CUSTOMER RELATIONSHIPS: ANTECEDENTS AND PERFORMANCE IMPLICATIONS By Feng Wang A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Business Administration Marketing Doctor of Philosophy 2018 ABSTRACT CUSTOMER RELATIONSHIPS: ANTECEDENTS AND PERFORMANCE IMPLICATIONS By Feng Wang This dissertation, using a two - essay format, focuses on a new avenue for cultivating customer relationships and extends the strategic implications of customer relationships. The first essay uncovers a novel driver of cultivating customer relationships from a product design perspective. Experiential design, the emerging XD design concept, not only m otivates purchases but also cultivates customer - brand relationships. It has a potential of extracting the emotional meanings of a design as well as transferring the affection from a product design to the focal brand, exclusive from traditional design dimen sions. Importantly, the impact of experiential design is contingent on other design dimensions (e.g., functional and aesthetic), marketing communication ambiguity, and customer subjectivity. The second essay, drawing on customer - company identification theo - relationships as well as the reinforcing value creation mechanism. Specifically, customer - relationships asset is - customer and customer - firm interactions to achieve organic growth in its market performance. Customer - relationships asset is proposed be a competitive advantage, supported by empirical evidence. Overall, this dissertation provides insights for marketing academics and practitioners on how experiential design dimension cultivates customer - brand relationships and how customer - relationships serve - driven customer interactions. Copyright by FENG WANG 2018 iv ACKNOWLEDGEMENTS This journey is exciting and challenging. First and foremost, to my parents: Thank you for unconditional support; Your love makes me fearless and finds the best part of myself. To Mini: I enjoy your everyday companion and I To my committee members, Dr. Calantone: Thank you for your guidance, support, and constant encouragement; Your mentorship helps me bec ome stronger in work and life. To Dr. Voorhees: Thank you for your honest guidance, sincerity, and genuine c aring throughout this journey. To Dr. Hult: Thank you for your support and insights shaping my dissertation. To Dr. Wooldridge: I am deeply gratefu l for your input and always ready - to - help. To Yufei Zhang, Hanieh Sardashti, and Angela Jones: Your friendship and emotional support are the most rewarding gift in this journey and my life. v TABLE OF CONTENTS LIST OF TABLES v i i LIST OF FIG URES ... v i i i INTRODUCTION ESSAY ONE : UNCOVERING THE PROMINENCE OF EXPERIENTIAL DESIGN ON PURCHASE DECISION AND CUSTOMER - BRAND RELATIONSHIP CULTIVATION CONCEPTUAL BACKGROUND Product Design Experiential Design, Imagery Experience, and Intersecting Contingencies 11 HYPOTHESES DEVELOPMENT 13 The Threshold Determined by Other Design Dimensions 14 The Functional Form of Experiential Design 16 The Roles of Marketing Communication Ambiguity 17 The Roles of Customer Imagery Subjectivity 19 Customer - Brand Relationships Cultivation 21 DATA AND MEASURES 22 Data Collection and Description. .. 22 Measures of Key Variables 23 ANALYSES 26 Model Specification 26 Addressing the Endogeneity of Experiential Design 27 Empirical Findings 28 The Effect of Experiential Design on Purchase Con version 28 The Effect of Experiential Design on Relationship Cultivation 32 33 ESSAY TWO: CUSTOMER - RELATIONSHIPS ASSET AS A COMPETITIVE ADVANTAGE AND ITS IMPACT ON ACQUISITION AND RETENTION DECISIONS 38 CONCEPTUAL BACKGROUND AND FRAMEWORK 41 Customer as Assets, Customer Equity, and Customer - Relationships as Assets 41 Customer - vi Customer - Firm Customer - Relationships Asset as a Competitive Advantage STUDY I: ASSESSING CUSTOMER - RELATIONSHIPS ASSET Modeling Approach Data Descrip tion Empirical Findings STUDY II: THE IMPACT ON CUSTOMER ACQUISITION AND RETENTION DECISIONS Hypotheses Development The Impact of Customer - Relationships Asset The Roles of Consideration Set Size The Roles of Customer Information Search : Objective versus Interactive Empirical Examination Data and Measures of Ke Model Specification Addressing the Endogeneity of Consideration Set Size Empirical Findings .64 DISCUSSION Theoretical Implications Customer - Relationships Asset as a Customer - Customer - Relationships Asset as a Competitive Advantage 68 Customer - Relationships Asset and Customer Base Analyses 6 9 Managerial Implications . 75 Appendix B: vii LIST OF TABLES Table 1.1: A Review of Key Dimensions in the Product Design Literature 75 Table 1.2: Customer Demographics 76 Table 1.3: Variables, Operationalization, and Data Sources 77 Table 1.4: Descriptive Statistics and Correlations 78 Table 1.5: The Results for Control Function Approach to Correct Endogeneity o f Experiential Design ... 79 Table 1.6: The Effect of Experiential Design on Purchase Conversion 80 Table 1.7: The Effect of Experiential Design on Customer - Brand Relationship Cultivation . 81 Table 2 . 1: Representative Studies on Customer Interactions and Value Creation 82 Table 2 . 2: Study I. Descriptive Statistics across 33 Brands .. 83 Table 2 . 3: Study I. Empirical Findings for Brands with Sig nificant Customer - Relationships Assets .. 84 Table 2 . 4: Study II. Descriptive Statistics .. 85 Table 2 . 5a: Study II. Model Results for Acquisition and Retention (Copulas Endogeneity Correction) .86 Table 2 . 5b: Study II. Results Validation (Modified Control Function Approach for Endogeneity Correction) .87 viii LIST OF FIGURES Figure 1.1: Conceptual Framework . .88 Figure 1.2: The Illustration of Experiential Design and Its Contingencies . ..89 Figure 1.3a: The Marginal Effects of Experiential Design at Levels of Marketing Communication Ambiguity . Figure 1.3b: The Marginal Effects of Experiential Design at Levels of Customer Imagery Subjectivity . Figure 2 .1 : Conceptual Model of Customer - Relationships Asset . Figure 2 .2 : Modeling Procedure of Study I ...93 Figure 2 .3a : Study I. Cumulative Impulse Response Functions for Brands with Positive Customer - Relationships Assets 94 Figure 2 .3b : Study I. Cumulative Impulse Response Functions for Brands with Negative Customer - Relationships Assets 96 1 INTRODUCTION This recognition starts from Srivastava, Shervani, and Fahey (1998) conceptualization that customer relationships are market - Essay One proposes a novel means to leverage and cultivat e customer relationships from the perspective of produ ct design. Product design bring s to light the emotional meaning products and services have, or could have, for consumers and extracts the high value of such emotional (Lojacono and Zaccai 2004). In the product design literature, most studi es have focused primarily on purchase - related outcomes with a few exceptions that highlight the hedonics from a design. Experiential product design, as the emerging design dimension, is able to embed the emotional connection with customers and evoke ima gery experience that is seductive (Hoch 2012) in the pre - purchase phase of a purchase journey. The emerging experiential design concept creates much more attractiveness to customers, relative to traditional design dimensions such as functional and aestheti c design. McKinsey & Company (2015) advocates that companies should build a design - driven culture, stating, the performance of the product itself. In fact, customer experience is becom ing a key source of Essay One demonstrates that experiential design is able not only to proceed the journey phase between consideration to actual purchase, but also to cultivate a customer - brand relationship. Importantly, uncovering and maximizing the prominent effect of experiential design requires the assistance of other design dimensions, marketing communication, and customer subjectivity. 2 Using unique data from multiple sources (e.g., design & imagery study, customer - level purchase, and objective third - party data), Essay One shows of experiential design, i.e., imager y experience, strongly increase s the probability of purchase conversion. This prominent positive effect emerges only when the threshold determined by the combination of functional and aesthetic design is exceeded. Moreover, the effect of experiential design shows an inverse S - Shaped fun ctional form and reaches the highest at higher levels of marketing communication ambiguity and moderate levels of customer imagery subjectivity. Ultimately, purchasing a product with superior experiential design significantly cultivates a customer - brand re lationship, whereas functional and aesthetic design, although motivate purchases, do not directly link to relationship cultivation. Recognizing experiential design as the novel antecedent of customer relationships, Essay Two focuses on the strategic impli cations of customer - relationships. F irms should strive to economic contributions as well as the contributions through their extra - role engagement. Studies on customer equity and customer lifetime values have extensively investigated the economic contributions of customers, considering their face value through c ustomer tenure, while assuming the independence among customers. Particularly, Essay Two probes the merits of rein forcing value creation through customer interdependencies (e.g., including both customer - customer and customer - firm interactions) and the development of customer - relationships asset as The current state of the art of custom er valuation is customer equity, which accounts literature has documented, the interactions among customers as well as between customers and 3 the focal firm (e.g., ref erral, word of mouth, customer participation, co - developing, etc.) can - typical likely to engage in many of these interactions, generating continuous value over time for organic growth. As a result, a firm can develop a reinforcing mechanism between its current installed customer base and market performance. We define this relationship - driven reinforcing - relationships asset. Formally, custome r - relationships asset - customer and customer - firm interactions to enhance the organic growth of market performance. Study I theorizes customer - relationships asset as a compet itive advantage and quantifies this concept by persistence modeling approach. Using longitudinal data in the U.S. automotive industry, we find that 7 (vs. 6) out of 33 brands possess positive (vs. negative) customer - relationships assets, whereas the rest d o not. To demonstrate that customer - relationships asset is a competitive advantage, Study II examines whether and how brand - level customer - relationships asset influences individual - level customer acquisition and retention decisions. We also find that a cus customer - relationships asset on acquisition and retention. 4 ESSAY ONE UNCOVERING THE PROMINENCE OF EXPERIENTIAL DESIGN ON PURCHASE DECISION AND CUSTOMER - BRAND RELATI ONSHIP CULTIVATION Product design has been one of the most central topics in the marketing literature and a critical source of competitive advantage, as it has the ability to lighten the embedded emotional meaning of a product and stimulate the emotional connections between a design and customers (Lojacono and Zaccai 2004). McKinsey & Company (2015) advocates that companies should build a design - driven culture, stating, product over the performance of the product itse lf. In fact, customer experience is becoming a The emerging concept, experience design (XD) 1 , has highlighted a trend that the key for companies to triumph in their verticals is no longer feature - focused product design, but the design of an art that can build a stunning first impression, provide feasible solutions, and evoke delightful sensory experience. Marketing communication has also underscored the importance of conveying a message that is immersive, humanistic, and experience - focused other than elaborates attribute - commercials have produced a series of multi - sensory and experience - focused themes, e customer attention in recent years. Although previous studies have demonstrated that imagery appeals in advertisement affect consumer choices (e.g., Petrov a and Cialdini 2005), the academic 1 (Wikipedia). 5 research on whether the imagery experience evoked by a product design affects customer purchase decision and how has been limited. Product experience is seductive, because it possesses the nature of engaging, nonpartisan , pseudodiagnostic, and endogenous that intrigues consumers with less guard and adapts product perceptions to their own interests and tastes (Hoch 2002). We maintain that experiential design experience design he reinafter to emphasize its imagery in the pre - purchase stage and distinguish it from actual experience ) is a captivating and of a design as well as the influenc e of marketing communication on design interpretations, the endogeneity of experiential design must be addressed. Since experience is certainly personal, it is experiential design as well as the roles of other design dimensions (e.g., functional and aesthetic), marketing communication, and the self as critical contingencies on purchase decision. As much as it is advocated to embed experience into a product desi gn, little is known whether and how product experiential design affects purchase decision at customer - level. Previous research has shown that a superior product design can delight customers through great experience in the post - consumption stage (e.g. Chitt uri, Raghunathan, and Mahajan 2008). Consumers make decision s g the actual experience with an alternative and assessing the desirability of the alternative according to the affe ctive response to this imagined (Keller and McGill 1994 ). The role of experiential design in the pre - purchase stage is even more valuable and long - lasting because it gives the initial impression that may profoundly influence ustomers are highly likely to transfer their emotional affection of a product design to the focal brand. 6 Hence, in addition to motivating purchase, experiential design also acts as a powerful catalyst to cultivate the relationship between a customer and th e focal brand. imagery experience (e.g., quasi - sensory and quasi - perceptual experience without actually using it) and subsequently examines its effect on purchase convers ion, from purchase consideration to . The reason for focusing on purchase conversion as our primary outcome is that theoretically, the gap between consideration and actual purchase is under - resea rched due to data scarcity , and it is an inevitable phase in purchase funnel 2 as well as the entire customer journey. It is worthwhile to examine the drivers that proceed distinctive stages of purchase funnel (e.g. , Hu, Du, and Damangir 2014). Practically, it is reasonable to expect that customers tend to with a product when they are interested in buying it, i.e., consideration . In addition to purchase conversion, this study also examines the effect of experiential design on custome r - brand relationship cultivation to provide evidence for the potential positive halo effect on the focal brand. We also examine the contingencies of experiential design that include the most extensively examined design dimensions (e.g., functional and aes thetic design), marketing communication, and customer characteristics to depict a holistic picture. A customer will interpret an experiential design positively only when he or she is confident about the product performance and when the aesthetics fits pers onal taste. The composite nature of experiential design complicates the interplay between experiential design and the other two design dimensions, since literature has shown asymmetric preferences of utilitarian and hedonic values 2 Note that, the outcome variable of this study is from purchase consideration to decision, which is theoretically different from intention - behavior consistency. Purchase consideration and decision are two separate stages of the classic purchase funnel model. 