• No results found

The influence of personalized direct mail on Value to Customer and Value to Firm

N/A
N/A
Protected

Academic year: 2021

Share "The influence of personalized direct mail on Value to Customer and Value to Firm"

Copied!
62
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Nynke de Vos | November 2010

(2)
(3)

Nynke de Vos

University of Groningen

Faculty of Economics & Business

Department of Marketing

Msc Business Administration: Marketing Management

Master thesis

Completion date November, 28, 2010

Author Nynke de Vos

Student number 1830783

Adress Curacaostraat 45B

8931 CH Leeuwarden 06 42053828

nynke_devos@live.nl First supervisor Prof. dr. J.C. Hoekstra Second supervisor Drs. G.F. Haanstra

(4)
(5)

MANAGEMENT SUMMARY

Customers nowadays are more demanding and customer data is easier available due to new data technology. Companies are moving from a product- or brand-centric view toward a customer-centric approach. Individual, or one-to-one marketing is becoming more popular. This research focuses on reaching individual customers by means of personalized direct mail. The problem statement of this study is:

‘What is the effect of a personalized direct mail on ‘value to customer’ and ‘value to firm?’

A personalized direct mail is directed to individual customers or to specific segments. Two segments are targeted during this study: families and empty nesters. The mailing includes special promotions and texts adjusted to this specific segment.

Value to consumer (V2C) includes three components: value equity, brand equity and relationship equity. Value equity is defined as ‘the customer’s objective assessment of the utility of the brand, based on what is given up for what is received’. Brand equity has to do with the customer’s ‘subjective and intangible assessment of the brand.’ Relationship equity is defined as the tendency of the customer to stick with the brand (Lemon et al., 2001).

Value to firm (V2F) also includes three components: the net promotor score (NPS), revenue and retention. For this study revenue can be defined as ‘the expenditures of the customer to the firm’ (Verhoef et al., 2010). NPS can be defined as ‘the likelihood that a customer will recommend a company to a friend or colleague’ (Reichheld, 2003). Retention can be defined as ‘the probability of a customer being ‘alive’, or repeats buying from a firm (Gupta & Zeithaml, 2006).

(6)

The following hypotheses are tested: Hypothesis

H1 A personalized direct mail has positive impact on 1a: value

equity; 1b: brand equity; 1c: relationship equity. Not Supported

H2

Value to customer positively influences value to firm in terms of 2a: revenue, 2b: the net promoter score, and 2c: customer retention.

Supported

H3

The more positive the attitude of customers toward

personalization, the stronger the impact of a personalized direct mail is on value to customer.

Not Supported

The results of this study show that the positive effect of a personalized direct mail on V2C (H1) is not supported, just as the moderating effect of attitude toward personalization (H3). On the other hand, the expected positive impact of V2C on V2F is significant (H2). This is consistent with previous findings.

Three control variables were taken into account: satisfaction, relationship quality and involvement. From the results it can be concluded that relationship quality and involvement do have a significant positive impact on both V2C and V2F, except from involvement in the family group when testing H3. Satisfaction only has a significant positive impact on V2C. This study cannot prove the positive effect of a personalized direct mail on V2C and V2F. However, the study underlines the importance of creating customer value, since this is a determinant of future firm value. Keeping customers satisfied, involved and creating a good customer relationship are proven to have an influence on V2C.

(7)

PREFACE

This paper includes my final assignment of the Master Business Administration Marketing Management at the Rijksuniversiteit Groningen. The research is about the effect of personalized direct mail on Value to Consumer and Value to Firm.

I want to thank my supervisor Janny Hoekstra for giving me professional guidance and feedback during this research. At the beginning we struggled with the research design, but I think we did a good job. I also want to thank my second supervisor Gert Haanstra for supporting me at the ending phase of this master thesis.

Nynke de Vos

(8)
(9)

TABLE OF CONTENTS

MANAGEMENT SUMMARY ... 5

PREFACE ... 7

1. INTRODUCTION ... 11

1.1.GENERAL INTRODUCTION ... 11

1.2.PROBLEM STATEMENT & RESEARCH QUESTIONS ... 13

1.3.RESEARCH METHOD ... 13

1.4.SCIENTIFIC AND MANAGERIAL RELEVANCE ... 14

1.4.1. Scientific relevance ... 14

1.4.2. Managerial relevance ... 15

1.5.STRUCTURE OF THE PAPER ... 16

2. CONCEPTUAL MODEL & HYPOTHESES ... 17

2.1.PREFERENCE FIT ... 18

2.1.1. Customization ... 18

2.1.2. Customer preference fit ... 20

2.1.3. Customized promotions ... 21 2.2.VALUE TO CUSTOMER ... 22 2.2.1. Value equity ... 22 2.2.2. Brand equity ... 23 2.2.3. Relationship equity ... 24 2.3.VALUE TO FIRM ... 25 2.3.1. Revenue ... 25

2.3.2. Net Promoter Score ... 26

2.3.3. Retention ... 26

2.4.ATTITUDE TOWARD PERSONALIZATION ... 27

3. RESEARCH DESIGN ... 29

3.1.INTRODUCTION TO COMPANY X ... 29

3.2.PROCEDURE ... 29

3.3.SAMPLE & DATA COLLECTION ... 32

3.4.MEASURES ... 33 3.5.PLAN OF ANALYSIS ... 37 4. RESULTS ... 41 4.1.DESCRIPTIVE STATISTICS ... 41 4.2.HYPOTHESES TESTING ... 44 5. DISCUSSION ... 47

5.1.DISCUSSION & CONCLUSION ... 47

5.2.LIMITATIONS AND DIRECTIONS FOR FURTHER RESEARCH ... 48

5.3.MANAGERIAL IMPLICATIONS ... 50

REFERENCES ... 51

(10)
(11)

1. INTRODUCTION 1.1. General introduction

“A company’s current customers provide the most reliable source of future revenues and profits” (Lemon et al., 2001).

The world is changing: customers are more demanding, individualistic and do not have much time (Reinartz & Kumar, 2006). Understanding individual customers is becoming increasingly important. Organizations are moving from a product- or brand-centric marketing toward a customer-centric approach (Reinartz et al., 2004). Customer-centric companies are externally oriented, use individual customer needs as starting point, are relationship-oriented and use customer metrics like satisfaction and CLV to assess performance (Shah et al., 2006). Building customer relationships can be seen as a key toward success. Due to these changes in the market environment, the concept of customer relationship management (CRM) has emerged rapidly. Successful CRM is not only about data technology, but about a companywide strategy focused on the value of its customers (Bohling et al, 2006). CRM acquisition and retention processes are positively related to economic performance (Reinartz et al., 2004).

