Effects of Web Retail Service Quality and Product Categories on Consumer. Behavior: A Research Model and Empirical Exploration

Transcription

1 Effects of Web Retail Service Quality and Product Categories on Consumer Behavior: A Research Model and Empirical Exploration Byounggu Choi a, Choongseok Lee b, Heeseok Lee b, and Mani Subramani a a Carlson School of Management, University of Minnesota bchoi1@csom.umn.edu; msubramani@csom.umn.edu b Department of Management Information Systems, Graduate School of Management, KAIST freeman@kgsm.kaist.ac.kr; hsl@kgsm.kaist.ac.kr Abstract This paper explores how web retail service quality at the point of purchase influences consumers perceptions of value and willingness to buy. In particular, this study highlights how web retail service quality has different effects on perceived product quality, value, and willingness to buy according to product categories. We provide a research model on the basis of service quality, product categories on the web, and marketing theories for consumers purchase behavior. Consumer perceptions of web retail service quality are collected from a sample of shoppers who actively search for products over the Internet. The results imply that functional web retail service quality has a direct effect on willingness to buy and technical web retail service quality influences consumer perceptions of product quality and value. Furthermore, product categories moderate the relationships among web retail service quality, perceived product quality, perceived value, and willingness to buy. Implications for Internet marketing and future research directions are discussed. 1. Introduction The Internet is increasingly an important channel for businesses to reach customers. It is predicted that the volume of online retailing will be nearly $105B in the USA by the 2007 [27]. Internet Retailing, an innovation pioneered in the USA, is being rapidly adopted worldwide. In some parts of the world, farsighted investments by national governments in creating a broadband technology infrastructure have spurred the growth of Internet commerce. For instance, Korea s total e-commerce revenue could explode from $4B in 2001 to as much as $24B by 2005 [10]. As a result, retail businesses in Korea have been aggressively adopting the Internet as a channel to address the needs of Korean customers. The number of web retailing stores in Korea reached to about 2600 and their sales in 2002 accounted for 11 percent of total sales of retail business [12]. However, firms face several problems in understanding the context of web based commerce. Current notions of service quality and delivering quality service are based largely on research and managerial experience in conventional contexts where customers can physically examine and interact with the products they purchase. In contrast, the nature of customer interactions with products as well as the cues they receive about the firm are likely to be significantly different online. Understanding how customers perceive service quality online and how such perceptions influence perceptions of value and willingness to purchase online therefore an important issue both for academics as well as for managers. Furthermore, our current understanding of the phenomenon of e-commerce phenomena comes largely from studies performed in the US context. With online retailing becoming increasingly mainstream across the globe, information on the phenomenon from an international context is particularly important. This paper addresses above issues and builds a theoretical framework relating online service quality to purchasing intent. It draws on marketing theories of /04 $17.00 (C) 2004 IEEE 1

2 consumers purchase behavior and prior work in the area to highlight how online retail service quality influences consumers perceptions of product quality, value, and willingness to buy. In particular, we examine the moderating effects of product characteristics on these relationships using data collected in South Korea. The outline of this paper is as follows: we survey prior literature on the topic in Section 2 and discuss the research model and construct inter-relationships in Section 3. Details of the study are presented in Section 4, followed by the results in Section 5. The limitations, implications and conclusions of our study are discussed in Section 6. Section 7 concludes the paper. 