How to build customer loyalty in an online environment?

(1)

How to build customer loyalty in

an online environment?

(2)

2

How to build customer loyalty in

an online environment?

Quantitative research on factors and mediators that play a role in the online

loyalty formation process in the Netherlands.

The Hague, October 2009

Faculty: Economics and Business

Specialization: Marketing Management

Author: Kim Pennings (F.J.H.)

Organisation: University of Groningen

Company: TNT Post

Supervisor 1: Dr. Wander Jager

(3)

Preface

At the end of my thesis period and internship at TNT Post, I am proud of the result that is in front of you. I had the opportunity to choose a subject that personally interests me and I have been able to execute a relevant and interesting study. The start was rather difficult and slow but after all, I look back on this period with positive feelings.

First of all, I was very happy to obtain a position as an intern at TNT Post and write my thesis at this company. I preferred to write my thesis at a company since I wanted to increase my professional experience at a large company and execute a research that would be used in practice. TNT Post provided a great opportunity to do so. For the purpose of the thesis I moved to The Hague since my department is settled at the head office of TNT Post in The Hague.

I have executed a study for the E-business development department of TNT Post. This department is engaged with the execution of new online business and management of existing online shops. It fits me because I already have a three year experience in online marketing and currently work for an online marketing agency. An e-commerce related topic therefore strongly interested me and provided lots of opportunities to apply my knowledge and experience.

However, the start of the thesis was rather difficult because there was not a clear research question or practical problem at the department. This is why it took a while to discover a relevant research problem and formulate a good research question. Finally, I managed to formulate a research question that was relevant and challenging and was also supported by all web shop managers of TNT Post. Once the research question was formulated the process took off and the time passed by quickly. My internship at TNT Post was great in several ways. First of all, the colleagues at the department were all very helpful and fun. My colleagues and the management treated me as a full employee which enabled me to gain relevant experience. For example, I was involved in the summer campaign from the Postal Card service, was responsible for a part of the project and got the opportunity to visit several events.

Nevertheless, all this experience has not been of negative influence on the result of my study. I have been able to finish my thesis within the predetermined deadline of six months and I am very pleased with the final result. I am convinced that the findings are relevant and innovative and therefore will be interesting for online practitioners and managers. This is why I strive to publish the results of my study in multiple online blogs or news websites, starting with Marketing Facts.

Finally, I will like to thank certain people. First of all, I would like to thank my supervisor from the university, Wander Jager and Peter Verhoef for his revision. Last but not least, I would like to thank my supervisor at TNT Post, Jeroen Siegerink, and all other colleagues that helped me with my thesis. In special, I would like to mention Bas Fontein, Huub Rulkens, Liesbeth Gouda and Ellen de Lange-Ros.

(4)

4

Figures and tables

Figures

Figure 1: conceptual part 1 ... 16

Figure 2: conceptual model to be tested ... 20

Figure 3: mediator analysis ... 36

Figure 4: mediator analysis of e-service quality and security on behavioural loyalty ... 37

Figure 5: mediator analysis of e-service quality and security on behavioural loyalty ... 37

Figure 6: mediator analysis e-service quality and security on behavioural loyalty ... 38

Figure 7: final model behavioural loyalty ... 39

Figure 8: mediator analysis e-service quality and security on attitudinal loyalty ... 40

Figure 9: mediator analysis e-service quality and security on attitudinal loyalty ... 40

Figure 10: mediator analysis of e-service quality and security on attitudinal loyalty ... 40

Figure 11: final model attitudinal loyalty ... 41

Tables

Table 1: overview questionnaire invitations and response rate….……….………...22

Table 2: characteristics of sample versus non sample………...22

Table 3: Measurement scales dependent variables……….…....24

Table 4: Measurement items of independent variable...………….………25

Table 5: Loyalty scores of sample………...…..…….…27

Table 6: details about loyal behaviour………28

Table 7: details about loyalty intentions……….28

Table 8: difference among groups of respondents………..29

Tabel 9: differences based on favorite web shop………31

(8)

8

Executive summary

In the last decade, e-commerce has become more and more important and the online expenditures have been increasing ever since online shopping was introduced. E-tailers are successful and the online shopping population continues to rise. It seems that companies are obliged to offer products,

information or services on the internet to stay into business. Moreover, the internet is characterized by many shops offering many similar products and therefore the power of the consumer is large. For the online shops of TNT Post, managers realize that they succeed in attracting customers but it is difficult to retain these customers and increase their loyalty.

This study therefore explores what factors and other variables are important for the loyalty formation process in the online environment. The research on loyal behavior has been limited to two important aspects of attitudinal loyalty which are propensity to recommend and propensity to continue

purchasing and behavioural loyalty measures as the percentage of total expenditures spent at the favorite shop. This study tried to investigate factors that are under direct control of managers and their effect on behavioural loyalty directly or via important mediators. These mediators are derived from previous research which found that overall satisfaction, commitment and trust are important mediators of customer loyalty.

The findings suggest that the mediators cannot be distinguished from the independent variables in the data. This is due to lack of a representative sample and design of the questionnaire. Overall

satisfaction and trust are evaluated on the same dimensions along with e-service quality aspects. Moreover, commitment seems to be heavily related to the attitudinal loyalty construct. Nevertheless, the data reveals that behavioural loyalty and attitudinal loyalty are separate constructs. The report provides two models that explain these constructs individually.

The findings confirm the general believe that satisfaction about the service quality plays a key role in loyalty formation and that this also applies to the online environment. It is surprising that trust is found to be evaluated on the same dimension e-service quality which implies that loyal customers do not distinguish between satisfaction about the service quality of the web shop and the trust placed in a web shop. Both satisfaction and trust are found to directly affect attitudinal and behavioural loyalty.

Moreover, this construct is also found to be a full mediator of the effect of interactivity, customization and relationship investment on both loyalty measures. This finding implies that these more advance activities of web shop do not directly affect loyal behaviour. The basic service quality of a web shop needs to be good in order to be able to affect loyalty by these activities. Moreover, the overall evaluation of the e-service quality is found to be a central construct in customer loyalty. In both attitudinal and behavioural loyalty this construct directly influential as well as mediating the effect of marketing tactics on customer loyalty. One marketing tactic in sovered by interactivity and is about the interactive opportunities that a web shop provides as well as the ease of getting in to contact.

Relationship investment is about rewards and e-mailings that are sent to customers whereas customization is about personalization of product as well as tailor made promotions and advertisements.

(9)

9

1. Introduction

“Every company’s greatest assets are its customers, because without customers there is no company” Michael LeBoeuf

The quote of Michael LeBoeuf (www.brainyquote.com), former management professor at the University of New Orleans, indicates the importance of customers and therefore a customer focus within marketing and businesses today. Customers are an important asset for a company which indicates a valuable resource and that it is a must for companies to invest in its relationship with customers. Customer loyalty can therefore be seen as one important objective for companies. Especially within the online environment, customer loyalty seems even more valuable but also more complex. Peelen (2001) argues that the establishment of loyalty in the sense of feeling committed, reliability and security might be more difficult when information technology is involved. It seems reasonable that loyalty within an online environment is difficult to achieve because of increased transparency, lower transaction costs and new transaction tools (e.g. auctions). Consumers have become more and more acquainted with the technology and the Web is used for diverse purposes and shopping is only a small part of it. When customers are willing to shop they can easily switch and compare products and offerings with just one click. These facts make it challenging to make a customer loyal to your online store.

