Social Media

A flow-based model of web site intentions when users customize products in business-to-consumer electronic commerce

Description
A flow-based model of web site intentions when users customize products in business-to-consumer electronic commerce
Categories
Published
of 13
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/220199012 A Flow-based model of web site intentionswhen users customize products in business-to-consumer electronic commerce  Article   in  Information Systems Frontiers · April 2010 DOI: 10.1007/s10796-008-9135-y · Source: DBLP CITATIONS 29 READS 204 3 authors: Arnold KamisSuffolk University 53   PUBLICATIONS   464   CITATIONS   SEE PROFILE Tziporah SternCIty University of New York - Kingsborough C… 7   PUBLICATIONS   155   CITATIONS   SEE PROFILE Daniel M. LadikSeton Hall University 14   PUBLICATIONS   143   CITATIONS   SEE PROFILE All content following this page was uploaded by Arnold Kamis on 03 December 2016. The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the srcinal documentand are linked to publications on ResearchGate, letting you access and read them immediately.  A flow-based model of web site intentions when userscustomize products in business-to-consumerelectronic commerce Arnold Kamis  &  Tziporah Stern  &  Daniel M. Ladik  # Springer Science + Business Media, LLC 2008 Abstract  User-customization is increasingly common inelectronic commerce, because both the buyer and seller  potentially benefit. The user interface to implement and theinfluence of the interface on various process and outcomemeasures, however, are not well understood. We developed aFlow-based model consisting of seven hypotheses regardingthe user interface and its consequents. We conducted a fieldexperiment to test an attribute-based interface vs. a question- based interface on three variables (perceived control,shopping enjoyment and choice satisfaction) as well as twoweb site intentions: intention to return and intention to purchase. Six of the seven hypotheses were supported in a parsimonious model. Variance explained was 16.3% for  perceived control, 45.6% for shopping enjoyment, 59.3% for choice satisfaction and 63.1% for web site intentions. Themain finding is that an attribute-based interface for retail e-shopping increasesthe shopper  ’ s sense of controland feelingof enjoyment in the process more than a question-basedinterface, and thereby increases satisfaction with the out-come. This combination of influences increases the intentionof the shopper to return to the web site and to purchase theitem. We discuss the results and suggest areas for futureresearch in user-customization, which may apply to manydifferent industries that engage in online commerce. Keywords  Business-to-consumerelectroniccommerce.Flow-basedmodel.Users 1 Introduction Onlinecustomersarenotjustattentivespectators,theyareoften active players as well, supplying labor andknowledge to the service creation process(Bowen1986). Thecustomerparticipationliteraturehasdevelopedtheroleof the consumer as a co-creator or co-producer of products or services(KelleyandSkinner 1990). Customer participation is “ the degree of consumer  ’ s effort, both mental and physical,necessary to participate in production and delivery of ser-vice ”  (Silpakit and Fiskin 1985). Due to the important rolecustomers can play, research has recommended that orga-nizations view customers as partial employees, extending the boundaries of the organization to include customers astemporary members of the firm (Lovelock and Young1979; Mills et al. 1983; Mills and Morris 1986). By self- serving, customers benefit by getting exactly what they want  Inf Syst Front DOI 10.1007/s10796-008-9135-y Note: Yogesh K. Dwivedi, Michael D. Williams and ViswanathVenkatesh were the guest editors accepting the article as part of thespecial issue on Adoption and Use of Information & CommunicationTechnologies (ICT) in the Residential/Household Context (seeDwivedi et al. 2008 for editorial).A. Kamis ( * )Department of Information Systems & Operations Management,Sawyer Business School, Suffolk University,8 Ashburton Place,Boston, MA 02108-2770, USAe-mail: akamis@suffolk.eduT. SternZicklin School of Business, Baruch College,The City University of New York,55 Lexington Avenue at East 24th Street, New York, NY 10010, USAe-mail: tzipk@yahoo.comD. M. Ladik Department of Marketing, Sawyer Business School,Suffolk University,8 Ashburton Place,Boston, MA 02108-2770, USAe-mail: dladik@suffolk.edu  in a process under their control, and the organization benefits by selling at a lower cost or a premium price. By offeringuser-led self-customization, businesses may increase searchefficiency (Diehl et al. 2003), reduce the user  ’ s cognitiveeffort  (Nielsen 2000) and improve decision quality (Häubl and Trifts 2000), thus improving the likelihood of actual purchase (Moe and Fader  2004). An important element of user-led customization is theelicitation of customer needs (Wind and Rangaswamy 2001;Zipkin 2001). It involves both the discovery by customersof their own needs as well as their communication of thoseneeds to the company. Therefore, a critical question for online companies is which type of interface will be moreappropriate for the needs elicitation process. The right user interface may result in increased sales and customer loyalty(Berman 2002), whereas an inappropriate user interface could result in