Within CRM, marketing instruments are used for building customer relationships and creating customer loyalty (de Wulf et al., 2001; Bolton et al., 2004). Relationship marketing has emerged and can be defined as ‘all marketing activities directed towards establishing, developing and maintaining successful relational exchanges’ (Morgan & Hunt, 1994). De Wulf et al. (2001) empirically tested different relationship marketing tactics and showed that these tactics have a positive impact on perceived relationship investment, relationship quality and behavioral loyalty. Besides, marketing tactics can have a positive effect on CLV (Bolton et al., 2004). It is therefore important to develop a good strategy for the implementation of marketing instruments and to also consider, besides the short-term effects, the long-term effects of such actions.

(12)

has emerged, which is about adjusting the companies’ marketing mix to the wishes of the individual customer (Peppers et al., 1999; Arora et al., 2008). Companies can start a ‘learning relationship’ with their customers. Individual customers can teach the company about their preferences and needs, which can lead to a high competitive advantage (Pine et al., 1995). Customized marketing solutions are useful for both acquisition and retention and can engender successful, long-term relationships (Ansari & Mela, 2003). Besides, customers are becoming more heterogeneous and customized promotions are proven to yield large increases in revenue and profits relative to uniform promotion policies or mass marketing (Khan et al., 2009). Customer preference fit is one of the key success factors of customized offers (Simonson, 2005; Franke et al., 2007, Kramer, 2007). Zhang & Wedel (2009) state that ‘customization is considered a key solution by marketers looking for ways to improve the effectiveness of promotions’.

This paper focuses on one aspect of customization: personalized direct mail. ‘Personalized’ means that the mailing is directed to one specific segment. Moreover, the mailing is personalized in a way that the customer's name is used in the beginning of the letter. The mailing is adjusted to the needs of the customer segment and includes personal information about the customer's retailer. The mailings used in this study are sent offline, by postal service.

Direct mail is a marketing tool which ‘can be employed to communicate directly with specific individuals and/or households in order to transmit direct marketing offers and messages’ (Vriens et al., 1998). It can also be defined as ‘an addressed, written, commercial message that is delivered to the addressee by a postal service, and may be used to accomplish three objectives: (1) cognitive effects (transferring information, brand awareness), (2) affective effects (image building) and (3) behavioral effects (sales or information inquiries) (Van der Scheer et al., 1996; Vriens et al., 1998).

This study further investigates the field of customization by testing the effect of a personalized direct mail, that fits customer preferences, on the drivers of customer equity (‘value to customer’, or V2C), as defined by Lemon et al. (2001): brand equity, value equity

(13)

Customer equity can be defined as ‘the total of the discounted lifetime values of all the firm’s customers’ (Lemon et al., 2001; 2004), so the sum of all CLV’s. Value equity is defined as ‘the customer’s objective assessment of the utility of the brand, based on what is given up for what is received.’ Brand equity has to do with the customer’s ‘subjective and intangible assessment of the brand.’ Relationship equity is defined as the tendency of the customer to stick with the brand (Lemon et al., 2001).

Customer equity is proven to be a determinant of firm value (Gupta et al., 2004; Gupta & Zeithaml, 2006; Gupta, 2009; Wiesel et al., 2008). Therefore it is expected that value to customer will have positive impact on value to firm. Revenue is about future expenditures of a customer to the company (Verhoef et al., 2010). Retention is one of the most critical variables that affect customers’ lifetime profit, because it has a large impact on customer- and firm value (Wiesel et al., 2008). Retention can be defined as ‘the probability that a customer is ‘alive’, or repeats buying from a firm’ (Gupta & Zeithaml, 2006). The net promoter score is about the likelihood that a customer recommends the company to friends and family (Reichheld, 2003).

The next sections address the problem statement and research questions, the research method, the scientific- and managerial relevance of this study, and finally the structure of the paper. 1.2. Problem statement & research questions

Based in the discussion of the previous section, the following problem statement is defined:

Sub questions which will be answered in this research are:

• How effective is a personalized direct mail compared to a general direct mail? • To what degree does a personalized direct mail contribute to long-term firm

performance? 1.3. Research method

This research is conducted in the x industry, for Company X. At this moment Company X makes use of a general direct mail tool. The mailings are the same for each customer and coupled to customers who own a loyalty card. With more personalized direct mailings adjusted to specific segments, Company X can create new opportunities to effectively

(14)

communicate with individual customers, and to take a new, decentralized, and customer-centric strategic view on marketing.

Together with the consultancy company XX and five entrepreneurs, a pilot about ‘local’ direct mail is conducted. Goal of the pilot was to test the effectiveness of a personalized- versus a general direct mail. The pilot is done by means of a field experiment in which three variants of direct mailings were sent to different customers. In total 9,500 mailings were sent. All customers who received a direct mailing plus a control group received a survey after sending the mailing. The total response on the survey was 17% (n=515), divided in four groups: control group (n=105), general (n=194), families (n=106), and empty nesters (n=110). 1.4. Scientific and managerial relevance

1.4.1. Scientific relevance

This research adds to the theory of Customer Relationship Management (CRM), since a personalized mail is a tool to use customer data in a more extensive way, and to increase the quality of the relationship with a company’s customers. Besides, personalized direct mail can be seen as a tool that underlines the overall importance of customers and customer value to a company.

The direct effect of a personalized direct mail on ‘value to customer’ and ‘value to firm’ is not studied in detail yet. Bolton et al (2004) gave propositions on the effect of direct mail on CLV, but did not empirically test this. Other research is about the effect of customer equity on firm value (Gupta et al., 2004; Gupta & Zeithaml 2006; Gupta, 2009, Wiesel et al., 2008), and the drivers of customer equity (Lemon et al., 2001). Besides, Leone et al. (2006) studied the relationship of customer equity with brand equity.

De Wulf et al. (2001) tested the effect of different relationship marketing tactics (including direct mail), and found a positive effect on relationship quality and behavioral loyalty. Lewis (2004) measured the effect of loyalty programs and other marketing instruments (e-mail coupons, fulfillment rates and shipping fees) on customer retention.

(15)

focuses on the effect of consumer preference insight on personalized product recommendations. Khan et al. (2009) tested the effectiveness of several promotional instruments, like discount coupons, free shipping offers and loyalty programs. Besides, most research focuses on direct mail as a price promotion to generate short-term sales (Bolton et al., 2004; Zhang & Wedel, 2009; Barone & Roy, 2010), while this study is about personalized direct mail that is aimed to give advice and customize promotions to a specific segment. Research specifically about direct mail is quite old (Vriens et al., 1998; van der Scheer et al., 1996). Nowadays, e-mail marketing is more popular, which is studied by Zhang & Wedel (2009) and Ansari & Mela (2003).

Thus, this research will contribute to the existing literature by testing the direct long-term effect of one marketing tactic, personalized direct mail, on value to customer and value to firm.

1.4.2. Managerial relevance

(16)

1.5. Structure of the paper

(17)

2. CONCEPTUAL MODEL & HYPOTHESES

This study is about the effect of a personalized direct mail on the constructs ‘value to customer’ (V2C) and ‘value to firm’ (V2F). Figure 2.1 presents these relationships in a conceptual model. The model assumes that preference fit has a positive relationship with V2C, while in turn value to customer is positively related to V2F. Attitude toward personalization moderates the relationship between preference fit and V2C. Finally, three control variables are included: satisfaction, relationship quality and involvement.