1.1 Introduction to the E-business development department at TNT Post

At TNT Post, the E-business development department was introduced two years ago. The department was formed out of a new business concept launch. The new business idea was initiated when mail volumes were declining and TNT Post realized it needed to invest in new businesses and concepts for future growth. The first step into the formation of new business was the establishment of the New Business Development (NBD) department. This department was initiated to think out of the box and come up with new business concepts. One step into new business was to focus e-commerce. This is where the department E-business development (EBD) was formed.

The department is charged with the exploitation of existing e-commerce initiatives as well as development of new e-commerce activities. TNT Post wants to invest in e-commerce because this market is still growing and has a high potential. The total revenue in online shopping in the Netherlands increased by 24% up to €4.85 billion in 2008. Moreover, an important trend in online shopping is that the total online shopping population and the frequency of experiences shoppers is increasing (Thuiswinkel Markt Monitor, thuiswinkel.org). TNT Post cooperates in this business by delivery and partnership with existing online shops and the exploitation of own web shops. In addition, TNT Post wants to be involved in facilitation of the whole chain. E-business development currently focuses on the existing web shops and services, on the integration of all the shops in one overall proposition and new initiatives to grow the current business. The web shops of TNT Post started with the online postcard-service (https://www.tntpostkaartenservice.nl) and then the online photo service (http://foto.tntpost.nl). Within the two years of existence, the online services for consumers have been extended to an online shop for presents (http://www.tntkadowinkel.nl), magazines (http://tijdschriften.tntpost.nl), mailing attributes for consumers

(www.tntpost.nl/onlinewinkel), companies (www.tntpost.nl/mailorder) and collector’s items

(www.tntpost.nl/collectclub).In addition, TNT hosts two important Dutch websites which is an online pharmacy-service apotheek.nl) and a drugstore-service

(10)

10

Nowadays, the organisation of TNT Post realizes that the online business needs to grow enormously in order to be able to survive in the future. The objective is to build sustainable and stable revenue in order to grow its profit in the future and to compensate for the decreasing mail market. Therefore, targets are sky-high which means that the revenue needs to grow by 200 % in five years. One way to attain this target is to launch new shops or services. Another way to achieve this target is to grow current business. Growing current business will include cross selling, product category extension and retention of customers. Until now, the focus of shop managers was on acquiring new customers, but to achieve these ambitious targets, retaining current customers will be valuable and contributes to an increase in revenue.

1.1.1 Problem statement

In order to contribute to the imposed targets for the E-business development of TNT Post by increasing customer loyalty, a better understanding of building customer loyalty in an online environment is required. The aim of this research is to describe the marketing efforts and measures that contribute to an increase in customer loyalty, either directly or indirectly through stronger customer relationships. Marketing efforts include marketing tactics that can be employed by shop managers in order to affect customer loyalty. Measures include adaptations in website design, content and communication that will help to improve customer experience in order to affect customer loyalty or improve customer relationships

How to build customer loyalty in an online environment?

- Which factors influence customer loyalty in an online environment? - Which constructs mediate the influence of factors on customer loyalty? - Which factors affect customer loyalty directly in an online environment? - What is the order of importance of factors that influence customer loyalty?

- What is the order of importance of factors that influence mediators of customer loyalty?

1.2 Managerial relevance

My research will provide an extensive overview about factors that are important to build customer loyalty in an online environment. This is valuable for online shop managers because it will provide insights in which aspects of the website, service, communication or marketing need to be changed in order to build sustainable customer relations. My study will focus on marketing tactics that are under direct control of managers as for example e-mailings and promotions. In addition, my study will identify factors in web site characteristics that play a role in building customer loyalty. Moreover, the effects of these factors on important mediating constructs between marketing tactics and customer loyalty separately will be further analyzed. This provides important insights in the differences in effect of these factors on the mediating constructs.

(11)

11

1.3 Academic relevance

My study will contribute to academic literature on building relationship in an online environment in three ways. First of all, my study will extend current research on antecedents or factors affecting customer loyalty in an online environment. A thorough understanding of factors that influence customer loyalty in an online context is still lacking. Secondly, it will reveal whether there are mediators between the factors and customer loyalty and what their impact is. Finally, the order of importance of factors influencing these mediators or customer loyalty will be further analyzed. Some researchers have addressed issues that influence several mediators in the online context but no overall research is available that provides any information about the order of importance of all factors. These findings fail to identify the effect of factors on these dimensions separately while this might be very valuable for management. Moreover, my study will combine findings of different studies which make the understanding of establishing customer loyalty more complete.

1.4 Structure overview

(12)

12

2. Customer loyalty theory

2.1 Customer loyalty definitions

Researchers distinguish between two types of loyalty, namely behavioural and attitudinal loyalty (Odin et al. 2001). Behavioural loyalty exists when a consumer repeatedly purchases a product or service, but does not necessarily have a favourable attitude towards the brand. This may be due to convenience, habit or because the barriers to change are too large. De Wulf, Odekerken-Schröder, & Iacobucci (2001) define behavioural loyalty as ‘a composite measure based on a consumer’s

purchasing frequency and amount spent at a retailer compared with the amount spent at other retailers from which the consumer buys’. Attitudinal loyalty is defined by Jacoby and Chestnut (1978) as ‘the consumer’s predisposition towards a brand as a function of psychological processes’. This includes ‘attitudinal preference and commitment towards the brand’. According to this separation in the definition of customer loyalty, I will also deal with these terms in different ways within my study. Behavioural loyalty without having a favourable attitude towards the firm or product results from habit, convenience or satisfaction. Behaviour of customers as a result of behavioural loyalty results in repeat purchasing behaviour and therefore is an important objective for managers. It is the most objective measurement of customer loyalty. In my study I will focus on both aspects of customer loyalty. Aspects of attitudinal loyalty are important because they represent the most sustainable characteristic of customer loyalty with customers. Achieving customer behavioural loyalty is only stable and desirable as a result of a positive attitude towards a store. Behavioural loyalty is valuable because it represents real behaviour and can be used as a good comparison tool and for calculation. Customer attitude and commitment to a brand or store are important cornerstones of behavioural loyalty. This is why I will measure behavioural loyalty not as a result of switching barriers or habit but in relation to positive attitude. Customer attitude will be captured by the mediators that are included in the model, satisfaction (see 2.3.1), trust (2.3.2) and affective commitment (see 2.3.3) as well as behavioural intentions of customers that will be discussed in section 2.2.1.