frustrated customers and lost sales.In this study, we deploy Flow as a theoretical lens toexamine a nomological network in the online shoppingcontext (Agarwal and Karahanna 2000; Koufaris 2002; Thatcher and Perrewe 2002; Pennington et al. 2003-2004). Flow is the phenomenon of deep engagement or immersionin a focused task, balancing challenges and skills whilehaving a sense of control and losing self- and time-awareness (Csikszentmihalyi 1975; Csikszentmihalyi 1977; Csikszentmihalyi and Csikszentmihalyi 1988). The Flowliterature has been used to describe and model the individualexperience in IT use (Agarwal and Karahanna 2000), e- commerce (Novak  2000; Koufaris 2002) and interactive entertainment (Polaine 2005). The voluntary online shoppingcontext is quite different from the mandatory, organizationalcontext in which the technology acceptance literature has become mature (Davis 1989; Venkatesh et al. 2003; Venkatesh and Agarwal 2006). In this study, we develop aFlow-based model which augments traditional User InterfaceDesign (UID) research by including three user-focusedvariables: perceived control, shopping enjoyment and choicesatisfaction. That is, our model focuses on the user  ’ s perceptions of different UIDs for decision support in theonline shopping context.We tested two types of user interfaces: an Attribute-BasedUID and a Question-Based UID. For an example of anAttribute-Based UID, consider the Dell configurator, whichenables users to self-customize computers, printers, tele-visions and other goods by supporting component- by-componentsubstitutions.ForexamplesofQuestion-BasedUID, consider travel recommender systems, software require-ments engineering or medical expert systems. Both Attribute-Based UID and Question-Based UID can be used for user-ledcustomizationandproduct personalization,a growingtrendinelectronic commerce (Pine 1993; Tu et al. 2004). An Attribute-Based UID is one in which the customer is presented with a list of attributes and the different values for each attribute. The customer then chooses a combination of the different values and is presented with a view of thecustomized product. Customers can iteratively choose differ-ent values and experiment until they feel they have developedan acceptable product. This has been an established form of customization for many products including sneakers, back- packs, watches, breakfast cereals and computers. In this typeof customizing experience, the customer can interact with thesystemaswellassee the resultsofhis/herefforts immediately.The essence of Question-Based UID is a natural languagedialogue between the system and the user to facilitatesearching the solution space in an intuitive way and toultimately recommend a product that will suit the customer  ’ sneeds. Hence, in essence, a Question-Based UID is arecommendersystem.Researchhasdemonstratedtheefficacyof recommender systems, particularly in e-commerce wherethey can help personalize the shopping experience andincrease sales (Schafer et al. 2001). Ricci and Werthner  (2006  –  2007) define recommender systems as  “ intelligent applications that assist users in a decision-making processwhen they do not have sufficient personal experience tochoose one item from an overwhelming set of alternative products or services ”  (p. 5). A classical recommendationsystem asks users about their preferences upfront and thenconstructs a query based on their responses to retrieve products that match the user  ’ s taste. It uses knowledge about the products and consumers ’  past buying behavior anddemographics to recommend products the customer might like in order to help reduce the complexity of choosing a product (Schafer et al. 2001; Ricci and Werthner  2006  –  2007). This type of system can be helpful in situations wherethere is an overwhelming number of options and the user may find it difficult to pinpoint which combination of attributes suits him or her. For example, one way recom-mender systems are used on the Web is to suggest movies or music to a customer based on a series of questions or ratings,e.g., Netflix. Question-Based UIDs can be used to recom-mend products that are difficult to represent on the web, suchas fragrance, e.g. Urbanbotanic.com and Gurl.com. In 1999,Proctor and Gamble ’ s Reflect.com was the first web-site tooffer customized cosmetics based on customers ’  response to personality, taste and feelings. Finally, Zafu.com recom-mends jeans based on customers ’  answers to a series of questions about styles and fit. Therefore, the researchquestion is: How does a naturalistic, intuitive UID, theQuestion-Based UID, compare to a sophisticated, interactiveuser interface, the Attribute-Based UID?In the next section, we introduce the research model, present the study ’ s variables, and develop our hypotheses.We follow with the description of the research design andmethods, data analysis, a discussion of the results, our study ’ s contributions and limitations, and possibilities for future research. Inf Syst Front   2 Research model and hypotheses This study integrates research in Information Systems,Human-Computer Interaction and Consumer Behavior toexamine the effects of Attribute-Based UID vs. Question-Based UID based on Flow (Detlor et al. 2003; Johnson andWiles 2003; Jiang and Benbasat  2005; Qiu and Benbasat  2005). Detlor et al. (2003) investigated the inducement of  Flow based on the different informational