This chapter will discuss the different constructs as presented in the conceptual model, and gives hypotheses for this study.

Figure 2.1.: Conceptual model

Preference fit

Value to customer Value to firm

(18)

2.1. Preference fit

In recent years, customers’ demand for customized products has increased (Gilmore & Pine, 1997). Customers’ preferences have become more heterogeneous and customer diversity is growing (Khan et al., 2009; Reinartz & Kumar, 2006). We have entered the age in which new technologies, increased competition, and more assertive customers are leading firms from standardization towards customization of their products and services (Lampel & Mintzberg, 1996). By offering consumers exactly what they want, companies will be able to charge a premium for quality and, at the same time, will be rewarded with greater customer loyalty (Pine et al., 1995). In this section, first different forms of customization are described. Next, the role of preference fit and customized promotions are discussed in more detail.

2.1.1. Customization

Customization is a general definition, since it has many forms. For instance, there is a difference between pure customization and mass customization. ‘Customization means manufacturing a product or delivering a service in response to particular customer’s needs, and mass customization means doing it in a cost-effective way’ (Pine et al., 1995). A more detailed distinction between different forms of customization is shown in Table 2.2.

Table 2.2.: Forms of customization

Standardized without customer involvement

Standardized with customer involvement

Customer completely involved

• Pure standardization • Segmented standardization (Lampel & Mintzberg, 1996) • Cosmetic customization • Transparent customization (Gilmore & Pine, 1997)

• Customized standardization • Tailored customization (Lampel & Mintzberg, 1996) • Adaptive customization (Gilmore & Pine, 1997)

• Pure customization (Lampel & Mintzberg, 1996) • Customerization

(Wind & Rangaswamy, 2001)

Based on: Hoekstra (2001)

(19)

Lampel & Mintzberg defined five types of customization: (1) pure standardization, (2) segmented standardization, (3) customized standardization, (4) tailored customization, and (5) pure customization. With pure standardization only one form of a product or service is the same for each customer. No distinctions between different customers exist. With segmented

standardization, products and services are tailored to different customer groups. Customers do

not have direct influence, but more choices are available to match the demand of different segments. With customized standardizations, customers can choose between a variety of standardized components. An example is Dell Computers: customers can compose their own computer, by choosing color, model, processor and so on. With tailored customization, a prototype product is presented to the customer and afterwards tailored to the customer’s specific wishes or needs. As already mentioned, with pure customization the customer is involved in the entire process (Lampel & Mintzberg, 1996).

Gilmore & Pine (1997) doubt about the low cost advantage of mass marketing, and concluded that mass customization can produce unnecessary cost and complexity. To overcome this, they formulated a finer definition of customization, by identifying four approaches to customization: (1) collaborative, (2) adaptive, (3) cosmetic, and (4) transparent. Collaborative customization is an extensive way of customization. Gilmore & Pine (1997) associate

collaborative customization with mass customization, however this type is far from

standardized, since the product is entirely adjusted to the customer’s wishes. A better comparison of collaborative customization therefore is with tailored customization (Lampel & Mintzberg, 1996), since the customer is involved in the process. Adaptive customization is more standardized. The product itself is fixed, but can be adapted to the customers’ wishes. With cosmetic customization, the product itself also does not change. However, the standard offering is packaged specially for each customer and can be personalized. Transparent

customization is a way of customizing products without direct involvement of the customer.

The customer is observed and products are tailored to specific needs, but the customer does not know the product or service is customized.

(20)

Wind & Rangaswamy (2001) came up with a new concept, representing the next generation of customization, called customerization. Customerization can be defined as ‘a buyer-centric company strategy that combines mass customization with customized marketing’. Marketing is redesigned based on customization approaches. Dell computers, for instance, developed specific websites, especially for employees of their business customers, to order computers which are already customized and approved for the company.

Wind & Rangaswamy (2001) also explain that customerization is an enhancement to one-to-one marketing and personalization. However, the most important difference between customization and customerization is that the customer has total control. The customer is co-producer in the process. One-to-one marketing and personalization are not by definition initiated by the customer. Customization is not a replacement of one-to-one marketing, personalization or mass customization, but goes further by relating these concepts to marketing strategy. Thus, customerization is an integrated approach of customization and marketing. When companies implement customerization, they definitely control customization, and are able to integrate it with marketing, operations, R&D, finance, and information. Customerization can also be interpreted as an extension of CRM. In both concepts the customer is starting point and the entire company strategy is built around the customer. Wind & Rangaswamy show how to integrate all the customization theories into a broader perspective, which can be seen as the next step for the future.

In this study, the personalized direct mail is partly customized, without direct customer involvement. The content of the mailing (e.g. promotions and text) is customized to specific customer segments, and to the wishes of the entrepreneur. The lay-out of the mailings for the different segments is the same, except from the pictures, because these are also customized to the segment and the entrepreneur. This way of customization can be related to segmented standardization of Lampel & Mintzberg (1996) or transparent customization of Gilmore & Pine (1997). When completely integrated with the strategy of the company, customerization could be the next step when developing a personalized direct mail procedure.

2.1.2. Customer preference fit

(21)

customer preference fit (Simonson, 2005; Kramer, 2007; Franke et al., 2009). Offers that are customized to individual customers’ preferences may provide superior value if marketers can uncover those preferences and if customers can recognize offers that indeed provide a fit to their preferences (Simonson, 2005). However, customer preferences are often ill defined. In many cases consumers do not have stable preferences or a good insight into their own preferences (Bettman et al., 1998; Simonson, 2005; Kramer, 2007; Franke et al., 2009).

Simonson (2005) suggests four basic steps of consumer response to a customized offer: (1) preference development, (2) evaluation of the offer, (3) acceptance or rejection of the offer, and (4) maintaining customized relationships. All steps are important to anticipate on when creating a customized offer. Overall, Simonson (2005) concludes that it is far from certain that offers that are customized to the preferences of consumers are always successful. The process is much more complex than it looks. The response to customized offers depends on cues regarding the offer’s fit and value, customers’ receptivity to certain type of offers, and on customers’ perception of one-to-one relationships.

While Simonson (2005) only gave propositions, Franke et al. (2009) empirically tested consumer preference fit of customized product offerings. They concluded that overall the benefit gain of a customized offer is high in terms of willingness to pay, purchase intention, and attitude toward the product. However, the benefit gain is higher when customer’s ability to express preferences and product involvement is high. Customers’ insight into their preferences is a prerequisite for a powerful customization strategy.

Both Simonson (2005) and Franke et al. (2009) mention the opposite side of customized offerings. A more standardized offer, which can also be provided by competitors, will be more successful when customers cannot define their preferences that well. When insight is higher, the more demanding and critical the customer will be, and the easier it is to select the best offer.