2.1.1 E-loyalty definition

E-loyalty is the type of customer loyalty referring to the online environment. Also in research on e-loyalty, researchers distinguish between the two types loyalty as described before. Anderson and Srinivasan (2003) define e-loyalty as ‘the customer’s favourable attitude toward an electronic business, resulting in repeat buying behaviour’. Their measurement of loyalty consists of a

(13)

13

2.2 Dimensions of loyalty

I approach behavioural loyalty is a construct that measures real behaviour of loyal customers. Behavioural loyalty will therefore involve the measurement of the percentage of total expenditures that customers spend at the shop they act loyal to. The loyalty construct is often used by researchers but the measurement differs on the facets of customer loyalty that are included. Facets of customer loyalty that are used are positive Word-of-Mouth (WOM), repeat purchase, continued interaction, low price sensitivity and willingness to pay more (Martin-Consuegra et al. 2007, Mukherjee and Nath 2007, Srinivasan et al. 2002).. For managers customer loyalty is interesting because it refers to customers willing to remain a customer and continue purchasing in the future. Another interesting behaviour of loyal customers is their capability to bring in more customers via recommendation to friends and family, the so called positive Word of Mouth (WOM). These aspects can lead to desirable behaviour of customers but in most cases the customers are asked about future behaviour and therefore about their intentions Customer judge statements like ‘I would like to ...’ (Mukherjee and Nath 2007) and ‘Will you continue....’ (Srinivasan et al. 2002) are used to measure a customer’s future intentions. In my study I will solely focus on two dimensions of customer loyalty which are positive WOM and expected continuity. Since I will measure these two aspects by intentions of the respondent I will refer to it as loyalty intentions. The following paragraph provides a theoretical justification for this choice.

2.2.1 Positive WOM and repeat purchase

Palmatier et al. (2006) argue that loyalty is defined and operationalized in many different ways but they conclude that expectation of continuity and positive WOM are the main outcomes of relationship marketing. Customer’s expectation of continuity, the customer’s intention to maintain the relationship in the future capturing the likelihood of continued purchases, as well as positive WOM, are found to be behavioural outcomes of relationship marketing. In addition, Söderlund (2006) acknowledges that most common application to measure loyalty is to include two loyalty facets, expected continuity and WOM intentions, in the same loyalty measure. They find that expected continuity and WOM can actually be seen as two discrete constructs but are both important outcomes of relationship marketing. Especially WOM is often considered to be an important outcome of relationship marketing. Henning-Thurau et al. (2002) and Srinivasan et al. (2002) both deal with these terms as separate constructs. Other researchers (Bansal et al. 2004, Henning-Thurau et al. 2002) measure customer loyalty as an aggregated construct of positive WOM and repurchase intentions. They consider these two factors as different dimensions of customer loyalty. The previous illustrates that repurchase intentions and positive WOM can be considered the main facets of customer loyalty, either regarded as separate constructs or on an aggregated level. Both constructs are important to managers and involve desirable behavior of loyal customers. This is the reason why I choose these two facets of customer loyalty to focus on in my study about customer behavioural loyalty. The measurement of customer loyalty therefore is limited to the facets of positive WOM and expected continuity and the behavioural aspect of the percentage of total expenditures spent at the web shop.

2.3 Three important mediators

(14)

14

relationships with customers in e-commerce. They found that the three important constructs also apply to the online environment. I will discuss their findings in the next paragraphs.

2.3.1 Satisfaction

Multiple researchers (Bansal et al. 2004, Ribbink et al. 2004, Shankar et al. 2003;) find that e-satisfaction directly affects e-loyalty. E-e-satisfaction is defined as an overall e-satisfaction with the merchant and their services. Shankar et al. (2003) compares the role of satisfaction in an online and offline context and find that satisfaction is even more important online than offline. Ribbink et al. (2004) measure e-loyalty on behavioural outcomes in positive Word of Mouth, expected continuity and website preference and find that e-satisfaction positively affects e-loyalty. Moreover, the empirical study of Bansal et al. (2004) finds that overall website satisfaction is positively related to retention, referral, conversion and stickiness of customers. This confirms the importance of overall satisfaction in behavioural intentions of customers in an online environment and can therefore be considered to be one important relational mediator in building customer loyalty, in line with findings in the offline environment. I specify the following hypothesis:

H1: satisfaction positively affects behavioural loyalty

2.3.2 Trust

Mukherjee and Nath (2007) examine the traditional trust-commitment theory of Morgan and Hunt (1994) in the online context and among consumers. The traditional commitment-trust theory of Morgan and Hunt (1994) shows that trust and relationship commitment are central to successful relationship marketing. They place trust and commitment as key mediating variables between five antecedents, marketing tactics and five business related outcomes. Mukherjee and Nath (2007) apply the same theory to the online environment to identify the importance of trust and commitment in online retailing. Mukherjee and Nath (2007) first acknowledge that online trust is different than offline trust because of several parameters. The online environment creates physical distance between buyer and seller since people buy from a website and not from a brick-and-mortar store. The absence of salespeople also contributes to a larger perceived distance because no real time personal contact is possible which hampers direct interaction. The online environment also means a separation between buyer and products (Yoon 2008). Customers are not able to feel, touch, smell or try the product before purchase. This experience is important for customers to make the right purchase decision. These factors make it different to build trust and commitment in the online environment.

Mukherjee and Nath (2007) define trust as the state when one party has confidence in an exchange partner’s reliability and integrity. They incorporate three key dimensions of trust which are customer’s propensity to trust the retailer, customer confidence in the website, and the customer’s trust in internet technology. Commitment is turn is defined as an enduring desire to maintain a valued relationship. Trust is hypothesized to be positively related to commitment in an online environment which in turn affects customer behavior. The customer behavior measurement consists of WOM, purchase intention and continued interaction. Their findings are that there is a strong positive linkage between trust and relationship commitment in the online environment. Their findings re-confirm the fundamental tenet of the commitment-trust theory proposed by Morgan and Hunt (1994). The trust that a customer places in a web shop is positively related to commitment. Their findings are supported by a study of De Wulf and Odekerken-Schröder (2003) that also confirms the positive effect of trust on relationship

commitment in a business-to-consumer context. Contrary to their expectation of trust being less important in a business-to-consumer context, they clearly found a significant relationship between the two constructs in Belgium and the Netherlands. Mukherjee and Nath (2007) then find that

(15)

15

affects behavioural loyalty in an online environment. This is contrary to the findings of Morgan and Hunt who did not find a significant and positive influence of trust on behavioural loyalty. More researchers have addressed this and trust is generally considered to be an important key element in building online loyalty. Ribbink et al. (2004) also finds that e-trust, defined as the degree of

confidence customers have in online exchanges, or in the online exchange channel, directly affects e-loyalty. Trust is therefore expected to affect behavioural loyalty directly and indirectly through commitment. Higher values of consumer trust leads to commitment in e-transactions. Behavioural intentions are in turn positively affected by higher level of commitment. I specify the following hypotheses:

H2: trust positively affects behavioural loyalty H3: trust positively affects commitment

2.3.3 Commitment

The study of Mukherjee and Nath (2007) confirms the importance of commitment in the establishment of customer loyalty in the online environment. Casaló et al. (2007) also acknowledge that commitment is a key factor in establishing long-term relationships between the consumer and businesses in the online context. Within literature, two types of commitment are distinguished (Geyskens et al. 1996): - Calculative commitment which is caused by the existence of sunk or switching costs.