2. Literature Review Service quality is a multi-dimensional construct [2, 3]. The SERVQUAL model [20] and the model proposed by Grönroos [8] are the two most influential models of service quality in conventional contexts. The original SERVQUAL model proposes five key dimensions influencing service quality perceptions: tangibles, reliability, responsiveness, assurance, and empathy. In contrast, the model of Grönroos emphasizes two-factors: functional quality and technical quality. These models are often identified as representing the American and the Nordic perspectives respectively [2]. Studies on assessing web-based retail service quality have recently begun to appear due to the different characteristics of web-based retail service. There is evidence that web-based and traditional retail services differ fundamentally on two dimensions: (i) the possibility of bi-directional many-to-many communication and (ii) the level of emphasis on intangibles [14]. These distinctions raise questions with regard dimensions that are salient in the perception of retail service quality on the web. For example, the minimal emphasis on the tangible dimension in webbased retailing as opposed to the considerable weight in the traditional retail environment indicates that traditional service quality dimensions may not be directly applied to web-based retail service quality. Prior studies examining online service quality can be broadly categorized into three groups: (i) SERVQUAL model, (ii) SERVQUAL-oriented model, and (iii) web-oriented model. Illustrative studies in each of these categories are listed in Table 1. The studies under the first category adopt simply apply SERVQUAL model to web-based shopping context. For instance, the WebQual model of Barnes and Vidgen [1] directly applies the SERVQUAL model to the online context. Similarly, Sullivan and Walstrom [23] employed SERVQUAL to evaluate service quality of a web-based retailer. Studies in the second category modify the SERVQUAL model to evaluate web service quality by incorporating factors like navigation [9] and quality of information [14]. However, the main dimensions of these studies still come from SERVQUAL model. For instance, in evaluating the online services of 23 travel agencies, Kaynama and Black [9] propose seven dimensions to establish a webbased retail service quality based on SERVQUAL Studies in the third category develop models where web characteristics like web design are central constructs [28]. For instance, Loiacono et al. [15] propose the WEBQUAL model with 12 dimensions And Wolfinbarger and Gilly [28] propose the four dimensional.comq framework. Yoo and Donthu [29] developed the SITEQUAL model viewing perceived quality of a web-based retail service as comprising four dimensions. Through a six focus group interviews with customers having some online experience, Zeithaml et al. [30] proposed eleven dimension of e-service quality. SERVQUAL model SERVQUALoriented Weboriented Table 1. Comparison of previous studies Study Barnes & Vidgen [1] Sullivan & Walstrom [23] Kaynama & Black [9] Li et al. [14] Lociacono et al. [15] Wolfinbarger & Gilly [28] Yoo & Donthu [29] Zeithaml et al. [30] Focus Applying SERVQUAL Applying SERVQUAL Modifying SERVQUAL Modifying SERVQUAL Developing WEBQUAL Developing.comQ Developing SITEQUAL Developing e-sq Industry Book Book Travel N/A CD, Book, Airline, Hotel N/A N/A N/A Dimension Tangibles; Reliability; Responsiveness; Assurance; Empathy Tangibles; Reliability; Responsiveness; Assurance; Empathy Responsiveness; Content & purpose; Accessibility; Navigation; Design & presentation; Background; Personalization & customisation Responsiveness; Competence; Quality of information Empathy; Web assistance; Call-back systems Informational fit to task; Interaction; Trust; Response time; Design; Intuitiveness; Visual appeal; Innovativeness; Flow (emotional appeal); Integrated communication; Business processes; Substitutability Website design; Fulfillment/reliability; Privacy/security; Customer service Ease of use; Aesthetic design; Processing speed; Security Reliability; Responsiveness; Access; Flexibility; Ease of navigation; Efficiency; Assurance/trust; Security/privacy; Price knowledge; Site aesthetics Customization/personalization /04 $17.00 (C) 2004 IEEE 2

3 While prior work highlighted various dimensions that are important in evaluating web-based service quality, few studies have considered how web-based retail service quality can affect consumers purchase behavior on the web. Second, the previous studies did not explore the effect of the product characteristics on service quality perceptions and purchase behavior. This study thus extends the literature in two important aspects: (i) it highlights how service quality affects consumer purchase behavior on the web and (ii) it examines if these effects are contingent on product characteristics. 