlueLight Always every day low pricing strategy to compete against Wal-mart, but without the deep pockets to sustain the strategy they have decided to focus more on branding (Cuneo, 2002). Similarly, more protable rms have a greater

3. Hypothesis development Our hypotheses involve two classes of comparisons. First we develop hypotheses comparing high and low

ARTICLE IN PRESS
C.D. Bodkin, M. Perry / Journal of Retailing and Consumer Services 11 (2004) 1929 21

ability to invest in and sustain extensive sales promotion activities. As a result, we would expect more protable rms to show a greater tendency toward the use of price and sales promotion. Thus we expect that: H2a: High income retailers use price more frequently on their web sites than low income retailers. H2b: High income retailers use sales promotion more frequently on their web sites than low income retailers. While more protable rms can invest more in marketing activities, they also have greater nancial resources to invest in web site technology. In May 2001 the cost for developing a small web site was estimated at $65,000 while a large site was estimated to cost $250,000 (Carmichael, 2001). The more protable rms would be capable of supporting the development of web specic elements associated with large web sites, such as visual content, email (an aspect of customer service), and the technology needed to develop extensive databases for searches (an aspect of web specic elements). More protable rms would have more nancial resources to apply towards the development of these types of web site functions, whether they are outsourced or developed inhouse. Thus: H3a: High income retailers use visuals more frequently on their web sites than low income retailers. H3b: High income retailers use web specic elements more frequently on their web sites than low income retailers. H3c: High income retailers use customer service more frequently on their web sites than low income retailers. While hypotheses 13 focus on the effects of retailers protability, hypotheses 36 focus on the effects of industry afliation. While we do not expect web site marketing communications to be unilaterally consistent with off-line marketing strategy, we expect considerable consistency by well-established incumbent rms (such as Fortune 500 retailers). Two related research streams support our expectations of consistency within a particular industry, although we are unaware that any specically address Internet strategies. First, well-established rms in a particular industry often develop or establish norms of behavior that reect, inuence or include their marketing choices (e.g. airline price wars). Secondly, well-established, incumbent rms within an industry are highly susceptible to organizational inertia. Ingrained processes and investment in infrastructure make incumbent rms less likely to recognize new opportunities that surface as a result of changes in the environment. More importantly, even when previously

successful incumbent rms recognize changes in the environment, they are likely to persist with their prior strategies (Audia et al., 2000). Organizational inertia results in information technology projects that represent the status quo rather than creative or new uses (Cooper, 2000). As a result of industry norms and organizational inertia we expect considerable consistency between Fortune 500 retailers off-line and on-line strategies. Consistency between off-line and on-line marketing communications may reect industry norms. The establishment of norms within a particular sector is indirectly supported by the ndings of Bush et al. (1998). Bush et al. (1998) found that only 41% of rms believed Internet marketing provided them with a competitive advantage. The uncertainty surrounding the Internet marketing environment is clear. DiMaggio and Powell (1983) suggest that when there is considerable uncertainty in the environment, rms will imitate comparable organizations. Given the considerable uncertainty surrounding the effectiveness of Internet marketing, it is likely incumbent rms will imitate the strategies of the competitors that they deem relevant. However, who are the relevant competitors? Markides (1999) suggests that traditional rms tend to look to traditional competitors. Similarly, Clark and Montgomery (1999) found that managers used supply based attributes (product offerings) to dene relevant competitors. Thus, other Fortune 500 goods retailers would consider other Fortune 500 goods retailers relevant competitors, and similarly for nancial and service retailers. Web sites are as easy for competitors to observe as customers. Retailers then develop similar marketing web site content and functionality once relevant competitors are observed. Just as mass merchandisers stores often have similar storefronts and layouts, certain web site norms may be established among competitors in a particular retail sector. Winter et al. (2000) exploratory study of corporate storefronts suggests that rms in the same industry develop norms for web site presentation and functions and that rms across industries differ considerably with respect to presentation (text and images). Thus, as web site norms are established, web sites within a particular retail sector will be similar, while web sites across retail sectors will differ. Web site norms are likely to reect off-line norms with respect to promotional activities. While this may not be true of every incumbent rm, this consistency is likely to be the dominant pattern of web site marketing communications. Inertia suggests that, in general, incumbent rms will continue their marketing communication patterns, even when a new communication channel becomes available (i.e. Internet). As Major (2000) proposes, goods retailers may simply extend their offline advertising practices to the on-line environment. Incumbent rms are experienced and invested in the