- Affective commitment which emerges because of the emotional feelings and closeness between the parties.

Calculative commitment will not be considered in my study since it is not a result of relationship marketing, marketing tactics or positive attitude. Creating sunk or switching costs is not a desired objective for TNT Post. Affective commitment is the type of commitment I will focus on. The research of Casaló et al. (2007) refers to affective commitment since more security is generated in the relationship. Affective commitment means that both parties involved in a relationship will be

interested in continuing it in the long term. Some authors suggest that affective commitment determines the consumer’s desire to continue with the relationship in the future (Casaló et al. 2007) and that is the most stable and secure type of commitment. The definition suggests that it is enduring, and it reflects a positive evaluation of a relationship. Many researchers therefore measure affective commitment on aspects like customers feeling loyal to a company (De Wulf and Oderkerken-Schröder 2003), feeling a sense of belonging (Mukherjee and Nath 2007) or caring (Moorman et al. 1992) and willingness to ‘go an extra mile’ to remain customer (De Wulf and Oderkerken-Schröder 2003). Affective commitment involves many aspects of attitude of customers towards an e-tailer. This constructs is therefore considered to capture the measurement of a customer’s attitude towards a company. Attitudinal loyalty was earlier described as ‘the consumer’s predisposition towards a brand as a function of psychological processes’ including attitudinal preference and commitment towards the brand’. The definitions of these two concepts, affective commitment and attitudinal loyalty, are very similar and therefore I suggest that many aspects of attitudinal loyalty are captured by the inclusion of affective commitment. Affective commitment is developed through time and therefore considered to be stable and secure. It measures a customer’s attitude towards an e-tailer or firm based on previous interactions and experience. This suggests that attitudinal loyalty is covered for a large part by this construct. I propose the following hypothesis:

(16)

16

2.3.4 Conceptual model part 1

The relationships that are discussed in the previous chapter are presented in the figure below, figure 1. In the subsequent chapters, I will extent this model and the final conceptual model is presented in chapter 3.4 (figure 1). Mediator Trust Mediator Satisfaction B e h a v io ra l l o y a lty H4 H1 H 3 H2 Mediator Commitment L o y a lty In te n tio n s

Figure 1: conceptual part 1

2.4 Conclusion

(17)

17

3. Factors affecting the mediators

The main focus of my research is which variables affect the mediators that are discussed in chapter 2. Generating insight into variables that influence the mediators will provide important information for managers. For each mediator, satisfaction, trust and affective commitment I will discuss previous research on variables influencing them. I will solely focus on research in the online environment since it is widely available and more relevant.

3.1 Factors affecting satisfaction

3.1.1 E-service quality

E-service quality is about aspects of the website and the web shop. These factors capture the quality of website visit and the web shop. Ribbink et al. (2004) find that e-satisfaction is largely explained by ease of use, web site design, customization, responsiveness and assurance. All these measures are e-service quality dimensions and will be further explained. Ease of use is an essential element of

consumer usage of computer technologies (Davis 1989) and is especially important for new users. It is decisive for customer satisfaction because it enhances the efficiency of using the service. It includes aspects such as functionality, accessibility of information, ease of ordering and navigation. Usability of a website and availability of information as important factors in determining customer satisfaction is also acknowledged by Bansal et al. (2004). Ease of use also includes information accessibility and depth which is addressed by Shankar et al. (2003) who find that ease of obtaining information and perceived depth of information are highly influential on customer’s satisfaction. Based on this previous consideration, I specify the following hypotheses:

H5: website usability contributes to overall customer satisfaction H6: information depth contributes to overall customer satisfaction

H7: ease of obtaining information contributes to overall customer satisfaction

Besides the easiness of use, a website should also be pleasant to the eye and this dimension is captured by website design or e-scape item. Ribbink et al (2004) find that e-scape positively affects customer satisfaction. This finding is confirmed by research of Srinivasan et al. (2003) who find that character defined as an overall image or personality that the e-tailer projects to consumers through the use of inputs such as text, style, graphics, colors and themes on the website. In general, this item is about the look and feel of the website projected through its design. I consider the following hypothesis:

H8: an attractive website design contributes to customer satisfaction

Customization refers to the extent to which the website can be personalized to the needs of consumers. This seems a challenging task because of the lack of human touch in the online context, as discussed before, but Ribbink et al. (2004) emphasize that the e-tailer should strive to customize the service to user’s individual needs e.g. based on past purchases and other information provided by customers. Srinivasan et al. (2002) also find that customization has a positive effect on e-loyalty. Based on the findings of Ribbink et al. (2004) I propose the following hypothesis:

H9: customization contributes to overall customer satisfaction

(18)

18

of websites on perceived relationship investment and in turn on behavioural loyalty. Interactivity is found to affect perceived relationship investment in two dimensions namely two way communication and synchronicity. Similar to Ribbink et al. (2004) two way communication is about the website’s efforts to gather feedback from customers. Synchronicity is about the processing and information speed of the website. Since these findings are rather similar I hypothesize that responsiveness of a website, that is the website’s efforts to obtain customer feedback, is positively related to satisfaction. H10: responsiveness contributes to overall customer satisfaction

3.1.2 Previous interactions

In addition, customer’s satisfaction is expected to be also highly effected by earlier experience of customers. Overall satisfaction of customers is a result of positive judgment about earlier experience with the web shop. This experience highly influences the overall judgment of a customer about the service and product quality. Empirical research of Bansal et al. (2004) confirms that, also in an online context, product selection, price, transaction duration, customer service and delivery affect a

customer’s overall satisfaction. Reassessment of the SERVQUAL scale for the online context by Cristobal et al. (2007) reconfirms this finding. They find that perceived quality is a multidimensional construct consisting of: web design, customer service, return policy, assurance and order management and that perceived quality influences on satisfaction. Moreover, Parasuraman et al. (2005) include another important item to the general analysis of perceived quality especially in the online

environment. They suggest that system availability is also an important dimension of perceived quality in the online context in general. In my study I consider this item especially valuable because usage of the photo service of TNT Post requires some downloading and uploading which is susceptible to technology failure. These findings confirm that customers’ previous interactions with an e-tailer affect satisfaction. Previous experience of the service quality in total of a web shop is expected to be an important construct in building customer loyalty. Since my study aims to uncover differences in importance of separate factors, I specify hypothesis for every aspect, factor separately. I will not specify hypotheses about the order of importance but I derive the following hypotheses to be able to identify differences in all these factors in their effect on overall satisfaction:

H11: previous experience in … positively influences overall satisfaction H11a: delivery

