Electronic service quality and its determinants under different levels of consumer product involvement.

Electronic service quality and its determinants under

different levels of consumer product involvement.

Electronic service quality and its determinants under

different levels of consumer product involvement.

University of Groningen

Faculty of Economics and Business

MscBA Marketing

Marketing Management

March 2011

Written by Supervised by

Maarten Jan Hermsen Prof. Dr. J.C. Hoekstra

Nieuwstad 39 / 5 Dr. J.A. Voerman

9711 JN Groningen Student number: 1334638 Tel: 06 - 20 30 27 76

Preface

When I started my thesis, it did not take long to find an interesting topic. However, to transform an interesting topic into the final report was quite harder to accomplish.

But after a few struggles, I can gladly present here my thesis. With this thesis, my time as a student finally comes to an end. Although I enjoyed my time as a student, I am very motivated to take the next step in my career.

I would like to thank the people who have helped me throughout this last phase of my study. At first I would like to thank Janny Hoekstra for her clear advices and her helpful feedback during the process. Her professional guidance and the feedback sessions were very instructive and helped me to finalize this thesis. Furthermore, I would like to thank my second supervisor Liane Voerman for here feedback in the final phase. Finally, I would like to thank all the people who have supported me and helped me during these last months of my study.

Management summary

Table of Contents CHAPTER 1: Introduction...11 1.1 Introduction ...11 1.2 Problem statement...13 1.3 Research method ...14 1.4 Relevance ...14

1.5 Structure of the report ...15

CHAPTER 2: Theoretical Framework ...16

2.1 Introduction ...16

2.2 Service quality...16

2.3 Electronic Service Quality (e-SQ) ...18

2.4 Measurements of e-SQ ...20

2.4.1 Main measurement scales: Dimensions ...21

2.4.2 Main measurements scales: Determinants ...22

2.5 Determinants of e-SQ ...23

2.5.1 Determinants which influence the pre-purchase service experience ...23

2.5.2 Determinants which influence the post-purchase experience ...26

2.6 Involvement and Consumer product involvement ...27

2.6.1 Involvement ...27

2.6.2 Consumer product involvement ...29

2.7 Involvement and hypotheses ...29

2.7.1 High and low involvement ...30

2.7.2 Ease of use ...30

2.7.3 Website design ...31

2.7.4 Privacy / security ...31

2.7.5 Fulfillment ...32

2.7.6 Customer Service ...33

2.8 Additional exploratory research ...33

2.9 Conceptual Model ...35

CHAPTER 3 Research Methodology ...36

3.1 Introduction ...36

3.2 Data collection method and sample design ...36

3.3 Operationalization of constructs / questionnaire...37

3.3.1 Outline of the questionnaire ...37

3.3.2. Involvement ...38 3.3.3. Ease of use ...39 3.3.4 Website ...39 3.3.5 Privacy / Security ...39 3.3.6. Fulfillment ...40 3.3.7. Customer service ...40

3.3.8. Electronic service quality (e-SQ) ...40

3.4 Reliability Analysis ...41

CHAPTER 4 Results ...46

4.1 Descriptive statistics ...46

4.2 Results regression analyses ...48

4.2.1 Relationships between e-SQ and the determinants: ease of use, website, privacy / security and fulfillment...48

4.2.2 Moderating effect of consumer product involvement ...49

4.2.3 Order of importance of the determinants...51

4.2.4 The influence of customer service ...52

CHAPTER 5 Conclusion and implications ...54

5.1 Conclusions ...54

5.2 Implications ...56

5.3 Limitations ...57

Literature ...59

CHAPTER 1: Introduction 1.1 Introduction

After a slow start at the end of the last century, internet sales form a significant part of the total consumption of consumers nowadays. According to Forrester Research (2010) the total spending of shopping online in the Western European countries was € 68 billion in 2009. This is a 12 % increase in contrast to the year 2008. Forrester also predicts this trend will continue over the next five years. This trend is also visible in the Dutch market. Research conducted by Blauw Research and thuiswinkel.org (2009) showed that internet sales in the Netherlands grew from € 3.91 billion in the year 2007 to € 4.85 billion in the year 2008. Also in the first six months of the year 2009 online retail sales grew again with 19 % in the Dutch Market to €2.7 billion.

Where companies like Amazon.com (worldwide) or Bol.com (the Netherlands) have taken the lead on internet sales, conventional businesses have also recognized the potential of the online channel. This resulted in the fact that nowadays most conventional retail organizations also have an internet sales channel next to their normal business channel (Hema.nl, AH.nl, etc.). This large increase in online shopping and the growth in the number of retailers online have created an extremely competitive marketplace.

determinants influence electronic service quality (e-SQ) perceptions. According to Zeithaml et al, (2002) e-SQ can be defined as the extent to which a web site facilitates efficient and effective shopping, purchasing and delivering of products or services.

The importance of traditional service quality and quality in general in the offline environment are widespread supported by several studies. Quality in general has multiple strategic benefits and contributes to market share and return on investment (Anderson & Zeithaml 1984; Philips et al 1983). Delivering service quality is also considered as an important strategy to be successful and competitive (Parasuraman et al 1985, 1996; Bolton & Drew 1991). Whether a company produces goods or services, delivering service quality is a critical factor both in attracting and retaining customers. The research of Rust & Zahorik (1993) revealed that increasing customer service has a positive influence on customer satisfaction, customer retention and market share. Parasuraman et al (1996) showed that service quality has a strong effect on behavioral intentions. This is supported by Cronin & Taylor (1991) who stated that service quality has a positive effect on the purchase intention of the consumer. In general all studies on service quality reveal the importance of this construct.

then be targeted with different strategies. According to Dholakia (2001) involvement can explain and moderate various aspects of behavior. Consumer behavior can be typified by different states of decision making and involvement plays an important role in this (Zaichkowsky, 1985). Earlier research has proved the influence of involvement on various marketing concepts like perceived risk (Dholakia, 2001; Chaudhuri, 2000), brand commitment (Coulter et al., 2003), brand loyalty (Quester & Lim, 2003), advertising (Petty & Cacioppo, 1983) and information search (Chaudhuri, 2000). More specific, studies also revealed the moderating role of involvement on various marketing concepts, like product fit and brand extensions (Nkwocha et al., 2005), brand attitudes (Suh & Yi, 2006), brand choice (Xue, 2008), product evaluations (Shen, 2005), and on advertising new products (Dens & De Pelsmacker, 2010).