ARTICLE IN PRESS
22 C.D. Bodkin, M. Perry / Journal of Retailing and Consumer Services 11 (2004) 1929

marketing tactics used off-line and can relatively quickly replicate those experiences in the on-line environment. In the off-line environment, goods retailers engage in extensive advertising and sales promotion activities for a number of reasons. The breadth and depth of the merchandise mix for goods retailers, such as Wal-Mart, necessitates sustained and frequent use of advertising and sales promotion. The importance of cross-merchandising promotions can have a considerable impact on goods retailer success (Dreze and Hoch, 1998). In addition, the same brands are often offered by different retailers, resulting in intense brand competition among goods retailers. Walters (1991) found that in-store promotions by goods retailers negatively affected competing retailer sales. Goods retailers may use sales promotions to encourage customer stockpiling (Taylor, 2001) and or liquidate inventories (Karande and Kumar, 1995). The intensity of the competitive environment for goods retailers may make the extensive use of advertising and sales promotions appropriate. In contrast, service retailers may have less need to engage in the same degree of advertising and sales promotion as goods retailers. Service retailers mix of services is likely to be a fraction of that of a goods retailer. Relative heterogeneity in the service environment may result in less substitutability across different service retailers (Reardon et al., 1996). Taylor (2001) found that coupons did not change the timing of service purchases and more importantly that coupons were not an effective inducement for non-users of a particular service retailer. Since services are perishable and cannot be inventoried (Zeithaml et al., 1985), service retailers lack many of the incentives for engaging in sales promotions. In summary, extensive sales promotion activities may have limited appeal or usefulness in the services context. Thus, web site promotional efforts would mirror the physical marketplace, so we expect goods retailers to be the most frequent users of advertising and sales promotions, with nancial and services retailers utilizing them much less. As a corollary to extensive advertising use, we would expect greater use of visuals when advertising is used more extensively. As primarily a visual medium, web sites provide the opportunity to display visuals in the form of graphics and pictures. In summary, service retailers would engage in advertising and sales promotion less, and by extension use visuals less than goods retailers. Thus we expect that: H4a: Goods retailers use advertising more frequently on their web sites than service or nancial retailers. H4b: Goods retailers use sales promotion more frequently on their web sites than service or nancial retailers. H4c: Goods retailers use visuals more frequently on their web sites than service or nancial retailers.

Differences in other marketing strategies between service and goods retailers arise given the inherent differences between goods and services (Zeithaml et al., 1985). While intangibles comprise some portion of a goods retailers offering, intangibles comprise the majority of the offering for nancial and service retailers. To minimize the perceived intangibility we would expect nancial and service retailers to focus on establishing a strong corporate image through their web site. Since web site content affects overall perceptions of an organization (Palmer and Grifth, 1998; Singh and Dalal, 1999; Winter et al., 2000), we expect a greater focus on content that emphasizes the organizations image. Content that focuses on the overall organizational image includes public relations, nancial (shareholder), and company information. Since development of a strong and favorable organizational image is more important for service than goods retailers, we would expect more frequent use of public relations, shareholder and company information on their web sites. Thus, H5a: Service and nancial retailers use corporate information more frequently on their web sites than goods retailers. H5b: Service and nancial retailers use public relations information more frequently on their web sites than goods retailers. H5c: Service and nancial retailers use shareholder information more frequently on their web sites than goods retailers.

While nancial and service retailers may use web site content to enhance their overall organizational image, they may be more inclined to use specic marketing activities as indicators of service quality. One of the determinants of service quality in the physical marketplace is responsiveness (Parasuraman et al., 1988). While service quality is most commonly thought of as a result of a service encounter, Clow et al. (1998) investigated the role of cues to service quality prior to a service encounter. They found that advertisements with cues to responsiveness signicantly reduced the perceived risk associated with professional services. Correspondingly, web site visitors (potential customers) are likely to consider services more risky than goods given the intangibility and heterogeneity of services. To reduce this perceived risk, service retailers would use customer service and interactive web elements on their web sites to signal the retailers responsiveness. Service retailers would need to provide more cues to indicate responsiveness because of the heterogeneity of services. Extensive customer service and interactive web specic elements would be cues to a retailers responsiveness.