From these studies is can be concluded that it depends on the company, the offer, the customer, and the competition what customization strategy fits best.

2.1.3. Customized promotions

(22)

Zhang & Wedel (2009) made a distinction between three levels of granularity: mass market, segment specific and individual specific, and two types of promotions: loyalty and competitive. An important finding of this study is that customized promotions are not by definition more effective. Online customized promotions do have more profit potential, but offline the differences between the levels of granularity are rather small.

Different from this study with Zhang & Wedel (2009) is the study of Khan et al. (2009), that focuses on multiple promotions like discount coupons, free shipping offers and reward programs. They state that online stores have an advantage over offline stores concerning customization, which is consistent with the study of Zhang & Wedel (2009). They concluded that overall, customized promotions lead to a significant increase in profits relative to the firm’s current practice of uniform promotions. Besides, they found heterogeneity in customer response, so firms can benefit from customizing promotions on individual level. These findings are consistent with Rossi et al., (1996), who conclude that customized coupons do have more profit potential than blanked coupons. They also conclude that direct marketing efforts become more profitable when firms make active use of purchase history and demographic customer data.

Barone & Roy (2010) examined whether consumers’ responsiveness to a targeted discount is influenced by their perceptions of deal exclusivity. They found that only males and customers with an independent self view favored exclusive deals, while females and those with interdependent construals react negatively to a targeted offer that is exclusive.

2.2. Value to customer

Lemon et al. (2001) determined three drivers of customer equity: value equity, brand equity and relationship equity. Verhoef et al. (2010) integrated this in one construct: Value to customer (V2C). The next paragraphs discuss the three drivers of customer equity in more detail.

2.2.1. Value equity

(23)

perceived price can be defined by consumers as ‘cheap or expensive’. Quality can also be interpreted objective or perceived. Objective quality is about the technical superiority of a product, while perceived quality can be defined as ‘the consumer’s judgment about a product’s overall excellence or superiority’ (Zeithaml, 1988). Convenience relates to actions that balance the customer’s time costs, search costs, and efforts to do business with the firm (Lemon et al., 2001). Convenience is an integral part to marketing of both goods and services. Nonmonetary costs are central to the convenience concept (Berry et al., 2002). Thus, value equity includes both monetary (price) and nonmonetary costs (convenience). Value equity is always important, but mostly in business-to-business and in highly competitive environments (Lemon et al., 2001).

2.2.2. Brand equity

According to Lemon et al. (2001) brand equity can be defined as ‘the customer’s subjective and intangible assessment of the brand, above and beyond its objectively perceived value’. They state that the key drivers of brand equity are brand awareness, customer brand attitudes and customer perception of brand ethics. Marketing communications can enhance brand awareness. Brand attitudes are more emotional and are influenced by direct marketing. Corporate ethics are about customers’ perception of the organization and their contribution to society.

The definition of brand equity by Lemon et al. (2001) is relatively narrow. Keller (2008) uses a broader definition of brand equity, that is conceptualized as ‘customer-based brand equity’

(CBBE). CBBE can be defined as ‘the differential effect that brand knowledge has on

consumer response to the marketing of that brand’. Sources of CBBE are brand image and brand awareness. Brand awareness is about brand recognition and recall, while brand image has to do with the strength, uniqueness and favorability of brand associations.

Keller (2008) differentiates between several building blocks, which are included in the CBBE-pyramid. The pyramid is unique because it includes both an emotional and rational route. Lemon et al. (2001) only includes the emotional components in their definition of brand equity, which, in Keller's (2008) terms, is about imagery and feelings of the brand. The rational route of the pyramid of Keller includes performance and judgments of the brand. Performance is about the satisfaction of functional needs. Judgments are about quality, reliability, consideration and superiority. Both performance and judgments can be related to

(24)

attachment, community and engagement. Brand resonance can be related to relationship

equity, which is explained in more detail below. Yoo & Donthu (2001) developed a brand

equity scale that also includes perceived quality and brand loyalty, and brand awareness. Leone et al. (2006) conclude that overall, the basic premise of brand equity is that the power of the brand lies in the minds of consumers and in their experiences. It is about creating added value, which is possible in many different ways.

Brand equity is considered to be more important in low-involvement consumer purchase decision processes when emotional and experiential aspects are important, consumption of the product is highly visible to others, it is difficult to assess the quality of a product before consumption (with credence goods), and when advertising is the primary form of communication (Lemon et al., 2001).

2.2.3. Relationship equity

Relationship equity (or retention equity) is defined as ‘the tendency of the customer to stick with the brand, above and beyond the customer’s objective and subjective assessments of the brand’. Key drivers of relationship equity are loyalty programs, special treatment and recognition, affinity programs, community-building programs, and knowledge-building programs. As mentioned, relationship equity can be related to Keller's (2008) brand resonance, which is positioned at the top of the CBBE-pyramid. Relationship equity includes loyalty and retention. This paragraph discusses the loyalty part, while retention is discussed in more detail in the next section.

Reichheld & Sasser (1990) state that loyal customers are more profitable, because loyal customers want to pay a price premium, generate word of mouth, are responsible for repeated purchases, and operating costs can be reduced. However, Reinartz & Kumar (2002) react on this study by stating that the relationship between loyalty and profitability is much weaker. They define four groups of customers based on short- and long-term customers and profitability, and conclude that short-term customers can also generate high profitability. Thus, marketers cannot take a positive relationship between loyalty and profitability for granted.

(25)

Hypothesis

Based on the previous discussion on preference fit and V2C, it is expected that a personalized direct mail that is customized and thus fits customer preferences, has positive impact on V2C. Overall, the effect of customization is positive (Franke et al., 2009), especially when consumer preferences are clearly defined (Simonson, 2005; Kramer, 2007, Franke et al., 2009). Besides, customized promotions do have more profit potential (Rossi et al., 1996; Khan et al., 2009).

When customers receive offers that fit their preferences, they are expected to respond more favorably to it. The offer is more relevant, which can result in a better price-quality perception (value equity). Besides, a personalized direct mail that fits customer preferences is expected to have a positive effect on brand awareness, brand image and attitudes, since the brand offers what the customer wants (brand equity). Finally, the customer may become more loyal, because when offers are customized, customers may be more satisfied and tend to stick with the brand earlier. Thus:

H1: A personalized direct mail has positive impact on 1a: value equity; 1b: brand equity;

1c: relationship equity.

2.3. Value to firm

Customer equity is an important determinant for the long-term financial value of firms (Lemon et al., 2001; Gupta et al., 2004, Gupta, 2009; Wiesel et al., 2008). Value to firm (V2F) can be measured in different ways. For this study the definition of Verhoef et al. (2010) is used. Three components are included: revenue, net promoter score, and retention. They based this on the studies of Gupta & Zeithaml (2008), Reichheld (2003) and Gupta et al. (2004). The components are explained in more detail below and at the end of this section the hypothesis is given.