3. A Research Model and Hypotheses 3.1 Research model The research model proposed in this study is illustrated in Figure Consumer behavior. Researchers in marketing drawing on work in sociology and psychology highlight a variety of influences on consumer purchase behavior [21]. They suggest a variety of relationships among perceived product quality, perceived relative price, perceived value, and willingness to buy. Two-factor Nordic Service quality model Functional Web service Quality H1 (+) Technical Web service Quality H3 (+) Product Quality Relative Price H8 (+) H4 (+) Value H10a H10c H10b H2 (+) H5 (+) H6 (-) H7 (-) H9 (+) Product classification Product Category Involvement High vs. Low Web Presentability High vs. Low Willingness to Buy Purchase behavior in marketing research Figure 1. A research model In this paper, we draw on these studies and theories of adoption [26] and diffusion [17]. Consumer perception is considered as the starting point for this paper since perception of product quality, price, and value constitutes an important dimension of willingness to buy. Typically, product quality and relative price are acknowledged to be two major antecedents of the perceived value for not only the traditional retail environment but also the web-based retail business. Price is critical for the success in web-based retail [4]. Price has a negative effect on perceived value and consumers willingness to buy while it has a positive effect on perceived product quality [6]. This dual role of price has been widely accepted [24]. product quality is the consumer s judgment about a product s overall excellence [4]. It has a significantly positive effect on perceived value. As perceptions of product quality increase, perceptions of value would increase. The most common definition of value is the ratio or trade off of quality to price [24]. Consistent with this definition, two strategies to improve perceptions of value were presented; (i) reducing prices while maintaining product quality and (ii) improving product quality while maintaining price. In addition, the perceived value has been regarded as antecedent to a consumers willingness to buy. For example, Sweeney et al. [24] suggested that monetary value would have a positive relationship with the perceived likelihood of purchase Web-based service quality. Service quality has been recognized as an important strategic retailing weapon [4]. Many successful product retailers differentiate themselves not through the product they sell but through the service they offer. Although it is not clear whether SERVQUAL can be applied to evaluating web-based service quality, the five SERVQUAL dimensions recur in online service quality models. For example, the dimensions of WEBQUAL [15] are similar to these of the SERVQUAL. However, the SERVQUAL model has been criticized by several researchers. They indicate that it overemphasizes the functional (process) perspective of service, and underemphasizes the technical (outcome) perspective. To respond to this /04 $17.00 (C) 2004 IEEE 3

4 criticism, we integrate SERVQUAL with Grönroos s [8] conceptualization of technical and functional quality. Grönroos suggested two service quality dimensions: functional quality and technical quality. Functional service quality deals with the process or the way in which service is delivered. That is, it defines customers perception of the interactions that happen during service delivery. Technical service quality reflects the outcome of the service or what is received. This integration may help identify the relationship between service quality and willingness to buy on the web. The quality of a web-based retail service depend on the perceived quality of the process of using the web for purchasing as well as the perceived quality of the outcome. Research has suggested that functional service quality influences consumers willingness to buy directly. For example, Sweeney et al. [24] suggested that functional service quality is main contributors of consumers willingness to buy. Functional service quality also influences technical service quality, which in turn influences product quality. The role of technical service quality is well recognized. It influences not only perception of product quality but also the personal construct of value. Competence (skill and expertise) or product knowledge is important for perceptions of technical service quality [19] Product category. Researchers suggest that consumers on the web do not regard all web service quality categories as equally important. Some web service quality factors differ from one product category to another while certain factors are regarded as highly important across all the product categories [16]. We draw from Lee et al. s [13] product classification framework and incorporate customers involvement and web presentability as moderators. Involvement refers to the degree of the psychological identification and emotional ties consumers have with a particular product [18]. For example, a ring bought by a customer for a boyfriend is a high involvement product. Emotional factors are likely to be important as the ring is associated with important rituals like the wedding ceremony. Web presentability refers to the extent to which a particular product can be assessed and evaluated over the web in isolation [7]. For example, books are under high web presentability because their characteristics can be assessed easily by the use of the web. Our research model incorporates the moderating effects of involvement and web presentability into the