ARTICLE IN PRESS
C.D. Bodkin, M. Perry / Journal of Retailing and Consumer Services 11 (2004) 1929 23

Goods retailers have less need to provide multiple cues to responsiveness than service retailers, so we would expect goods retailers to use customer service and interactive web specic elements less than service retailers. In fact Grifth and Krampf (1998) found that in their sample of 64 goods retailers that few fully utilized customer service aspects (p. 20.). We expect less utilization of customer service by goods retailers because goods retailers are less likely to have the experience and infrastructure to support extensive interactions with individual customers. Interpersonal exchange between goods retailers and their customers is relatively limited, while services require considerable interpersonal exchange. Contrast the limited amount of interpersonal exchange involved in a purchase at Wal-Mart with the extensive interpersonal exchange involved in the purchase of life insurance. The considerable interpersonal exchange is required because service delivery usually requires interaction between the person delivering the service and the customer (simultaneous production and consumption). Financial services are often complicated and customers may have relatively low levels of knowledge (nancial literacy as described by Daniels et al., 1988) requiring extensive interaction between buyer and seller. As a result, nancial and service retailers have experience and investments in the human resources and infrastructure necessary for extensive interaction with customers on an individual basis in the physical marketplace. Service retailers could more easily support extensive customer service and interactive web specic elements, such as email. In contrast, supporting extensive customer service and interactive web specic elements would require signicant changes and investments by goods retailers. Thus, H6a: Financial retailers use customer service most frequently on their web sites, followed by service retailers, and then goods retailers. H6b: Financial retailers use web specic elements most frequently on their web sites, followed by service retailers, and then goods retailers.

4. Research design The sampling frame and associated URLs of the top 500 revenue producing companies were taken from Fortune magazines web site from May to July 1999 (Fortune, 1999). Included in the current study were 152 companies classied as retailers (54); nancial institutions and real estate (70), or services (28) as dened by the standard industrial classication index (i.e., Divisions G, H, and I respectively).

Content analysis has been successfully used in a variety of marketing research settings (Hanssens and Weitz, 1980; Naccarato and Neuendorf, 1998) including research on web sites (Kagan et al., 2000; Philport and Arbittier, 1997). In the current study a coding sheet was developed for a content analysis of corporate web site components through analysis of 50 randomly selected Fortune 500 corporate web sites. Only the rst screen of the web site (home page) was analyzed since a goal of the current study was to identify immediate perceptions of a web site. After independently reviewing the sites, the authors identied 63 different web site components. Inter-coder reliability for categorizing each web site item into one of the 63 categories was 94%. Once the 63 different web site components were identied they were organized into comprehensive categories using a two-step process. First, 36 e-business graduate students classied each of the 63 items into ten categories. Inclusion of an item in a category was based on frequency of response. The ten categories included advertising, company specic, customer service, prices, public relations, sales promotions, shareholder information, visuals, Internet only interactive activities, and Internet only passive activities. This resulted in the authors identifying one of the ten categories entitled Internet onlypassive activities that was used to classify only one of the 63 web site components (i.e., educational). Based on this feedback the category Internet onlypassive activities was removed from the list and the category Internet onlyinteractive activities was renamed as web specic. A second set of 40 e-business graduate students classied the 63 items into the nine subsequent categories. For an overview of the components associated with each category and the percentage of the companies using that particular component see Table 1. Once the categories were developed analysis of variance was used to test the hypotheses. Since each of the nine categories had a varying number of items per category (i.e., advertising 7 components, public relations 4 components) a mean usage rate per category (i.e., the mean number of items used by a company within each category) was calculated for each company. For example the mean usage rate for Wal-Mart on the advertising category was 43%. This was calculated by summing the total number of items in the advertising category used by Wal-Mart (i.e., 3 items) and dividing by the total number of items for that category (i.e., 7 total items in the advertising category). Hence the mean usage rate for Wal-Mart in the advertising category was determined to be 43% (i.e., 3/7=43%). To test hypotheses 1, 2, and 3 the 152 companies were divided into two groups based on their net incomes. The median net income was $416.25 million. Companies with a net income of $416.25 million and higher were placed in the high net income group. To test hypotheses