2.3.1. Revenue

(26)

2.3.2. Net Promoter Score

The Net Promoter Score (NPS) can be defined as ‘the likelihood that a customer will recommend a company to a friend or colleague’. A strong correlation is found between a company’s growth and the number of promoters (Reichheld, 2003). Luo & Homburg (2008) state that customer complaints have a larger negative impact on the stock value gap (the shortfall of a firm’s actual market value from its optimal market value) than customer satisfaction. Besides, the long-term financial harm of negative word-of-mouth becomes more destructive in magnitude, kicks in more quickly, and haunts investors longer (Luo, 2009). It is therefore important to stimulate word-of-mouth, and the NPS, to prevent financial harm and to increase V2F.

2.3.3. Retention

Retention can be defined as ‘the probability of a customer being ‘alive’, or repeats buying from a firm (Gupta & Zeithaml, 2006). Customer retention is one of the key drivers of CLV and firm profitability (Gupta & Zeithaml, 2008). Reichheld & Sasser (1990) show that 5% improvement in retention could improve the overall profitability of service companies from 25% to 85%. Gupta et al. (2004) also found that retention has a significant impact on customer value: 1% improvement in retention increases customer value by 3% - 7%. This is much higher than when investing in acquisition activities. Bolton et al. (2004) propose that the length of the customer relationship positively affects financial outcomes of the firm (revenues and CLV). Besides, they propose that the length of the relationship is an effect of marketing actions and relationship perceptions (like price perceptions, satisfaction and commitment).

Hypothesis

The three components of V2C (value equity, brand equity and relationship equity) are stated to be the drivers of customer equity (Lemon et al., 2001). In turn, customer equity is an important determinant for long-term firm value (Lemon et al., 2001; Gupta, 2004; Gupta & Zeithaml, 2006; Gupta, 2009; Wiesel et al., 2008). It is therefore expected that V2C is positively related to V2F.:

H2: Value to customer positively influences value to firm in terms of 2a: revenue, 2b: the

(27)

2.4. Attitude toward personalization

We expect that customer attitude toward personalization will moderate the relationship between preference fit and V2C. This section first discusses the concept and ends with the accompanying hypothesis

To send personalized direct mail, customer information is needed. The quality and the amount of customer data determines whether customization strategies will work (Khan et al., 2009). However, to let personalization work, customers have to accept that a company uses their personal information for commercial purposes. Privacy-concerns become more and more an issue, especially because of the increasing use of the internet (Chellappa & Shivendu, 2008, Rapp et al., 2009). Okazaki et al. (2009) studied consumer privacy concerns in mobile advertising. They found that when customers do have the feeling a company violates the rights they have over the customer’s information, they are ‘reluctant to disclose personal information, will not respond to advertising offers, and may seek even stricter regulatory control over mobile advertising’. It is therefore important to be careful with personal customer information, so customers will accept the use of it for advertising purposes. Since personalized direct mail is concerned with consumer privacy issues, it is important to take this into account (Rapp et al., 2009).

Hypothesis

When customers trust the company and accept their information to be used, they will respond more positive towards advertisements (Okazaki et al., 2009). This will in turn positively influence the effect of a personalized direct mail on V2C, since consumers process the advertisements in a better way and will have a more positive attitude toward the personalized direct mail:

H3: The more positive the attitude of customers toward personalization, the stronger the

(28)

Furthermore, we control for satisfaction, relationship quality, and involvement.

Satisfaction can be defined as ‘the emotional state that occurs as a result of the customer’s interactions with the firm over time’ (Verhoef, 2003). Relationship quality is about the overall assessment of the strength of the relationship. Involvement considers the importance consumers attach to a product category, based on the consumer’s inherent needs, values and interests (de Wulf et al., 2001).

When consumer satisfaction, relationship quality and involvement become higher, this can have a positive impact on both V2C and V2F. When consumers are satisfied, have a good relationship with the company, and/or are involved with the product category, it is possible that they have more positive attitudes toward the company, are more loyal, and want to spend more. Therefore this study will control for these variables.

(29)

3. RESEARCH DESIGN

This chapter describes the research design, which ‘specifies the details of the procedures necessary for obtaining the information needed to structure and/or solve marketing research problems’ (Malhotra, 2007). This study can be defined as a combination of exploratory and conclusive research (Malhotra, 2007; Blumberg et al., 2008). Exploratory research is done by understanding the problem behind the research, from a theoretical as well as a practical point of view, as described in chapter 1 and 2. The main objective of conclusive research is to test hypotheses and to examine specific relationships. Conclusive research can be divided in descriptive and causal research (Malhotra, 2007). This study includes mainly causal research, since we want to understand the effect of a personalized direct mail on V2C and V2F by an experiment.

This chapter will further describe the research design for the conclusive research. The research is done for Company X. The next sections will give an introduction to Company X, and discuss the sample, procedure, measures, and finally the plan of analysis.

3.1. Introduction to Company X

Since 2009 Company X applies a general direct mail tool, addressed to customers who own a customer card. However, company X is only in the beginning phase of using direct mail as a marketing tool. Not every entrepreneur has developed a large customer database and understands why customer data is important for marketing purposes. Besides, the current direct mails are not targeted to specific segments and therefore not relevant to every customer. Company X understands the importance of using customer data in a more extensive way, and therefore wants to experiment with more segment specific, local marketing. They started a pilot to test the possibilities of local marketing. This pilot is done in cooperation with XX, and five entrepreneurs.

This study is conducted within this pilot and is aimed to test the effects of a personalized

direct mail, targeted to specific segments and adjusted to the wishes of the independent

entrepreneurs, compared to a general direct mail. 3.2. Procedure

(30)

direct mail, without knowing they were part of an experiment. Besides, the promotions included in the direct mailings were real and could be handed in at a participating store.

Variants

Three different versions of direct mailings were developed. In addition to the mailing, each customer received a questionnaire. Surveys are the most flexible way of obtaining data, and can reach a large number of respondents. Besides surveys can be best used, among others, to test advertising effectiveness (Malhotra, 2007). For this study we want to test the effectiveness of a direct mail, which can also be seen as a way of advertising. The measures used in the survey are described in section 3.4.

In order to test the effectiveness of a general versus a personalized direct mail, three variants of direct mailings were developed, directed to two target groups: families and empty nesters. These target groups were chosen, since they are the main target groups of company X. In total 9500 mailings were sent. The three variants are:

• General direct mail, directed to both families and empty nesters (4750 mailings) • Personal direct mail directed to families (2375 mailings)

• Personal direct mail directed to empty nesters (2375 mailings) The three variants differed in pictures, texts and promotions.

The different variants were similar to each other in a way that the graphical design (title, colors, size, order of content) of every mailing was the same. Moreover, the text used in the mailings was about the same subject, but some adjustments were made in the text to make it more appealing to the segment the mailing was aimed for.

The general direct mail was sent in the name of Company X, and not directed to a specific segment. Both families and empty nesters were targeted for this direct mailing. On the other hand, the personalized direct mails are directed to a specific segment (families or empty nesters), and sent in the name of the independent entrepreneur who is closest to the address of the consumer.