relationships among web service quality, perceived product quality, perceived value, and willingness to buy. Because most customers are likely to need product information that can explain various aspects of the product in high involvement categories, responsiveness and empathy assurance are likely to be important for service quality perceptions. Therefore, functional web service quality is likely to affect customers willingness to buy in high involvement categories. In contrast, for low involvement products, this may not be the case. For products with low web presentability, customers are likely to have difficulty in judging product attributes and tend to spend more time in the purchase process; technical web service quality such as complete knowledge about products of web sites may be important. value may also vary. It is easy for customers to understand product characteristics in high web presentability categories [7]. Therefore, technical web service quality may not be critical for value perceptions in this category. 3.2 Hypotheses The preceding conceptualization suggests direct relationships among web service quality, product quality, relative price, perceived value, and willingness to buy. We posit direct relationships between functional web service quality and technical web service quality as well as between functional web service quality and willingness to buy (H1-H2). Furthermore, we expect direct relationships between technical web service quality and product quality as well as between technical web service quality and perceived value (H3- H4). Similarly, direct relationships among relative price and product quality, value, and willingness to buy are suggested (H5-H7). Relationships between product quality and value as well as between value and /04 $17.00 (C) 2004 IEEE 4

5 willingness to buy are also suggested (H8-H9). Finally, the moderating effects of product categories are suggested (H10). The hypotheses are summarized in Table 2. Table 2. Hypotheses Hypotheses H1: The functional and technical dimensions of web service quality are positively related. H2: Functional service quality is positively related to the willingness to buy. H3: Technical web service quality is positively related to perceived product quality. H4: Technical web service quality is positively related to perceived value. H5: Relative price is positively related to perceived product quality. H6: Relative price is negatively related to perceived value. H7: Relative price is negatively related to the willingness to buy. H8: Product quality is positively related to perceived value. H9: value is positively related to willingness to buy. H10a: Product involvement moderates the relationship between functional web service quality and willingness to buy. H10b: Web presentability moderates the relationship between technical web service quality and product quality. H10c: Web presentability moderates the relationship between technical web service quality and value. 4. Methodology 4.1 Sample and data collection To assess our research model as shown in Figure 1, we employ an experimental survey to collect data from a group of senior undergraduate students majoring in MIS in two Korean universities. The participants performed four shopping activities: (i) buying a ring for their boy/girl friend, (ii) buying a CD, (iii) buying a notebook computer, and (iv) buying toys for a younger brother or sister. After each of the four shopping activities, the participants were asked to complete questionnaires. Four hundred and twenty four questionnaires were collected over two months in Due to incomplete data, four questionnaires were eliminated. Two products were chosen from each criterion that are popularly sold over the web in Korea. We selected rings and music CDs for high and low involvement respectively; computers and toys for high and low web presentability respectively. For the study, we selected four Korean sites where these products are sold online: Coupletown.co.kr for rings, Tubemusic.com for music CDs, Compuzone.co.kr for notebook computers, and Mytoyworld.co.kr for toys. Table 3 presents the demographics of the participants. These characteristics reflect the relative youth of our sample. Almost all participants had previous exposure to the web percent have positive attitude towards web browsing. Table 3. Characteristics of participants Female Gender Male Unanswered Average age Ranges from 20 to 42 Average years of prior web use 3.59 Frequency of web usage Less than 1 hour a week 1 to below 5 hours a week 5 to below 10 hours a week 10 to below 15 hours a week 15 to below 20 hours a week 20 to below 25 hours a week 25 or more a week Unanswered Attitude towards browsing the web Very positive Positive Neither Negative Very negative Unanswered 4.2 Research constructs 35.8 % 63.2 % 1.0 % 2.3 % 2.3 % 3.5 % 9.3 % 23.3 % 39.5 % 18.8 % 1.0 % 31.4 % 52.5 % 15.1 % 0 % 0 % 1.0 % The questionnaire was developed based on previous literature by the use of existing scales where possible. Measures related to web service quality were developed on the basis of the items of previous research [1, 24, 29]. Measures of value, willingness to buy, and product quality were adopted from Dodds et al. [6]. Measures of price were developed based on Sweeney et al. s [24] perceived relative price concept. It was defined as the consumer s perception of product s price compared to other brands of the same product with similar specifications. Responses were obtained on a five point Likert scale anchored at the end points by very low and very high. Table 4 provides details on the source and descriptions of the constructs /04 $17.00 (C) 2004 IEEE 5