ARTICLE IN PRESS
24 C.D. Bodkin, M. Perry / Journal of Retailing and Consumer Services 11 (2004) 1929 Table 1 Percentage of companies using the web site component Advertising Logos Tag Lines Buttons for different market segments Brand names Kids Links to nearest dealers Advertising unrelated to the company Company specic About us Careers Product list buttons News related to the company Buttons for corporate divisions History Directory Customer service Email/contact us Nearest dealer search Privacy policy Catalogs/on-line ordering Educational Customer service On-line account information Apply for an account Help button Request materials FAQ Signups Research Apply for a password Explanation of buttons Surveys Prices Prices 92% 58% 46% 29% 3% 1% 1% Public relations Press releases Causes Public relationsother Health/safety/environment Sales promotions General promotion (hot deals) Highlights (whats new, cool) Games/sweepstakes Free gift Coupons Shareholder information Legal button Shareholder informationother Financials Annual report Stock quotes Visuals Continuous visuals Visualsfemales Visualsmales Visualsother Visualsproducts Visualsashing words Moving words Welcome button Gallery of photographs Web-specic Search Site map Links to other sites Select a country Email the webmaster Specialize the screen Download A place to enter a userid Quizzes 38% 15% 9% 5%

84% 74% 60% 57% 44% 11% 8%

20% 13% 13% 5% 2%

68% 51% 34% 20% 20%

70% 49% 43% 37% 22% 22% 21% 15% 14% 9% 8% 8% 7% 3% 1% 1%

57% 49% 49% 39% 28% 26% 20% 5% 4%

14%

45% 41% 34% 20% 12% 11% 8% 8% 1%

4, 5, and 6 the companies were divided into groups based on their standard industrial classication division.

5. Results Hypothesis 1 stated that highly protable rms would be more likely to communicate company specic information (H1a), shareholder information (H1b), and public relations information (H1c). The analysis of variance indicated that protability does affect two of the three sub-hypotheses (i.e., H1a and b). Both company specic and shareholder information were found to be signicantly different between higher net income and lower net income rms. Higher net income rms used on average 51.6% of the company specic items compared to 45.4% for lower net income rms (p 0:03; see Table 2). In addition, shareholder

information use is greater for high net income rms (48.6% versus 29.2% for low net income rms; p 0:00). No signicant difference was found between high and low net income rms use of public relations information (p 0:13). The second hypothesis was based on the premise that protability would impact the rms ability to spend on particular marketing strategies. The two categories used to test this hypothesis included price (H2a) and sales promotion (H2b). The results did not support this hypothesis. Firm protability was not signicantly related to the use of either pricing or sales promotion. Hypothesis three stated that nancial resources would impact a rms ability to invest in web site technology. For this study three categories (i.e., visuals, web specic, and customer service) were used to examine this hypothesis. The results indicated that nancial resources did not affect the use of visuals (H3a; p 0:13) but did

ARTICLE IN PRESS
C.D. Bodkin, M. Perry / Journal of Retailing and Consumer Services 11 (2004) 1929 Table 2 ANOVA results for web site components by net income Lower net incomes Hypothesis 1 a. Company specic b. Shareholder information c. Public relations Hypothesis 2 a. Price b. Sales promotion Hypothesis 3 a. Visuals b. Web specic c. Customer service 0.454 0.292 0.142 Higher net incomes 0.516 0.486 0.191 Level of signicance 0.030 0.000 0.128 25

0.097 0.111

0.194 0.111

0.100 1.000

0.296 0.131 0.175

0.338 0.199 0.241

0.126 0.003 0.005

affect both web specic (H3b; p 0:00) and customer service (H3c: p 0:00) categories. Companies with higher net incomes on average used approximately 20% of the components in the web specic category and 24% of the components in the customer service category compared to lower net income companies that used 13% of the web specic and 17.5% of the customer service categories. While the rst three hypotheses focused on the impact of protability, hypotheses four through six address industry afliation (see Table 3). The purpose of these hypotheses was to examine differences between goods, nancial, and service retailers. Hypothesis four states that nancial and service retailers would engage in less advertising (H4a), sales promotion (H4b), and use fewer visuals (H4c) relative to goods retailers. The results supported hypothesis 4b in the proposed direction. Goods retailers used more sales promotion components compared to both nancial and service retailers (p 0:00). Hypothesis 4a was signicant (p 0:00) but in the opposite direction. Financial and service retailers used more advertising components (i.e., approximately 39% and 31%, respectively) on their web sites relative to goods retailers (i.e., 29%). A post hoc analysis (i.e., tukey-test) found signicant differences between nancial and goods retailers. Finally, the impact of industry afliation and use of visual components was not signicant (p 0:48). Hypothesis ve addressed the impact of industry afliation and the use of web site content. We hypothesized that service and nancial retailers would use more company specic (H5a), public relations (H5b), and shareholder information (H5c). The results indicate support for hypotheses 5a and 5c (p 0:00 for both hypotheses), but not for hypothesis 5b (p 0:43). First, service retailers on average used 56% of the company specic components compared to 48% for nancial retailers and 44% for goods retailers. The post