1.2 Problem statement

This study will try to link involvement with the e-SQ domain and assume that involvement has a moderating role on the relationships of the determinants with e-SQ. Therefore the central question of this research will be:

How do the relationships between the determinants and perceived electronic service quality (e-SQ) differ for different levels of consumer product involvement?

The goals of this research will be 1) identify the determinants of e-SQ, 2) identify how involvement influence the strengths of the relationships between the determinants and e-SQ and 3) identify how involvement influence the order of importance of the determinants. To achieve these objectives, this study will try to answer the following research questions:

- Which determinants influence the perceived electronic service quality (e-SQ)? - What is the effect of consumer product involvement on the relations between the

- What is the order of importance of the determinants of e-SQ under different levels of consumer product involvement?

1.3 Research method

This study will investigate the moderating role of consumer product involvement on the relationships between the determinants and e-SQ. In this study, the e-SQ literature and the involvement literature will be investigated to develop a conceptual model that will explain the relationships between the different constructs. Based on the theoretical framework the various determinants of e-SQ will be proposed and hypotheses will be constructed to investigate the moderating role of involvement. The conceptual model will be empirically tested with an online questionnaire. The data will be collected among customers of the online retailer Bol.com. In this research, Bol.com is chosen, because it is very well known in the Dutch market and many consumers have used this online retailer in the past.

1.4 Relevance

1.5 Structure of the report

CHAPTER 2: Theoretical Framework 2.1 Introduction

For companies in the online environment electronic service quality offers a good opportunity to differentiate themselves from competitors in this very competitive market. Academic evidence exists that many companies in the past failed to deliver adequate service quality through the internet (Ahmad, 2002; Lennon & Harris, 2002). Lennon & Harris (2002) stated that because the quality of customer service at online firms seems to be consistently poor, there is plenty of room for companies to compete on the service dimensions.

To gain a better understanding of e-SQ, it is needed to first discuss the “traditional” service quality. This will be followed by a review of the e-SQ literature and the different measurements of e-SQ. This analysis will give insight in the determinants of e-SQ and their influence on e-SQ. Furthermore, an overview of the concept of involvement will be given. Based on the literature hypotheses will derive and the conceptual model will be presented.

2.2 Service quality

To gain a better understanding of service quality, first the concept quality will be defined. Quality itself has been studied by many academics and has proved to be an important construct in the marketing literature. Crosby (1979) defined quality as conformance to fully understand the requirements. Nevertheless this definition of quality is based on research of the quality of goods. However according to Parasuraman et al. (1985) knowledge of quality of goods is not sufficient to understand the quality of services. Early research on this topic revealed that (1) service quality is more difficult to evaluate than goods quality, (2) service quality perceptions are a result of the discrepancy between expected and actual performance and (3) the quality evaluations are not merely focused on the outcome of the service, but also the process of the service has to be taken into account (Gronroos, 1982; Parasuraman et al., 1985). From this Parasuraman et al. (1985) constructed the following definition for service quality: Service quality can be defined as the customer’s impression of the service provided and refers to the quality of all customer interactions and experiences with companies.

To understand service quality, researchers and practitioners were especially interested in how service quality could be conceptualized. The first research in conceptualizing service quality had its starting point in the disconfirmation theory. The disconfirmation theory suggests that quality results from a comparison of perceived performance with expected performance (Gronroos 1982, 1984). In this context, service quality can be defined as the degree and direction of the discrepancy between customer’s service perception and expectations (Parasuraman et al., 1985). One of the first models that conceptualized service quality is the model of Parasuraman et al. (1985, 1988), named SERVQUAL. This gap-model indicates service quality as the discrepancy between the expected level of service and the customers perceptions of the level received. According the SERVQUAL model, consumers evaluate perceived service quality on five dimensions:

Tangibles: The appearance of physical facilities, equipment and personnel. Responsiveness: The willingness to help customers and provide prompt services. Reliability: The ability to perform the promised service accurately and

dependably.

Assurance: The knowledge and courtesy of employees and their ability to inspire trust and confidence.

Although the SERVQUAL model is / was very popular and is widely used over the years by many researchers and practitioners in the field of service quality, there is also a wide range of criticism on the model.

The most criticism on the SERVQUAL model concerns the dimensionality of the model and the fact that it uses the disconfirmation theory. The dimensionality of the model is questioned, because studies revealed that the dimensions are dependent on the particular service being offered (Babakus & Boller, 1992; Cronin & Taylor, 1992), while Parasuraman et al. (1991) claimed that the model is generic across service contexts. Also, several studies proved that there is little evidence that customers assess service quality in terms of the Perceptions-Expectations gap (Babakus & Boller, 1992; Teas, 1994; Gronroos, 1993). In addition, Cronin & Taylor (1994) conceptualized a model, called SERVPERF, which is based on performance only instead of expectations. They state that their performance based scale is efficient in comparison to the SERVQUAL model and is a useful tool for conceptualizing service quality. Later studies also showed support for the performance based view above the expectation based view (Brady & Cronin, 2001; Dabholkar et al., 2000).

In the understanding and development of measurements of e-SQ, many researchers progressed on this earlier work on service quality, and tried to make it applicable for the internet based channel. The following section will continue on this.

2.3 Electronic Service Quality (e-SQ)

(Fassnacht & Köse, 2007). These findings support the importance for companies to understand the underlying constructs of e-SQ.

Where “traditional” service quality refers to the quality of all (non-internet-based) customer interactions and experiences with companies (Parasuraman et al., 1985), electronic service quality (e-SQ) can be defined as the extent to which a web site facilitates efficient and effective shopping, purchasing and delivering of products or services (Zeithaml et al., 2002). This definition implies that the meaning of service includes both pre- and post – purchase experiences. Collier & Bienstock (2006) support this by stating that E-SQ relates to consumers perceptions of the outcome of the service along with recovery perceptions if a problem should occur. The pre-purchase experiences are influenced by aspects like the ease of use of the website and the availability of product information, ordering information and security information. The post-purchase experiences are influenced by aspects like the delivery of the products and the fulfillment of the order, the delivered customer service and the return policy.

As mentioned earlier, traditional service quality has been subject of many researchers in the last decades. However, the service provided in the internet based environment differs from the traditional retail context in various ways. The main differences will be given below:

Convenience: Shopping online offers consumers great convenience in terms of comparing prices or technical features of products efficiently (Santos, 2003). The wide selection of products in the online channel is also seen as more convenient compared to the traditional retail context (Szymanski & Hise, 2000; Wolfinbarger & Gilly, 2003).