H11b: price

H11c: customer service H11d: return policy H11e: system availability H11f: assortment

3.2 Factors affecting trust

3.2.1 Reputation

(19)

19

3.2.1 Security

In the study of Mukherjee and Nath (2007), privacy turns out to be the most significant determinant of consumer trust. It is an expected result as privacy is the key factor affecting the trustworthiness of an online retailer. Security is the second most important determinant of consumer trust. Use of secured payment mechanisms, no leakage of credit card information, and past attempt on website hacking all played significant role in developing a sense of security in the consumers mind. In their study, privacy and security are measured very extensively and on many dimensions. Privacy addresses the issue of the protection of individually identifiable information on the internet. They suggest that the privacy policies of an online retailer involve the adoption and implementation of a privacy policy, notice, disclosure, and choice/consent of consumers. The measurement is about credibility of the online retailer not to sell customer information or send unsolicited mails. Privacy in online transactions in the Netherlands has been already captured by national regulations and laws and therefore is considered to be a less distinctive item for e-tailers in current business. Privacy will therefore not be incorporated in my study. The following hypothesis specifies the expected effect of security.

H13: security contributes to trust in an online environment

3.2.3 Contact interactivity

Contact interactivity is addressed by Srinivasan et al. (2002) and involves important aspects for the online environment. Research of Capgemini (Trends in Retail 2008, www.nl.capgemini.com) revealed that customers get more and more power and require much information before purchasing. They expect the e-tailer to take them by the hand and enable them to make the right decision. Interactivity of the websites that enables customer to search for products, have a good look at it and compare with other offers is important. This is considered to affect e-loyalty trough trust since it facilitates customers to make a good decision. I specify the following hypothesis:

H14: contact interactivity positively affects trust

3.3 Factors affecting commitment

3.3.1 Perceived relationship investment

In the research of Yoon et al. (2008) a new construct within the establishment of customer relationships is perceived relationship investment. This construct is introduced by De Wulf et al. (2001). They define perceived relationship investment as a consumer’s perception of the extent to which a retailer devotes resources, efforts, and attention aimed at maintaining or enhancing

relationships with regular customers that do not have outside value and cannot be recovered if these relationships are terminated (Smith 1998). De Wulf et al. (2001) and Yoon et al. (2008) all find that perceived relationship investment positively affects an aggregated measurement of satisfaction, trust and commitment, which in turn affects behavioural loyalty. I consider this concept to be an important mediator between marketing tactics and customer commitment. The definition of perceived

relationship investment shows that it is a construct that represents important emotional and attitudinal value of consumers. Since affective commitment also emerges because of the emotional feelings and closeness between the parties (Geykens et al. 1996), perceived relationship investment is considered to be a mediator between marketing tactics and commitment. The following hypothesis is based on this consideration:

H15: perceived relationship investment positively affects commitment

3.3.2 Factors affecting perceived relationship investment

(20)

20

is a fundamental building block of a relationship’. The definition of interactivity is adapted from Liu (2003) as a construct consisting of three distinct dimensions: active control, two-way communication, and synchronicity. The interpersonal communication antecedent which was addressed by De Wulf et al. (2001) was dropped but other marketing tactics like direct mail, preferential treatment and tangible rewards are included. The direct mail component was replaced by emailing. They found that the marketing tactics of direct mail, rewards and two dimensions of website interactivity (namely two-way communication and synchronicity) positively affect perceived relationship investment which in turn positively affects relationship quality and behavioural loyalty in an online context. Two-way

communication is already captured by the contact interactivity concept and synchronicity is captured by the responsiveness. Both factors are expected to mainly affect customer satisfaction and will not be hypothesized to affect trust. Nevertheless, personal rewards and communication in the sense of e-mailings are taken into account and the following hypotheses are derived:

H16: rewards contribute to an increase in perceived relationship investment H17: communication contribute to an increase in perceived relationship investment

3.4 Final Conceptual model

The previous hypotheses can be combined to one model including all the factors and mediators. In this conceptual model I hypothesize that the factors as discussed before affect the mediators which in turn contribute to customer loyalty. The research will finally reveal whether these mediators and the relationship between factors and mediators are present or that some factors affect customer loyalty directly. The conceptual model demonstrates the proposed relationships within my study.

Mediator Commitment Mediator Trust Mediator Satisfaction Factors Website usability (H5) Information depth (H6)

Ease of obtaining information (H7) Website design (H8)

Customization (H9) Responsiveness (H10)

Previous interactions (H11a - H11f) Factors Reputation (H12) Security (H13) Contact interactivity (H14) H12, H13, H14 H5, H6, H7, H8, H9, H10, H11 Factors Economic rewards (H16) Communication (H17) B e h a v io ra l l o y a lty Perceived relationship investment H4 H1 H 3 H2 H16, H17 H15

(21)

21

4. Methodology

The E-business department of TNT Post incorporates several online shops which were briefly discussed before. My study will try to explain which factors are important to customers in order to become loyal to an online store. The main part of my study therefore consists of a survey among online shoppers about this topic. To test the proposed hypotheses I will distribute a questionnaire to online shoppers about a self reported online shop they are loyal to. The questionnaire is distributed by e-mail with a direct link to an online questionnaire. The sample is drawn from current customers of two TNT Post web shops, Kadowinkel and Fotoservice.

4.1 Sample and procedure

The advantage of using existing customers from the two TNT shops is twofold. First of all, it provides me with a large customer base in order to obtain a large sample size. Secondly, the current online customers can be considered to have experience in online shopping. I consider it an important

condition for the purpose of my study that the sample consists of merely experienced online shoppers. The total customer database consists of 180.000 customers of Fotoservice and 20.000 customers of Kadowinkel. The first step was to obtain all email-addresses of future respondents. In order to obtain a representative and sufficient sample size, a total of 2000 customers of both shops were selected to participate in my study. An online survey was sent out and since the customers under study are considered to be familiar with the online medium and online shopping, an online survey that is sent by email is appropriate. Customer can fill out the questionnaire directly via the link and the answers will be saved and sent to a database. The software that is used for this online questionnaire is provided by TNT Post and is licensed by Survey World. In order to double-check a respondent’s internet

experience, the first question of the questionnaire is included to check this. To stimulate response I put ten gift vouchers up for raffle for completing the questionnaire.

The questionnaire for this group of customers is used to gather data to test the proposed hypotheses. My study wants to uncover the factors that are important to customers to become or stay loyal to a web shop. The main condition is that customers need to provide information about any shop towards which they act loyal and therefore is not beforehand about any of the TNT Post shops. Because the sample is extracted from current customers of the two TNT Post shops, the sample might be biased.