hoc analysis indicated that service retailers used signicantly more company specic components compared to goods retailers. Second, while the use of shareholder information components was found to be signicantly different among the three types of retailers, nancial retailers used signicantly more of these components (i.e., 47%) compared to either service retailers (i.e., 32%) or goods retailers (i.e., 31%). Hypothesis six addressed the responsiveness of rms by examining their use of customer service and web specic components according to industry afliation. To be specic we proposed that customer service (H6a) and web specic components (H6b) would be used most frequently by nancial retailers, less frequently by service retailers and least frequently by goods retailers. Hypothesis 6a was not supported. We found no signicant difference in the use of customer service components across industries. However, Hypothesis 6b was partially supported by the data. Specically goods retailers used signicantly fewer web specic components (i.e., 11.7%) than either nancial (i.e., 17.9%) or service retailers (i.e., 20.2%). However, contrary to our hypothesis, service retailers actually used more web specic elements than nancial retailers.

6. Discussion The purpose of our study was to examine the degree to which retailers exploit web site opportunities in ways similar to how they market in the physical marketplace. We anticipated differences in web site marketing strategies associated with rm characteristics, such as net income, and industry afliation, such as goods, nancial and services, to parallel those found in the physical marketplace. As part of our initial analysis we identied nine categories of web site components. For ve of the nine categories some of the individual items were used by more than half of the rms in the study. For example, in the advertising category more than 50% of the rms used logos and tag lines on their web sites. Similarly, in the company specic category four items (about us, careers, product list buttons, news related to the company) were used by more than half of the rms. Contrary to expectations, fewer rms in the study made use of the sales promotion and public relations components. The sales promotion component used most often by rms in the study was that of general promotions (hot deals) and this particular component was used by only 20% of the rms. In addition, press releases were found on only 38% of the rms web sites home page. Overall our results suggest that retailers are only scratching the surface with respect to possible marketing communications activities.

ARTICLE IN PRESS
26 C.D. Bodkin, M. Perry / Journal of Retailing and Consumer Services 11 (2004) 1929 Table 3 ANOVA results for web site components by type of retailer Goods retailers Hypothesis 4 a. Advertising b. Sales promotion c. Visuals Hypothesis 5 a. Company specic b. Public relations c. Shareholder information Hypothesis 6 a. Customer service b. Web specic 0.286 0.182 0.329 Financial retailers 0.369 0.057 0.302 Service retailers 0.311 0.086 0.286 Level of signicance 0.004 0.000 0.477

0.439 0.139 0.307

0.482 0.179 0.471

0.561 0.188 0.321

0.009 0.428 0.000

0.211 0.117

0.216 0.179

0.174 0.202

0.414 0.010

6.1. Protability In this study we examine three hypotheses regarding the importance of rm size on web site content. Using net income as a measure of protability we posited that more protable rms would be more motivated and able to use specic types of marketing communications on their web sites. Our analysis provides partial support for the hypotheses. The results indicate that company specic and shareholder communications were found to differ based on rm protability. Based on our sample it could be implied that larger rms are more likely to use their web site as a communication tool targeted toward investors desiring information about the rms nancial strength. The fact that public relations components were not found to differ based on protability suggests that all rms in this study recognize the usefulness of web sites to communicate with a variety of stakeholders. The second hypothesis is not supported by this study, which suggests that the use of pricing strategies (i.e., pricing information and sales promotions) is not related to rm protability. While it was proposed that rms with deep pockets could sustain prolonged price competition the results suggest that both large and small rms may recognize the negative impact that such a pricing strategy would have on their long-term protability, at least in the Internet channel. Irrespective of overall protability, few retailers focused on price via their web sites. It may be that traditional retailers adapt their communications in the Internet channel to avoid the potentially detrimental effects of intense price competition and focus on other non-price ways to compete. The third hypothesis addressed the impact of rm protability and the use of web site technologies. The study found partial support for this hypothesis. It appears that more protable rms are likely to incorporate additional web specic and customer service