7 of a product design under various design formats (e.g., Chit turi, Raghunathan, and Mahajan 2007, 2008). The second critical contingency is marketing communication ambiguity measured by the across different customers. It is a product - level characteristic and signals the overall quality of marketing communication. Integrated marketing communications, such as advertising, PR, and consumer word - of - mouth, collectively provide product information and, thus, directly affect how a customer perceives a product design. Marketing communication ambiguity has been shown to have a two - interpreting marketing message, resulti ng in less favorable interpretation of experiential design. However, marketers can also appreciate such ambiguity because it is the fundamental reason that product e xperience is seductive (Hoch 200 2). Marketing communication ambiguity has both a negative e ffect on imagery experience formation and a positive moderating effect, such that the positive effect of experiential design on purchase conversion is stronger when marketing communication ambiguity is higher. As experience is subjective, customer characte ristics are highly relevant in adjusting the effect of experiential design on purchase decision. Specifically, customer imagery subjectivity the average of othe because it may potentially have two opposite effects. Customer imagery subjectivity might have a positive moderating effect, since a customer who has unique processing may perc eive appreciate more of the imagery experience because it is personal. However, too much 8 subjectivity may potentially cause cognitive dissonance, particularly when a customer realizes Using unique multi - source data included a design & imagery study, customer purchase, and third - party objective data, we find that experiential design is endogenous and has profound effects on both purchase conversion and customer - brand relationship cultiva tion. In addition, our empirical findings also reveal the critical roles of contingencies in activating and maximizing the effect of experiential design. Specifically, at lower levels of functional and aesthetic design, the effect of experiential design is not significant, so consumers rely fully on functional and aesthetic design for decision - making. When functional and aesthetic design exceeds the threshold, the effect of experiential design becomes prominent (e.g., much stronger than the effects of eithe r functional or aesthetic design) and exhibits an inverse S - Shaped functional form. Moreover, the effect of experiential design achieves the highest at higher levels of marketing communication ambiguity and moderate levels of customer imagery subjectivity. Finally, a superior experiential design not only increases the probability of converting consideration to actual purchase, but also cultivates a customer - brand relationship. The linkage of design and customer relationships is found to be exclusive by the experiential dimension but not functional and aesthetic dimensions. This study highlights the importance of experiential design, considers the unique nature of experience, and reveals its multifaceted contingencies (e.g., other design dimensions, marketing communication, and customer characteristics), providing direct managerial guidance on how to reveal and maximize the prominence of experiential design. 9 CONCEPTUAL BACKGROUND Product Design Product design refers to the gestalt ve elements of a product that consumers perceive and organize as a multidimensional construct comprising the three perception (Homburg, Schwemmle, and Kuehnl 2 015) . We review the key constituent aspects of product design and classify the representative empirical studies in Table 1.1. The most fundamental aspect of product design is its function, as it is the initial promise objective product performance and utilitarian benefits (e.g., Bloch 1995; Chitturi, Raghunathan, and Mahajan 2007 ). The customer - evaluation, which is a function of objective product attributes and customer idiosyncrasy. Product attributes and subjective evaluation joi Kannan, and Ratchford 2008). Although not explicitly, Luo, Kannan, and Ratchford (2008) used perceived comfort as the measure of subjective evaluation, implying the critical role of experience in design and its dependence on functional design and customer characteristics. influences affective and behavioral outcomes. One sub - 2011; Desmet and Hekkert 2007). Aesthetic product design conveys hedonic values and elicits both attitudinal and behavioral outcomes (Patrick and Hagtvedt 2011; Wu et al . 2017). A beautiful design can arouse both cognitive and emotional reactions and then ultimately influence - 10 generation, brand, or segment expectations. Atypical product design arouses aesthetic preference and, at the aggregate level, increases sales and sales growth ( Landwehr, Labroo, and Herrmann 2011 ; Landwehr, Wentzel , and Herrmann 2013 ; Liu et al. 2017). The distinctive processing mechanisms and the interplays of function and aesthetics have also been documented in the product design literature. Chitturi, Raghunathan, and Mahajan (2007) found that product function corr esponds to a prevention focus, whereas form (e.g. aesthetics) corresponds to a promotion focus. As a result, customers are more attracted to appealing design, when function and form meet the expectations; conversely, consumers tend to select a superior fun ctional product, tho ugh still appreciate its beauty, when either functional or form expectations is not met. Similarly, in the post - consumption stage, a superior design with higher hedonic benefits from aesthetic design can delight customers and enhance cu stomer loyalty (e.g., word of mouth recommendation and repurchase intention), whereas utilitarian benefits from functional design increase satisfaction (Chitturi, Raghunathan, and Mahajan 2008). s rooted in the hedonic value of a product, but the utilitarian benefits should not be neglected because they set up the acceptance requirement for purchase decision. Design dimensions are not independent and their interplays are important in purchase deci sions. The product design literature has also been paying attention to experiential design. ly with little effort and lead to positive product evaluation. More recently, Jindal et al. (2016) showed that ergonomics is a user - and experience - centric dimension of product design that influences market 11 share. These two studies signify the extrusive po purchase decision. However, the former uses product evaluation as the outcome that is not necessarily purchase while the latter does not indicate whether ergonomics refers to imagery - based experience that dire ctly leads to purchase decision or actual experience that affects aggregate - level market performance through indirect mechanism, e.g., consumer WOMs. This study differs from the extant experience - focused design study by examining the direct effect of exper iential design that closes the funnel gap between consideration and actual purchase. Moreover, we extract the emotional meaning of experiential design by including usage pleasure and social value of a product design in addition to usability (Mishra 2016) i n the measure of experiential design and specify that experiential design is a function of fact - based design dimensions (functional and aesthetic design), customer characteristics (Luo, Kannan, and Ratchford 2008), and marketing communication, given its en dogenous nature. Experiential Design, Imagery Experience, and Intersecting Contingencies n product experience in the pre - consumption stage, most likely through elaborated imagery (MacInnis and Price 1987). Mental imagery refers to the quasi - sensory and quasi - perceptual experiences i) of which people are self - consciously aware, ii) which exist in the absence of stimulus conditions that are known to produce genuine sensory or perceptual counterparts, and iii) which may be expected to yield different consequences from sensory or perceptual counterparts (Richardson 1969, pp. 2 - 3). Therefore, colloq 12 Imagery experience can be evoked from product design and advertisements that contain semantic information and multi - sensory stimulation. Given that the context of this study is experience. One common factor that stands in both product design and imagery literature is processing fluency. In the product design literature, Brakus, Schmitt and Zhang (2014) found that processing fluency affects the effectiveness of experiential attributes but not functional attributes, whereas i n the imagery literature, Petrova and Cialdini (2005) showed that low processing fluency not only inactivates positive imagery appeals but even leads to backfire effects. Fluency has been recognized as a critical element to generate imagery and to regulate the effectiveness of imagery. fluency with which customers interpret experiential design and form imagery experience. Customers encounter variou s types of product inf ormation, which jointly elicit image ry experience. The complexity and ambiguity of marketing information may result in difficulties in information processing that impede the formation of design interpretation and imagery experien ce . One manifestation of less effective marketing experience tha t can become positive is if marketing communication were effective and the variation were due to individual specific preferences and heterogeneity. In other words, this high variation reflects an adequate level of assists self - referencing, where a consumer relates the self 13 with the message in marketing communication, inferring that the marketing communication is more persuaded (Burnkrant and Unnava 1995). Both product design and imagery literature have highlighted the relevance of marketing communication an d, more precisely, its level of ambiguity that may shape the effect of experiential design and imagery experience, respectively . In addition, consumers naturally present differential design interpretations and thus imagery experience because experience is experiences and memories also contribute to the formulation of mental imagery (MacInnis and Price 1987). Contrary to cognitive elaboration from discursive processing, imagery experience results from processing that involves imagistic and sensory representations (Epstein 1994). Hence, imagery experience is more romantic and self - oriented rather than stimuli - based. In the product design literature, customer identification is also a critical factor in interpretatio n (Homburg, Schwemmle, and Kuehnl 2015 ), and customer idiosyncrasy affects subjective evaluation and ultimately purchase decision (Luo, Kannan, and Ratchford 2008). Therefore, the relevance of customer uniqueness has been recognized in both product design and imagery literature. HYPOTHESES DEVELOPMENT The primary interest of this study is to explore the effect of experiential design on purchase conversion. As discussed, experiential design is endogenous; it is a function of and interacts with other design dimensions, marketing communication, and customer characteristics. Specifically, we are interested in functional and aesthetic design, marketing communication ambiguity, and customer imagery subjectivity, respectively. 14 Following the prior studies, functio nal and aesthetic design dimensions refer to consumer evaluations of utilitarian benefits (e.g., quality) and appearance hedonics (e.g., sleek design), interpretations of experiential design. Customer imagery subjectivity is a customer characteristic These three factors not only influence experiential design perception, but also m oderate its effect to customer - brand relationship cultivation. Figure 1.1 exhibits the conceptual framework and hypothesized relationships. The Threshold Deter mined by Other Design Dimensions Intuitively, all design dimensions, functional, aesthetic, and experiential, are expected to etation of a design gestalt. Previous research has suggested that consumers adapt asymmetrical weights in seeking utilitarian and hedonic value and that the priority can be shifted in different contexts (e.g., Chitturi, Raghunathan, and Mahajan 2007, 2008) . Particularly, meeting utilitarian needs is an indispensable requirement to make a choice, and fulfilling hedonic benefits provides more differentiating value. Additionally, Noseworthy and Trudel (2011) showed that before appreciating hedonic consumption, studies imply that consumers may engage in a hierarchical decision process (Tversky, Sattath, and Slovic 1988) when interpreting the value from different design dimensions. 15 Hierarchical decision - makin g posits that the decision - maker looks for dominance relationship, decisive advantage, and prominent factor sequentially. Because of the composite nature of experiential design, a consumer is more likely to follow the dominance - prominence decision - making r ule (e.g., Evangelidis and Levav 2013) and skip the second stage. Experiential design is composite, wherein utilitarian and hedonic benefits are both integrated, e.g., a customer will perceive a positive experiential value of a design only if the product i s expected to perform well and provide visual satisfaction. This is one of the key reasons that experiential design differs from aesthetic design even though both deliver hedonic values. When either functional or aesthetic benefit is low, customers do not sense the value of experiential design, e.g., it is absent rather than negative. Simply, when customer encounters a poorly performing or designed product, e.g. a car, he/she is less likely to fantasize bad scenarios in mind such, as a car accident. Hence, when functional and aesthetic benefits are limited, functional and aesthetic design predominates decision - making and the effect of experiential design is absent. Once functional and aesthetic design delivers adequate value to evoke positive imagery exper ience, the seduction of experiential design takes off. That is, both functional and aesthetic design determine a threshold effect to reveal the positive effect of experiential design. When the values of functional, aesthetic, and experiential design dimens ions appear simultaneously, a customer is more likely to be attracted to experiential design. Due to the composite nature, experiential value assures that at least the minimum requirement for utilitarian benefits and the personal aesthetic preference are m et. Thus, customer prioritizes experiential design much more than functional and aesthetic design in this context. 16 H1a: When the threshold ( determined by functional and aesthetic design ) is not exceeded, functional and aesthetic design dominantly affect p urchase conversion; whereas the threshold is exceeded, experiential design prominently affects purchase conversion. The Functional Form of Experiential Design We posit that the prominent effect of experiential design on purchase conversion shows an inverse S - Shaped functional form, assuming that functional and aesthetic design exceeds the threshold. The continuum going from concrete to abstract design dimensions follows: functional, aesthetic, and experiential. Concrete information is clear, objective, and evidence - based, whereas abstract information is ambiguous, subjective, and romantic. Thinking concretely or abstractly e, but also design dimensions on which they could ascetic, and experie ntial, when experiential design is perceived at lower levels. Consumers tend factors, e.g., functional and aesthetic benefits. Thus, at this stage, consumers may subconsciously seek balance among functional, aesthetic, and experiential designs. The effect of experiential design is monotonically positive with a decreasing rate because consumers may constantly remind themselves to value concrete design dimension s, particularly when the threshold determined by functional and aesthetic design is exceeded initially. As the perceived value of experiential design increases even higher, customers rely more since the m inimum requirement for functional and aesthetic design is already assured. In addition, when overall experiential design perception 17 increases, the positive feelings of usability, usage pleasant, and social value may synergize and enhance the effect of each sub - element, e.g., 1+1+1>3. The emotional satisfaction and delight accumulate. Subsequently, the positive effect of experiential design indicates an increasing rate ex periential design is rated at higher levels. In summary, the positive effect of experiential design on purchase conversion shows a decreasing positive slope initially and then turns into increasing margins, so called inverse S - Shape, which i s first concave and then convex. H1b: When the threshold is exceeded, the positive effect of experiential design on purchase conversion exhibits an inverse S - Shape d functional form . The Roles of Marketing Communication Ambiguity Both product design and imagery literature have pointed out the importance of marketing communication reflects the clarity (e.g., less ambiguous) of marketing content. Following this logic, we measured marketing communication ambiguity regarding product experience by the experien ce) across different customers. Two aspects cause high dispersion of imagery experience. The first one is that the level of ratings on imagery experience of a product tends to approach the extreme ends, e.g., many consumers rate it too high and low simulta neously. We rule out this situation because our empirical setting targeted customers who already consider purchasing the target product. When a customer has already considered purchasing the target product, he or she is more likely to project 18 a pleasant ex perience overall, but such pleasure might come from different sources. Hence, high dispersion in imagery experience indicates customer heterogeneous preference and different weights in the sub - dimensions of imagery experience, e.g., usability, usage pleasu re, and social value. For example, a pragmatic customer may prefer usability, whereas a more emotional - driven customer may care more about usage pleasant and social value. Therefore, we predict that marketing communication ambiguity has a positive effect o n purchase conversion. H2a: Marketing communication ambiguity has a positive effect on purchase conversion. Common wisdom holds that marketing message should be clear to deliver the intended meaning to ensure that customers are not biased or that they do not misinterpret it (Keller 1993). Yet, another stream of studies has shown that ambiguity might favor customer perceptions (Hagtvedt 2011; Miller and Kahn 2005; Peracchio and Meyers - Levy 1994), who potentially interacts with experiential design positivel y. This is because when ambiguity occurs, individuals may seek further information to fill the perceptual gap, and such perceptual engagement stimulates the positive effect of experiential design. To resolve the potential contradicting influences of market ing communication ambiguity on the effect of experiential design, we assert that whether marketing communication ambiguity benefits or harms experiential design In the stage of processing experien tial design and forming imagery experience, marketing communication ambiguity is expected to have a negative influence. In this stage, customers are more willing to grasp concrete and clear information to ease the processing. Marketing communication ambigu ity decreases the fluency in generating experiential attributes (Brakus, Schimtt, and Zhang 2014)) and imagery formation (Petrova and Cialdini 2005; Schwartz 2004 ). 