(31)

pictures were chosen. For families a picture of a happy family was portrayed, while for the empty nesters an older, but active couple was chosen to represent them.

The texts were written by a professional text office which is a fixed partner of Company X. The mailing started with an introduction letter. For the personalized direct mail, the customer’s name was included at the beginning of the letter. For the general direct mail only the title was included, without personalizing the letter by name. The content of the letters was almost the same, however some adjustments were made in the personalized direct mails to make the introduction more interesting to the segment (families or empty nesters).

Concerning the promotions used in the mailings, the general direct mail included the same promotions for every entrepreneur and every customer. For the personalized direct mail another approach was used. Company X wanted the personalized direct mails to be developed on a local level, to adjust the mailing even more to their local target group. To keep it controllable, a list of pre-selected options for both empty nesters and families was developed. Entrepreneurs could choose a maximum of four options for every personalized mailing. Furthermore, the pictures and promotion texts of the products were also adjusted to the segment.

Finally, the direct mailings included seven tips. The tips for families and empty nesters were adjusted to the target group. The general direct mail included tips that may be interesting to everyone.

The mailings were developed in cooperation with the Marketing department of Company X, XX and five participating entrepreneurs. These parties assessed the mailings on attractiveness, from a graphical as well as from a target group perspective. Based on their comments the final mailings were created.

Manipulation check

A manipulation check is done to test if the personalized direct mail is indeed more attractive and relevant to the customer compared to the general direct mail. The variables promotion

attractiveness and content relevance were analyzed. The items that were used for measuring

(32)

Table 3.1. Manipulation check t df Sig. (1-tailed) Mean difference St. Error Difference 95% Confidence Interval of the Difference Lower Upper Promotion attractiveness (equal variances: F=.110, p >.05) -3,130 196 ,001 -,49440 ,15794 -,80589 -,18291 Content relevance (equal variances: F= 3,834, p >.05) -2,156 196 ,016 -,34680 ,16087 -,66405 -,02955

From the results we can infer that the personalized direct mails are indeed more attractive and relevant to the segment the mailing was aimed for, compared to the general direct mail. (p<.05).

Survey

A week after sending the direct mailings, respondents received the survey and were asked to participate in this research. Receivers had two weeks to respond. To increase response, the survey was both offered offline and online, which was explained in an accompanying (personalized) letter. Offline surveys could be sent back to Company X for free. To stimulate response even more, all respondents received a little present if they responded and filled in their personal details. Besides, 10 giftcards of € 25 were randomly given away.

3.3. Sample & data collection

For this research four groups of respondents were asked to fill in the questionnaire. In total 3,000 surveys were sent and 17% (n=515) responded. Table 3.2. shows the groups used in this study, including the number of surveys sent and the response.

Table 3.2: Response per group

Group Nr of surveys % Expected N Observed N %

Receivers of a general direct mail (families and empty nesters) 1250 41,7 214,8 194 37,7 Receivers of a personal direct mail directed to families 625 20,8 107,1 106 20,6 Receivers of a personal direct mail directed to empty nesters 625 20,8 107,1 110 21,3

Control group (families and empty nesters) 500 16,7 86,0 105 20,4

Total 3000 100 515 515 100

Chi-Square 6,290a

df 3

(33)

To check if the sample represents the sample frame, a Chi-Square test is conducted. From this test we can infer that the observed- and expected N do not significantly differ from each other (p>.05). It can therefore be concluded that the sample represents the sample frame.

For the data collection the customer databases of the five participating entrepreneurs was used. XX received all data and sent it to Experian, who enriched the data with target group information by means of ‘Mosaic’ profiles. Experian is a company specialized in consumer segmentation, and is able to optimize customer data. With Mosaic all households in the Netherlands are classified and divided into 44 groups, combined in 10 segments, based on demographics, psychographics, geographics and lifestyle. The customer data of the five entrepreneurs is merged with Mosaic. In this way it could be determined to which segment (families or empty nesters) the data belongs. For the general direct mail, both families and empty nesters were randomly selected. For the family segment, customers with children between 0 and 19 years old who are living with their parents at home were selected. For the empty nester segment, customers of 55 years and older who are living alone or with a partner and without children were selected.

The control group did not receive a direct mail, but did receive a questionnaire. This group was used to conduct a zero measurement for this study. The letter and the survey included some small changes, since the questions about promotion attractiveness and content relevance of the direct mail were deleted (see appendix C). Data was collected from the customer databases of the participating entrepreneurs. To keep the groups comparable to each other, receivers were also families and empty nesters and were randomly selected to make the groups comparable to each other.

More detailed sample characteristics can be found in chapter 4. 3.4. Measures

The survey tests the constructs as described in the conceptual model of figure 2.1. All measures are based on extant literature in marketing. Table 3.2. gives overview of the measures used in this study. The scales used in this research consist of multiple items. According to Malhotra (2007), a multi-item scale should be evaluated for accuracy and applicability.

(34)
(35)

Table 3.3: Measures

Construct Alpha Items Source

.918 Value to Customer*

Value equity .938 • X is reasonably priced

• The quality/brand ratio at X is good • X gives me my money’s worth • It is easy for me to go to X (deleted)

Verhoef, Langerak, & Donkers (2007)

Rust, Lemon, & Zeithaml (2004) Brand equity .880 • X is a strong brand

• X is a well-known brand • X is a unique brand • X is renewing Verhoef, Langerak, & Donkers (2007) Relationship equity

.853 • I communicate back and forth with X • I am often in dialogue with X

• I have the feeling X knows exactly what I want • I feel at home at X

• I feel connected with X

Verhoef & Lemon (2009) Bügel, Verhoef, Buunk (2009) .643 Value to Firm Revenue (deleted)

- • How often, on average, do you visit X a year?

• How much €, on average, do you spend per visit at X? Rust, Lemon & Zeithaml, 2004 NPS* - • I would like to recommend X to my friends and family

• I am likely to make negative comments about X to my friends and family (reversed and deleted)

Nguyen & Leblanc (2001)

Retention* - • In the near future, I intend to visit X more often. • If competitors have better discounts than X, I will go to

the competitor (reversed and deleted)

• As long as I live in this neighborhood, I do not foresee myself switching to another store than X (deleted) • I will stay a customer at X in the future (deleted)

Nguyen & Leblanc (2001)

Moderator**

Attitude toward personalization

.706 • Can you indicate how much you would appreciate the offers below?

o Offers that are adjusted to my personal circumstances.

o Offers based on information X has gained automatically, with which I as an individual am not traceable.

o Offers based on information X has gained automatically, with which I as an individual am traceable.

o Offers based on information that has been provided by myself, with which I as an individual am not traceable.

o Offers based on information that has been provided by myself, with which I as an individual am traceable.