technologies into their home pages. Web site components such as email, dealer searches, site searches, and site maps build upon a larger rms abilities to develop and maintain extensive databases whether they be outsourced or developed in house. The use of visual components did not differ with respect to protability. Our results suggest that rms have come to recognize the importance of the Internet as a visual medium. What once may have been a tool for web site differentiation may no longer provide the desired competitive advantage. Overall we found support for select differences in rms web site marketing communications strategies based on protability. More successful rms seem to recognize consumers desire for facts rather than hype in the on-line environment. More successful rms seem to focus on communicating the strength of their organization to potential stakeholders. Moreover, more successful rms seem to be reinvesting their prots in the use of advanced technologies which they can better afford. 6.2. Industry afliation Hypotheses four through six address the impact of industry afliation on differences in web site marketing strategies. Previous research has suggested that rms within a particular industry sector may establish norms regarding their marketing communications (Bush et al., 1988). In addition, rms tend to perceive their competition as those with similar product offerings. By examining web site components across industry afliation, we can assess whether competing rms (i.e., rms with similar product offerings) use their web sites to differentiate themselves just as rms attempt to differentiate themselves in the physical marketplace. The fourth hypothesis examines the extent to which goods retailers make use of intensive marketing strategies such as advertising, sales promotions, and visuals

ARTICLE IN PRESS
C.D. Bodkin, M. Perry / Journal of Retailing and Consumer Services 11 (2004) 1929 27

relative to other types of retailers (i.e., nancial and service retailers). Based on observations in the physical marketplace, we hypothesized that goods retailers were most likely to use advertising on their web sites, but we found the opposite. Perhaps because goods retailers are capable of selling their goods online, they are less focused on branding, and more concerned with providing immediate incentives to buy via sales promotion. In fact we found that goods retailers were least likely to use advertising but more likely to use sales promotions to move their merchandise. In general it appears that goods retailers may be reacting to the intense brand competition that is also evident in their off-line environment. It might be that since on-line competitors are likely to carry similar brands, goods retailers need to resort to strategies with the greatest potential to drive customer trafc to their site. Rotating sales promotions offers this potential. The fth hypothesis examined rms use of web site content to project their public image. Due to the intangible nature of nancial and service retailers offerings we proposed that these types of institutions would be more likely to use their web sites to enhance their corporate communications strategy and image. Specically since previous research suggested that web site content affects consumers perceptions of an organization (Palmer and Grifth, 1998; Winter et al., 2000), nancial and service retailers would be more likely to use company specic, shareholder, and public relations information on their sites. The results suggest that nancial retailers are likely to use shareholder information and service retailers are likely to use company specic information. By providing legal information, nancial statements, annual reports, and stock quotes nancial retailers may be attempting to overcome the intangibility of their offering by developing trust between the rm and the consumer. Using web sites these rms can communicate messages of nancial strength and safety to their customers, thus reducing consumer risk and by extension increasing consumer preference for the rm. Service retailers also use their web sites to communicate specic messages. In this case their message could be characterized as let us tell you who we are. With emphasis on the use of about us, careers, and product list buttons service retailers are able to build awareness of the depth of their offerings. The last hypothesis was also partially supported. We found that nancial retailers, followed by service retailers, were most likely to use customer service and web specic components due to the highly interactive and interpersonal nature of exchanges that take place off-line in these industries. While the results indicated no differences on customer service components across industries, both nancial and services retailers were signicantly more likely to provide web specic features, such as customization, on their web sites. This nding

supports our earlier supposition that service providers have greater experience and investments in the human resources and infrastructure needed for extensive interaction with customers on an individual basis. Such extensive experience gives service providers more elaborate databases that are advantageous to web site customization.