needs of searching vs. browsing. Johnson and Wiles (2003) took  the domain of computer games as one conducive tounderstanding Flow, so that one can improve affectivedesign in other domains. Jiang and Benbasat (2005) tested a user interface in terms of virtual and functional control withthe aim of increasing the likelihood of Flow. Qiu andBenbasat (2005) studied different user interfaces, specifi-cally text-to-speech and three-dimensional avatars, asmeans of enhancing Flow. An important missing piece inthis stream of literature, however, is the examination of therelationships among process and outcome variables, whileusing different UIDs, which influence two web siteintentions: intention to return and intention to purchase.Studies of preferential choice tasks, such as this one, allowusers to create a product that satisfies their needs or desires by processing multiple types of information (Payne 1976; Toddand Benbasat  1992; Zigurs and Buckland 1998). The decision-making process can have multiple alternativesolutions, which the users may compare and evaluate beforeselecting the most desirable one. Based on Flow, we expect that a technology which is more interactive and engagingwill result in a more strongly preferred, i.e., more desirable,outcome. Research has found that customers who use anAttribute-Based UID should feel that they have all theavailable information for them to make a decision, therebyincreasing their feeling of control (Kahn 1998). One reason is that with an Attribute-Based UID, users can see the products ’  attributes, e.g., the watch face design, or watch- band types and their values, e.g. leather or stainless steel,immediately. Hence, they can clearly see what the customiz-able elements of the product are, choose the customizationthey want and see the results interactively. In contrast, with aQuestion-Based UID, the system first requires the user toanswer a series of questions, e.g., about style preferences,and then recommends a choice of customized products fromwhich the users could only infer the customizable attributesalong with the different possible values for each one.The experience of Flow often occurs in the presence of important facilitators such as perceived control, personalenjoyment, and satisfaction. The first of these, perceivedcontrol, is widely accepted as a human driving force and hasoften been defined as the need to demonstrate one ’ scompetence and mastery over the environment (White1959). In dynamic environments where the user is guided by self-motivated goals, Bandura maintains that performanceon a task varies as a function of the individual ’ s perceivedcontrol regarding that task (Bandura 1982; Bandura 1986). When people perceive that they have a high degree of control, they exert cognitive effort and persist in the face of failures and setbacks; they evince interest, optimism andsustained attention while problem solving. When peopleexperience a loss of control, they withdraw, retreat, escape,or otherwise become passive. In this sense, people will seek out enjoyable situations they feel they are capable of managing and avoid uncomfortable situations they feel are beyond their abilities. This pattern of reactions forms thecornerstone for all major theories of perceived control.Perceived control in the e-commerce context is defined asthe degree to which customers or service providers perceivethat they are able to influence the process and outcome. Withmore control, customers feel more responsibility for, andsatisfaction with, the process and outcomes. In addition, perceived control may be more pertinent in a self-servicesetting in the absence of any front-line service personnel, as theconsumer must perform the entire service for him- or her-self.Serviceencounterresearchconsidersperceivedcontrolacrucialelement of e-service quality (Zeithaml et al. 2002). Thus, H1 Perceived control is greater for an Attribute-BasedUID than for a Question-Based UID.Many different user interfaces have been developed tosupport the e-shopper in making choices, including example- based (Pu and Kumar  2004), list-based and matrix-basedrepresentations (Häubl and Trifts 2000; Hong et al. 2005) and recommender systems (Ziegler and Golbeck  2007). Thetype of interface, e.g., three dimensional representations of  products or sales agents such as avatars can increase choicesatisfaction (Holzwarth et al. 2006; Yen and Ng 2007). Interface designs vary from search engines, which focus on price-comparison for homogenous products (Bakos 1997;Brynjolfsson and Smith 2000), to user-supporting tech-niques, which offer greater user-control and decisiontransparency for comparing heterogeneous products (Puand Chen 2005; Wujin et al. 2005). For homogeneous  products, a simple list of alternatives may be an appropriatedesign, because price may be the only differentiator. For heterogeneous products, however, the interface may have to be richer to convey several attributes simultaneously or to provide the user support in examining attribute tradeoffs(Lingyun and Benbasat  2005; Jahng et al. 2006). Some research has divided the choice making process intomultiple, sequential stages, supporting each one differently(Gilbride and Allenby 2004; Gilbride and Allenby 2006; Kamis 2006). The success of product configurators and user-led customization demonstrates that the sequential stages can be many and brief  (Pine 1993; Wan 2000). Overall, research