6 Table 4. Constructs and their sources Construct Source Items Functional web service quality [1, 24] 1. This site has attractive appearance (FWSERV) 2. This site makes it easy to give feedback or contact the organization. 3. This site provides content tailored to me 4. This site has fast navigation to information. 5. This site provides information to me at an appropriate level of detail. Technical web service quality [1, 24, 1. This site provides right answer or information to my questions. (TWSERV) 29] 2. This site gives a professional and competent image. Perceptions of product quality [6] 1. This product would be reliable. (PPQ) 2. This product would be durable. 3. This product would be dependable. Relative price (PRI) [24] 1. Compared to other product type with similar features, the price of this product offered by this site is high. 2. Considering 1 as the least expensive product type with similar features available and 5 as the most expensive product type with similar features available, how would you rate the product type you looked at or purchased? Perceptions of value (VALUE) [6] 1. This product is good value for money. 2. At the price shown this product is economical. 3. This product is considered to be a good buy. 4. This product appears to be a bargain. Willingness to buy (WILL) [6] 1. I would consider buying this product at this site. 2. I will purchase this product at this site 3.There is a strong likelihood that I will buy this product at this site 5. Research Results The SEM (Structural Equation Models) incorporated in LISREL 8.30 was used for our analysis. 5.1 Measurement properties The values of Cronbach s alpha, range from (functional web service quality) to (perception of value), suggesting adequate measurement reliability. This study follows the Straub [22] s process for validating instruments in MIS research to test construct validity in terms of convergent and discriminant validity. Convergent validity is based on the idea that items for valid measures of the same concept should correlate rather highly with one another or yield similar results even though they are different instruments. It was assessed by computing item-to-total correlation scores. Items with item-to-total correlations less than 0.4 should be considered for rejection [11]. Therefore, items having item-to-total correlation scores lower than 0.4 were dropped from further analysis. All items surpass the recommended level. Table 5. Reliability and convergent validity results Construct Item Reliability Convergent Validity (Correlation of item with (Cronbach alpha) total score-item) Functional web service quality ; 0.491; 0.546; 0.511; (FWSERV) Technical web service quality ; (TWSERV) Perceptions of product quality ; 0.727; (PPQ) Relative price (PRI) ; Perceptions of value (VALUE) ; 0.696; 0.849; Willingness to buy (WILL) ; 0.610; Table 5 shows the results of reliability and convergent test results. We examined the level of discriminant validity using chi-square difference tests (see Chin et al. [5] for details). The results presented in Table 6. The significant chi-square values indicate adequate discrimination between pairs of closely related constructs. Table 6. Chi-square difference test results for discriminant validity χ 5.2 Hypothesis test We tested the hypotheses. To identify goodness of fit, we examine (i) the Goodness of Fit Index (GFI), (ii) Root Mean Square Residual (RMSR), (iii) Root Mean Square Error of Approximation (RMSEA), (iv) Comparative Fit Index (CFI), and (v) Chisquare/degree of freedom. Previous studies suggest that a good fit is indicated by CFI above 0.8, RMSR less than 0.1, RMSEA less than 0.08, CFI greater than 0.9, and chi-square/ degree of freedom less than 5. Table 7 summarizes the fit indices and the parameter estimates for the hypothesized structural equation model. The results indicate fairly reasonable goodness-offit indexes; the hypothesized model fits the data well. In addition, examinations of the estimated coefficients of the model are significant; except for H5, hypotheses 1 through 9 are supported. Functional web service quality positively affects technical web service quality (γ=0.66, t=8.12, p<0.01) and willingness to buy (γ=0.13, t=2.42, p<0.05). Technical web service quality /04 $17.00 (C) 2004 IEEE 6