Chellappa et al. (2005)

(36)

Satisfaction .916 • Overall, how satisfied are you with your X store? • How satisfied are you about:

o The personal attention of X

o The willingness of giving advice at X o The service quality of X

o How X deals with complaints

o The knowledge and expertise of the personnel of X

Ganesh, Arnold, & Reynolds (2000)

Verhoef (2003)

Relationship quality

.900 • I have a high quality relationship with X (deleted) • I am happy with the efforts X makes towards me as a

customer

• I am satisfied about the relationship I have with my X store

De Wulf,

Odekerken-Schröder & Iabucci (2001)

Involvement .929 • In general I have a strong interest in drugstore products • x products are important to me

• x products matter a lot to me

• I get bored when other people talk about x products(reversed and deleted)

Beatty & Talpade (1994)

Manipulation check*

Promotion attractiveness

.926 • This is a good newsletter • I like the newsletter

• I feel positive toward this newsletter

• This newsletter is awful (reversed and deleted) • This newsletter is pleasant

• This newsletter is attractive • I approve this newsletter

Shamdasani, Stanaland & J.Tan (2001)

Leclerc, Smitt & Dubé (1994)

Content relevance

.728 • This newsletter is useful for me

• This newsletter specifically relates to my needs • This newsletter was meaningful to me

• This newsletter did not have anything to do with me or my needs

• This newsletter gave me a good idea

• When reading this newsletter I thought of reasons why or why not to take action on it

Lastovicka (1983)

* 7-point Likert scale (1= totally disagree, 7= totally agree) ** 7-point Likert scale (1= totally not, 7 = totally

*** 7-point Likert scale (1= unsatisfied, 7= satisfied)

(37)

Preference fit is manipulated by the different versions of the direct mailings: general versus

personalized, as described in section 3.2. The personalized direct mails are used for the preference fit condition, while the general direct mail is used for the no-preference fit condition (0/1 variable).

Most measures included a 7-point Likert scale, ranging from (1) strongly disagree, to (7) strongly agree, or for satisfaction ranging from (1) totally dissatisfied to (7) totally satisfied. For attitude toward personalization a scale ranging from (1) totally not, to (7) totally is used. We used Likert scales, since analyses can easily be made and the scale may be interpreted as interval. Moreover, for respondents the scales are easy to understand (Malhotra, 2007). An odd scale was used, since respondents are allowed to be neutral (Malhotra, 2007).

Revenue is measured by multiplying the number of visits per year and the average spending of the customer per visit. However, revenue was excluded from further analysis, because when including revenue in V2F internal reliability was not sufficient.

3.5. Plan of analysis

Pre-test

To check the soundness of the survey, a pre-test was conducted. For this test a convenience sample of six persons was asked to look and fill in the survey, and to express unclarities. Based on their comments some adjustments were done.

Control group

(38)

Table 3.4. Check Control Group Variable t df Sig. (2-tailed) Mean difference St. Error Difference 95% Confidence Interval of the Difference Lower Upper V2C (equal variances: F= ,124, p >.05) ,858 462 ,392 ,10306 ,12018 -,13310 ,33923 V2F (equal variances: F= ,473, p >.05) ,268 483 ,789 ,03522 ,13128 -,22274 ,29318 Satisfaction (equal variances: F= ,194, p >.05) ,112 450 ,911 ,01248 ,11124 -,20613 ,23109 Relationship quality (equal variances: F=1,007, p >.05) ,493 488 ,622 ,07492 ,15198 -,22369 ,37353 Involvement (equal variances: F= ,070, p >.05) -,141 479 ,888 -,02275 ,16104 -,33919 ,29368 Attitude toward personalization (equal variances: F= ,642, p >.05) -1,199 431 ,231 -,16881 ,14084 -,44563 ,10800

We did not find significant differences between the groups (p>.05). The fact that there are no significant differences between the control group and the basic group confirms that we study a homogeneous group. In spite of this, within the experimental group we can study mutual differences.

Groups

(39)

Table 3.5. Group check Variable t df Sig. (2-tailed) Mean difference St. Error Difference 95% Confidence Interval of the Difference Lower Upper V2C (unequal variances: F= 11,755, p <.05) -1,439 163,043 ,152 -,23603 ,16401 -,55988 ,08782 V2F (equal variances: F= 3,292, p >.05) -,247 199 ,805 -,04342 ,17556 -,38962 ,30278 Satisfaction (equal variances: F= 7,324, p >.05) -1,529 163,156 ,128 -,20549 ,13438 -,47085 ,05986 Relationship quality (unequal variances: F= 5,432, p <.05) -,762 184,136 ,447 -,14932 ,19586 -,53573 ,23709 Involvement (unequal variances: F= 5,016, p <.05) ,308 183,605 ,758 ,06279 ,20385 -,33939 ,46497 Attitude toward personalization (unequal variances: F= 9,391, p <.05) -,922 156,166 ,358 -,16786 ,18198 -,52732 ,19161

From these tests it can be concluded that for all constructs no significant differences exist between the groups (p>.05). This implies that families and empty nesters do not significantly differ from each other

To be sure the combined group did not have a different effect on the hypotheses compared to the separate groups, we also tested the effects on the hypotheses for these different groups. Since most of the effects did not differ between the groups compared to the combined group, we decided to keep one group (the preference fit condition). Only some small results differ, which are discussed in the results chapter as well.

The results we found for the separate groups can be found in appendix E.

Hypotheses & control variables

To test the hypotheses and control variables, regression analysis is applied.

For each hypothesis a separate regression model is used, to prevent multicollinearity between the variables. This is a very high state of intercorrelations among independent variables (Malhotra, 2007). When using separate models, the VIF scores are between 1 and 2. This implies that there is no need to take multicollinearity into account, since this is the case when the VIF does not exceed a score of 10 (Hill et al., 2001).

(40)

testing H2 the dependent variable is V2F and the independent variable V2C. Finally, for testing H3, the dependent variable also is V2C, while the independent variable is preference fit*attitude toward personalization, since we expect that attitude toward personalization moderates the relationship between preference fit and V2C. Besides, attitude toward personalization is added to the model as an independent variable.

In all regression models the control variables (satisfaction, relationship quality and involvement) are included as independent variables, since we expect a possible effect of those variables on both V2C and V2F.

The following equations will be estimated:

(1) (2)

(3)

V2C = Value to Customer V2F = Value to Firm FIT = Preference fit

ATFIT = Attitude toward Personalization * Preference fit SAT = Satisfaction

(41)

4. RESULTS

This chapter describes the results of this study. The response on the survey was 17% (n=515), as explained in chapter 3 (table 3.1.). The data included some missing values, but we decided to keep all respondents, since we did not want to lose data. As explained in chapter 3 it can be concluded that the data is reliable and valid, and therefore can be used for further analysis. The first section will give some more detailed information about the sample used in this study. The next section will further explain the regression analysis used to test the hypotheses as explained in chapter 2.

4.1. Descriptive statistics

As explained in chapter 3, four groups of respondents can be distinguished. However, families and empty nesters are combined in one group who represent the ‘preference fit condition', since these respondents received the personalized direct mail and do not significantly differ from each other on the constructs of the conceptual model used in this study.