19 Effective marketing communication should be able to not only ensure the vividness and concreteness to form imagery experience, but also stimulate experience that is relatable to different customers. High dispersion of imagery experience may suggest that the messages conveyed by marketing communication contain rich and plentiful associations that increase self - referencing and imagery value ( and and C ian 2014). Explicitly, Hoch (200 2) stated that ambiguity leaves room for individuals to be open about various interpretations, and it is the fundamental reason that experience is seductiv e. Therefore, although marketing communication ambiguity limits the processing of experiential design, it provides opportunities to better satisfy heterogeneous experience preferences and strengthens the positive effect of experiential design. H2b: The po sitive effect of experiential design on purchase conversion is stronger when marketing communication ambiguity is at higher levels . The Roles of Customer Imagery Subjectivity retation of experiential design (e.g., the evoked imagery experience) is different from the average of others. Previous research has shown that customers have innate needs for uniqueness (Tian, Bearden, and Hunter 2001). Unique or niche customers who have fairly distinctive preferences are more difficult to satisfy using firm - initiated marketing communication. Moreover, a customer with customers tend to be either very c ertain with what they want or unsure with what they want when aware of their differences. Customers with either high or low certainty about their own attitudes resist social influence (Mourali and Yang 2013), e.g., customer - exchanged information, such as W OMs. Therefore, a customer with high subjectivity is less likely to be interested in learning 20 either firm - initiated or customer - exchanged information, which results in decreased engagement. The lack of information seeking and absorbing then reduces the pro bability of purchase conversion. H3a: Customer imagery subjectivity has a negative effect on purchase conversion. A lthough proposing a negative relationship between customer imagery subjectivity and purchase conversion, we expect that the effect of exp eriential design on purchase conversion is at the highest when the level of customer imagery subjectivity is at moderate levels. Initially, like the symbolism dimension of product design (Homburg, Koschate, and Hoyer 2015), experiential design conveys an i - image the need for uniqueness and a desire to express the self in product experience. Therefore, a synergistic effect between customer imagery subjectivity and experiential design on purchase conversion is expected when the level of customer imagery subjectivity is within a reasonable level that does not cause concerns. However, customers cognitive dissonance, resulting in uncertainty and concern. Too much of imagery subjectivity m ay encourage a customer to shift the focus back on fact - based design dimensions, such as functional and/or aesthetic. Then, the effect of experiential design on purchase conversion is attenuated at higher levels of customer imagery subjectivity. Overall, t he maximum effect of experiential design is expected to be found at moderate levels of customer imagery subjectivity. 21 H3b: The positive effect of imagery experience on purchase conversion is strongest when customer imagery subjectivity is at moderate level s . Customer - Brand Relationships Cultivation In addition to the transactional outcome (e.g., purchase conversion), the product design literature has also demonstrated that product design can delight customers, yielding relational outcomes at both product level, such as word of mouth recommendation ( Chit turi, Raghunathan, and Mahajan 2008 ; Homburg, Schwemmle, and Kuehnl 2015) and affection (Wu et al. 2017 ) , as well as brand level, such as customer - based brand equity (Mishra 2016) . The modern view of product design has also recognized the emotional connection. Relationship outcomes are critical because customer - flow. In this study, we emphasize customer - brand relationship because c ustomers are more likely to be bonded with a brand rather than a product. Customer relationship with the focal brand is dynamic in nature. Drawing on self - enhancement theory, individuals seek to maintain and cultivate positive feelings of the self when pu purchased brand. We maintain that the emotional bond from experiential design can be transferred to the post consumption stage and extended to the target brand. Such emotio n transferring mechanism can cause a noticeable (e.g., discontinuous) positive change in customer - brand relationship. H4: A customer who purchases a product with superior experiential design is more bonded with the focal brand, compared to if he/she did n ot. 22 DATA AND MEASURES Data Collection and Descripti on This study entails multiple data sources design & imagery study, customer purchase, and third - party data for empirical examination. We use the U.S. automotive industry across 39 brands as the represent ative product category, because product design plays the most fundamental role as the stimuli for imagery experience, which is the direct customer measure of experiential design. Additionally, cars are a high - involvement product in that customers are inter ested in both utilitarian and hedonic value (Lapersonne, Laurent, and Le Goff 1995; Stahl et al. 2012). In addition, the gap between purchase consideration and purchase decision is meaningful and observable. The final sample comprises 7619 respondents who were asked to evaluate up to 24 car models, resulting in 35,974 observations (after filtering out observations that respondents are not familiar with the car model , do not consider for purchase, and drove the car model before to eliminate actual instead of imagery experience) for analyses. Table 1.2 shows the demographic profiles of the respondents. In total, 5,397 purchase conversion take places, comprising about 15 % of the observations in the sample. The first part of the data was collected by an international marketing research firm that has been conducting a longitudinal study on design & imagery across the 39 brands in the U.S. market. For the purpose of this st udy, we extracted the information on customer evaluations of product (functional and aesthetical) attributes, imagery experience based on design (as the measure of experiential design), customer demographic information, current car usage, whether marketing messages have reached the respondents, customer engagement activities, and typical questions to measure customer - brand relationships. The effective sample comprises the primary decision makers of vehicle purchase who are not working as automobile manufact ure/dealer or 23 marketing/marketing research/advertising. T o exclude the data on actual product experience, we also removed the respondents who indicated that they drove the target car model, e.g., test drive. The second part of the data was obtained from mu ltiple panel companies who had a contract with the marketing research firm and had access to customer sales data with car dealers. The marketing research firm merged the data of customer purchase data and the design & imagery study (i.e. survey - based), fro m February to December of 2009. This sample contained observations of only those customers who indicated that they are familiar with 3 and consider purchasing the target car model. We also collected third party data for control variables. Price information , including the base and highest premium price of each car model, was collected at cars.usnews.com. We also collected quarterly Young and Rubicam customer - reactions under brand - level marketing activities to account for time - varying overarching brand effects. Table 1.3 summarizes variables, operat ionalization, and data sources. Measures of Key Variables Experiential Design. We follow Mishra (2016) and include items that capture the three sub - dimensions of experientia l value: usability, usage pleasure, and usability to measure sum of the twelve indicates from the design & imagery study: fun to drive, excellent ride, effor tless performance, comfortable, versatile/good for any occasion, feel fun, feel happy, feel energized, feel invigorated, feel proud, feel confident, and feel respected by others. 3 We only k irresponsive ratings. 24 Functional & Aesthetic Design . According to usage dominance framework (Deighton, usage. The ordered values that measure product attributes are stronger and more prevalent than actual values in choice models (Drolet, Simons, and Tverskey 2000; Bodapati and Drolet 2005). A consumer is more likely to replace a car that has some advantages over the current driving car. To address the heterogeneity of current driving, we use the sums of product variants (advantage) in function/aesthetics indic ators , relative to the currently driving car model, to measure a , , where i refers to the ta rget car model of purchase and r for car model to be replaced. Functional indicators include whether this car model is high quality, good value for the money, good resale value, roomy interior, and good gas mileage. Aesthetic indicators include good looking, a leader in exterior design, a leader in interior design, progressive design, bold design, and sleek design. Marketing Communication Ambiguity. Marketing communication ambiguity is a product - experiential design (e.g. the evoked imagery experie nce) across different customers regarding the target car model. The higher is the dispersion, the higher level of marketing communication ambiguity. Since imagery experience is measured by a set of indicators, we use binary entropy function to measure the level of dispersion: 25 where n is the number that respondents select a particular imagery experience indicator k and N is the total number of cus tomers that evaluates this car model i . We validate this binary entropy measure with the averaged variance score of 12 binary imagery experience indicators : , where p is the probability that the imagery experience indicator k i s selected across customers of the target car model i. The correlation between the binary entropy function measure and the averaged variance score is .96 (p<.001), indicating high consistency. Customer Imagery Subjectivity. C ustomer imagery subjectivity is a customer - level model. We me asure customer imagery subjectivity by evaluation on each imagery experience indicator k from the probability that this indicator k is chosen by all customers towards the focal car model: (1.4) where k stands for specific imagery experience indicator of the target car model i. Outcome Variables . Purchase conversion is operationalized as a binary variable that equals to 1 if a customer purchases the target car model and 0 otherwise, given that he or she considers buying this car. The other dependent variable is customer - brand relationship, measured by the average score of three relationship - based questions to account for relationship richness (Palmatier, Dant, and Grewal 2007). Specif ically, the questions (scaled from 1 - 10) are: consi ; The Cronbach alpha is .87, suggesting a satisfactory construct reliability. 26 Control V ariables. The control variables include customer demographic information (e.g. gender, age, marital status, and household income), product level marketing mix (e.g. price, product variety, and marketing effort, customer - based brand equity, and brand & time dummies. Product level marketing effort is measured by a set of indicators whether a customer 1) knows about the car model by receiving mail from manufacture, 2) hearing/reading TV/Radio news of newspaper, 3) seeing any national advertising, and 4) seeing /reading any local advertising. We also include product price and variety, proxies the relative price range of the target car model. To partial out the confounding time - varying brand effects, we control for the four pillars of customer - based brand equity ( e.g. quarterly measure) and scale them into the competitive measure (Stahl et al. 2012): , where h refers to one of the four pillars, i refers to the focal brand, and t refers to quarter. We also incl ude brand dummies to control for time invariant brand effect and quarter dummies to control for market level unobservables. Table 1.4 displays descriptive statistics and variable correlations. ANALYSES Model Specification Equation 1.5 specifies the model to examine the effect of experiential design and its contingencies on purchase conversion. Because brand effect tends to influence customer judgments prevalently, we relax the assumption that the parameters of experiential design are equal across all the brands and use heterogeneous choice model, e.g. heteroskedastic probit model to test Equation 1.5. We examine the moderating effects of marketing communication ambiguity and customer imagery subjectivity by drawing the marginal effects instead of conventio nal interaction terms to avoid potential problematic estimation (Williams 2009). 