7. Limitations and future research While we have condence in our study, we do recognize that some limitations exist which point to future research. First, we assessed only the home page of these retailers web sites. Certainly many site visitors will enter at the home page. However, some will certainly enter the web site via other pages when they create permanent links (bookmarking) to specic pages and search engines that link to pages embedded in the site. Primary data that captures exactly where particular groups of customers arrive at the web site could be combined with a fuller examination of retailers web sites to reveal additional aspects of retailers web site communication strategy. The second, and more important, limitation is the lack of measures of web site effectiveness. While we appropriately applied a measure of overall company performance (net income), we recognize that given our focus on web sites that a more specic measure would provide more guidance for managers. However, we suggest that future research on web site effectiveness take a broad view of effectiveness. In particular, any measure of web site effectiveness should certainly capture the effects within the Internet channel, but also in the physical channel. For instance, site visitors may use the web site to gather information and subsequently make the purchase at a physical store. For retailers that use multiple channels for marketing communications, integrating communication across channels may distinguish the successful from the less successful retailers. In other words, any investigation of web site effectiveness must capture effects of the web site across different channels.

8. Conclusion We focused on two main issues that lead to differences in web site marketing strategies. First, we found some support for the existence of a relationship between retailer protability and web site marketing strategies. More protable retailers used their web page as an information source for company and shareholder information as well as to enhance consumer relationships through technology. Specically more protable retailers were more likely to provide customer services

ARTICLE IN PRESS
28 C.D. Bodkin, M. Perry / Journal of Retailing and Consumer Services 11 (2004) 1929

and web specic elements such as email, and database searches. What these elements have in common is their ability to foster a positive relationship with consumers through enhanced web site utility and usability. With a stronger focus on consumers, more protable retailers will likely continue to be more protable well into the future as more and more consumers turn to the Internet. Moreover, our study provides direction for less protable retailers. These retailers should reect on how more protable retailers make use of such utility enhancing elements and nd similar, but perhaps lower cost, ways to enhance consumer relationships in the Internet environment. Based on our comparison of Internet web site marketing strategies, we can conclude that meaningful differences do exist between more and less protable incumbent retailers. The second issue we address is the relationship between industry afliation and web site marketing strategies. Previous research would suggest that retailers in a particular industry sector establish norms regarding their web site strategies. We found some support for differential use of promotional, informational, and technological strategies. Financial and service retailers web pages were distinguishable from goods retailers based on promotional and technological strategies. Financial and services retailers used advertising and web specic elements more frequently in the design of their web sites, whereas goods retailers were noted for more frequent use of sales promotion. Such a differential emphasis is logical given the relative heterogeneity of services compared to goods. Given the intangibility of their offerings, nancial and service retailers have an increased need to establish a positive company image with consumers but a lesser need to compete via shortterm sales promotion tactics which may ultimately diminish their image. In addition to the goods versus services comparison, we also found differences in marketing strategies between nancial and services retailers. Financial retailers focused on shareholder information compared to services retailers that focused on more general company specic information. Thus we can conclude that relatively strong intra-industry norms do exist. Our study provides retailers with an understanding of how industry afliation related to web site marketing strategies. Whether or not an individual retailer can be effective by simply following such intra-industry norms remains to be seen. However, in order to develop appropriate competitive strategies it is important that retailers recognize that these norms do exist for their particular industry sector. Retailers can more effectively direct their resources to differentiating their web sites if they know which unique components are being emphasized within their respective industry.