7 has a significant effect on perceptions of product quality (β=0.70, t=7.78, p<0.01) and value (β=0.33, t=3.61, p<0.01). Relative price has a negative effect on perceived value (γ=-0.33, t=-6.24, p<0.01) and willingness to buy (γ=0.64, t=-6.98, p<0.01). However, no direct relationship is found between relative price and perceived product quality (γ=0.00, t=0.05, p>0.1). Product quality positively affects perceived value (β=0.22, t=2.55, p<0.05) and value positively affects willingness to buy (β=0.34, t=5.94, p<0.01). Table 7. Results of estimated SEM (N=420) Endogenous Constructs Technical web service quality (η1) Product quality (η2) value (η3) Willingness to buy (η4) Technical web service quality (η1) Endogenous Constructs Product value (η3) quality (η2) Willingness to buy (η4) Exogenous Constructs Functional web service quality (ξ1) (8.12***) 0.70 (7.78***) 0.33 (3.61***) relative price ((ξ2) (0.05) 0.22 (2.55**) (5.94***) χ 2 =346.8, d.f.=142, χ 2 /d.f.=2.44, GFI=0.92, RMSR=0.038, RMSEA=0.059, CFI=0.94 Supported hypothesis in boldface type. ***: p < 0.01, **: p < 0.05 t values in parentheses (6.24***) (2.42**) (6.98***) Structural equation fit (R 2 ) Other important issues are the moderating effects of product categories. To identify moderating effects of product categories, two-group structural equation model of involvement (high involvement and low involvement) and web presentability (high web presentability and low web presentability) are examined. Table 8 reports fit indices of the two-group model. The result shows that our model fits with the data. Table 8. Fit indices of the two-group model Two-group model Involvement Web presentability Recommended Indices High Low High Low value (ring) N=106 (CD) N=105 (notebook) N=105 (toy) N=104 Goodness of Fit Index (GFI) >0.80 Root Mean Square Residual (RMSR) <0.1 Root Mean Square Error of Approximation (RMSEA) <0.08 Comparative Fit Index (CFI) >0.90 chi-square/degree of freedom / /142 =1.78 = /142 = /142 =1.19 To test the moderating effect of high and low involvement product categories on the relationship <5 between functional web service quality and willingness to buy, we compared this baseline two-group model of involvement with a constrained model in which a pair of path coefficients are assumed to be equal. In a similar way, we test the moderating effect of high and low web presentability product categories on the relationship between technical web service quality and perceived product quality, between technical web service quality and perceived value. Table 9(a) shows the path coefficients for two groups (high involvement and low involvement) from baseline two-group model and χ 2 differences between the baseline model and the constrained model. Similarly, Table 9(b) reports the path coefficients for high and low web presentability groups. From Functional web service quality From Technical web service quality Technical web service quality Table 9. Results of moderating effects of Path To Path To W illingness to buy product quality value (a) Involvement Parameter Standardized Coefficients (t statistics) High Involvement Low Involvement? χ 2 (df) (Ring) (CD) γ (1.91) 0.18 (2.77) 2.90 (1)* (b) Web presentability Parameter Standardized Coefficients (t statistics) High Web Low Web?χ 2 (df) presentability presentability (Notebook) (Toy) β (2.14) 0.51 (2.75) 3.60 (1)* β (3.35) 0.80 (3.82) 3.05 (1)* * p< 0.1; Chi-square values greater than 2.71 are significant at the 90 percent confidence level. Table 9(a) represents that the positive effect of the functional web service quality on willingness to buy is significantly higher in the high involvement product (e.g., ring) than the low involvement products (e.g., CD) ( χ2 =2.90, p<0.1). The positive effect of the technical web service quality on perceived product quality is significantly higher in the low web presentability products (e.g., toy) than the high web presentability products (e.g., notebook) ( χ2 =3. 60, p<0.1). In addition, the technical web service quality has significant higher effect on perceived value in the low web presentability products (e.g., toy) than the high web presentability products (e.g., notebook) ( χ /04 $17.00 (C) 2004 IEEE 7