27 where is a binary dependent variable that indicates a customer purchases car i of brand j at the t th month of year 2009. refers to imagery experience as the measure of experiential design and the cubic function captures the inverse S - Shaped. and represent the functional and aesthetic design, respectively. stands for marketing communication ambiguity and for customer imagery subjectivity. is a vector of the four pillars, indicated by h, of customer - based brand equity, relative to the quarterly industry average. A set of demographic variables, brand dummies, and quarter dummies are also included, where brand dummies take account for time - invariant brand he terogeneity and time dummies for unobserved market fluctuation. Addressing the Endogeneity of Experiential Design We use the control function approach (Petri and Train 2010) to correct the endogeneity of experiential design. The control function method i s generally preferable for models nonlinear in endogenous variables (Wooldridge 2007). We use two instrument variables. The first one is brand knowledge engagement the brand refers to the re placed car rather than the target car model. This instrument variable model. Thus, it is exogenous. The other instrument variable is the aggregate - level objecti ve product evaluations, the sum of functional and aesthetic design, as the objective measure, e.g. experiential design. Because we include the individual - specif ic measures of both functional and 28 aesthetic design in the main model, this aggregate level IV is not expected to directly influence Table 1.5 reports the first stage results of con trol function approach to correct the endogeneity of experiential design, with clustered the residual variance across different brands. positively associated with experiential design. As expected, - .639, p < .001) is negatively associated with associated with imagery experience. This implies that in the stage of forming imagery experience, marketing communication ambiguity is more likely to represent low clarity and fluency in processing product design. Also, a customer who has higher need for uniqueness is mo re likely to interpret experiential design positively. The R - square is 82.33%, indicating a good model fit. Empirical Findings The Effect of Experiential Design on Purchase Conversion Maximum - likelihood heteroscedastic probit model is used to test Equation 1.5. We again cluster the residual variance across brands to adjust the variance function and obtain robust standard errors. Table 1.6 exhibits three model results. Model I is when th e threshold of functional and aesthetic design is not exceeded; Model II shows when the threshold of functional and aesthetic design is exceeded and specifies a linear relationship between experiential design and purchase conversion probability. Model III is the final one which the threshold is exceeded and an inverse S - Shaped relationship (e.g. indicated by the cubic terms) between experiential 29 design and purchase conversion probability is specified. The threshold of functional and aesthetic design (the su m of aggregate evaluation on functional and aesthetic design) ranges from 0 to 5.16 and we detect the threshold at its value at 2. 4 Figure 1.2 illustrates the relationship between experiential design and purchase conversion probability under contingencies. Model I presents the results when the threshold is not exceeded, i.e. the sum of functional & aesthetic design is lower than 2. The number of observations is 23,719 and 2,366 purchase conversion take place under this condition. Before exceeding the thre shold, the effect of significant. That is, when the th reshold of product benefits is not achieved, functional and aesthetic design dominantly affect purchase conversion. The effect of functional design is much stronger. Model II reports the results when the threshold is exceeded and the relationship between experiential design and purchase conversion probability is linear. Once the threshold is achieved, the number of observations is 10,774 and purchase conversion occurs 2,807 . The effect of - val ue < .001). Similarly, both - - value < .05) design dimensions are positive and significant. We discuss other effects in the model III interpretations. Model III adds the squared and cubic terms to specify the functional form of experiential design on purchase conversion. The parameter estimates of first - order, squared, and cubic terms of experiential design are all significant ( =.429, p - value < .001; = - .015, p - value < .05; 4 We estimate Equation 1.5 at levels of summed (objective) functional and aesthetic design. We set up the conditions where it is increased by .1 consecutively. When the level of summed (objective) functional and aesthetic design exceeds 2, the effect of exp eriential design becomes positive and significant. 30 =.001, p - value < .05). Following Homburg, Koschate, and Hoyer (2005), we conclude that there is an inverse S - Shaped relationship between experiential design and purchase conversion at Model III is preferred ( ). In addition, the effect of functional and aesthetic design dimensions on purchase conversion also indicate significant and positive ( =.083, p - value < .001; = .034, p - value < .05) effects. To compare the effects of three design dimensions, we draw the average partial effects: experiential, functional, and aesthetic dim ensions show =.111 (p - value < .001), = .024 (p - value < .05), and =. 010 (p - value < .05), respectively. The effect of aesthetic design is weaker than functional design. We speculate this is because the effect of aesthetic design i s likely to be mediated by experiential design, since both share the hedonic benefits and are more abstract. Importantly, the effect of experiential design is highest among the three dimensions, indicating its prominent effect on purchase conversion. To s ummarize, when the threshold determined by functional and aesthetic design is not achieved, the positive effect of functional and aesthetic product benefits are dominant and the effect of experiential design is insignificant (Model I); when the threshold i s exceeded, the positive effect of experiential design is prominent , much stronger than that of either functional or aesthetic design (Model II and Model III) and exhibits an inverse S - Shaped functional form (Model III). Therefore, Hypothesis 1a and 1b are supported. Model II and III find very consistent results that the effect of marketing communication ambiguity does not have a significant main effect on purchase conversion. It implies that y exist, neutralizing the main 31 effect of marketing communication ambiguity. Hypothesis 2a is not supported. Next, we examine the marginal effects of experiential design across levels of marketing communication ambiguity (the variable ranges from 0 5.581) f rom 0 to 6 with one - unit interval. Initially, the effect of experiential design is not significant until the level of marketing communication ambiguity achieves 2. When the level of marketing communication ambiguity is above 2, the positive effect of exper iential design shows a monotonically increasing trend (Figure 1.3a). Thus, Hypothesis 2b is supported that marketing communication ambiguity strengthens the positive effect of experiential design on purchase conversion, given the threshold is exceeded. Sim ilarly, Model II and Model III consistently report that customer imagery subjectivity has a negative main effect on purchase conversion ( = - 1.213, p - value < .001; = - 1.221, p - value < .001). Thus, Hypothesis 3a is support ed. We examine the effects of experiential design across levels of customer imagery subjectivity (the variable ranges from 0 3.319) from 0 to 3 with a half unit interval. The positive effect of experiential design is significant across all levels of custom er imagery subjectivity and its effect size indicates an inverse - U shaped (Figure 1.3b). That is, the positive and prominent effect of experiential design on purchase conversion reaches the highest when customer imagery subjectivity is at moderate levels. Hypothesis 3b is supported. The residual term derived from the control function approach is significant (p - value < .001) in Model II and III, indicating that experiential design is statistically endogenous. Regarding customer demographics, male custome rs and high - income customers are more likely to realize a purchase conversion, all else being equal. Marketing effort has a strong and positive effect, while product variety has a negative impact on purchase conversion. The reason might be 32 that cars with a larger range in product variety are more likely to be premium cars. Purchase decisions of high priced products are generally longer. The Effect of Experiential Design on Relationship Cultivation Hypothesis 4 states that a customer who purchases a car with superior experiential design is more bonded with the focal brand, compared to if he/she did not. To examine this hypothesis, we use the treatment effect model with the combination of inverse - probability - weighted and regression - adjustment (IPWRA) estim - include all the covariates in Equation 1.5 to ensure unconfoundeness. After merging purchase - level survey that contains less observ ations and comprises customer - brand relationship questions), we end up having 6,244 observations for the examination of Hypothesis 4. Note that our focus is the change of customer - brand relationship instead of the level, we are primarily interested in cus tomer - brand relationship cultivation, a within - person comparison. Thus, we use the counterfactual framework, where the parameter of the treatment effect on the treated (ATET) indicates the change in a customer - brand relationship, comparing if the customer purchases the product to if he or she did not. The dependent variable is the average score of the three variables ranging from 1 - Given that our sample includes respondents who show strong purchase interests, t he average customer - brand relationship score is 8.38 out 10 and its distribution is left - skewed. We use bootstrap estimation with 1000 resampling to test the treatment effect model. The treatment effect of treated (ATET) is .123 with a standard error of .052 and a 95% confidence interval of 33 [.0 22, .224] (Table 1.7). Experiential design has positive effects on customer - brand relationship in both purchase and non - purchase groups, where the parameter estimate is .198 (p - value <.001) for the purchase group and .158 (p - value <.001) for the non - purcha se group, respectively. A customer who rates high on experiential design not only convert a purchase but also increase the emotional bond with the focal brand. However, we do not find significant effect of either functional or aesthetic design in increasin g a customer - brand relationship. This implies that although all design dimensions: functional, aesthetic, and experiential are expected to positively influence purchase probability, experiential design exclusively extend the emotional affection from a prod uct design to the brand. Thus, a customer who purchases a product with superior experiential design is more bonded with the focal brand. Hypothesis 4 is supported. DISCUSSION The concept of product design has evolved from simply providing solutions to de liver a great experience for customers. Customers perceive a product design as a unitary whole of all elements that develops into psychological perception (Homburg, Schwemmle, and Kuehnl 2015). We assert that this psychological perception differentiates a product design from others and embeds it with emotional affection. Most studies on product experience have focused on actual experience ( e.g. , Chitturi, Raghunathan, and Mahajan 2008 ). Our study explores how a nterpretation of a product (experiential) design, influences his or her purchase decision and cultivates the brand relationship. To the best of our purchase decision and relationship cultivation in the marketing literature. 34 When the new version of Jeep Wrangler was introduced, many non - - and experience - dimension in the product design literature. Two academic studies have addressed this emerging design dimension, specifically, Brakus, Schmitt, and Zhang (2014) and Jind al et al. (2016). The former used the outcome of product evaluation, and the latter investigated aggregate market share. Our study also fills the purchase journey gap by using purchase conversion, e.g. consideration to purchase, as the primary outcome. In addition, consistent with the theoretical discussion in Hoch (2002), this study empirically demonstrates various unique characteristics of product experience, such as endogenous nature, seductive, and ambiguous etc. Specifically, we show that experiential design is not only influenced by but also interacts with other design dimensions (e.g., functional and aesthetic), marketing communication ambiguity, and customer imagery subjectivity. The product design literature has not yet explored whether the design d imensions, the self, and the marketing message interact together synergistically to evoke a The seduction, i.e. the strong and long - lasting positive effect, of experiential design occurs only w hen there is initial promise from functional and aesthetic design. We show the interplays among design dimensions in the hierarchical decision process at the customer level. The positive effect of experiential design is not significant until the combined b enefit from functional and aesthetic design achieves an acceptable level. Once the threshold determined by the benefits delivered by both functional and aesthetic design, is exceeded, the effect of experiential design is much stronger than that of the othe r two design dimensions and exhibits an inverse S - shaped function form. The dominant - prominent interplays among design dimensions 35 asymmetric. Our results recognize the necessity to provide a decent functional and aesthetic concept and neglect the primary promise of good product performance to customers. Customers will not react favorably to experience stimuli (e.g., experiential design) or other marketing messages (e.g., advertisements), without an acceptable expected performance and v isual satisfaction from a product. extends both design and imagery literature by confirming the importance of marketing communication. Previous research has shown tha t integrated marketing communications, such as product presentation ( Yoo and Kim, 2014 ), logo shapes ( Jiang et al., 2015 ), acoustic pitch ( Lowe and Haws 2017 ), and the like, affect imagery effectiveness. Particularly, we find opposing effects of marketing interpretations of experiential design. In the interpretation stage of imagery experience, we find a negative effect of marketing communication ambiguity. This relationship is consistent with the studies on vividness and concreteness in imagery - related studies (e.g., Argyr iou 2012; Miller and Stocia 2004 ; Schlosser 2003; Shiv and Huber 2000 ; Walte rs, Sparks, and Herington 2007 ). We also show that marketing communication ambiguity can be be neficial and maximize the positive in that customers can develop tailored experience that is quite personal. This is consistent with n that ambiguity is the reason that product experience is seductive. 36 Managers should rely on analytical guidance to monitor marketing communication effectiveness Customer imagery subject ivity is another moderator that shapes the relationship between experiential design and purchase conversion. Customer imagery subjectivity reflects the extent of . The inclusion of customer specific characteristics has been drawing more attention in the product design literature (e.g., Luo, Kannan, and Ratchford 2008). Our findings show that although it is more difficult to satisfy a unique customer or a niche segm ent, additional emotional affection from experiential design is higher for customers who show moderate level of subjectivity. Cooper, Mercedes Benz, and a num ber of other automotive pioneers actively offer different In most cases, customers pay for tailored components. It is a win - win for both customers and marketers because customizati Finally, the two outcomes used this study, i.e., purchase conversion and customer - brand relationship cultivation, are novel in the product design literature. Marketing literature has u of the combination of survey - based and purchase data consideration to actual purchase behavior are rare. In practice, such at tempts would entail effortful marketing strategies and tactics. It is also more effective and efficient to invest marketing resources and actions in customers who are interested in purchase, i.e. holds a positive purchase consideration, rather than who are not. Finally, this study also demonstrates that 37 the focal brand. Customer relationship is dynamic in nature. The change in relationship is more salient com pared to the level of relationship (Huston et al. 2001 ; Palmatier et al. 2013; Zhang et al. 2016 ). Firms engage in and dedicate significant resources to customer relationship management (CRM) activities. It is important for marketing researchers and practi tioners to learn how to cultivate relationships with customers, targeting at the change not the level. Our findings show that experiential design can enhance customer - brand relationship directly , but neither functional nor aesthetic design boost customer r elationships, implying the exclusive value of experiential design. More importantly, if a customer purchases a product with superior experiential design, he or she is highly likely to extend the emotional affection to the brand, causing a potential positiv e halo effect and loyalty tendency. Given that our sample targeted customers who showed significant purchase interests, the brand rating was high (average of 8.38 out of 10). Our finding shows that experiential design has the potential to break the ceiling effect of marketing offerings on brand love. 38 ESSAY TWO CUSTOMER - RELATIONSHIPS ASSET AS A COMPETITIVE