Table 4.1. and figure 4.1. give the main sample characteristics per group. In this figure families and empty nesters are separated, since we wanted to check if they really represent the segment they belong to.

Table 4.1. Sample characteristics

Group % women % men Average age

Control group 96,1% 2,9% 50 General 95,3% 4,7% 48 Families 95,2% 4,8% 39 Empty nesters 95,4% 4,6% 64 Total 95,5% 4,3% 50

(42)

Table 4.1. shows that the majority of the sample are women. From figure 4.1. it can be concluded that both the control group and the general direct mail receivers include a good spread of different living situations. The average age in these groups is about 50 years old, just as for the total sample. Families have a lower average age of 39 years old, and their main living situation is with partner and children (73,8%). Empty nesters have an average age of 64 and most of them are living together with a partner (61,7%). Thus, it can be concluded that the segments represent the right target group, as expected.

Table 4.2. gives the means and standard deviations of the constructs as presented in the conceptual model (figure 2.1.) for the total sample, and for the different groups.

(43)

Table 4.2. Means and standard deviations

Total General Personal Control

group Construct N Mean Std. Deviation N Mean Std. Deviation N Mean Std. Deviation N Mean Std. Deviation V2C 464 4,93 1,06 169 4,83 0,99 195 4,98 1,12 100 5,01 1,07 Value equity 495 5,44 1,22 183 5,35 1,22 208 5,48 1,25 104 5,48 1,16 Brand equity 488 5,31 1,18 183 5,24 1,09 204 5,35 1,28 101 5,39 1,14 Relationship equity 474 4,14 1,35 177 4,01 1,25 197 4,22 1,39 100 4,22 1,43 V2F 485 5,59 1,17 183 5,47 1,10 201 5,69 1,24 101 5,62 1,15 NPS 490 5,25 1,49 185 5,12 1,36 203 5,32 1,61 102 5,33 1,45 Retention 487 4,56 1,53 185 4,56 1,41 201 4.56 1,59 101 4,55 1,65

Attitude toward personalization 433 4,42 1,19 166 4,42 1,14 176 4,49 1,20 91 4,29 1,29

Satisfaction 452 6,00 0,96 176 5,91 0,99 183 6,09 0,91 93 6,01 0,98

Relationship Quality 490 5,56 1,36 186 5,42 1,36 203 5,66 1,38 101 5,62 1,30

Involvement 481 5,10 1,44 180 5,03 1,46 200 5,18 1,43 101 5,08 1,42

Promotion attractiveness 198 4,67 1,14 98 4,42 1,12 100 4,92 1,10 - - -

(44)

4.2. Hypotheses testing

As already explained, regression analysis is used to test H1, 2 and 3. In section 3.5. the equations to be estimated are displayed. The control group was excluded from these analyses, since we wanted to test the effect on receivers of the direct mailings.

Equation 1

The results of equation 1 can be found in table 4.2.

Table 4.2. Results equation 1 (H1)

ANOVA

Model Sum of squares df Mean square F Sig.

Regression 174.787 4 43.697 78.057 .000a

Residual 182.496 326 .560

Total 357.283 330

Variable B St. Error Beta t Sig.

Constant .581 .280 - 2.076 .039 Preference fit -.020 .083 -.009 -.237 .813 Satisfaction .321 .057 .290 5.618 .000 Relationship quality .287 .041 .367 6.924 .000 Involvement .165 .031 .220 5.239 .000 Dependent variable: V2C

The R² of this regression model is .489, and the adjusted R² .483. This implies that the model explains about 49% of the total variance. From the ANOVA it can be concluded that the overall regression model is significant (F=78.057, p<.001).

(45)

Equation 2

The results of equation 2 can be found in table 4.3.

Table 4.3. Results equation 2 (H3)

ANOVA

Model Sum of squares df Mean square F Sig.

Regression 157.957 5 31.591 57.474 .000a

Residual 159.403 290 .550

Total 317.360 295

Variable B St. Error Beta t Sig.

Constant .351 .319 1.100 .272

Attitude*fit -.008 .019 -.019 -.439 .661

Satisfaction .322 .060 .284 5.318 .000

Relationship quality .303 .043 .391 7.038 .000

Involvement .110 .035 .145 3.128 .002

Attitude toward personalization .095 .042 .102 2.245 .026

Dependent variable: V2C

The R² of this regression model is .498, and the adjusted R² .489. This implies that this model explains about 49% of the total variance. From the ANOVA it can be concluded that the overall regression model is significant (F=57.474, p<.001).

The effect of attitude*fit is slightly negative and not significant (p>.05). Thus, H3 is not supported. The control variables satisfaction, relationship quality and involvement again have a significant effect on V2C (p<.01). The effect of the control variables is positive. When looking at the beta's, relationship quality has the highest positive effect (.391). Finally, attitude toward personalization has a positive influence on V2C (.102), and is significant (p<.05).

(46)

Equation 3

The results of equation 3 can be found in table 4.4.

Table 4.4. Results equation 3 (H2)

ANOVA

Model Sum of squares df Mean square F Sig.

Regression 270,006 4 67,501 137,558 ,000a

Residual 158,010 322 ,491

Total 428,015 326

Variable B St. Error Beta t Sig.

Constant ,711 ,264 2,686 ,008 V2C ,428 ,052 ,387 8,223 ,000 Satisfaction ,062 ,056 ,051 1,101 ,272 Relationship quality ,328 ,042 ,382 7,899 ,000 Involvement ,113 ,031 ,137 3,670 ,000 Dependent variable: V2F

The R² of this regression model is .631, and the adjusted R² .626. This implies that this model explains 63% of the total variance. From the ANOVA it can be concluded that the overall regression model is significant (F=137.558, p<.001).

The effect of V2C on V2F is positive and significant, thus H2 is supported.

V2C has the highest positive effect with a beta of .387. Relationship quality again also has a high beta (.382). The control variables all have positive influence on V2F, however only relationship quality and involvement are significant (p<.001).

Referenties

GERELATEERDE DOCUMENTEN

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of

Besides investigating the overall effect of the five different customer experience dimensions (cognitive, emotional, sensorial, social, and behavioural) on customer loyalty, I

Based on this expected effect and the research of Liu and Brock (2007) the assumption is that relationship length positively moderates the expected effect

The effect size statistics derived from the churn driver variables in this analysis are influenced by metadata which can be classified in five categories: driver

Since no individual customer data is available, the eight customer equity strategies are inferred from the aggregated customer metrics data, specifically the

Trying to examine the effect of awareness amongst consumers in online legal music purchasing on their ethical judgement and perceived value could lead to

Comparing the transition matrix for journeys where affiliates were used (Figure 4) to the journeys without any FIC, we notice some positive differences in the probabilities

A more direct way of measuring the effect of recommendations on the degree of loyalty is by using own transaction information, like average spendings, number of different