8 =3. 05, p<0.1). Although the p-value is relatively low, these results imply that product categories moderate the effect of technical web service quality on value. In sum, hypothesis 10 is supported. 6. Discussion 6.1 Limitations The findings of this study are interesting, but they should be considered in the light of its inherent limitations. First, subjects in this study were senior undergraduate students majoring in management information systems. They were frequent users of the web; their positive experience with the web - this may have biased the results. For example, consumers who were infrequent users of the web or had negative experience might think of reputation and familiarity with a retailer s physical store as important factors for perceived product quality in web retailing. Second, factors such as willingness and security are uncontrolled. Future works incorporating additional factors are therefore of further interest. 6.2 Contributions and Implications The paper makes several theoretical contributions. First, it proposes a model identifying the link between service quality and other factors influencing online consumers purchase intents. Most previous studies have focused on what dimensions are important in web service quality. By contrast, our model highlights the influence of web service quality on consumer purchase behavior. Our study shows that web service quality has both direct and indirect effects on perceived value and willingness to buy. Furthermore, our study highlights how factors like product quality and price affect willingness to buy. Second, this study focuses on the two-factor Nordic model [8] that has received relatively little attention and examines the relationship between functional and technical web service quality. Third, this study investigates the moderating effects of product categories on the relationship between web service quality and perception of product quality, value, and willingness to buy. Although many researchers have proposed that web service quality should be influenced by product characteristics [16], our study provides the earliest empirical evidence on this issue. Finally, this study provides data on service quality from a Korean context and supplements the overall understanding of the phenomenon from work based on the US context. The study has several implications for practice. Our findings indicate that functional web service quality is positively associated with technical web service quality. i.e., well designed websites influence customer perceptions of technical quality. Functional web service quality also has a significant effect on a customer s willingness to buy regardless of perceived quality and value of product. Responsiveness, well organized information, and visual appealing are clearly vital. In particular, web sites that deal with high involvement products like jewelry should be more carefully designed to enhance functional web service quality (see Table 9(a)). Technical web service quality is important for perceived product quality and value, influencing customers willingness to buy indirectly. Our analysis confirms that technical web service quality such as reliability has a significant effect on the perceived product quality and value. Furthermore, customer needs for technical web service quality are influenced by product characteristics (see Table 9(b)). For example, complete product or service descriptions with visual appeal is important for shopping malls that deal with low web presentability products like toys. Because customers have difficulty in assessing product characteristics for such products online, they require comprehensive information. To provide proper product information, most business webs attempt to display products in a variety of ways. Some use text with picture and others use 3D or virtual reality techniques. However, most web retailing sites adopt the same implementations for all products. Our results suggest that retailing websites should be driven by product characteristics. For example, customers need strong credibility for high involvement products. Therefore, lock-in and branding /04 $17.00 (C) 2004 IEEE 8

9 may be effective to increase customers attractiveness in these products. Customer-generated information (e.g. the comments after using products) is also important. The insignificant findings in this study also bear some examination. Our study shows no direct relationship between price and perceived product quality. This rather intriguing result may be explained in part by other extrinsic cues like brand name or packaging. When price is the only extrinsic cue available, the consumers perceive product quality positively. However, this positive relationship may not hold if more cues are available [4, 6]. 7. Conclusion This paper highlights the important role of functional and technical dimensions of service quality and their link to perceptions of value and the willingness to buy. The results provide evidence of the moderating effects of product characteristics on the relationships of web service quality to perceived product quality, value, and willingness to buy. The results using data from Korea provide a broader picture of these relationships in an international context, supplementing prior examination of these constructs that have largely been based on studies in the US context. The study highlights several avenues for future research. The current study examined a single site in isolation within each product category. Future research examining multiple web sites in each product category can sharpen our results. In addition, understanding how culture