ADVANTAGE AND ITS IMPACT ON ACQUISITION AND RETENTION DECISIONS Firms have imperatively become customer centric, a fact evidenced by the trend among the chief executive officers (CEOs) of many successful firms to prioritize their customers at the top of marketing strategy. Jeff Bezos, Founder and CEO of Amazon, articul ated this customer - centric Customers are certainly the lifeblood of every business. As an intangible asset and ultimately a scarce resource (Peppers and Rogers 2005), customer base and its quality are inextricably linked to firm value and future prospects. The state of the art of the valuation of customer equity, - based, customer - driven, competitor - cognizant, and - term success (Rust, Lemon, and Z eithaml, 2004). Customer equity has been acknowledged as one of the most critical marketing metrics. It proxies firm value (Gupta, Lehmann, and Stuart 2004), formulates customers as accountable assets (Rust, Lemon, and Zeithaml 2004), supports forward - look ing (Zeitham et al. 2006), and facilitates long - term decision making (Kumar and Shah 2015, p. 7). Current customer valuation, e.g. , customer equity, has focused predominantly on the direct face value of individual customers without accounting for the creation of indirect value from customer - customer and customer - firm interactions. This omission leads to biased estimates customer relationships with the focal c ompany. Specifically, customer - c ompany identification (CCI) theory 39 and that such a connection forms the foundation for a mutually beneficial customer - company relationship (B - role behaviors (purchasing and using products), customer - company identification sheds light on extra - role - of - mouth, product improvement suggestions, recruiting other custom Gruen 2005). Relationship marketing has identified and documented many of these relationship - driven behaviors (e.g., referral, positive word - of - mouth, customer participa tion, etc.) and their profound consequences. Yet most of the empirical research has focused on specific behaviors rather than creat ing value among customer s, who are expected to engage in many of these behaviors. While examining the effects of s pecific behaviors is certainly noteworthy, aggregating a - role behaviors directly helps in assessing - enhancement theory, a typical c ustomer tends to engage in multiple favorable or destructive behaviors. The challenge is that relationship - driven behaviors might take place spontaneously and in any format. It is impossible to exhaustively track and assess the impacts of all of a customer - driven behaviors. Instead of appraising the activity level, this study examines relationship - driven value customer, who engages in multiple relation ship - driven behaviors, can contribute to its organic growth over time. This essay starts by summarizing the value creation mechanism of customer - customer and customer - firm interactions. Specifically, the value creation of customer - customer interactions ha s two distinctive processes: (i) customer - customer interactions directly influence 40 prospective market performance, and (ii) customer - customer interactions synergistically strengthen other growth forces such as marketing effort, and consequently contribute to market performance. Secondly, customer - adaptive capability of its resource allocation; in the sec ond stage, the firm employs such improvements to achieve organic growth. Overall , both types of value creation develop into a reinforcing mechanism between current installed customer base and prospective market performance . We theorize that custom er - value creation through customer - customer and customer - firm interactions to achieve organic growth in market performance and propose that this is a competitive advantage. Empirically, we use p - relationships asset in Study I. To demonstrate the profound impacts of customer - relationships asset, we examine whether and how it affects individual customer acquisition and retention decisions. Usi ng the empirical findings from Study I, e.g., the CIRFs (impulse response functions ) as measures of customer - relationships asset for 33 brands, we find that customer - relationships asset has positive impacts on both customer acquisition and retention decisi ons, controlling for all other brand - level characteristics. To the best of our knowledge, few empirical studies have investigated how a firm - or brand - demand for data support. We are also interested in the contingencies that shape the effect of customer - consideration set size is found to have negative main effects on both acquisition and retention decisions, it strengthens the positive effect of customer - relationships asset on acquisition decisions. Additionally, marketing efforts as well as interactive and objective information search 41 selectively enhance the positive effects of customer - relationsh ips asset, depending on either acquisition or retention as the outcome. These findings offer managerial guidance on how to strategically direct different information sources to customers under different contexts, e.g., whether the goal is acquisition or re tention; whether a firm possesses (positive vs. negative) customer - relationships asset or not. CONCEPTUAL BACKGROUND AND FRAMEWORK Customers as Assets, Customer Equity, and Customer - Relationships as Assets Fundamentally, the notion of customers as assets can be traced back to Srivastava, - based assets, in which customers are off - balance, intangible, and a type of relational (vs. intellectual) asset. Rel ational market - based assets are outcomes of the relationship between a firm and its key external stakeholders, among which customers are the most important ones for most businesses. Therefore, the concept that r - The state of the art of customer valuation customer equity is recognized as the most critical off - balance - (Gupta, Lehmann, and Stuart, 2004). Customer equity has thr ee main drivers: value equity, brand equity, and relationship equity ( Lemon, Rust, and Zeithaml 2001 ). Vogel, Evanschitzky, and Ramaseshan (2008) find that the three customer equity drivers customer perceptions of value, brand, and relationship, simultaneo usly affect loyalty intentions and future sales. On the impacts of customer equity, the marketing literature has acknowledged that customer equity increases the return on marketing investment (Rust, Lemon, and Zeithaml 2004), and outperforms other proxies 42 Stuart 2004). Since then, customer equity has been bridged with market capitalization (Kumar and Shah 2009; Srinivasan and Hanssens 2009; Wiesel, Skier, and Villanueva 2008) and shareholder value (Schulze, Skiera and Wiesel 2012). Marketing scholars advocate that customer equity be incorporated in financial statements (Wiesel, Skiera, and Villanueva 2008; Skiera, Bermes, and Horn 2011). Customer equity is defined as the sum total of customer lifetime values (CLVs). CLV is profitabili ty, captured by profit margin and purchase frequency. Regardless of the various ways of computing CLVs and customer equity (e.g., see detailed discussions in Kumar and George d overlook the interrelatedness among customers. the environment guid es the firm in choosing which entities to align with, and how to do so, and ce its long - term performance. To this end, instead of examining the effects of specific customer - customer or customer - overall indirect (e.g., relationship - driven activities rather than purchase) contributions. We propose that a customer can generate value through customer - customer and customer - with the focal firm. In the next section, we summarize customers - customers and customer - firm 43 interactions, the value creation paths and framework, and theorize customer - relationships asset as a competitive advantage. Customer - company identification theory (CCI) posits that there is a connection betw een a mutually beneficial customer - company relationship (Bhattacharya and Sen 2003). In addition to purchases as in - role behavior, customers engage in extra roles such as suggesting product improvements, positive word - of - mouth, and proactive communication of anticipated problems (Ahearne, Bhattacharya, and Gruen 2005). The relationship marketing literature has numerous studies showing that customer - customer and cust omer - firm interactions lead to profound impacts - term performance and organic growth. Table 2. 1 summarizes representative customer - customer and customer - firm interactions. Customer - Customer Interactions and Value Creation . customers are connected and interact in various ways. For example, studies on referral programs as a form of incentive - motivated word - of - mouth activities, show that customer referral not only serves as an effective acquisition tool but also generates a lo ng - term influence on both referring and referred customers. Studies show that among new customers, customers that are acquired by a referral program are at least 16% higher in terms of customer value than non - referred customers, given similar demographics and acquisition timing (Schmitt, Skiera, and Van den Bulte. 2011). Also, customers who participated in a referral program became more loyal and increased revenue growth by 11.4% ( Garnefeld et al. 2013). The connection between referring and referred custome rs can produce a chain effect in which a consumer who uses WOM referral in a purchase decision will also recommend to friends (Yang et al. 2012). This may result in a reinforcing mechanism between customers who refer others and prospective customers. Expli citly, 44 during their tenure, and customer acquisition and firm performance are reinforced by the buzz effects (Villanueva, Yoo, and Hanssens 2008). Additionally, Trus ov, Bucklin, and Pauwels (2009) find strong carryover effects for WOM referral on customer acquisition and, more importantly, these carryover effects are the key drivers for long - responsiveness to marketing efforts (elasticities). Th erefore, customer referrals generate value for - term performance in two ways: direct reinforcement and indirect enhancement of marketing effectiveness. We maintain that these are the two fundamental value creation paths that are applicable to other forms of customer - customer interactions, which we discuss briefly as follows. In addition to referral, customers interact with others in different ways. Generally, customers communicate with others in daily conversations, blogs, social media, etc. W ord - of - mouth communication has become the fifth marketing mix and potentially the most powerful one ( Armelini and Villanueva 2011). Regardless of whether it is online or offline, word of mouth ker, Donthu, and Kumar 2016; Bansal and Voyer 2000; Berger, Sorensen, and Rasmussen 2010; East, Hammond, and Lomax 2008; conversations are expected to follow the two val ue creation paths: increasing brand awareness to attract more purchases directly and enhancing the effectiveness of marketing efforts indirectly. As a platform, brand community provides opportunities for customer - customer interactions. The concept of bra nd community was introd ) as - geographically bound community, based on a structured set of social 45 experienced an d novice customers, regardless of whether the communities reside on company - owned or independently owned websites (Adjei, Noble, and Noble 2010). A strong brand community can turn into a natural shield for competitors, increase the relational switching cos t for non - community members, and even engender a sense of oppositional loyalty toward competing brands (Thompson and Shiha 2008). Interestingly, customers may have a psychological sense of brand community so that individuals perceive relational bonds without necessarily knowing each other (Carlson, Suter, and Brown 2008). Brand community identification not only leads to purchase but also stimulates other extra behaviors such as recommendations and participation (Algesheimer, Dholakia, and Herrmann 2005 ). Not all customer - customer connectedness or interactions are desirable. For example, in linked with an 80% increase in the defection hazard (Nitzan and Lib defection decision may lead to the churn behavior of socially connected others who share the same mobile provider (Haenlein 2013). Regardless of whether the effects of customer - customer interactions are positive or negative, thes e effects are dynamic and potentially cause a reinforcing mechanism. In summary, customer - customer interactions occur in various forms and have the synergistic effect with marketing effectiveness. Customer - Firm Interactions and Value Creation . Interactions between customers and becoming a forum for conversation and interact ions between consumers, consumer communities, 46 increases both its customer - based profit and relational performance 5 (Ramani and Kumar 2008). Customer - firm interactions can also take place in different formats (e.g., suggested service improvements, answering a satisfaction survey, complaints), and these interactions generate value in two stages. In the first stage, a firm improves its customer knowledge and adaptive capa bility of resource allocation through interactions, and in the second stage, these improvements drive to long - term success. Specifically, customers add value to companies by participating in the knowledge development process on customer preferences (Joshi and Sharma, 2004). Customer participation can be information providers and/or co - developers in the NPD process (Fang 2008; Fang, Palmertier, and Evans 2008; Lengnick - Hall 1996). The service - dominant framework (Vargo and Lusch, 2004) also advocates custome r co - creation. Customer knowledge can be internalized and that serve its own customers. Arnold, Fang, and Palmatier (2011) show that acquisition and retention orientations increase customer knowledge development ( depth and diversity) and the configuration of resource allocation (exploration and exploitation). As a result, a firm with superior customer knowledge can foresee the evolution of customer preference and thus outperform in the marketplace. In addition to i mproving customer knowledge, customer - firm interactions also facilitate the adaptive capability of resource allocation. Among the most challenging marketing decisions are who to invest in, how much to invest, how to invest, and where to invest (Venkatesan and Kumar 2004; Reinartz, Thomas, Kumar 2005). Managers are intrinsically motivated to maximize 5 In Ramani and Kumar (2008), customer - based profit performance includes the identification of profitable customers, the acquisition and retention of profitable customers, and conversion of unprofitable cu stomer s to profitable ones, whereas customer - based relational performance refers to increased customer satisfaction levels, increased custom er ownership, and positive word - of - mouth. 47 analytics programs diagnose the balance between the benefits an d cost of marketing actions to allocate marketing resources differently to customers based on customer lifetime value (Venkatesan and Kumar 2004), but also other no n - transactional behaviors, such as product returns. (Peterson and Kumar 2009). In the long term, a dynamic and competitive market requires a firm to adjust its resource allocation in a routine manner and operate with adaptive capability, e.g. vigilant mark et learning and adaptive market experimentation (Day 2011). Customer - firm interactions provide great opportunities for companies to gather customer feedback, which is not revealed by transactions. In summary, customer - firm interactions create value by impr in the first stage. In the second stage, firms subsequently employ these improvements to thrive in the accelerating complexity of their market. Customer - Relationships Asset as a Competitive Advantage Drawing on self - enhancement theory, a customer can engage in multiple extra - role behaviors, which can be spontaneous, unsupervised, and untraceable. These behaviors are driven l firm. Value creation is realized through customer - customer and customer - long - term success. We conceptualize and define customer - to utilize the value cr eation through customer - customer and customer - firm interactions to achieve organic growth in market performance. The most critical characteristic of this concept is 48 - term performance is not readily observed and measurable. We prop ose that customer - relationships asset is a competitive advantage. The value creation elements (e.g., customer base, knowledge development, and adaptive capability of resource allocation) present profound difficulties to competitors who want to develop simi lar substitute be its resources, while improving customer knowledge and resource allocation capability can be considered as its quality of marketing deployment. Resourc Moorman, and Inman 2003). In effect, customer - relationships asset indicates not only that a firm owns customers and relationships as strategic resources, but also that the firm deploys these customer - relationships effectively and profitably. The presence of customers as resources and value creation through interactions as effective deployment affirms customer - relationships asset as a c ompetitive advantage. Figure 2. 