Security / Privacy: In the online environment, security and privacy issues play a greater role than in the offline context. According to Ladhari (2010) participating in the online environment involves users in distinctive issues regarding privacy, safety, and confidentiality.

who shop online interact with a technical interface and contact with the employees of an online retailer is never physical.

Absence of physical products: In the online environment the consumer is not able to see, feel or experience a product prior to the purchase (Zeithaml et al., 2002). Also, because of the absence of the physical product, consumers have to wait till the order is delivered before they can use it.

In the past decade, a lot of research has been conducted to identify the number and nature of the dimensions of e-SQ. However still no consensus about the dimensions of e-SQ have been established (Collier & Bienstock, 2006; Ladhari, 2010). Another problem also arise, because where some studies tried to indentify the dimensions of e-SQ (Zeithaml et al., 2005; Francis & White, 2004, 2009; Li et al., 2002; Yoo & Donthu, 2001; Loiacono et al. 2002), other investigated the determinants of e-SQ and focused on their relationship (Wolfinbarger & Gilly, 2003; Collier & Bienstock, 2006; Critobal et al. 2007; Szymanski & Hise, 2000). This means that some focused more on how e-SQ could be dimensionalized, while others tried to measure the determinants and their influence on e-SQ. A short summary of the different measurement scales will be given in the following section 2.4.

2.4 Measurements of e-SQ

if people for example perceive the ease of use of an online retailer as high, the e-SQ of that online retailer is also perceived higher and vice versa. Furthermore, there is also still no consensus about the number and nature of dimensions and / or determinants (Ladhari, 2010; Collier & Bienstock, 2006). However, a lot of similar dimensions can be drawn from the different scales. In the following the main scales will be discussed briefly to give insight in these differences.

2.4.1 Main measurement scales: Dimensions

The first research in conceptualizing e-SQ focused mostly on the interaction of the customer with a website (e.g. Yoo & Donthu, 2001; Loiacono et al. 2002). These studies revealed the dimensions of website quality, and formed the basis for further research. Loiacono et al (2000; revisited 2002) constructed WEBQUAL to measure how consumers evaluate the website quality. They identified twelve dimensions on which a customer evaluates the website quality (for these dimensions see Appendix 1). Yoo & Donthu (2001) made a somewhat similar but simplified model to measure website quality. They discovered that website quality can be measured by a 9 -item scale (SITEQUAL) which consists of four dimensions; ease of use, processing speed, aesthetic design and interactive responsiveness. The SITEQUAL model shows great similarity with Gounaris et al. (2005) who identified also four dimensions on which consumers evaluate a website, namely information, user friendliness, interaction and aesthetics.

they developed different scales for specific internet retailing concepts. Francis & White (2002); Francis (2009) identified four specific categories of internet retailing. According to them the way that consumers select, pay and obtain products can differ depending on the type of product that is purchased (goods or services) and the method of product delivery (offline or services). These four categories are 1) the offline goods : Consumers purchase tangible goods that are delivered to them via post or courier (e.g. books, groceries, CDs, Personal Computers, etc); 2) offline services: Consumers purchase or reserve a service then travel to an offline delivery location to consume the core service (e.g. airline travel, accommodation, holiday, etc); 3) electronic goods: Consumers purchase and download digital goods from a website (e.g. software, music files); and 4) electronic services: Consumers purchase, co-produce and consume a service via a website (e.g. brokerage, chat/dating sites, insurances, etc). In the offline goods category (which is used in this research), Francis (2009) discovered four dimensions on which customers evaluate e-SQ; Website, Exchange, Customer service and Security.

Li et al. (2002) tried to build a framework for e-SQ based on the well known SERVQUAL model of Parasuraman et al. (1988). Li et al (2002) tried to apply the SERVQUAL dimensions in the online environment. There study showed that this frequently used model in the service industry needed to be modified to be useful in web -based service context. Their modified SERVQUAL model for the online environment identified seven dimensions; tangibles, reliability, responsiveness, integration of communication, assurance, quality of information and empathy

2.4.2 Main measurements scales: Determinants

management. They also state that perceived e-SQ has a positive effect on customer satisfaction, and that customer satisfaction is a driver for customer loyalty and repeat purchases in the online context. In this context Collier & Bienstock (2006) also developed a measurement scale for e-SQ. They found empirical support that e-SQ can be measured by three scales, namely process quality (5 determinants), outcome quality (3 determinants) and recovery (3 determinants). Although Szymanski & Hise (2000) did not developed a measurement tool for e-SQ, they found empirical evidence that convenience, site design and financial security are strong determinants of how customers evaluate the service delivered through a website.

2.5 Determinants of e-SQ

Although the distinction named above, a lot of similarity can be drawn between the dimensions and the determinants of e-SQ. As can be seen in Appendix 1, all studies show different numbers of dimensions / determinants and the nature of the dimensions / determinants also differ, however some dimensions and or determinants seems to be consistent over most studies. Ladhari (2010) state that the following criteria which consumers use to evaluate e-SQ appear to be consistent over most studies, namely 1) ease of use or usability, 2) website design, 3 privacy / security, 4) fulfillment or reliability and 5) customer service or responsiveness. These criteria named above will be used as determinants of e-SQ and will be categorized as pre- (ease of use, website design, privacy / security) and post (fulfillment, customer service) purchase experiences with an online retailer.

2.5.1 Determinants which influence the pre-purchase service experience

e-SQ. Ease of use can be defined as how easy it is for customers to use a website (Yoo & Donthu, 2001; Collier & Bienstock, 2006). Ease of use includes aspects like the functionality of the site, the ease of the order process, the ease of navigation through the site and the accessibility of information. The ease of access to information about prizes and products is seen as an important reason why customers shop online (Critobal et al., 2007; Santos, 2003; Wolfinbarger & Gilly, 2003; Van de Poel at al., 1999). The internet based web shop offers customers the ability to reduce their search cost and makes it easy to compare products. Based on this, it can be assumed that ease of use is a determinant of e-SQ and that the usability of a website has a positive influence on e-SQ perceptions. This results in the following hypothesis:

H1: The higher the perceived usability of the website, the higher the perceived e-SQ.

Website design: The second common criteria customers use to evaluate e-SQ is website design. Studies have proven that how the site looks and works has influence on how customers evaluate e-SQ (Zeithaml et al., 2005; Ariely, 2000) and that website design is one of the most important determent how consumers evaluate the website quality (Yoo & Donthu, 2001; Loiacono et al. 2002). Where ease of use refers to how the site works and how easy it is for customers to place an order, website design is more focused on how the site looks. In this context, website design refers to the visual appearance and all applications on the site. This includes variables like the animations, pictures, contents / texts, uses of colors, and the overall layout of the site. The website of an online retailer can be seen as similar as a physical store environment in the normal retail context and therefore has a similar influence on customer perceptions and behavioral intentions. According to Ariely (2000) the graphic design of the website (colors, layout, pictures etc.) has indeed a significant influence on the purchase process of an online consumer.

It can be assumed that website design quality is an important determinant for consumers to evaluate the e-SQ of a retailer and therefore the next hypothesis will be:

H2: The higher the perceived website design quality, the higher the perceived e-SQ.

Privacy / Security: The security and privacy of online transactions is a widely spread research topic. Previous research showed that security covers two main aspects. Privacy (the protection of personal information) and security (the protection of customers from risk of fraud and financial loss) are seen as important factors of influencing e-SQ (Zeithaml et al, 2005; Wolfinbarger & Gilly, 2003). Li et al. (2002) defined the perceived security as the ability of a web-based system to convey trust and confidence of the customer. Following from this privacy /security can be summarized as the degree to which the website is safe and protects customer information. It includes factors like not sharing personal information with third parties and the protection of the shared information (e.g. credit card information) between the customer and the e-retailer.

Research on the topic of security and privacy has indicated that online customers are often concerned about sharing their personal information and credit card information on the internet (Davis et al., 2000) and that e-retailers should manage these information discreet to gain and retain the trust of the customer (Miyazaki & Fernandez, 2001). Cristobal et al. (2007) even stated the lack of confidence in the security and privacy on the internet is the most important obstacle for internet companies to handle. Therefore, the perceived security and privacy can be seen as an important determent of e-SQ, and this is supported by many academics (Zeithaml et al, 2005; Wolfinbarger & Gilly, 2003; Collier & Bienstock, 2006; Francis, 2007; Janda et al., 2002; and others). This results in the following hypothesis:

2.5.2 Determinants which influence the post-purchase experience

Fulfillment / Reliability: Several studies recognized fulfillment or reliability as an important criteria on which consumers evaluate e-SQ (Zeithaml et al., 2005; Wolfinbarger & Gilly, 2003; Collier & Bienstock, 2006; Janda et al. 2002; Francis, 2009). Wolfinbarger & Gilly (2003) even found that fulfillment / reliability is the strongest predictor of customer’s satisfaction and perceived quality in a web-based context and that fulfillment has a strong influence on the loyalty intensions of a customer and its willingness to repurchase on the site. In this context, fulfillment can be defined as the extent to which the website promises about order delivery and item availability are fulfilled (Zeithaml et al., 2005). It can be stated that fulfillment refers to the outcome of the perceived service experience. According to Collier & Bienstock (2006) the outcome quality of the service is dependent on three variables, namely 1) order timeliness (refers to receiving the service within an expected amount of time), 2) order accuracy (refers to how the process after an order meets the customer expectations) and 3) order condition (refers to the product being free from damage and decay). In general customers evaluate the perceived fulfillment as to what extent an online retailer can meet the customer’s expectations regarding the fulfillment of an order. This includes that the customer receives the correct products at the expected time and that the service meets the customer’s expectations. From this, it can be assumed that the determinant fulfillment has a positive influence on e-SQ perceptions, which results in the following hypothesis:

H4: The higher the perceived fulfillment of the order, the higher the perceived e-SQ.

company to offer services that responds to customer inquiries quickly. This can be the case before a purchase is done (for example a question about the product before purchase) or it can be the case if a problem has occurred during the process and the customer needs contact with the company. The research of Zeithaml et al (2002) proved that customer service is a key element for achieving good results in internet retailing. Customer service includes aspects like ease to contact service personal, quick reply to e-mails, personal communication, helpfulness on questions, information on delays and the company’s willingness to help and fix problems. Although customer service also includes interactivity before a purchase is done, in most cases customer service will only influence e-SQ when a consumer has a non-routine encounter with the website/ e-retailer (Zeithaml et al, 2005; Collier & Bienstock, 2006). Handling problems correct and efficient has been proved to be a key indicator of the evaluation of the e-SQ delivered by an online retailer (Yang & Jun, 2002). Based on this, the following hypothesis can be derived:

H5: The higher the perceived customer service, the higher the perceived e-SQ. 2.6 Involvement and Consumer product involvement

2.6.1 Involvement

Derived from research on involvement, Zaichkowski (1986) conceptualized three mayor antecedents’ factors that influence consumer involvement. The first factor relates to the characteristics of the person (needs, importance, interest and values), the second relates to the characteristics of the stimulus (the differentiation of alternatives, source of communication and the content of the communication), the third factor relates to characteristics of the situation (the purchase and occasion). According to Zaichkowski (1986) one or more of these factors could affect the level of involvement. While these factors influence the level of involvement, involvement itself can be classified in three main types or forms. Michaelidou & Dibb (2008) state (based on research of Houston & Rothschild, 1979) that involvement can be categorized into 1) enduring involvement, 2) situational involvement and 3) response involvement.

Enduring involvement represents the long-term attachment of an individual with a specific product class (Richins and Bloch (1986). This means that after a purchase is done, the enduring involvement will stay, or even increase.

Situational involvement represents the a short-term phenomenon where an individual becomes involved with a “situation”, usually a purchase decision (Mittal, 1989). This means that once the purchase is done, the situational involvement will decrease.

Response involvement refers to a behavioral orientation which involves information acquisition and decision processes (Leavitt et al., 1981). This means for example that the extent of information search or the total acquisition time represent the level of involvement of the person.

(Petty et al; 1983; Zaichkowsky 1994), involvement with products and / or product categories (Quester & Smart, 1998; Laurent & Kapferer, 1985; Mittal & Lee, 1988), involvement with the purchase decision (Mittal, 1989; Slama & Tashchian, 1985), involvement with a service (Keaveney & Parthasarathy, 2001; Kinard & Capella, 2006) and involvement with an activity or an event (Neelamegham & Jain, 1999; Speed & Thompson, 2000).

2.6.2 Consumer product involvement

Because in this research the focus will be on consumer’s product involvement, the following will discuss this briefly. The concept of product involvement is a recognition that a particular product category may be more or less central to people’s lives, their sense of identity, and their relationship with the rest of the world (Traylor, 1981). Zaichkowsky (1985) study showed that there was a significant difference in the product involvement level of various products. For example, she found low level product involvement for coffee or breakfast cereal and high product involvement for calculators and automobiles. Laurent & Kapferer (1986) also identified various product categories with different product involvement levels. Although according to these findings some products could be classified as high or low involvements products, Zaichkowski (1986) also states that involvement is an individual variable. In other words, it is not the product that is high or low involved, but it is the consumer (Zaichkowski, 1986; Harari & Hornik, 2010). Therefore, classifying products in high or low involvement products is not encouraged, but instead the involvement of the consumer in a certain product or product category is mostly used.

2.7 Involvement and hypotheses

of e-SQ will be described, which will lead to five hypotheses on the influence of involvement on the relationship between e-SQ and its determinants.

2.7.1 High and low involvement

From section 2.6 can be drawn that consumers can be classified according to their level of involvement into either low involvement or high involvement (for the purpose of this research moderate involvement will not be taken into account). In general it can be concluded that in low product involvement situations, the product is not seen as very important for the consumers and the product has not many relevance to the consumers (Zaichkowski, 1986). In low involvement situations the perceived risk of obtaining the product is also perceived as low (Chaudhuri, 2000) and the consumer can be characterized as little motivated to expend much effort and time in the purchase (Schiffman & Kanuk, 1991). High product involvement situations however, cause a high level of motivation, arousal and interest of the consumer (Nelmapius et al., 2005) and can result in greater search efforts, information processing and decision-making by the consumer. Therefore, high product involvement means that the product is important and very relevant to the consumer.

2.7.2 Ease of use

of use could be more important than for low involved consumers. In other words, the usability of the website will have a stronger influence on the e-SQ perceptions of a consumer in high involvement situations. This results in the following hypothesis:

H6: Under high level of consumer product involvement the usability of the website will have a stronger positive impact on e-SQ than under low level of consumer product involvement.

2.7.3 Website design

In this context, website design refers to the visual appearance and all applications on the site. Petty et al. (1983) found empirical results that the visual appearance of an add has different influence on the behavior of a consumer under different levels of involvement. They state that in case of low level of involvement, the consumer will be influenced more by simple peripheral cues (like brand names, endorsers or pictures), while in case of high involvement the consumer will be more influenced by argument quality (like accurate product information). From this reasoning it can be assumed that this will also apply for a website and that how the site looks (all peripheral cues on the site) will have a stronger impact on the behavior of low involved consumers then on the behavior of high involved consumers. Therefore how the website looks and is designed, might perceived more important for low involved consumers. This results in the following hypothesis:

H7: Under low level of consumer product involvement the perceived website design quality will have a stronger positive impact on e-SQ than under high level of consumer product involvement.

2.7.4 Privacy / security

perceived risk of a consumer. In the involvement literature involvement have been linked to the perceived risk of a consumer as well (Dholakia, 2001; Chaudhuri, 2000). These studies reveal that if consumers are high involved, they are more likely to reduce the perceived risk they obtain. According to Laurent & Kapferer (1985) high involvement decisions can be characterized by a certain degree of emotional or financial risk. Implicit this means that when consumers are high involved, the emotional and financial risk should be reduced. Therefore it can be assumed that in case of high consumer product involvement, the determinant privacy / security will be more important than for low product involvement. This results in the following hypothesis:

H8: Under high level of consumer product involvement the perceived privacy / security will have a stronger positive impact on e-SQ than under low level of consumer product involvement

2.7.5 Fulfillment

The fulfillment of an order refers to the outcome of the perceived service experience. Research on fulfillment and e-SQ has proved that fulfillment can predict customer satisfaction, customer loyalty and perceived quality (Wolfingbarger & Gilly, 2003). In the context of product involvement, Prenshaw et al. (2006) found empirical evidence that performance evaluations will have a greater influence on consumers’ satisfaction in high involvement situations than in low involvement situations. This means that the performance outcomes are highly relevant for high involved consumers. From this it can be reasoned that if a person is high involved it would value the fulfillment determinant of e-SQ as more important than when the person is low involved. Therefore the following hypothesis is:

2.7.6 Customer Service

As mentioned earlier customer service can be defined as the responsiveness, helpfulness and willingness of the company to offer services that responds to customer inquiries quickly (Wolfingbarger & Gilly, 2003). Because in most cases the customer service will only affect e-SQ after a purchase is done, customer service can be seen as some kind of a performance outcome as well. This means, that (like fulfillment) customer service can be seen as more important in high involvement situations. This results in the last hypothesis:

H10: Under high level of consumer product involvement the perceived customer service will have a stronger positive impact on e-SQ than under high level of consumer product involvement.

In conclusion, it is assumed that the determinants ease of use, privacy / security, fulfillment and customer service should be more important for high involved consumers than for low involved consumers and that the determinant website design should be more important for low involved consumers than for high involved consumers.

2.8 Additional exploratory research

high involvement situations may differ with another determinant in the high involvement situation. This assumption implicate that the order of importance might differ for the different groups of respondents (high or low involved) and that maybe a certain group should be targeted differently to maximize its perceived e-SQ.

In section 2.7, it is assumed that the relationships between ease of use, privacy / security, fulfillment, customer service and e-SQ are stronger under high consumer product involvement, and that the relationship between website design and e-SQ is stronger under low consumer product involvement. From this, it can also be assumed that, when looking at the order of importance of the determinants, ease of use, privacy / security, fulfillment and customer service are more important for the evaluation of e-SQ in high involvement situations, while website design might just be the most important determinant of e-SQ in low involvement situations. As theory in this field is lacking, this part of the research will be of an exploratory nature.

2.9 Conceptual Model

The goal of this research is to gain insight in the influence of involvement on the relationships between the determinants and e-SQ. Based on the theoretical framework of e-SQ and the product involvement of the consumers the following conceptual model is constructed, which shows the proposed relationships between e-SQ and its determinants and the moderating role of product involvement on this relationship. In the following chapters, these relationships will be empirically tested.

Ease of use

Website

Design

Privacy /

Security

Fulfillment

Customer

Service

E-SQ

Product

Involvement

Ease of use

Website

Design

Privacy /

Security

Fulfillment

Customer

Service

E-SQ

Product

Involvement

CHAPTER 3 Research Methodology 3.1 Introduction

In this chapter, the research methodology of this study will be presented. To test the hypotheses constructed in chapter 2 an empirical study will be conducted. First, the data collection method will be explained, followed by how the questionnaire is constructed and how the constructs are operationalized. Finally the research procedure will explain which measurements will be used.

Because this research investigates the total online purchase experience of the consumer and not only the interaction with the website, an e-retailer is chosen which sells products (instead of services) and where the consumers receive a package (instead of for example software). Such an e-retailer makes it possible to measure all aspect of perceived e-SQ, from ordering the product through the website, to delivering the product and handling any complaints. For this study, an online CD/DVD and book vendor is chosen since these vendors are transactional, while the CD/DVD and books market is widely used in earlier research on e-SQ and e-commerce (like Yoo & Donthu, 2001; Barnes & Vidgen, 2002; Francis & White, 2002; Francis 2009; Wolfinbarger & Gilly, 2003; Ribbink et al., 2004). Another reason for the CD/DVD and book market is that it has established itself as a strong online market. In the Netherlands the website Bol.com is well known and many people have used this website for online purchases. Therefore this website will be used as reference in this research.

3.2 Data collection method and sample design

friendliness. The downside of an online questionnaire is however that in most cases the response rate is low.

The people who were recruited to participate for this study were selected by using convenience sampling and snowball sampling. The main target population of this study are the Dutch citizens between the ages of 20-55. According to Collier & Bienstock (2006) the young adults and early internet users are most likely to conduct purchases on the internet. The survey was distributed by an online mailing list to 267 addresses. From the 267 people (and the people who were contacted by snowball sampling), 228 people filled in the questionnaire. With the extraction of people who did not fill in the questionnaire completely and who never bought an item at Bol.com, 181 respondents of the survey remained. A general description of the population will be given in chapter 4. From the 181 respondents who remained in the survey, a high proportion of missing data was found on the determinant customer service. Approximately a quarter (26 %) did respond to these items. The respondents, who did not respond to this item, presumably did not have experienced the issues measured by the items of customer service. Section 3.5 will elaborate on this.

3.3 Operationalization of constructs / questionnaire

3.3.1 Outline of the questionnaire

In appendix II the Dutch version of the questionnaire can be found. Next to some questions about the background information of the respondents, the questionnaire consist of a set of questions that will measure the seven different constructs (the five determinants of e-SQ: ease of use, website, privacy / security, fulfillment and customer service; the variable: the involvement level of the customer and the dependent variable e-SQ).

from which e-retailers; and finally if they ever had purchased an item at Bol.com and what product. If the respondent never has done a purchase at Bol.com, that person does not have to fill in the rest of the questionnaire and will be excluded from this research.

After this the respondents will be asked about the different constructs from the conceptual model. The model consist of six independent variables and one dependent variable. The five independent variables (ease of use, website, privacy / security fulfillment, and customer service) are expected to have a direct effect on the dependent variable e-SQ, while the level of involvement of the consumer is expected to have a moderating effect on the relationship between the five determinants and e-SQ. To measure these constructs validated scales are used from earlier academic research. This will be discussed in the following sections and is summarized in table 3.1. Each of the independent factors (ease of use, website, privacy / security, fulfillment, customer service and the involvement level of the customer) will be measured by multiple items. The dependent variable e-SQ will be measured by two items.

3.3.2. Involvement

agree). For the additional exploratory research the respondents will be separated by their level of involvement. To make a distinction between high and low involvement, this research will regard respondents who score 4 or lower (1 -4) as low involved and respondents who score higher than 4 (>4 - 7) as high involved.

3.3.3. Ease of use

The determinant ease of use can be defined as how easy it is for customers to use the website (Collier & Bienstock; 2006, Yoo & Donthu; 2001). To measure if the determinant ease of use is a good predictor of e-SQ, eight items (with a Cronbach Alpha of 0.94) are adapted from research from Zeithaml et al. (2005). Ease of use includes aspects like the functionality of the site, the ease of the order process, the ease of navigation through the site and the accessibility of information. A seven points Likert scale (1 totally disagree, 7 totally agree) is used to capture the responses of the respondents.

3.3.4 Website

The determinant website in this context refers to the visual appearance and all the applications on the site. This includes variables like the animations, pictures, contents/ texts, uses of colors, and the layout of the site. The five items which are selected to measure this determinant are adapted from the research of Collier & Bienstock (2006). Their scale with a Cronbach Alpha of 0.71 measures the determinant website on a 7 point Likert scale (1 totally disagree, 7 totally agree).

3.3.5 Privacy / Security

five items on a 7 point Likert scale (1 totally disagree, 7 totally agree) are used to measure the determinant privacy / security.

3.3.6. Fulfillment

In this study, the determinant fulfillment refers to the extent to which the website promises about order delivery and item availability are fulfilled (Zeithaml et al, 2005). Seven items are used to operationalize fulfillment. The items are adapted from the research of Zeithaml et al. (2005). This scales with a reported Cronbach alpha of 0.93,

measures fulfillment on a 7 point Likert scale (1 totally disagree, 7 totally agree).

3.3.7. Customer service

Wolfinbarger & Gilly (2003) define customer service as the responsiveness, helpfulness and willingness of the company to offer services that responds to customer inquiries quickly. This means that this factor is only applicable when the consumer has a non routine encounter with the e-retailer and made contact with the company. Because it is expected that in most cases this is probably not happening, it could be that respondents cannot answer to the questions about the customer service level of Bol.com. Therefore, respondents who did not have an encounter with the customer service of Bol.com can skip these questions.

To measure the determinant customer service for the respondents who did have an encounter with the customer service, five items (on a 7-point Likert scale) are used from the research of Wolfinbarger & Gilly (2003) (Cronbach Alpha 0.84).

3.3.8. Electronic service quality (e-SQ)

To measure the electronic service quality respondents were asked to evaluate their last

purchase and offered service of their last purchase at Bol.com. The items to measure e-SQ are extracted from the research of Wolfinbarger and Gilly

service can be measured by two items, namely the overall quality of the purchase experience at the site (with the endpoints 1, very poor and 7, excellent) and the overall feeling towards the site (with the endpoints 1, very dissatisfied and 7, very satisfied). 3.4 Reliability Analysis

To assess the reliability of the different constructs, a reliability analysis is conducted. A reliability analysis (using Cronbach alpha) will check the internal consistency of the scales. The alpha measures to what extent the items measure the same construct and can range from 0 to 1. The closer the alpha is to the 1, the greater the internal consistency of the items (Hair et al., 2006). According to Hair et al. a Cronbach alpha of 0.6 and above generally indicates a satisfactory internal consistency. The Cronbach alphas of the variables in this study are displayed in table 3.1. As can be seen in the table, all scales show satisfactory results. The reliability analyses also indicated that no items needed to be deleted.

Table 3.1: Summary Scale Items

Constructs Source Items Cronbach

Alpha Product

Involvement

Mittal & Lee (1988)

In general I have a strong interest in this product.

0.904 This product is very important for me

The product matters a lot to me

I get bored when other people talk to me about this product.

7 point Likert scale (1 totally disagree, 7 totally agree)

Ease of use Zeithaml, Parasuraman & Malhotra (2005)

This site makes it easy to find what I need. 0.728 This site makes it easy to get anywhere on

the site

This sites is simple to use

This site enables me to get on it quickly This site is well organized

7 point Likert scale (1 totally disagree, 7 totally agree)

Website Design

Collier & Bienstock (2006)

This e-retailers website is visually pleasing

0.631 This e-retailer’s website design is

innovative

I’m able to see the graphics clearly on this site

This e-retailer’s website does not have fine print that is difficult to read

I don’t have to scroll from side to side to adequately see this e-retailers site 7 point Likert scale (1 totally disagree, 7 totally agree)

Privacy / Security

Wolfinbarger & Gilly (2003),

I feel my privacy is protected by the site 0.712 I feel safe in my transactions with this

website

The website has adequate security features I feel that I can trust the website

7 point Likert scale (1 totally disagree, 7 totally agree)

Fulfillment Zeithaml, Parasuraman & Malhotra (2005)

This company delivers orders when promised

0.742 This company makes items available for

delivery within a suitable time frame This company quickly delivers what is ordered

This company send out the items ordered This company has in stock the items it claims to have

This company is truthful about it offerings This company makes accurate promises about delivery of products

7 point Likert scale (1 totally disagree, 7 totally agree)

Customer Service

Wolfinbarger & Gilly (2003)

The company is willing and ready to respond to customers’ needs

shows a sincere interest in solving it Inquiries are answered promptly Customer service personal is always willing to help you

Returning items is relatively straightforward

7 point Likert scale (1 totally disagree, 7 totally agree)

E-SQ Wolfinbarger & Gilly (2003)

The overall quality of my purchase experience with this website is

0.673 7 point Likert scale (1 Very poor, 7

excellent)

My overall feeling towards this website is 7 point Likert scale (1 very dissatisfied, 7 very satisfied)

3.5 Analysis method

To measure the relationships, several regression analyses will be conducted. A regression analysis will test if the independent variables (determinants) explain a significant variation in the dependent variable, in other words whether a relationship exists and the strength of the relationship (Malhotra, 2004).

Model 1: H1 - H4

To test if a relationship exist between e-SQ and the four determinants a linear regression model is used. This model captures the relationship between e-SQ and its determinants as followed:

(1) ESQ = α + β 1EoU + β 2Web + β 3P/S + β 4Ful +ɛ

Model 2: H6 - H9

The hypotheses 6-9 suggest that the relationship between e-SQ and its determinants depends on the level of consumer product involvement. In other words, it is assumed that consumer product involvement moderates the effect of the determinants on e-SQ. A moderation effect occurs when one variable influences the relationship between two other variables (Hair et al, 2006). To test the these assumed moderating effect, a moderated multiple regression model is used. This regression model is as followed:

(2) ESQ = α + β1EoU + β 2Web + β 3P/S + β 4Ful + β 5PI + β 6EoUxPI + β 7WebxPI +

β 8P/SxPI + β 9FulxPI +ɛ

Model 3: Order of importance of the determinants

To gain insight in the order of importance of the determinants under different level of involvement, the respondents of the survey are split into a high and a low involvement group. To test if there is a difference, the following regression models will be used: (3a (High)) ESQ = α + β1EoU + β2Web + β3P/S + β4Ful +ɛ

Model 4 / 5 / 6a / 6b: H6 & H10

These models will test the hypotheses concerning the customer service hypotheses. These regression analyses will only be conducted on the respondents that had an encounter with the customer’s service of Bol.com. These models are as follows:

(4) ESQ = α + β1EoU + β2Web + β3P/S + β4Ful + β5CS +ɛ

(5) ESQ = α + β1EoU + β2Web + β3P/S + β4Ful + β5CS + β6PI + β7EoUxPI +

β8WebxPI + β9P/SxPI + β10FulxPI + β11CSxPI +ɛ

(6a (High)) ESQ = α + β1EoU + β2Web + β3P/S + β4Ful + β5CS +ɛ (6b (Low)) ESQ = α + β1EoU + β2Web + β3P/S + β4Ful + β5CS +ɛ

Legend models:

ESQ = Electronic service quality EoU = Ease of use

Web = Website

P/S = Privacy / Security Ful = Fulfillment CS = Customer service

CHAPTER 4 Results

In this chapter the results of the analyses will be discussed. First, a general description of the population of the survey and a general description of the data will be discussed. After this, the results of the several regression analyses (which will test the hypotheses and determine whether the order of importance of the determinants differ under different levels of consumer product involvement) will be given.

4.1 Descriptive statistics

As stated earlier, 181 respondents from the survey remained in the sample. The sample consists of 109 men (60%) and 72 women (40%) with age varying between 18 and 65 (see appendix III for tables). The average age of the sample is 29.9 and most respondents fall within the category 21-30 years old (62%). This indicates a mostly young adult population, which is the main target group of this research. Most of the respondents purchases something from the internet once every six months (35%), once every three month (23%) or once a year (23%). This indicates that most respondents are familiar with purchasing through the online channel. Furthermore, books, study books and film DVD’s are the most purchased items.

Table 4.1: Means and standard deviations variables Means and standard deviations variables Variable Number of respondents Mean Standard deviation Minimum Maximum PI 181 4.48 1.356 1.75 7.00 EoU 181 5.70 0.532 3.88 7.00 Web 181 5.34 0.624 3.40 6.80 P/S 181 5.44 0.669 3.50 7.00 Ful 181 5.48 0.824 4.10 7.00 CS 47 5.20 0.541 3.20 7.00 E-SQ 181 5.73 0.587 4.00 7.00

For model 3 and 6, the respondents will be separated into a high and a low involvement group (1 - 4 = low involved, > 4 - 7 = high involved). Of the 181 respondents, 106 (59%) were high involved and 75 (41%) were low involved. The average consumer product involvement of the high involvement group is 5.51 (with a standard deviation of 0.583), while the average consumer product involvement of the low involvement group is 3.02 (with a standard deviation of 0.562). Table 4.2 shows an overview of the main statistics for these different groups.

Table 4.2: Means and standard deviations variables High / Low Means and standard deviations variables

4.2 Results regression analyses

In this section, the several regression analyses will be presented. First I will identify how the determinants (EoU, Web, P/S and Ful) are related to e-SQ. Second, I will test whether the level of consumer product involvement moderates the relationship between e-SQ and the determinants. After this, the order of importance of the determinants for the different involvement groups will be investigated. Finally, the same analyses will be conducted as above with the inclusion of the determinant customer service.

4.2.1 Relationships between e-SQ and the determinants: ease of use, website, privacy / security and fulfillment

To test the hypotheses H1, H2, H3 and H4, model 1 (see section 3.5) is used. The average scores on ease of use, website, privacy / security and fulfillment are the independent variables, and e-SQ is the dependent variable. The results of the regression analysis can be seen in table 4.3

Table 4.3: Linear regression models 1/2/3

Linear regression models 1/2/3, e-SQ as dependent variable Variable Hypothes

is

Model 1 Model 2 Model 3a (High Inv) Model 3b (Low Inv) (b unstandardized) Constant (β0) 0.733** 1.350 0.709 0.625 EoU (β1) 1 0.340* -0.133 0.428* 0.169*** Web (β2) 2 0.185* 0.609* 0.090 0.332* P/S (β3) 3 0.187* 0.120 0.214* 0.177** Ful (β4) 4 0.193* 0.233 0.167*** 0.257** CS (β5) 5 PI (β6) -0.142 PI x EoU (β7) 6 0.101** PI x Web(β8) 7 -0.097** PI x P/S (β9) 8 0.018 PI x Ful (β10) 9 -0.005 PI x CS (β11) 10 R2 0.473 0.496 0.462 0.490 (Adj R2) 0.461 0.470 0.441 0.461 F-Value 39.452* 18.732* 21.689* 16.792*

p-value < 0,01; ** p-value < 0,05; *** p-value <0,10

4.2.2 Moderating effect of consumer product involvement

The hypotheses which deal with the moderating role of consumer product involvement on the relationship between e-SQ and its determinants are investigated by measuring the interaction effects between product involvement and the determinants. To test the hypotheses H6, H7, H8 and H9, model 2 is used (see section 3.5).

4.2.3 Order of importance of the determinants

To investigate the order of importance of the determinants under different levels of consumer product involvement, model 3 is used for both the high involvement group (3a) and the low involvement group (3b). The results of the regression analyses are shown in table 4.3. Given the fact that these analysis are conducted with a relatively small number of observations (i.e. high involved: 106 observations and low involved: 75 observations), a critical p-value of 0.10 will be accepted here. From this analysis, it can be concluded that for the high involvement group (model 3a) ease of use is the strongest predictor of e-SQ, followed by privacy / security and finally fulfillment. Website design in this model has a p-value of 0.254 and is therefore not significant. When we look at the low involvement group however, it can be concluded that the order of importance is different than in the high involvement group (see table 4.4). In this case website design is the strongest predictor of e-SQ, followed by fulfillment, privacy / security and finally ease of use. Consumers in the high involvement group, might value ease of use higher, because they will be longer on the site, and therefore will “experience” the ease of use more than people who are less involved. Low involved consumers might be more influenced by peripheral cues (which includes website design). In these models, multicollinearity causes no problems, because all the VIF values are below 2.

Table 4.4: Order of importance determinants Order of importance

High Involvement Low Involvement

1 Ease of Use Website Design

2 Privacy / Security Fulfillment

3 Fulfillment Privacy / Security

4.2.4 The influence of customer service

To test hypotheses H5 and H10, the determinant customer service will be included. However, the regression analyses with customer service will only include the respondents who responded to the customer service items (47 in total). Models 4, 5, 6a and 6b (see section 3.5) are used. The results from these analyses are shown in table 4.5.

Table 4.5: Linear regression models 4/5/6

Linear regression models 4/5/6 e-SQ as dependent variable Variable Hypothes

is

Model 4 Model 5 Model 6a (High Inv) Model 6b (Low Inv) (b unstandardized) Constant (β0) -0.216 2.823 -0.524 1.281 EoU (β1) 1 _{0.371* -0.226 0.625** 0.163} Web (β2) 2 _{0.157 -0.009 0.115 0.243} P/S (β3) 3 _{0.151 0.151 0.191*** 0.088} Ful (β4) 4 _{0.262** 0.650 0.031 0.336} CS (β5) 5 _{0.126 -0.017 0.167 -0.035} PI (β6) -0.673 PI x EoU (β7) 6 _0.146 PI x Web(β8) 7 _0.038 PI x P/S (β9) 8 _0.016 PI x Ful (β10) 9 _-0.108 PI x CS (β11) 10 _0.023 R2 _{0.679 0.624 0.632 0.372} (Adj R2) _{0.640 0.537 0.594 0.130} F-Value _{17.362* 8.351* 21.863 1.539}

CHAPTER 5 Conclusion and implications

In this chapter the conclusion of the study will be presented. First, the main findings of the analysis will be given. Conclusions based on the hypotheses and the findings of the different relationships will be discussed. This will be followed by the implications this study provides for researchers and practitioners. Finally, the limitations of this research will be discussed briefly.

5.1 Conclusions

The goal of this research was to gain insight in the influence of involvement on the relationships between the determinants and e-SQ. To achieve this, this study tried to answer the following questions:

1) Which determinants influence perceived electronic service quality (e-SQ)?

2) What is the effect of consumer product involvement on the relations between the determinants and perceived electronic service quality?

3) What is the order of importance of the determinants of e-SQ under different levels of consumer product involvement?

To answer question 1 and 2, 10 hypotheses were constructed. An overview of the accepted and rejected hypotheses is listed in table 5.1. The first question was used to identify the determinants which consumers use to evaluate the perceived e-SQ of an online retailer (in this case Bol.com). The results show that consumers see ease of use, website design, privacy / security and fulfillment as determinants of perceived e-SQ. The results show a positive relationship between these determinants and e-SQ, which indicates that the higher the values on the determinants, the higher the perceived e-SQ. The results also show that ease of use is the strongest predictor of e-SQ. This study did not find support for customer service to be a determinant of e-SQ.