References
Asthana, S.C., Mishra, B.K., 2001. The differential information hypothesis, rm size, and earnings information transfer: an empirical investigation. Journal of Business Research 53 (1), 3747. Audia, P.G., Locke, E.A., Smith, K.G., 2000. The paradox of success: an archival and a laboratory study of strategic persistence following radical environmental change. Academy of Management Journal 43 (5), 837853. Bennett, P.D., 1995. Dictionary of Marketing Terms. American Marketing Association, Chicago. Budman, M., 1998. Why are so many web sites so bad? Across the Board 35 (9), 2935. Bush, A.J., Bush, V., Harris, S., 1998. Advertiser perceptions of the Internet as a marketing communications tool. Journal of Advertising Research 38, 1727. Carmichael, M., 2001. 2001s median prices for full-site development. B to B. Accessed August 2002. http://www.btobonline.com/webPriceIndex/. Clark, B.H., Montgomery, D.B., 1999. Managerial identication of competitors. Journal of Marketing 63, 6783. Clow, K.E., Baack, D., Fogliasso, C., 1998. Reducing perceived risk through advertising service quality cues. Journal of Professional Services Marketing 16 (2), 151162. Cooper, R.B., 2000. Information technology development creativity: a case study of attempted radical change. MIS Quarterly 24 (2), 245277. Council For Public Relations Firms and Thomas Harris/Impulse Research, Corporate communications spending and reputations of Fortune 500 companies, white paper accessed August 2002, http://www.prrms.org/resources/research/harris/default.asp. Cuneo, A.Z., 2002. Ailing Kmart surrenders in price war. Advertising Age 73 (3), 1. Daniels, P.W., Leyshon, A., Thrift, N.J., 1988. Large accountancy rms in the UK: operational adaptation A. The Service Industries Journal 8 (3), 3171347. Dimaggio, P.J., Powell, W.W., 1983. The iron cage revisitedinstitutional isomorphism and collective rationality in organizational elds. American Sociological Review 48 (2), 147160. Dreze, X., Hoch, S.J., 1998. Exploiting the installed base using crossmerchandising and category destination programs. International Journal of Research in Marketing 15 (5), 459471. Fortune Magazine web site (1999). Accessed MayJune 1999. http:// cgi.pathnder.com/fortune/fortune500/500list.html. Gardener, E., Trivedi, M., 1998. A communications framework to evaluate sales promotion strategies. Journal of Advertising Research 38, 6771. Ghose, S., Dou, W., 1998. Interactive functions and their impacts on the appeal of Internet presence sites. Journal of Advertising Research 38 (2), 2943. Grifth, D.A., Krampf, R.F., 1998. An examination of the web-based strategies of the top 100 US retailers. Journal of Marketing Theory and Practice 6 (3), 1223. Hanssens, D., Weitz, B., 1980. The effectiveness of industrial print advertisements across product categories. Journal of Marketing Research 17 (3), 294305. Innis, D., LaLonde, B.J., 1994. Customer Service: the key to customer satisfaction, customer loyalty, and market share. Journal of Business Logistics 15 (1), 127. Kagan, A., Post, G., Messier Noel, C., 2000. Selling the web: web features used by retailers. Journal of Applied Business Research 16 (1), 1525. Karande, K.W., Kumar, V., 1995. The effect of brand characteristics and retailer policies on response to retail price promotions: implications for retailers. Journal of Retailing 71 (3), 249279.

ARTICLE IN PRESS
C.D. Bodkin, M. Perry / Journal of Retailing and Consumer Services 11 (2004) 1929 Maddox, L.M., Mehta, D., Daubek, H.G., 1997. The role and effect of web addresses in advertising. Journal of Advertising Research 37 (2), 4759. Major, M., 2000. Adding steak to the sizzle. Supermarket Business 55 (2), 99. Markides, C.C., 1999. A dynamic view of strategy. Sloan Management Review 40 (3), 55. Naccarato, J., Neuendorf, K., 1998. Content analysis as a predictive methodology: recall, readership, and evaluations of business-tobusiness print advertising. Journal of Advertising Research 38 (3), 1933. Palmer, J.W., Grifth, D.A., 1998. An emerging model of web site design for marketing. Communications of the ACM 41 (3), 4451. Parasuraman, A., Zeithaml, V.A., Berry, L.L., 1988. Servqual: a multiple-item scale for measuring consumer perception. Journal of Retailing 64 (1), 1241. Perrault, W., Russ, F., 1976. Physical distribution service in industrial purchase decisions. Journal of Marketing 40, 310. Philport, J.C., Arbittier, J., 1997. Advertising: brand communications styles in established media and the internet. Journal of Advertising Research 37 (2), 6876. 29 Reardon, J., Miller, C., Hasty, R., Waguespack Jr., B., 1996. A comparison of alternative theories of services marketing. Journal of Marketing Theory and Practice 4 (4), 6171. Sheth, J., 1973. A model of industrial buyer behavior. Journal of Marketing 37, 5056. Singh, S.N., Dalal, N.P., 1999. Web pages as advertisements. Communications of the ACM 42 (8), 9198. Taylor, G.A., 2001. Coupon response in services. Journal of Retailing 77 (1), 139151. van Waterschoot, W., Van den Botte, C., 1992. The 4P classication of the marketing mix revisited. Journal of Marketing 56, 8393. Walters, R.G., 1991. Assessing the impact of retail price promotions on product substitution, complementary purchase, and interstore sales displacement. Journal of Marketing 55 (2), 1729. Winter, S.J., Saunders, C., Hart, P., 2000. Electronic window dressing? impression management on the Internet. Working paper, UNC Charlotte. Zeithaml, V.A., Parasuraman, A., Berry, L.L., 1985. Problems and strategies in services marketing. Journal of Marketing 49 (2), 3347.