1 exhibits the conceptual model of customer - relationships asset. Customer - company identification motivates customers to engage in customer - customer and customer - firm interactions. The value creation of customer - customer interactions incorporates two paths. A performance, and/or (ii) indirectly, enhance the effectiveness of marketing efforts. The value creation of customer - knowledge and adaptive capability of resource allocation, and then ii) employing th ese improvements to better serve prospective customers and achieve superior market performance. 49 The reinforcing mechanism develops into organic growth, evolving into customer - relationships STUDY I ASSESSING CUSTOM ER - RELATIONSHIPS ASSET Modeling Approach Assessing the customer - relationships asset requires a careful examination that captures long - term market performance. Th e reinforcing mechanism may involve a complex system with interactions and a feedback loop between customers and market performance. The recent literature has suggested that persistence modeling approach is appropriate for quantifying the value generation - term reinforcement (Trusov, Bucklin, and Pauwels 2009; Villanueva, Yoo, and Hanssens 2008; Zhang et al. 2012) Persistence modeling approach addresses the problem of long - term market response by reinforcing mechanism an - run modeling approach to the short - and long - run marketing effectiveness of advertising (e.g., Caval iere and Tassinari 2001; Kim and Hanssens 2017), promotions (Pauwels et al. 2002; Pauwels, et al. 2004; Slotegraaf and Pauwels, 2008; Pauwels et al. 2003), distribution changes (e.g., Bronnenbert et al. 2000), a systematic approach to monitor customer reac tions (e.g . , Dekimpe et al. 2005), and interplay between marketing and firm valuation (Pauwels et a l. 2004; Joshi and Hanssens 2004 ). More recently, Osinga, Leeflang, and Wieringa (2010) have extended a time - varying parameter model to compare persistent and transient marketing efforts. 50 Technically, persistence modeling approach comprises multi - procedure analyses that include: (ia) un it root testing, (ib) cointegration testing, if applicable, (ii) vector autoregressive model: VAR in levels or differences, or vector error correction model (VECM) if equilibrium exists among variables, and (iii) a derived (cumulative) impulse response fun ction that measures how much one unit increase in the impulse variable, e.g., customer base, will lead to a fluctuation in the response variable, e.g., market share, over time. Although our conceptual framework is developed at the firm level, following Sta hl et al. (2012), we formulate the assessment of customer - relationships asset and its impact (Study II) at brand level. It has been widely accepted that brand has its individual operation in practice and customers naturally react to a brand, instead of a f irm, in their mindsets (Keller 2008; Keller and Lehmann 2003). customer base and its (unit) market share. To avoid marketing - induced interruptions in market share, we inc lude changes of advertising intensity as the covariates. In addition, we control customer profiles for each brand by the percentage of male and married customers: = + + Data Description The sampling scheme for this study needs to satisfy the following criteria. First, the level of customer - relationships asset varies across brands, e.g., high, low, or absent. Focusing on one entire industry is advantageous for comparing brands and examinin g the effects of customer - relationships asset in Study II. Next, the product category should have high customer involvement and meaningful customer - brand relationships so that customer - customer and customer - firm interactions occur frequently. Our conceptua l framework is drawn from customer 51 conditions for the manifestation of the theory are: (i) it is important enough to the customer to make the company salient to t he customer (Bhattacharya and Sen 2003); (ii) the organizations relationships can be reflected by frequent engagement and interactions with the company and its agents (R ao, Davis, and Ward, 2000; Bergami and Bagozzi, 2000). Third, we want to focus on the category, which have low frequency in repeat purchases so that the increase in market performance is driven mainly by customer interactions. Given the aforementioned crit eria, we use the U.S. automotive industry as the sampling frame, as cars are high - involvement products in terms of interest, risk, and the combination of symbolic and hedonic value (Lapersonne, Laurent, and Le Goff 1995; Stahl et al. 2012). The data contai n two major sources, over the period August 2006 to March 2012 (68 months). One is objective market performance data, e.g., unit market share, from WardsAuto. The second part is supported by a leading international marketing research firm, which collects c ustomer - level data in the automotive industry across all brands. This archive data report current car usage to and demographic information. The number of resp ondents ranges from 110,112 to 162,319 across the 68 - month period. The sample is adequate for reflecting actual brand usage in percentage terms. report they are currently driving the focal brand. It is a brand - level aggregation by month. We calculate market share based on objective sales data. Advertising intensity is measured by the share of voice (Datta, Foubert, and Heerde 2015), where respondents were asked to report 52 i ndicating whether they recalled seeing any automotive advertising during the previous three months. Customer profiles include the percentage of males and the percentage of married subjects (vs. other marital categories that include single, living with a si gnificant other, separated/divorced but not remarried, or widowed). We end up having 33 brands that have sufficient data inputs to support our time series analyses. Table 2 .2 summarizes the descriptive statistics of Study I. Estimation Procedure Stationarity Tests. The estimation procedure starts with the stationarity tests regarding the two focal endogenous variables to determine whether the series is evolving or stationary by the augmented Dickey - Fuller (ADF) unit root test. We use the iterative pro cedure suggested by Enders (2008 ), which examines whether to include a deterministic trend in the ADF test. We then use KPSS tests to validate the ADF tests for each brand and decide the stationarity of the focal endogenous variables. The results from the ADF and KPSS tests suggest high consistency, where 31 out of 33 brands are indicated with the same results relating to whether the series is stationary or not. We rely on the ADF test for a final decision on the brands that indicate inconsistent result s between the ADF and KPSS tests. We do not find any case with both endogenous variables to be non - stationary. Thus, there is no need for a further cointegration test. Series Transformations. We first difference the endogenous variables that are non - stat ionary. Then, we test the stationarity of the first differenced terms, and the results suggest that all the first differenced terms, if applicable, are stationary. As for the endogenous variables that are trend stationary, we detrend the original series. F inally, we insert the endogenous variables directly if the original series is stationary, the differenced terms if the original series is non - 53 stationary and the first differenced terms are stationary, and the detrend terms if the original series is trend - s tationary into the VAR models for each brand. Lags Selections and VAR Model. Following the literature, we use both AIC (Akaike lengths that need to be included into the VAR system for each brand. The results indicate that 22 out of 33 brands are con sistent with the same lag selections. Since SBIC tends to select a smaller model, we rely on AIC to select the lag lengths to avoid omitting the potential carryover effects ics indicate different lag selections. Our results also indicate that for all brands, the lag selection by AIC is consistent with the FPF (final prediction effort) and HQIC (Hannan and Quinn information criterion). We then specify the number of lags into t he VAR model for each brand. Figure 2. 2 depicts the procedures of Study I. Empirical Findings We test the above procedures, e.g., the stationarity tests, transformations if applicable, lag selections, and VAR model repeatedly for each of the 33 brands. T able 2. 3 summarizes the empirical findings for brands that have significant customer - relationships asset. Specifically, we find 10 out of 33 (e.g., 7 positive and 6 negative) brands having significant 6 customer - relationships asset, indicated by the signifi cant parameter estimates that a lagged customer base leads to market share. Next, we estimate the orthogonalized IRFs (impulse response functions) and CIRFs (cumulative impulse response functions), using customer base as the impulse variable and market sha re as the response variable. 6 In order to avoid the potential threats of missing practical relationshi ps between customer base and market share, we use the significance level at .10, given that our sample for each brand has only 68 months periods. 54 The IRFs trace the change in market share when there is an unexpected shock to a - run effects, the CIRFs capture the total impact of one r base on market share over time. The primary typical customer can influence its organic growth in market performance over time. The interpretation of CIRFs refers to, on average, how muc h more (or less) market share is generated by one typical customer of the focal brand throughout his or her usage tenure. Thus, we use CIRFs as the measure of customer - relationships asset for each brand. Figure 2. 3 presents the CIRFs graphs, using customer base as the impulse variable and market share as the response variable for the brands that have significant customer - relationships assets. Using Brand 7 to illustrate, there is a significant and positive effect of the lagged customer base on market share. Over time, this additional (unexpected) customer, through his or her tenure of usage, can generate .02 58% (indicated by CIRF) market share for Brand 7. Building a regression model (e.g., regress unit sales on market share and time dummies) to test the association between unit sales and share, controlling for current market demand for Brand 7, one increase in market share leads to about 10,332 (B=10332.27, p - value<.001, R - Square=97.92%) additional unit sales. That is, this additional customer can contribute to more than two additional purchases (10332x.0258%=2.66) through his or her usage tenure. The average price of the car models of Brand 7 in 2012, the ending period of the time window for this study, is $42,668.65. Therefore, typical revenue, without considering his or her own purchases. 55 To conclude, Study I summarizes value creation through customer - customer and customer - firm interactions and theorizes customer - value creation through customer - customer and customer - firm interactions to achieve organic growth in market performance. Customer - relationships asset is assessed by the persistence modeling approach and measured as the CIRFs, using longitudinal data of 33 brands in the automobile industry. To demonstrate that customer - relationships asset is a competiti ve advantage, Study II proceeds to demonstrate whether and how customer - relationships asset, as an inter - brand characteristic, affects individual - level customer acquisition and retention decisions. 56 STUDY II: THE IMPACT ON CUSTOMER ACQUISITION AND RETEN TION DECISIONS To examine whether customer - relationships asset is a competitive advantage and how it functions, Study II tests its impact on individual - level acquisition and retention decisions. Consideration set size and external information search are selected as moderators because these two are highly relevant in consumer purchase decision making. Specifically, consideration set - relationships asset is expected to ou tperform to a greater extent when a customer has a larger - relationships asset is not readily observed, further su b - categorized into interactive and objective search activities. These two types of search activities differ and selectively moderate the effect of customer - relationships asset across acquisition and retention decisions. HYPOTHESES DEVELOPMENT The Impact of Customer - relationships Asset Fundamentally, customer - relationships asset is developed from customer interactions and the dynamic creation of value for organic growth. We propose that customer - relationships asset is a brand - level competitive advantage, a nd thus positively affects individual - level customer acquisition and retention decisions. Previous research has shown that value created through customer interaction benefits acquisition and retention. Wangenheim and Bayon (2007) show that satisfaction inc reases word - of - mouth referrals, and then prospective customers receive a higher level of referrals, resulting in higher customer acquisition. Similarly, Choi, Bell, and 57 Lodish (2010) show that interdependence among target customers, defined by physical cus tomer density and word - of - mouth consecutiveness, creates a synergistic effect, e.g. it serves as a social mode. Studies have found that acquisition under referra l impact induces greater retention and values (Villanueva, Yoo, and Hanssens 2008; Schmitt, Skiera, and Van den Bulte 2011). Therefore, customer connections play an important role in both acquisition and retention decisions. In addition, the interactions between customers and a firm also increase the likelihood of customer acquisition and retention decision. A firm with positive customer - relationships asset is expected to have better customer knowledge and adaptive capability of resource allocation. Custo mer knowledge development helps firms select prospective customers more effectively and efficiently. When it comes to the task of retention, a firm with positive customer - relationships asset is able not only to accurately identify profitable customers but also to design suitable marketing communication strategies. Thus, customer - relationships asset is expected to have a positive impact on both customer acquisition and retention decisions. - relationships asset increases the probability of (a) customer acquisition and (b) customer retention decisions, all else being equal. The Roles of Consideration Set Size Following the mainstream of marketing literature on choice models, we assume a two - stage choice decision process, in which a cons umer selects a group of favorable options to form a consideration set, and then makes a final purchase decision based on evaluations among the 58 options in the consideration set. The size of the consideration set refers to the number of brands 7 existing in a In general, the size of a consideration set captures the level of brand competition intensity will be less focused on any par ticular brand. Analytically, if we assume fixed total dedication of information collection and evaluation per purchase, the effort and time a customer can dedicate will be less per option. This may lead to a lack of sufficient information to realize the pu rchase, on average. Nevertheless, if a customer purposely maintains the same level of dedication on each option, he or she would have to collect an extensive load of information in total. Branco , Sun and Villas - Boas (2015) find that when too much informati on is provided, the average results in a smaller likelihood of purchasing the target option, regardless of which decision - making strategy (fixed dedication in t otal vs. per option) the customer adopts. H2: Consideration set size has a negative effect on the probability of (a) customer acquisition and (b) customer retention de cisions, all else being equal. Although consi deration set size negatively affects the p acquisition and retention decision, we propose that a larger consideration set sizes facilitates the positive effect of customer - relationships asset on both acquisition and retention decisions. Compared to other brand - level marketing metrics, e.g., brand equity, customer - relationships asset is less observed. Consideration set size can function as a memory - based cue for a customer - relationships asset to make decisions. 