4.2.2 Procedure

The questionnaire was sent out to two groups of customers. The first group was used to test the response rate and the second group was used as a backup when the response was low. Customers received the e-mail on a Thursday since customers are likely to fill out a questionnaire during the weekends. The reminder, then, was sent out after one week. During the first period the questionnaire was sent to 1000 customers in total of which 700 customers of Fotoservice and 300 customers of Kadowinkel. The response rate was 17% and in one week 170 respondents completed the

(22)

22

Invitation # E-mailaddresses # Respondents Response rate

Kadowinkel Fotoservice Kadowinkel Fotoservice Kadowinkel Fotoservice

First group 300 700 98 72 32,67% 10,29%

Second group 500 500 72 68 14,40% 13,60%

800 1100 170 140 21,25% 11,67%

Total 2000 310 15,81%

Table 1: overview questionnaire invitations and response rate

The total sample consists of 198 females (63,8%) and 112 males (36,2%). The education level of the sample is medium up to high and most respondents are in a family with one or more children.

4.2 Representativeness of the sample

The representativeness of the sample can be tested by the non response bias. In order to test the non response bias, the differences between the respondent group and the non respondent group are

analyzed. The final sample of respondents consists of customers from the Fotoservice and Kadowinkel of TNT Post as well as the non respondent group. However, the response rate of customers from the Kadowinkel is 21,25% (170 out of 800) and the response rate of the Fotoservice customers is 140 out of 1200 (11,7%). The total sample therefore contains more customers from the Kadowinkel than from the Fotoservice which might be less representative. The distribution of females and males in the non response sample is not much different from the sample, 61% females and 39% males, as well as the characteristics in education level and family situation. Since the sample consists of more females than males, the sample might be less representative. Nevertheless, the experience of the online shops from TNT Post is that for all the shops, females are the majority of customers. The sample therefore is representative for the online activities of TNT Post.

Sample Non sample

Female 64% 61%

Male 36% 39%

Education level

Elementary school 0.5% 1,0% Low vocational education 1.5% 1,5% Average general education 8,0% 6,0% Average vocational education 18,0% 20,0% Higher general education 16,5% 14,0% Higher vocational eductaion 34,0% 32,5%

University 23,5% 25,0%

Family situation

Family without children 23% 28%

Single person 55+ 8% 6%

Single person < 55 24,5% 22% Family with children with youngest < 12 34% 29% Family with children with youngest > 12 10% 15% Table 2: characteristics of sample versus non sample

The large response of customers from the Kadowinkel might indicate that the involvement of these customers is different from the Fotoservice which might also indicate a self-selection bias.

(23)

23

respondents filled out the questionnaire on a non TNT Post web shop suggests that the self selection bias does not affect generalizability. The customers from the Kadowinkel might be more involved in or attached to their TNT web shop but since they completed the questionnaire on a self-reported favorite shop it does not hurt representativeness of the sample for my study.

Another aspect that can affect representativeness is captured by the response bias. This occurs when customers answer questions in the way they think the researcher wants them to answer rather than according to their true beliefs. Since respondents answer the questions about a web shop which is not from the sender of the survey, I consider the response bias to be minimal. Respondents do not have a reason to answer socially correct answers because they were free to choose their own online shop. In the e-mail and the introduction text respondents were made clear that the intention of the research was to obtain an objective view about the evaluation of web shops in general and that we did not expect them to consider a TNT Post web shop as their favorite web shop at any price. The response bias thanks to morally right answers is minimized.

Moreover, response bias might occur when the researcher is obviously angling for an answer. I consider this problem to be minimized as well since the questionnaire involves many questions about characteristics and activities of the web shop which can be objectively evaluated by the respondents. Nevertheless, since I asked the respondents to fill out the questionnaire on their favorite web shop and the subjects in the questionnaire are considered to be of importance in the loyalty process, the

assessment of these subjects are inclined toward the higher end scores. Moreover, the questions in the questionnaire were almost all posed in a positive way on a 7 point Likert scale and this can lead to Common Method Variance. This occurs when answers of customers are influenced by the way of posing questions. On the one hand solely asking 7 Point Likert scale questions might cause bias since customers are likely to fill out an answer by habit or without reading the questions. On the other hand, posing solely positively formulated questions might be more reliable when it is difficult for

respondents in an online questionnaire to switch to a negatively posed question and will therefore cause misinterpretation. I have tried to reduce the Common Method Variance by including an open question before a series of 7 Point Likert scale in order to test for unexpected variables that could be of impact on the loyalty process which was not addressed by the questionnaire. In addition, I have tried to reduce the questionnaire as much as possible to lower fatigue. Moreover, I have added some multiple choice and yes/no questions. Nevertheless, I asked loyal customers in my study which yields rather positive scores on most subjects. This fact will make it more difficult to identify and explain differences. I consider this a limitation of the study.

4.3 Operationalization of constructs

For all variables in my study, measures were borrowed from previous studies. This section provides a justification for the use of these measurement items. The measurement items were turned into

questions in the questionnaire which is presented in Appendix A. I will first discuss the dependent variables in the model which are the mediators and the measurement items of behavioural loyalty and loyalty intentions. An overview is presented in table 1.

4.3.1 Dependent variables

(24)

24

For the general survey I need to incorporate a question to uncover the online shop towards which customers are loyal. Since this can be any online shop, this question is important because the questionnaire will continue on this web shop. The following question is formulated: ‘what is your favorite online shop where you shop often and gladly’ (QUESTION 2) and is based on research of Casaló et al. (2008). Customers are not allowed to answer auction websites. Question 3 then measures the frequency of buying at the favorite web shop. It is based on principle to measure the percentage of total expenditures on one product or product category that is spend in the one store or brand under study (De Wulf et al. 2001, De Wulf and Odekerken-Schröder 2003). This question is a measurement of behavioural loyalty as well as a good instrument for comparison in future and adapted from De Wulf et al. (2001). Question 4 is an open question where respondents can express why a certain web shop is their favorite web shop and why they act loyal to it. Answers to this question can provide important insights on factors that were not included in the study. The questions to measure the

mediating variables and the dependent variable are presented in table 1 and will be further explained in the following paragraphs.

The dependent variable consists of two dimensions of behavioral intentions resulting from customer loyalty. As discussed before I will focus on WOM and repeat purchase. Question 5.1 measures WOM and is based on research of Henning-Thurau et al. (2002). Question 5.2 measures repeat purchase intention and is based on studies as for example Ribbink et al. (2004).

4.3.2 Mediators

In many surveys the mediators, satisfaction, commitment and trust are often measured by three questions. Since the main focus of my study is to reveal important but many factors that influence these constructs, the questionnaire will be rather lengthy. In order not to bother the respondent with similar questions about the same construct, I chose to reduce the measurement scale to a single item for all three constructs. Overall satisfaction is adapted from Shankar et al. (2003) who measured overall satisfaction by the expressed level of overall satisfaction with the service provider. For my study, overall satisfaction (QUESTION 6.1) is measured by one single item; ‘how satisfied are you about this web shop‘. The measurement of trust and commitment by is often done by three questions. An example of the measurement of trust is proposed by De Wulf et al. (2001) and involves questions like ‘this store gives me a feeling of trust’, ‘I have trust in this store’ and ‘this store gives me a trustworthy impression’. Inspired by Shankar et al. (2003) who reduced the measurement of

satisfaction to a single item I decided to do the same for trust. The one item that remains is ‘the store gives me a trustworthy impression’ (QUESTION 9.2). Measurement of affective commitment is based on Verhoef (2003) which uses the questions ‘I am a loyal customer of this store’, ‘I feel strong

attachment to this store’ and ‘I feel a sense of belonging with this store’. Measurement is reduced to one question, ‘I consider myself a loyal customer of this store’ (QUESTION 5.3) based on Verhoef (2003) and Henning-Thurau et al. (2002). The final mediating variable is perceived relationship investment. This variable is considered to mediate the effect of the independent variable of economic rewards and communication on commitment. This variable is captured by question 8.7.