affects satisfaction and perceived service quality is an interesting avenue for research [25]. The Finally, certain potential antecedents of customers purchase behaviors like customers buying decision style are uncontrolled because of the limitation of our laboratory setting. Incorporating these effects may yield useful insights. Acknowledgement This work was supported by Korea Research Foundation Grant. (KRF B00016) Reference [1] Barnes, S.J., and Vidgen, R. An Evaluation of Cyber-Bookshops: The WebQual Method, International Journal of Electronic Commerce (6:1), 2001, pp [2] Brady, M.K., and Cronin, J. Some New Thoughts on Conceptualizing Service Quality: A Hierarchical Approach, Journal of Marketing (65:3), 2001, pp [3] Carman, J.M. Consumer Perceptions of Service Quality: An Assessment of SERVQUAL Dimensions, Journal of Retailing (66:1), 1990, pp [4] Chen, Z., and Dubinsky, A.J. A Conceptual Model of Customer Value in E-Commerce: A Preliminary Investigation, Psychology & Marketing (20:4), 2003, pp [5] Chin, W.W., Gopal, A., and Salisbury, W.D. Advancing the Theory of Adaptive Structuration: The Development of a Scale to Measure Faithfulness of Appropriation, Information Systems Research (8:4), 1997, pp [6] Dodds, W.B., Monroe, K.B., and Grewal, D. Effects of Price, Brand, and Store Information on Buyers Product Evaluation, Journal of Marketing Research (28), 1991, pp [7] Figueriedo, J.M. Finding Sustainable Profitability in Electronic Commerce, Sloan Management Review, 2000, pp [8] Grönroos, C. Service Management and Marketing: Managing the Moments of Truth in Service Competition, Lexington Books, Lexington, MA, [9] Kaynama, S.A., and Black, C.I. A proposal to Assess the Service Quality of Online Travel Agencies: An Exploratory Study, Journal of Professional Services Marketing (21:1), 2000, pp [10] Kim, S.B. Korea s e-commerce: Present and Future, Asia-Pacific Review (8:1), 2001, pp [11] Kline, P. A Handbook of Test Construction, Methuen, London, [12] Korea National Statistical Office. Survey on Cyber Shopping Mall in 2001, Korea National Statistical Office, [13] Lee, H., Choi, S., and Choi, B. Internet Business /04 $17.00 (C) 2004 IEEE 9

10 Implementation Guidelines for Retailing Using Product Classification Framework, in Proceedings of 2001 KORMS International Conference, 2001, pp [14] Li, Y.N., Tan, K.C., and Xie, M. Measuring Webbased Service Quality, Total Quality Management (13:5), 2002, pp [15] Loiacono, E.T., Watson, R.T., Goodhue, D.L. WebQual: A Website Quality Instrument, Working Paper , University of Georgia, [16] Lowengart, O., and Tractinsky, N. Differential Effects of Product Category on Shopping Selection of Web-Based Stores: A Probabilistic Modeling Approach, Journal of Electronic Commerce Research (2:4), 2001, pp [17] Mahajan, V., Muller, E., and Bass, F.M., Diffusion of New Products: Empirical Generalizations and Managerial Uses, Marketing Science (14:3), 1995, pp [18] Martin, C.L. Relationship Marketing: A High- Involvement Product Attribute Approach, Journal of Product Brand Management (7), 1998, pp [19] Mersha, T, and Adlakha, V. Attributes of Service Quality: The Consumers Perspective, International Journal of Service Industry Management (3:3), 1992, pp [20] Parasuraman, A., Zeithaml, V.A., and Berry, L.L. SERVQUAL: A Mutiple-Item Scale for Measuring Consumer Perceptions of Service Quality, Journal of Retailing (64:1), 1988, pp [21] Robertson, T.S., Zielinski, J., and Ward, S. Consumer Behavior, Scott, Foresman and Company, Glenview, IL, [22] Straub, D.W. "Validating Instruments in MIS Research ", MIS Quarterly(13:2), 1989, pp [23] Sullivan, J.R., and Walstrom, K.A. Consumer Perspectives on Service Quality of Electronic Commerce Web Sites, Journal of Computer Information Systems (41:3), 2001, pp [24] Sweeney, J.C., Soutar, G.N., and Johnson, L.W. The Role of Risk in the Quality-Value Relationship: A Study in a Retail Environment, Journal of Retailing (75:1), 1999, pp [25] Tsikriktsis, N. Does Culture Influence Website Quality Expectations? An Empirical Study, DSI 2002 Proceddings, 2002, pp [26] Tornatzky, L.G., and Klein, K.J. Innovation Characteristics and Innovation Adoption- Implementation: A Meta-Analysis of Findings, IEEE Transactions on Engineering Management (EM-29:1), 1982, pp [27] VNU emedia, Web Effect, Potentials (36:2), 2003, p. 7. [28] Wolfinbarger, M.F., and Gilly, M.C..comQ: Dimensionalizing, Measuring and Predicting Quality of the E-tail Experience, Working Paper No , Marketing Science Institute, Cambridge, MA, [29]Yoo, B., and Donthu, N. Developing a Scale to Measure the Quality of an Internet Shopping Site (SITEQUAL), Quarterly Journal of Electronic Commerce (2:1), 2001, pp [30] Zeithaml, V.A., Parasuraman, A., and Malhotra, A. A Conceptual Framework for Understanding E- Service Quality: Implications for Future Research and Managerial Practice, Working Paper No , Marketing Science Institute, Cambridge, MA, /04 $17.00 (C) 2004 IEEE 10