7 We select brand varie ty, instead of product variety, to measure , since t his study purports to compare brand - level competitive advantage. 59 In addit ion, assessing customer - relationships asset may need analytical reasoning and a comparative evaluation in the decision - making process. Previous research has shown that a larger consideration set stimulates the comparative evaluation pro cess (Olsen 2002; He et al. 2016 ) and comparative processing enhances the fluency of relative comparisons, and consequently purchase intention ( Newman, Howlett, and Burton 2015). Moreover, Oakely et al. (2007 ) find that the existence of salient competition may lead to polarized evaluations, e.g., a good one is perceived to be better and a bad one is perceived to be worse. This finding implies that comparative processing eases the decision task by adding the fa vorableness of the focal - relationships asset is more appreciated when consumers have a larger consideration set size. H3: Consideration set size strengthens the positiv - relationships asset on customer (a) acquisition and (b) retention decisions, all else being equal. The Roles of Customer Information Search: Objective versus Interactive Customers engage in inf ormation search in various way s for both acquisition and retention decisions. Searching for information plays an important role in purchase decisions to reduce perceived risk and uncertainty, compare alternatives, and mitigate opportunity costs. Compared to the dissemination of markete r - initiated information, consumer information search is stronger in predicting the probability of purchase, because it is self - engaged and consumers mostly initiate a search when they are genuinely interested in a product. Moreover, the information collect a purchase task, since the consumer can select what aspects of information are needed to fill their 60 knowledge gap. Thus, we expect that consumer information search positively rel ates to acquisition and retention decisions in general. The types of information search matter. Consumer information search activities include proactively collecting product - related information from media, e.g., magazines, websites, third parties, or tal king with family or friends. Following the marketing literature, we further classify information search activities into objective information search and interactive information - base d information, - party reviews, and referencing consumer reports. Such information usually contains informative details on attributes, well - established evaluation criteria, an d straightforward presentation. Interactive customer search also refers to interpersonal information exchange ( Lampert and Rosenberg 1975; Gilly et al. 1998 ), such as receiving product information by talking with others. This type of search makes the conte nt more subjective and personalized. Compared to objective information search, the moderating effect of interactive information search is more prevalent in both acquisition and retention decisions. acquisition or adoption behaviors (e.g., Choi, Bell, and Lodish 2010; Nam, Manchanda, and Chintaguta 2010; Libai, Muller, and Peres 2013; Wangeheim and Bayon 2007) and in retention or switching costs (e.g. Jones et al. 2007). The information content from interactive search can be customized and self - relevant. Also, interactive information search requires a higher level of engagement and hedonic experience. Regulatory engage ment theory suggests that hedonic experiences play an important role in shaping perceived value as a force of attraction to or repulsion from the target (Higgins, 2006; Higgins and Scholer, 2009). Thus, we anticipate that 61 interactive information search inc reases the attractiveness of the brand for both new and existing customers. We expect that objective information search is more valuable for acquisition, but not for retention decisions. Customers have a higher level of uncertainty in adopting a new brand . Objective information search involves standard and credible information, which is particularly valuable for new customers. However, for existing customers, broad, fact - based information does not add value in updating their beliefs, because brand percepti on is long lasting. Customers are less likely to search for brand information from objective sources but may participate in an objective search only for some fairly specific product information. For example, customers may be interested in some particular f eatures that are attractive and new on the market. Thus, objective information is more accessible, credible, and thus valuable. Given that customer - relationships asset is a brand - level characteristic, objective information search does not necessarily enhan ce its effect when a customer is making a retention decision. H4 a , b : Interactive information search strengthens the effect of customer - relationships asset on (a) acquisition and (b) retention decisions, all else being equal. H4 c: O bjective information s earch strengthens the effect of customer - relationships asset on acquisition decision, all else being equal. EMPIRICAL EXAMINATION Data and Measures of Key Variables We use the parameter estimates from Study I (e.g., CIRFs) as the measure of customer - relationships asset for each brand and add customer - level data from a combination of survey and purchase data. The purchase data are provided by six contracted panel compa nies that have 62 access to sales data at dealers. To match the data from Study I, Study II uses the first quarter of 2012, which follows the time window of Study I (e.g., from August 2006 to March 2012). The survey data include demographic information on re spondents, current usage, information search activities, and an indication of whether the evaluated car model is in the consideration set. Specifically, in addition to demographic information, respondents were asked what car they were driving before the pu rchase, whether they considered purchasing a set of car models, and what search activities they were engaged in during the purchase decision. Each respondent was asked up to 25 car models. We merge the CIRFs results of Study I as the measures of customer - r elationships asset across 33 brands with the purchase data. The retention sample contains respondents who purchased the same brand as the cars that they were replacing, and the acquisition sample includes those who purchased different brands between replac ed and replacing cars. The acquisition sample has 8,192 respondents and 6,705 (8.06%) purchase s out of 83,184 observations; the retention sample has 3,086 respondents and 2,633 (8.08%) purchases out of 32,567 observations. Table 2. 4 reports the descriptive statistics. Consideration set size is measured by the count of unique brands a respondent considers consideration stage. Interactive search is measured by an indicator d etermining whether a respondent has discussed the focal car model with friends or family. Objective search is the sum of a set of indicators that determine whether a respondent has engaged in: (i) a search for information online at third - party automotive s ites, e.g., cars.com, carbuyer.com. Edmunds.com; 63 and household incom e). Marketing mix variables include marketing reach, price, product variety, and te st drive as control variables. Model Specification This data contains a combination of customer - level variables (e.g., demographics and consideration set size) and product - level responses. We use the mixed - effects probit models and allow random effects on the intercepts to account for customer heterogeneity in acquisition and retention samples. In addition, because marketing reach is also a source of product informat ion, we include the interaction between customer - relationships asset and marketing reach in the model as well. The probabilities of customer acquisition and retention are modeled with binomial probit models. Each customer k, indicates the purchase of the c ar model i of brand j, at month t. The utility formulation is specified as follows: + + + where CR _ A sset stands for customer - relationships asset, SetSize represents consideration set size, Search_interactive is interactive information search, and Search_objective is objective information search. To remove all the other brand - level characteristics, we control brand dummies in Equation 2.2. We also control customer demographics, such as age, g ender, and household income, and month dummies for market fluctuation. We test the mixed - effects probit model, using Equation 2.2 as the utility function, for both acquisition and retention samples. 64 Addressing the Endogeneity of Consideration Set Size Th e marketing literature has suggested that consideration set size is an endogenous variable (e.g., Wu and Rangaswamy 2003). It is challenging to find an exogenous instrument variable that is a relevant consideration set size but not purchase decisions. Thus , we use the copulas approach to correct the endogeneity of consideration set size (Park and Gupta 2012). The copulas approach directly models the correlation between the endogenous variable and the error term. It is an instrument - free approach and imposes a strict assumption that the endogenous variable should be non - normally distributed. Practically, consideration set size is a count variable that is not expected to be normally distributed. Statistically, the Shapiro - Wilk tests show that consideration set size is non - normally distributed in either acquisition (W=.995, p - value<.001) or retention samples (W=.988, p - value<.001). We then insert the copulas repressors of consideration set size for both samples: , where is the cumulative distribution function of consideration set size. Empirical Findings Table 2. 5a reports the parameter estimates and overall model statistics of the mixed - effects probit models for both acquisition and retention samples. In both samples, we find significant positive main effects of customer - relationships asset, where the parameter e stimates are 128.433 (p - value<.001) for acquisition and 136.651 (p - value<.001) for retention decisions. H1a and 1b are supported. Note that we control all the other time - constant brand characteristics by brand dummies and market fluc tuation by month dummies. These 65 significant parameters demonstrate that customer - relationships asset is a brand - level advantage that leads to both customer acquisition and retention decisions. As expected, consideration set size has negative effects for b oth acquisition ( - .040, p - value < .05) and retention decisions ( - .078, p - value < .001). Thus, regardless of acquisition or retention, larger consideration set size increases brand competition in pur chas e probabilities. H2a and 2b are supported. We find a significant interaction effect between customer - relationships asset and consideration set size in the acquisition sample ( = .074, p - value<.05), while we do not find such significant interac tion effect in the retention sample. H3a is supported, while 3b is not. Existing customers might already sense customer - relationships asset when they adopted the brand in the first place. Given the long - term orientation of customer - relationships asset, the re might not be noticeable updates for existing customers, although larger consideration set size drives them into an analytical process. Thus, the positive moderation effect of consideration set size only holds for customer acquisition but not retention. Our results show that all types of information, e.g., marketer initiated ( .210, p - value < .001; .188, p - value < .001), interactive information search ( = .734, p - value < .001; .641, p - value < .001), and objective information search ( .259, p - value < .001; .184, p - value < .001) positively affect to purchase decisions. However, they selectively moderate the effect of customer - relationships asset. Specifically, in the acquisition sample, both inte ractive ( = 3.155, p - value < .05) and objective ( = 2.048, p - value < .05) information search strengthen the positive influence of customer - relationships asset. In the retention sample, only interactive information search = 4.48 1, p - value < .05) interacts significantly with customer - relationships asset, while marketing reach 66 marginally ( = 1.482, p - value < .10) interacts with customer - relationships asset. These results show that interactive information search is more pre valent in strengthening the effect of customer - relationships asset in both acquisition and retention decisions, while objective information search only strengthens customer - relationships asset in acquisition but not retention decisions. Thus, H4a, 4b, and 4c are supported. To validate, we use a modified control function approach to correct the endogeneity of consideration set size, in which we standardize the residuals from the first stage, without using any instrument, based on the variance from a Poisson distribution and then insert the scaled term into Equation 2. 2. We find fairly consistent results (Table 2. 5b) and statistical evidence that consideration set size is endogenous. DISCUSSION Theoretical Implications Customers are most critical assets. The notion that customers, and more accurately, customer relationships are assets , has been discussed for a few decades ever since Srivastava, Shervani, and Fahey (1998) formally stated that customers are a must be cultivated and leveraged . The m arketing literature has subsequently recognized that customers are intangible, accountable, forward looking, and capitalization. In this regard , customer equity, as the primary quantification approach o f customers as assets, strategically valuate s alternative customer valuation approach that emphasizes customer relationships and value creation through customer - customer and customer - firm interactions. We defi ne the focal concept customer - relationships asset as a - customer and customer - firm interactions to achieve organic growth in market performance. 