Dependent variable Nr. Question Loyalty intentions

Propensity to

recommend 5.1 I am likely to recommend this web shop to friends or family

Henning-Thurau et al. (2002)

Propensity to continue

purchasing 5.2 It is likely that I continue purchasing from this web shop Ribbink et al. (2004). Behavioural loyalty

Percentage of

expenditures 3

Of the 10 times you the type of products like in your favorite store, how

(25)

25

4.3.3 Factors

The model that I will test contains 12 independent variables, the factors. The independent variables are expected to simultaneously affect one of the mediators separately. In contrary to the mediators, the independent variables are not supposed to become a dependent variable in the next stage of analysis. Some independent variables will be measured by a single item whereas other variables are measured by multiple items to measure a construct. Reduction of the amount of measurement items is considered necessary in order to obtain an acceptable length of the questionnaire. For independent variables that could reasonable be reduced to a single item without losing information were reduced to a single item scale. Other variables that are considered to be less easy assessable by one single item and contained important additional measurement scales are still measured by multiple items. Table 1 provides a complete overview of the measurement scales for the independent variables and constructs that will be used in the questionnaire and analysis.

Independent variable

(factor) Nr. Question Adapted from:

Website usability 8.1 I find the web shop user friendly Ribbink et al. (2004), Bansal et al. (2004)

Information depth 8.3 I think the quality of information at the web shop is good Shankar et al. (2003) Ease of obtaining

information 8.4

I think it is easy to find the right information at the web

shop Shankar et al. (2003)

Website design 8.5 the website is appealing to me Ribbink et al. (2004) Customization 7.3 The web shop offers me the possibility to personalize

products Srinivasan et al. (2002)

7.4 The web shop offers me interesting recommendations on

the website or in e-mailings Srinivasan et al. (2002) 7.5 The web shop offers me advertisement and promotions

that are tailor-made for me Srinivasan et al. (2002) Responsiveness 9.3 The web shop appreciates visitor’s feedback Ribbink et al. (2004).

9.5 The web shop makes it easy to get into contact with the

web shop Ribbink et al. (2004).

9.6 The web shop facilitates the sharing of information and

experiences about products Ribbink et al. (2004). Previous experience 6.2 I am satisfied about the customer service Parasuraman et al. (2005)

6.3 I am satisfied about the return policy of the web shop Parasuraman et al. (2005) 6.4 Delivery is always on time Parasuraman et al. (2005) 8.10 Delivery is very quick Parasuraman et al. (2005) 6.5 I am satisfied about the process of the web shop Parasuraman et al. (2005) 9.1 The website never had any technological failures Parasuraman et al. (2005)

7.1 The web shop offers me the right assortment Srinivasan et al. (2002), Bansal et al. (2004)

7.2 The web shop offers me a large assortment Srinivasan et al. (2002), Bansal et al. (2004)

Reputation 8.5 I think the web shop has a good reputation Srinivasan et al. (2003) Security 8.6 I think that the security measures are up to date Mukherjee and Nath (2007) Mediators

Overall satisfaction 6.1 how satisfied are you about this web shop Shankar et al. (2003) Trust 9.1 the web shop gives me a trustworthy impression De Wulf et al. (2001), Shankar et al. (2003) Commitment 5.3 I consider myself a loyal customer of this store Verhoef (2003),

Henning-Thurau et al. (2002). Perceived relationship

(26)

26

Contact interactivity 8.9 I find it easy to compare products Srinivasan et al. (2002) 9.4 The web shop provides tools to view products from

different angles Srinivasan et al. (2002) Rewards 8.8 I find that the web shop rewards regular customers De Wulf et al. (2001). Communication 9.7 The web shop regularly informs me De Wulf et al. (2001)

10 The web shop sends me e-mailings? De Wulf et al. (2001) 11 The web shop regularly sends me other information De Wulf et al. (2001) E-mailings 12.1 I am satisfied about the e-mailings Vriens et al. (1998)

I like the offerings in the e-mailings Vriens et al. (1998) I like the design of the e-mailings Vriens et al. (1998) I receive the e-mailings on a relevant moment Vriens et al. (1998) I like the information in the e-mailings Vriens et al. (1998) Table 4: Measurement items of independent variable

4.3.4 Specified relationships

The measurement items in the questionnaire will address all hypotheses that are stated before in several stages. In order to assess the relationships among independent variables, mediators and the dependent variable I will use multiple regression analysis and factor analysis. In my study, the model consists of multiple independent variables, the factors which are expected to affect the dependent variable through three important mediators. The three mediators between independent variables and customer loyalty are overall satisfaction, trust and commitment. Moreover, perceived relationship investment is also a mediating variable between independent variables and commitment.

Satisfaction is a mediator between previous interactions, customization, responsiveness, website design, website usability, information depth and ease of obtaining information as independent

variables, or factors and the dependent construct of loyalty. Trust is a mediating variable in the model between contact interactivity, reputation and security as independent variables and the dependent construct of loyalty. Commitment in turn is dependent of trust directly and a mediator between perceived relationship investment and the dependent construct. The effect of economic rewards, communication on commitment is mediated by perceived relationship investment.

4.5 Conclusion

(27)

27

5. Results and conclusions

This chapter discusses the results of my study. First I will discuss the outline of the data that is collected on 310 respondents. Secondly, I will discuss the analysis on the data and the findings. These findings will address the hypotheses that are specified before. Thirdly, I will specify the final models based on the findings.

5.1 Data description

The total sample equals 310 respondents. The representativeness of the sample has already been discussed and will therefore not be addressed in this section. The top three favorite web shops that were mentioned by respondents are Bol.com (N=85, 29%), Wehkamp (N=29, 10%) and H&M (N=25, 9%). Other web shops that were mentioned more frequently are Esprit.com (N=18, 5.8%),

Neckermann.com (N=9, 2.9%). Analyses of the open ended question about the reason why

respondents are loyal to the web shop do not show any surprising results. This question was inserted in order to test for factors that are important to customers in the loyalty process which were not

addressed in the questionnaire. The open answers reveal that respondents mention five topics as being important for staying loyal. These five topics are the convenience of online shopping, the size and characteristics of the assortment, usability of the website, and price (products and delivery) and service of the web shop as main reason to continue buying at their favorite shop. Appendix B contains all the open answers of customers. Many respondents mention more than one of these five topics as a reason to continue buying at a certain web shop. Since all answers of respondents can be assigned to one of the five topics, this analysis suggests that the questionnaire addressed the main facets of customer loyalty and that no important facet is lacking. This finding implies that the theoretical foundation for the study and the questionnaire is solid.