shows that users benefit from UIDs which describe Inf Syst Front   alternatives according to meaningful attributes that can beeasily narrowed to the ones most relevant to the user. AnAttribute-Based UID makes the different customizableattributes clear to the user, whereas the process is moreopaque with a Question-Based UID. Thus,H2 Choice Satisfaction is greater for an Attribute-BasedUID than for a Question-Based UID.Since users are boundedly rational, they tend to satisficerather than optimize when making decisions (Simon 1972;Simon 1982). They need control over the process to feelable to assess progress to a satisfactory outcome, after which they will likely stop. They do not typically want toexamine every piece of information and every possiblechoice, achieving the optimal choice, because they wouldeither feel information overload or it would simply take toomuch time and effort, i.e., not be worth it. To aid progress,they may choose to use some kind of recommendationagent, e.g., by responding to a Question-Based UID (Wangand Benbasat  2005; Bo and Benbasat  2007), which is  presumed to ask appropriate questions. Alternatively, anenjoyable, Attribute-Based UID requires only incrementaleffort for each iteration, and offers the user a great deal of control. After each iteration, the user is able to reflect uponthe current choice, decide how satisfied he or she is, assesshow much control he has to improve the outcome, anddecide whether to continue (Urbany 1986) or whether to defer choice, i.e., choose not to choose (Dhar  1996), or  whether to terminate the process with a satisfactory overalloutcome, i.e., the choice. Thus, regardless of the UID,H3 Perceived control is positively related to choicesatisfaction.Shopping enjoyment, also based on Flow, has been usedas a hedonic measure of the individual experience whenusing information technology (Agarwal and Karahanna2000), in web environments (Koufaris 2002) and in games (Sweetser and Wyeth 2005; Ryan et al. 2006). Shopping enjoyment is defined as the intrinsic enjoyment of theinteraction with the web site and it is an important predictor of technology acceptance in Business-to-Consumer e-commerce (Koufaris et al. 2001  –  2002). Web shopping that is game-like will be one in which the user feels in control,losing self- and time-awareness, and according to the Flowliterature, experiencing commensurate enjoyment. Thus,regardless of the UID,H4 Perceived control is positively related to shoppingenjoyment.According to the Flow literature, enjoyment of the process influences the user  ’ s perceptions, subsequent evaluations of the outcomes and the overall technologyused to produce them (Hoffman and Novak  1996).Research which examines both utilitarian and hedonicfactors within the context of e-service quality shows that  process enjoyment has a significant positive impact onchoice satisfaction and customer lifetime value (Bauer et al.2006). Other e-commerce research shows that shoppingenjoyment is a key component of socio-psychologicalvalue, which is itself a predictor of choice satisfaction(Lee et al. 2003). Research on recommender systems has shown that enjoyment influences the subsequent evaluationof the perceived fit of the recommendation, i.e., the choicesatisfaction (Gretzel and Fesenmaier  2006). Some e- commerce research which separates shopping enjoyment from process value found that the process value contributesto choice satisfaction, but that shopping enjoyment does not (Cai and Xu 2006). We argue that enjoyment is a key component of the process, and therefore should not beconsidered in isolation. An enjoyable process confers a positive feeling to whatever the process produces. Thus,regardless of the UID,H5 Shopping enjoyment is positively related to choicesatisfaction.The online shopper uses technology not for its own sake, butfor a specific purpose, to make a purchase. The likelihoodof purchase, defined as the propensity or probability of making an actual purchase, may vary depending on urgencyof the desire to transact, the helpfulness of the technologysupport, the extent of the user  ’ s product knowledge and other factors. Assuming the desire to purchase is present, thetechnology is helpful rather than obstructive, and the user  ’ s product knowledge is sufficient, the intention will besignificant, if not sufficient to execute the purchase. Whenthe user is satisfied with the choice and no longer wishes tocontinue searching for something better, he has, according tothe Theory of Planned Behavior (Fishbein and Ajzen 1975;Pavlou and Fygenson 2006), some positive web siteintentions: to make the purchase immediately or to returnto the web site later to possibly make the purchase. Thus,regardless of the UID,H6 Choice satisfaction is positively related to web siteintentions.Shopping enjoyment has been found to be a significant  predictor of a user  ’ s behavioral intentions, including theuser  ’ s intention to return to a web site (Koufaris 2002), intention to use it  (Van der Heijden 2004), intention to  purchase (Yen and Ng 2007) and intention to repurchase (Bauer et al. 2006). Thus, regardless of the UID, H7 Shopping enjoyment is positively related to web siteintentions.The overall nomological net for the theorized ante-cedents of web site intentions are presented in Fig. 1. Inf Syst Front 
Search
Similar documents
View more...
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks