Personalized Services in Online Shopping: Enjoyment and Privacy

Transcription

1 Personalized Services in Online Shopping: Enjoyment and Privacy Ilias O. Pappas, Michail N. Giannakos, Vassilios Chrissikopoulos Department of Informatics Ionian University Corfu, Greece Abstract-Enjoyment and privacy are essential ingredients for successful personalization. However, the current understanding of the influence of personalization is limited. This study extends personalization literature into the area of enjoyment and privacy issues related to intention to purchase and into the context of online shopping. Responses from 148 online customers were used to examine the effects of personalization on enjoyment, privacy issues and intention to purchase. Results show that personalization affects positively enjoyment and intention, but has no effect on privacy. Additionally, privacy affects negatively both enjoyment and purchase intentions, while enjoyment has a positive influence on purchase intentions. This study suggests future research directions including the relation of trust with privacy, as well as different types of emotions, hedonic and utilitarian dimensions that relate with personalized services. Keywords-personalization; enjoyment; privacy; online shopping; intention I. INTRODUCTION Marketing and advertizing use personalization in an attempt to communicate in various levels and ways with the customers, as all the collected data can be used to offer better services. The use of transactional, demographic and behavioral data makes it a great way to offer personalized services to different customer categories. Online personalized services influence customers' reactions, making personalization a reliable mean that will lead customers into buying more often. Online personalized environments may affect positively customer's experience and increase his loyalty [I]. Online personalization can help companies to build strong relations with their customers that will last for a long time. Retaining a customer is much more profitable for a company than finding a new one and keeping a customer happy by satistying his need increases his intentions to purchase again [2][3]. Personalization is based on customer's personal information. Hence, the more information a company knows about the customer, the better services it can offer. The goal of every company using such systems is not just to fulfill consumers' needs but also to create needs that the same company will fulfill in time. Therefore, they need to know as much as possible for their customers. Nevertheless, at this point privacy concerns start to affect people in a significant way, resulting in negative relation between privacy and purchase behavior [4]. Privacy concenrs when shopping online, like the perception of loosing control, are related with anxiety which means less shopping enjoyment for the customer [5]. This paper offers evidence on the use of personalized services in online shopping, if and how they raise customers' privacy issues, how they affect customers' enjoyment and their intention to shop online based on them. Moreover, the influence of privacy on enjoyment and intention to purchase is examined, as well as the relation of enjoyment with intention. It is very interesting to study personalization as it may aid in solving the problem of choice; customers who shop online have to choose among a vast number of products. Although personalization technologies have been adopted in electronic commerce over the past years [6], the roles of personalization, enjoyment and privacy together in online shopping are understudied. The objective of this paper is to explore the relationships between personalization, enjoyment, privacy and customer intention to purchase online. In particular, we developed a research model, which examines the relations among personalization, enjoyment and privacy and how every one of these factors influences customers' intention. The remainder of the paper is organized as follows. In the next section we review the existing literature on personalization, enjoyment, privacy and intention to purchase. In section 3, we present the theoretical foundation of the research model. In section 4, we present the methodology and the measures adopted for collecting data on the online shopping behavior. The next section presents the empirical results derived. In the last section of the paper, we discuss the [mdings and conclude by providing theoretical and practical implications and make recommendations for future research. II. LITERATURE REVIEW Personalization is defined, among others, as "the ability to provide content and services tailored to individuals based on knowledge about their preferences and behavior" and "the use of technology and customer information to tailor electronic commerce interactions between a business and each individual customer" [7]. Nowadays, personalization has an important role for online vendors when they plan their strategies regarding the services that they offer. It is important to create a customer relationship 168

2 that will give them the chance to communicate with the customers directly and personally. Trying to offer personalization in e-commerce means dealing with three problems: identitying a customer, gathering information about the customer, and processing data to create a service which will be customized for a certain customer (e.g. recommender systems) [8]. Personalization's goal is to indentity and satisty a special customer or user need. E-commerce, and its transactions which involve selling and buying, is a field of application offered for using personalization methods and techniques. Hence, e commerce beyond eshops, includes recommendation systems and comparison agents. The broad field of personalization has been widely researched [8][9]. It has focused, among others, on recommender systems [10]. Most recommendation tools ask too much information from the customer. Although these tools seem like a very good solution to the problems of too much information and too many choices, they are not widely accepted from consumers [11]. Particularly, these tools will first ask a number of questions in order to learn the customers' needs and preferences and eventually provide the proper recommendation. However, gathering and using consumers' information has a downside; it may be considered as an invasion to the consumers' privacy. Salomann et al.[12], pointed out the importance of privacy and security aspects in the field of customer relationship management (CRM). Data from consumers can be collected either by directly asking them (explicitly) or by inferring information about them without their knowledge or consent (implicitly) [13]. When customers are asked for their personal data it is very likely to share them with the company in order to gain access to its service, although, many companies do not ask them and customers realize the privacy violation only when the first personalized s arrive [14]. Researchers have tried to understand how customers evaluate their privacy and how they decide to trade it for tangible or intangible benefits [15]. Such benefits are those offered through personalization and recommendation tools. Murray et al.[i 6], argue that no matter the great usefulness of any recommendation tool, their ease of use seems to be more important for the customer. Research has focused on making recommendation agents more useful in an attempt to offer more specific and personalized services, because it is clear that the more useful a technology is the more likely is to be adopted. Moreover, until now there is limited research on the acceptance of recommendation systems from customers and their effect on customers' loyalty [17]. Shopping enjoyment, and by extent enjoyment when receiving personalized services, plays an important role in customers behavior towards online shopping [18]. Enjoyment is related to many types of shopping, while the experience of the shopping procedure is defined by the customers' beliefs and affects the way they feel about it [19]. Some people may shop when they are in a bad mood as they enjoy the process and the result, while others maybe thrilled when they find new deals or acquire products that will make them feel popular. TTT. HYPOTHESES AND MODEL DEVELOPMENT The aim of the study is threefold; first the study investigates the influence of personalization on privacy, enjoyment and intention to purchase, then the relationships between privacy, enjoyment and intention to purchase and finally the relationship between enjoyment and intention. Tn our study, grounded on the view of the literature we adopt the respective constructs and develop the research model. A. Personalization Personalization is important for a customer that his ultimate goal is to create a relationship with an online vendor. As stated before personalized services aim to aid the customers decide when it comes to online purchases by recommending certain products or services. Komiak and Benbasat [20], posit that higher personalized recommendation agents can offer better services to a customer because they understand him and his preferences in a better way. Also, personalization attitudes, depending on the provided service, have a positive effect on customers' intention to purchase [21]. Based on the fact that personalization requires a lot of private data from the customer, while its goal is to provide a solution to his problem of choice, the customer may develop mixed emotions towards it and towards the services that are provided [22], while shopping online is a procedure that customers enjoy [19]. Nevertheless, asking a customer for personal information may lead to privacy violations. Customers have to choose between their privacy and personalized services, but eventually they have to trade it in order to shop online [15]. Hence, we propose that: Hi: Personalization will have a negative impact on privacy. H2: Personalization will have a positive impact on enjoyment. H3: Personalization will have a positive impact on intention to purchase. B. Privacy The paradox of privacy in online shopping is that the better services the customer wants, based on his needs, the more personal information he has to reveal. Previous studies revealed that customers feel that their privacy has been violated when they receive recommendations based on previous purchases or even their browsing habits [23]. Feeling such a violation can diminish the enjoyment that occurs from purchasing online. It has been found that privacy concerns influence negatively customers' online behavior and attitude [24]. propose that: H4: Privacy will have a negative impact on enjoyment. Hence, we H5: Privacy will have a negative impact on intention to purchase. 169

3 Privacy H5 Intention to Purchase HI H6 Personal ization H2 Enjoyment Figure I. Research Model C. Enjoyment Various customer emotions arise during consumption and affect their behavior. In online shopping, positive emotions such as enjoyment, are important because they affect customers' future behavior [25]. The enjoyment that a customer feels when shopping online, affects positively his attitude towards online shopping. Moreover, previous studies have showed that enjoyment has a strong positive effect on customers' intentions [26]. Shopping enjoyment has an important effect on customers behavior [18]. Hence, propose that: H6: Enjoyment will have a positive impact on intention to purchase. In Fig. 1, we present the research model based on the aforementioned theoretical and empirical research. In addition Fig. 1 clearly states the set of hypotheses that have been formulated and examined in this paper. A. Sampling IV. METHODOLOGY Our research methodology included a survey conducted through the delivery and collection of individual questionnaires. A number of different methods were recruited for attracting respondents; questionnaires distributed in various places (universities, public areas) and s were sent to different mailing lists. It was made clear that there was no reward for the respondents and the participation was voluntary. The survey was open during the last two weeks of November We aimed at about 600 Greek users of online shopping, 148 of which fmally responded. As Table I shows, the sample of respondents was composed of almost equally men (50.7%) and women (49.3%). In terms of age, the majority of the respondents (34.5%) were between 25 and 29, while the second more frequent age group (26.4%) involved people between 18 and 24. Finally, the vast majority of the respondents (84.4%) included graduates or post-graduate students. we TABLE 1. Users' demographic profile Demographic Profile No % B. Measures Gender Male % Female % Marital Status Single % Married % Divorced 5 3.4% Age % % % % % Edncation Middle School I 0.7% High School % University % Post Graduate 38.5% The questionnaire was divided into two parts. The first part included questions on the demographics of the sample (age, gender, education). The second part included measures of the various constructs identified in the literature review section. Table IT lists the operational definitions of the constructs in this theoretical model as well as the studies from which the measures were adopted. In all cases, 7-point Likert scales, in which 1 indicates "completely disagree" and 7 "completely agree", were used to measure the model's variables. TABLE II. Constrnct 57 Construct definition and instrument development Operational Definition Sonrce Measuring the customers' privacy Privacy (PR) issues when using personalized [6] services. ITailoring content and services to Personalization (PER) atch the buyer's personal [27] interests or preferences. Measuring the customers' Enjoyment (ENJ) njoyment when using [26] personalized services. Intention to Purchase onsumers' intention to shop pnline based on personalized (INT) services. [28][29] 170

4 TABLE III. Descriptive statistics and correlations of latent variables C. Data Analysis Structural equation modeling was conducted using AMOS version 18.0 software, based on [30]. At first, a measurement model was created based on a confirmatory factor analysis, and then the structural model was built in order to test the hypothesized relationships. 1) Reliability and validity: The constructs used in this research were evaluated in terms of reliability and validity. Reliability was examined based on the Cronbach alpha indicator which needs to be higher than 0.7 for every factor. Validation analysis consists of convergent and discriminant validity. Establishing validity requires that Average Variance Extracted (AVE) should be greater than 0.50 [31], the correlation between the different variables in the confirmatory models should not exceed 0.8 points, as this would suggest low discrimination, and that the square root of each factor's average variance extracted (AVE) is larger than its correlations with other factors [30]. 2) Goodness of fit: Goodness of fit indices describe how well the model fits its data. Here, several fit indices were used to assess model-data fit. Root mean square error of afproximation (RMSEA), comparative fit index (CFI) and X /df ratio were all used to evaluate model-data fit [27]. RMSEA less than 0.05 suggests good model-data fit; between 0.05 and 0.08 it suggests reasonable model-data fit and between 0.08 and 0.01 suggests acceptable model data fit. CFI indices greater than 0.90 suggest good model-data fit and greater than 0.80 suggest adequate model-data fit. A l/df ratio less than 3 is acceptable. V. FINDINGS First, an analysis of reliability and validity was carried out. Reliability testing, based on the Cronbach alpha indicator, as seen on Table Ill, shows acceptable indices of internal consistency because they all exceed the cut-off level of Validation testing results can be seen on table 3. AVE for all constructs is higher from the level of 0.50, all correlations are lower than 0.80 and square root AVEs for all constructs are larger from their correlations. Constrnct Construct Mean SD CR AVE PER PR ENJ INT PER PR ENJ INT Note: Diagonal elements (in bold) are the square root of the average variance extracted (AVE). Offiagonal elements are the correlations among constructs. For discriminant validity, diagonal element should be larger than off-diagonal elements. PER, Personalization; PR, Privacy; ENJ, Enjoyment; INT, ntention to Purchase The fit indices of the research model are presented on table 4. All values are within the recommended range. SpecifIcally, X 2 /df= 1.52, CFT = 0.98 and RMSEA = TABLE IV. Overall model fit indices for the model Model fit indices Results Recommended Value X 2 /df 1.52 (X 2 = 72.9; df = 48) <= 0.3 CFI 0.98 >= 0.9 RMSEA 0.06 <= 0.08 The estimated path coefficients of the structural model were examined in order to evaluate the hypotheses. Figure 2 exhibits the analysis of the research model. The results are as follows. Personalization has a significant effect only on enjoyment and intention to purchase. It affects positively both enjoyment supporting H2, and intention to purchase, but with less significance, supporting H3. On the other hand, personalization has no significant effect on privacy rejecting HI. Regarding the effects derived from privacy, we found that a significant effect on both enjoyment and intention to purchase is presented. Privacy (R 2 =0.00) -0.16** I ntenti on to Purchase (R 2 =0.72) *** Personalization I I ***p<o.ool; **p<o.ol; *p<o.l 0.65** Figure 2. SEM Analysis of the research model Enjoyment (R 2 =0.37) 171

5 This effect is negative on both enjoyment, supporting H4, and intention to purchase, supporting H5 (with higher significance). Finally, the effect of enjoyment on intention to purchase was found significant as well. Tn particular, enjoyment affects intention to purchase positively, supporting H6. Square multiple correlations (R 2 ) are presented on figure 2 as well. The R 2 of privacy was zero as its only predictor had no effect on it. The R 2 of enjoyment was 0.37 and that of intention to purchase was Values higher than 0.26, imply high effect of the predictors of enjoyment and intention to purchase respectively. VI. DISCUSSION AND CONCLUSIONS Tn order to understand customers' intentions towards online shopping we explored the relations between personalization, privacy, enjoyment and intention to purchase. Our findings indicate that while personalized services do not affect customers' privacy issues, they influence their enjoyment as well as their intention to purchase. Tn particular, personalization does not influence privacy, but has a positive effect on both enjoyment and intention to purchase. Additionally, we found that privacy has negative effects on enjoyment and intention to purchase, while enjoyment has a positive effect on intention. Our results further demonstrate that customers shopping online will raise their enjoyment if they use personalized services (HI). This may be due to the way these services work, offering the right product at the right time, making the process easier and more enjoyable. Similarly, personalization will raise the intention of customers to shop online (H3), probably because it answers to their needs by providing the right services or products, similar to the findings of [21]. Nevertheless, we found that personalization has no significant effect on the privacy issues of the customers (HI), maybe because when individuals choose to use personalized services they accept and expect some privacy violations, for which they probably have already been informed. Moreover, studies have showed that when customers are informed properly and assured for their privacy they are willing to share they personal information in order to receive personalized services [32]. Our results contradict with previous studies that have found negative relation between privacy and personalization [24]. Regarding privacy, it has a negative effect on both enjoyment (H4) and intention to purchase (H5), which was expected as having high privacy concerns, being unwilling to share information while it is the only way to shop online, will reduce the positive feelings that occur when shopping online reducing enjoyment and future intentions. Finally, enjoyment has a significant effect on intention to purchase (H6), which was expected because the better a customer feels with a service the more likely he is to use it. Our results concur with previous studies, which mention that shopping enjoyment has an important effect on customers' behavior [18]. As such, all our hypotheses except HI are supported. Our empirical research has addressed several shortcomings of previous studies in the area. Specifically, [32] study the relations between privacy and personalization when the customer reveals willingly his personal information, but do not take into account how that affects his shopping enjoyment. Moreover, [26] examine the effects of enjoyment on online purchasing intensity, but they do not study the effects of privacy issues on enjoyment or online shopping intentions. Lee and Park [21], study the influence of online service personalization on purchase intentions taking into account the effects of subjective norms on customers' attitudes. Nevertheless, they do not measure shopping enjoyment, which is related with subjective norms. Our study is one of the few so far that includes personalization as a predictor variable, while examining its relation with privacy and enjoyment. Our findings provide useful insights for both practitioners and researchers. The results point out the importance of personalization on customers shopping enjoyment, which in tum affects strongly intention to purchase. Hence, providing personalized services is important to increase enjoyment and purchase intentions. Personalization was not found to have a significant effect on privacy issues, adding to the statement that when personalized services are offered properly and the individual is well informed, using such services does not affect his privacy issues. Nevertheless, the issue remains that with high privacy comes low enjoyment and intention to purchase. We provide evidence that using personalization as an online marketing tool, which will act as it promises without exploiting customers' personal data, will help practitioners to keep their customers happy, which may eventually lead to spending more money and making more purchases. As with all empirical studies, there are some limitations. First, the empirical data used in this study include mostly Greek online shopping customers. Second, the findings are based on self-reported data; other methods (e.g. depth interviews, observations) could provide a complimentary picture of the findings. Third, we only investigate enjoyment while, some other aspects of individual emotions, such as anxiety, anger or happiness that have been proved to affect people's behavior [33] were not included in the study. Further research on these aspects would valuable contribute on the understanding of customers' emotions on personalized services. Tn addition, it would be interesting to see how privacy would affect trust issues towards the online vendors that offer personalized services and if and how intention to purchase leads to actual purchases. Hedonic and utilitarian shopping motivations could be included as well. ACKNOWLEDGMENT The authors wish to thank the respondents to the survey who kindly spent their time and effort in sharing their valuable experience. REFERENCES [ I] L. Zhou, L. Dai, and D. Zhang, "A critical survey of consumer factors in online shopping, " Marketing Journal, vol. 8, pp , [2] C. Liao, P. Palvia, and H. N. Lin, "The roles of habit and web site quality in e-commerce, " International Journal of Information Management, vol. 26, pp , [3] M. N. Giannakos, A. G. Pateli, and I. O. Pappas, "Identifying the direct eflect of experience and the moderating effect of satisfaction in the 172

6 greek online market, " International Journal of E-Services and Mobile Applications, vol. 2, pp , 201l. [4] 1. Castaneda, and F. Montoro, "The effect of internet in general privacy concern on customer behavior, " Electronic Commerce Research, vol. 7, pp , [5] Y. Hwang, and 1. D. Kim, "Customer self-service systems: The effects of perceived Web quality with service contents on enjoyment, anxiety, and e-trust, " Decision Support Systems, vol. 43, pp , [6] S. Y. Ho, and K. K. W. Ho, "The Eflects of web personalization on influencing users' switching decisions to a new website, " PACIS 2008 Proceedings. Paper 67, [7] G. Adomavicius, and A Tuzhilin, "Personalization techniques: a process-oriented perspective, " Communications of the ACM, vol. 48, pp , [8] C. Adolphs, and A Winkelmann, "Personalization research in e commerce - a state of the art review ( ), " Journal of Electronic Commerce Research, vol. II, pp , [9] A V. Bodapati, "Recommendation systems with purchase data, " Journal of Marketing Research, " vol. 45, pp , [10] X. Bo, and I. Benbasat, "E-commerce product recommendation agents: Use, characteristics, and impact, " MIS Quarterly, vol. 31, pp , [11] K. B. Murray, and G. Haubl, "Interactive consumer decision aids, " In B. Wierenga (Ed.), Handbook of Marketing Decision Models. New York, NY: Springer Science + Business Media, 2008, pp [12] H. Salomann, M. Dous, L. Kolbe, and W. Brenner, "Rejuvenating customer management: How to make knowledge for, from and about customers work, " European Management Journal, vol. 23, pp , [13] 1. Cranor, P. Guduru, and M. Arjula, "User interfaces for privacy agents, " ACM Transansactions Computer-Human Interaction, vol. 13, pp , [14] K. M. Sheehan, "Towards a Typology of Internet Users and Online Privacy Concerns, " The Information Society, vol. 18, pp , [15] J. Y. Tsai, S. Egelman, 1. Cranor, A. Acquisti, "The effect of online privacy information on purchasing behavior: an experimental study, " Information Systems Research, vol. 22, pp , [16] K. B. Murray, and G. Haubl, "Personalization without interrogation: Towards more eflective interactions between consumers and featurebased recommendation agents, " Journal of Interactive Marketing, vol. 23,pp , [17] D. Baier, and E. Stuber, "Acceptance of recommendations to buy in online retailing, " Journal of Retailing and Consumer Services, vol. 17, pp , [18] S. Ha, and L. Stoel, "Consumer e-shopping acceptance: Antecedents in a technology acceptance model, " Journal of Business Research, vol. 62, pp , [19] A G. Close, and M. Kukar-Kinney, "Beyond buying: Motivations behind consumers' online shopping cart use, " Journal of Business Research, vol. 63, pp , [20] S. Y. X. Komiak, and I. Benbasat, "The effects of personalization and familiriaty on trust and adoption of recommendation agents, " MIS Quarterly, vol. 30, pp , [21] E. J. Lee, and J. K. Park, "Online service personalization for apparel shopping, " Journal of Retailing and Consumer Services, vol. 16, pp , [22] Y. C. Chen, R. A. Shang, and C. Y. Kao, "The effects of information overload on consumers' subjective state towards buying decision in the internet shopping environment, " Electronic Commerce Research and Applications, vol. 8, pp , [23] G. Roussos et ai., "A Case Study in Pervasive Retail, " Proc. 2nd Int "l Workshop Mobile Commerce (MOBICOM WMC 02), ACM Press, pp , [24] A N. Joinson, and C. B. Paine, "Self-Disclosure, Privacy and the Internet, " In AN Joinson, K.Y.A McKenna, T. Postmes and U-D. Reips Eds. Oxford Handbook of Internet Psychology, Oxford University Press, 2007, pp [25] S. Menon, and B. Kahn, "Cross-category eflects of induced arousal and pleasure on the internet shopping experience " Journal of Retailing, vol. 78, pp , [26] C. Liu and S. Forsythe, "Examining drivers of online purchase intensity: Moderating role of adoption duration in sustaining post-adoption online shopping, " Journal of Retailing and Consumer Services, vol. 18, pp , [27] H. Xu, X. Luo, J. Carroll, and M. B. Rosson, "The personalization privacy paradox: An exploratory study of decision making process for location-aware marketing, " Decision Support Systems, vol. 51, pp , [28] Y. T. Chen, and T. Y. Chou, "Exploring the continuance intentions of consumers for B2C online shopping: perspectives of fairness and trust " Online Information Review, vol. 36, pp , [29] Y. Lu, Y. Cao, B. Wang, and S. Yang, "A study on factors that affect users' behavioral intention to transfer usage from the offline to the online channel " Computers in Human Behavior, vol. 27, pp , 201l. [30] B. Byrne, Structural Equation Modeling with AMOS (Multivariate Applications Series). Routledge Academic; 2 edition, New York, [31] C. Fornell, and D. Larcker, "Structural equation models with unobservable variables and measurement error, " Journal of Marketing Research, vol. 18, pp , 198 I. [32] C. H. Lee, and D. A Cranage, "Personalisation-privacy paradox: The eflects of personalisation and privacy assurance on customer responses to travel Web sites, " Tourism Management, vol. 32, pp , [33] M. N. Giannakos, K. Chorianopoulos, K. Inkpen, H. Du, and P. Johns, "Understanding Children's Behavior in an Asynchronous Video Mediated Communication Environment, " Personal and Ubiquitous Computing, (In Press), doi: /s