The mean scores of the respondents on the three items measuring customer loyalty are high. Table 3 shows the average scores for all three variables. The standard deviation is rather small as well which indicates low variance in the dependent variables. This result is logical since all respondents in the sample are loyal customers of the web shop which are likely to positively evaluate their favorite store. The loyalty measures in the questionnaire therefore score high but the low variance will make

explanation through regression more difficult.

N Minimum Maximum Mean Std. Deviation

Percentage expenditures 296 0 10 6,43 2,763

Propensity to recommend 291 1 7 6,05 1,320

Propensity to continue purchasing 294 1 7 6,31 1,133

Table 5: loyalty scores of sample

Loyal customers indicate that they buy, on average, 6.5 out of 10 times at theit favorite shop. Further analysis of the frequency of the answers in the sample reveals that the largest part of respondents indicates to buy 8 out of 10 times at their favorite web shop when buying this type of products. Surprisingly, almost 17% of the respondents indicate to buy always at their favorite web shop. These results are interesting for managers since it can be used in calculation of the revenue that can arise from an increase in loyalty.

(28)

28 indicate a number between 0 and 10

Frequency Percent Cumulative Percent

Valid 0 5 1,6 1,7 1 7 2,3 4,1 2 21 6,8 11,1 3 23 7,4 18,9 4 23 7,4 26,7 5 32 10,3 37,5 6 20 6,5 44,3 7 28 9,0 53,7 8 65 21,0 75,7 9 22 7,1 83,1 10 50 16,1 100,0 Total 296 95,5 Missing 14 4,5 Total 310 100,0

Table 6: details about loyal behaviour

Question 5.1 and 5.2 ask about the loyalty intentions of respondents regarding their favorite web shop. Respondents were asked to indicate to what extent they agree with the statements given on a 7 point Likert scale. Table 5 shows the frequencies of response for the sample. Similar to table 3, the review of data reveals that customers are more likely to continue purchasing than recommending the web shop to friends and family. Nevertheless, the mean scores for both loyalty intentions are above 6 and the majority of respondents answer the maximum score for the propensity to continue purchasing and propensity to recommend the web shop to friends and family.

Question 5.1 :It is likely that I recommend this web shop to friends and family Question 5.2 : It is likely that I continue purchasing from this web shop

Frequency Q 5.1 Percent Q5.1 Frequency Q5.2 Percent Q5.2 Valid 1 2 0,6% 3 1% 2 6 1,9% 4 1,3% 3 11 3,5% 4 1,3% 4 19 6,1% 5 1,6% 5 32 10,3% 33 10,6% 6 68 21,9% 69 22,3% 7 153 49,4% 176 56,8% Total 291 93,9% 294 94,8% Missing System 19 6,1% 16 5,2% Total 310 100,0 310

Table 7: details about loyalty intentions

(29)

29

answers among respondents do not differ much from the mean score. This low variance in the data will make further analysis difficult.

5.1.3 Differences among respondents

The optimal objective for an increase in loyalty for managers is that customers shop more frequently at their shop whenever they need the kind of product they are offering. This is about real behaviour of customers. Therefore it is interesting to go into more detail about the difference between respondents based on their indication of loyal behaviour. For this purpose I will execute a one way ANOVA test that will reveal whether there exists significant difference among the groups of respondents based on their answer on question 3 (the percentage of expenditures at the favorite web shop).

The one way ANOVA test reveals that respondents differ significantly on 14 out of 32 questions in the questionnaire. Respondents differ significantly on a 0.05 significance level on questions 5.1, 6.1, 6.3, 6.4, 7.1, 7.2, 7.4, 8.4, 8.7, 9.1, 9.2, 9.3, 9.4 and 9.6. For these items respondents that shop less frequently at their favorite shop are significantly less positive than other respondents.

Difference in groups of respondents

Sig. Question 5.1 : It is likely that I recommend this web shop to friends and family ,000

Question 5.2 : It is likely that I continue purchasing from this web shop ,000

Question 5.3: I consider myself a loyal customer of this web shop ,000

Question 6.1: I am satisfied about this web shop in general ,001

Question 6.3: I am satisfied about the return policy of the web shop ,017

Question 6.4: I am satisfied about the prices of the web shop ,003

Question 7.1: The web shop offers me the right assortment ,019

Question 7.2: The web shop offers me a large assortment (for this type of products) ,025 Question 7.4: The web shop offer me interesting recommendations on the website or in e-mailings ,008

Question 8.4: I find the quality of information good ,044

Question 8.7 : I find the security measures up to date ,029

Question 9.1 : The website never has any technological failures ,044

Question 9.2 : The web shop gives me a trustworthy impression ,009

Question 9.3 : The web shop appreciates visitor’s feedback ,000

Question 9.4 : The web shop provides tools to view products from different angles ,040 Question 9.6 : The webs hop facilitates the sharing of information and experiences about products ,024

Table 8: difference among groups of respondents

First of all, groups of respondents based on purchase frequency seem to differ significantly on loyalty intentions. In general are higher scores on loyalty intentions reported by respondents wit a high purchase frequency. Moreover, respondents seem to differ on the level of commitment to a web shop, covered by question 5.3, overall satisfaction covered by question 6.1 and trust covered by question 9.2. Moreover respondents mainly differ significantly on factors that are expected to affect overall

satisfaction (question 6.3 up to 8.4 and 9.1) and on questions that measure the interactivity of websites (question 9.3 up to 9.6). These factors seem to be the most distinctive variables in the explanation of loyal behaviour. Subsequent steps in analysis will further investigate the relationships between

How to build customer loyalty in an online environment?

How to build customer loyalty in

an online environment?

How to build customer loyalty in

an online environment?

Quantitative research on factors and mediators that play a role in the online

loyalty formation process in the Netherlands.

The Hague, October 2009

Faculty: Economics and Business

Specialization: Marketing Management

Author: Kim Pennings (F.J.H.)

Organisation: University of Groningen

Company: TNT Post

Supervisor 1: Dr. Wander Jager

Preface

Table of contents

Figures and tables

Figures

Tables

Executive summary

1. Introduction

1.1 Introduction to the E-business development department at TNT Post

1.2 Managerial relevance

1.3 Academic relevance

1.4 Structure overview

2. Customer loyalty theory

2.1 Customer loyalty definitions

2.2 Dimensions of loyalty

2.3 Three important mediators

2.4 Conclusion

3. Factors affecting the mediators

3.1 Factors affecting satisfaction

3.2 Factors affecting trust

3.3 Factors affecting commitment

3.4 Final Conceptual model

4. Methodology

4.1 Sample and procedure

4.2 Representativeness of the sample

4.3 Operationalization of constructs

5. Results and conclusions

5.1 Data description