67 Customer - Relationships Asset as a Customer - Based Metric . Customer - relationships asset provides an alternative perspective distinctive from customer equity, but also related to customer equity. First, customer equity directly me - relationships asset measures the indirect creation of value from customer - customer and customer - firm interactions. While customer equity quantifies the direct dyadic relationship between a c ustomer and the firm, customer - relationships asset captures the net contribution of a customer who interacts with others and whose feedback is internalized to achieve organic growth in market performance. Second, this study also offers a quantification app roach that can account for the dynamic reinforcement between customer base and market performance over time. The empirical findings suggest brands can have positive, negative, and absent customer - relationships asset. This is intriguing, because a firm can have "good" or "bad" customers who engage in extra - role ive market performance. Due to the specification of its formula, customer equity only assigns positive values to customers, completely overlook ing The literature has shown that firing bad customers through customized high pricing is profitable (Shin, Sudhir, and Yoon 2012). From a different angle, based on the indirect influences driven by customer relatio nships, we also show the potential downside of maintaining bad customers. To maximize profits, leveraging the trade - potential indirect harm can be a fruitful avenue for future studies on strategic resource allocation. Customer - relationships asset and customer equity share some unique characteristics and are expected to be interrelated. Both customer - relationships asset and customer equity are long - term oriented, where the former accoun t er for 68 Though customer - relationships asset focuses on the indirect value creation, its linkage with market performance suggests financial implications. C ustomer - relationships asset an d customer equity capture t he indirect and direct customer contribution to a firm's future cash flow, respectively. Thus, customer - relationships asset is expected to share some merits that the customer equity metric has, e.g. , accountable, forward looking, and serving as a proxy for In addition, while the relationship value captured by customer equity can be reflected by profit margins and retention rate, the relationship value captured by customer - relationships a sset can be shown as favorable customer interactions and extra roles toward the focal firm. Overall, a customer should have relationship value in both aspects. It is reasonable to assume that a good quality customer who has higher purchase frequency and/or retention probability is more likely engaged in extra - role activities. Schweidel, Park, and Jamal (2014) find that customers are more likely to simultaneously engage in contributive activities (cross sub - brands purchasing and participating in user - generated contents) rather than engage in one activity and stay inactive in oth ers. Thus, customer equity and customer - relationships asset are expected to be intertwined . To conclude, customer equity and customer - relationships asset are two distinctive and inter - related customer valuation approach es and customer - based metrics. Customer - Relationships Asset as a Competitive Advantage . We propose that customer - relationships asset is competitive advantage, because this concept incorporates non - substitutable resources (e.g. , customers) and marketing deployment (e.g. , value c reation through customer - customer and customer - firm interactions). Study II shows that customer - relationships asset affects individual acquisition and retention, controlling for all the other brand - level characteristics. Empirically, we demonstrate that cu stomer - relationships asset is 69 competitive advantage. To this end, most studies on customer acquisition and retention decision s focus mainly on marketing tactics, such as acquisition mode, channel selection, loyalty and referral reward program s , promotions, and free shipping , etc. (e.g. , Anderson and Simester 2004; Datta, Foubert, and van Heerde 2015; Garnefeld et al. 2013; Iyengar et al. ; 2011 Lewis 2006; Schmitt, Skiera, and Van den Bulte 2011; Verhoef 2003; Verhoef and Donkers 2005; Voss and Voss 2008). T o the best of our knowledge, this essay is the first to empirically examine how a marketing metric or brand - level advantage affects customer - level acquisition and retention decisions . That said, i t is noteworthy to understand how inter - firm strategic resources affect customer - level decisions to understand the underpinning (Stahl et al. 2012). Customer - Relationships Asset a nd Customer Base Analyses . At the micro level, customer base valuation is directly and strategically relevan t to a fi valuation focus of this study complements the CLV - extra - role behaviors, which may facilitate addressing managerial flexibility and consumer learning (Lewis 2005). Moreover, Schweidel, salient limitations of ext a nt research in customer base analysis is that it primarily focuses on a single type of transactional activity , - activity analytical approach has proved t - transaction al engagement. Shah et al. (2014 ) also extend customer valuation by integrating customer repetitive habit - based behavior, including purchase s , promotion, return s , and low - margin habit. Our study extends the current transactional - focused customer base analyses by showing that customers also generate value from non - transaction al activities. Although we conduct the analyses at the brand level, marketing scholars and practitioners can easily use our model set - ups for customer base analyses across segments within a firm. 70 Managerial Implications T he motivation for extra - role activities may be associate d with significant marketing efforts, e.g. , loyalty, referral program s , community club s , etc. It is necessary for firms to weigh the benefits and costs that encourage customers to engage in these extra - role activities. A firm can conduct experiments to find the best suitable marketing actions as an empowe r the reinforcement of good quality customers and, consequently, increase customer - relationships asset. In addition, to monitor customers and their profitability, firms should adopt customer - based KPI s (key performance indices) that indicate network effects and customer - firm interactions . This expands the database marketing that not only transactional data, the non - transactional data based on any form of touch points should be tracked and analyzed to predict market performance. Linking the level of ma rketing investments and types of marketing actions to customer - based KPIs directly assesses the ROI (return on investment) of marketing tactics, facilitating the maximization of profitability in business operations. Moreover, our findings show that negati ve customer - relationships asset does exist. This may be caused by any kind of failures in marketing offerings or touch points. Active c ustomer recoveries might be the most effective way to deal with these failures , e.g. understand customer problems and cus tomize the solutions. Inactivat ing the negative extra - role behaviors (e.g. , speak ing ill of the company to other customers or remaining inert in the face of costly marketing actions) of bad quality customers is fairly c hallenging. Research shows that facin g and facilitating complaints from dissatisfied customers can drastically reduce their negative word - of - mouth behavior (Nyer and Gopinath 2005). An efficient c ustomer service team is able not only to provide solutions for failure but also to relieve the po tential adverse effects of customers' unhappiness and their ongoing dissatis faction . 71 Furthermore, our valuation approach might be sensitive to product category, business model , and market conditions . Because of the scope of this study, we use the automot ive industry, which is characterized by less frequent purchase s (discrete purchase s ), the non - contractual nature of the industry , and the fact that most brands are relatively mature. For products that enjoy a higher purchase frequency, e.g. , apparel, assessing economic value and relationship - driven value are both important. Moreover, our approach might be sensitive to the time window and, in other words, dynamic fluctuation of base, especially when the business mod el is contractual, e.g. , subscription - based , which (McC arthy, Fader, and Hardie 2016) as the primary f inancial capability for implement ing marketing actions. A shorter time window for the assessment might be more desirable. Another possible extension is that, similar to CLV - based customer equity models, careful timely recalibrations are needed when the business is going through radical innovations (Zhang 2016). T he empirical findings of S tudy II also provide strat egic guidance on acquisition and retention strategies as well as customer decision journeys. Particularly, consideration set was traditionally the start ing point of the classic purchase funnel and customer decision journey. In McKinsey & Company (2009) r eport, automobile shoppers added, on average, 2.2 brands into their initial c onsideration set of 3.8 brands. This is fairly consistent with our sample that the average consideration set size of a customer is around 6. Although the automobile industry is tr aditionally thought of as a high - invo lvement and high - loyalty industry, b rand managers can no longer take loyal customers for granted. C ompeting brand interruptions may be unexpectedly strong and cause customer s to switch brands . In the 2015 report , he new consumer decision journey McKinsey & Company advocate s that companies that are able to compress the 72 consideration and evaluation phases, and even eliminate them, make the purchase journeys becom e . Consideration set size seems to be a peril for brand managers. Study II finds that consideration set size increases acquisition probability if a brand possess es positive customer - relationships asset, while this effect does not hold in customer retention de cisions. When the goal is to explore and recruit new customers, motivating a larger consideration set for customers who are currently served by competing brand s is the key. For example, in the automobile industry, many third part ies, such as Edmunds , provi de comparisons of twinned vehicles, e.g oth are supposedly able to satisfy similar customer core needs and preferences. Man a gers could utilize this opportunity to directly compete with the target - competing model . Brands wit h positive customer - relationships asset are particularly expected to outperform in this battle. To do this, m anagers can promote a variety of added - on features of the twinned car model , potentially stimulating the audience to enlarge the size of considerat ion set size and be ing amenable to accepting the focal brand. Our findings also suggest that the source of information, company - initiated, customer - initiated, or customer interactions selectively enhance s - relationships asset. Customer i nteractive search seems to be more prevalent for both acquisition s and retention s . For brands that own positive customer - relationships asset, managers should initiate and help with the platforms to enhance customer connections, e.g. , brand community, event. Managers should be creative in connect ing customers. One successful example is that mobile game companies reward extra free lives to motivate users to team up and stay connected . Although, less prevalent, c ustomer - initiated search for objective information is also beneficial for enhanc ing the positive effect of customer - relationships asset, especially to acquire new customers . Most objectiv e 73 information nowadays is from website and social media partners. Not only because the search cost is low, but because search can be fun and involve hedonic experiences . The skills to operate and manage SEO, Google search advertising, and third - part social sites enable a firm to succeed in customer acquisition and excel in the dynamic market place . 74 APPENDICES 75 Appendix A: Tables Table 1.1 : A Review of Key Dimensions in the Product Design Literature Representative Empirical Study Functional Physical Experiential Interplay among Design Aspects Interacting with Marketing Communication Interacting with Customer Characteristics Outcomes Aesthetic (Form) Atypical Chitturi, Raghunathan, and Mahajan (2007) yes yes yes yes Product preference Chitturi, Raghunathan, and Mahajan (2008) yes yes yes WOM recommendation and repurchase intention Luo, Kannan, and Ratchford (2008) yes yes yes Purchase intention Patrick and Hagtvedt (2011) yes Willingness to buy more & return Townsend and Sood (2012) yes yes Product choice Landwehr, Wentzel, and Herrmann (2013) yes yes Aesthetics preference and sales Brakus, Schmitt, and Zhang (2014) yes yes Product evaluation Rubera (2015) yes yes Sales & sales growth Homburg, Schwemmle, and Kuehnl (2015) yes yes Purchase intention & Relational WOM Mishra (2016) yes yes yes Customer - based brand equity Jindal et al. (2016) yes yes yes yes Market share Liu et al. (2017) yes yes yes Purchase decision Wu et al. (2017) yes Usage & emotion This study yes yes yes yes yes yes Purchase conversion & customer - brand relationship cultivation 76 Table 1.2 : Customer Demographics Demographics Description Percentage Gender Male 56.39% Age 18 - 30 16.58% 30 - 40 21.26% 40 - 50 21.21% 50 - 60 23.83% 60 - 70 14.28% >70 2.11% Marital Status Single 17.85% Living with significant other 9.41% Married 62.54% Separated/divorced, but not remarried 8.41% widowed 1.79% Household Income <$25,000 4.53% $25,001 - $50,000 15.13% $50,001 - $75,000 16.01% $75,001 - $100,000 14.24% $100,001 - $125,000 9.25% $125,001 - $150,000 7.30% $150,001 - $175,000 7.11% $175,001 - $200,000 5.55% $200,001 - $225,000 3.22% $225,001 - $250,000 2.09% [250,300] >$250,000 6.50% Prefer not to answer 9.07% 77 Table 1.3 : Variables, Operationalization, and Data Sources Variable Operationalization Data Source Main variables of interest Experiential design experiential design, which include usability, usage pleasant, and social value dimensions, of the new purchased car i: ,where k stands for imagery experience indicators. Design & Imagery Study Functional design , relative to the replaced one r : , where a stands for 5 functional indicators. Design & Imagery Study Aesthetic design , relative to the replaced one r : , where b stands for 6 aesthetic indicators. Des ign & Imagery Study Marketing communication ambiguity target car i across different customers. It is measured by binary entropy function, where n is the number that customers select a particular imagery experience indicator k and N is the total number of customers that evaluates this car model i. Design & Imagery Study Customer imagery subjectivity The ove imagery experience indicator k from the probabilit y that this indicator is chosen by all customers: . Design & Imagery Study Purchase conversion 1 if a customer purchase car i or 0 otherwise, given that the customer has reported that he/she considered purchasi ng this car less than a year ago. Firm Clinic Data Customer - brand relationship The average score of three items that measure the overall customer - Design & Imagery Study Control variables Customer demographics Gender, age, marital status and household income. Design & Imagery Study Marketing mix variables Product price refers to the base price for the focal car model. cars.usnews.com Product variety refers to package range of the focal car model. It is measured by the relative price range that offers by the manufacture: . cars.usnews.com Marketing effort is measured by the sum of four indicators, whether a customer: 1) received mail from manufacture or dealer, 2) heard or read TV/radio/newspaper article, 3) saw or heard any advertising, and 4) saw or read any local newspaper advertising. Design & Imagery Study Time - varying brand effects Relative customer - based brand equity: brand knowledge, differentiation, esteem, and relevance with regards to the four pillars of the industry average: Young & Rubicam 78 Table 1.4 : Descriptive Statistics and Correlations Variables Mean S.D. (1) (2) (3) (4) (5) (6) (7) (8) 1. Experiential design 2.90 3.21 2. Functional design - 1.17 1.95 .40*** 3. Aesthetic design - .90 1.91 .37*** .67*** 4. MKT communication ambiguity 3.81 1.34 .35*** .19*** .10*** 5. Customer imagery subjectivity 1.99 1.91 .89*** .39*** .36*** .46*** 6. Age 45.03 13.88 .07*** - .03*** - .00 .08*** .07*** 7. Income 12.51 25.79 - .01 .01** .01** .04*** - .00 .11*** 8. Price 33309.44 17054.81 .01** .09*** - .01 .29*** .14*** .10*** .05*** 9. Marketing effort .14 .57 .04*** .04*** .03*** .09*** .07*** .08*** - .00 - .08*** 10. Product variety 38.08 29.78 .08*** - .03*** - .00 .18*** .08*** - .01** - .02*** .10*** 11. Relative brand differentiation - 5.24 24.26 .09*** .12*** .05*** .27*** .10* .01 .03*** .21*** 12. Relative brand relevance 62.26 75.51 .05*** .01** .05*** .27*** .10*** - .04*** - .04*** - .26*** 13. Relative brand esteem 22.15 28.69 .14*** .10*** .07*** .45*** .19*** .00 - .00 .02*** 14. Relative brand knowledge 12.90 15.06 .03*** - .01*** .04*** .16*** .06*** - .03*** .04*** - .18*** 15. Brand knowledge engagement 7.64 2.49 .04*** - .11*** - .14*** - .02** .01* - .02** - .03*** .05*** 16. Objective product evaluation 1.55 .98 .32*** .18*** .11*** .83*** .37*** .03*** .02** .09*** 17. Customer - brand relationship 8.38 1.86 .28*** .04** .01 - .10*** .23*** .01 .02* .09*** Notes: Correlations are reported listwise. *** indicates p - value < .001; ** indicates p - value < .05; * indicates p - value < .10. N=34,493 for correlations among variable 1 - 16 and N=6,244 for correlation with variable 17. Variables Mean S.D. (9) (10) (11) (12) (13) (14) (15) (16) 10. Product variety 38.08 29.78 .08*** 11. Relative brand differentiation - 5.24 24.26 .05*** - .16*** 12. Relative brand relevance 62.26 75.51 .21*** .20*** - .03*** 13. Relative brand esteem 22.15 28.69 .19*** .10*** .38*** .79*** 14. Relative brand knowledge 12.90 15.06 .18*** .23*** - .05*** .90*** .77*** 15. Brand knowledge engagement 7.64 2.49 - .02*** - .00 - .01** - .09*** - .04** - .05*** 16. Objective product evaluation 1.55 .98 .11*** .32*** .23*** .41*** .51*** .33*** - .04*** 17. Customer - brand relationship 8.38 1.86 - .02 - .00 .10*** .02 .08*** .03** .68*** .07*** Notes: Correlations are reported listwise. *** indicates p - value < .001; ** indicates p - value < .05; * indicates p - value < .10. N=34,493 for correlations among variable 1 - 16 and N=6,244 for correlation with variable 17. 79 Table 1.5 : The Results for Control Function Approach to Correct Endogeneity of Experiential Design Coefficient Robust Std. Err. Functional design .069 *** .012 Aesthetic design .040 ** .015 MKT communication ambiguity - .693 *** .049 Customer imagery subjectivity 3.920 *** .178 Age .003 .002 Gender - .001 .016 Marital Status - .003 .009 Income - .001 ** .000 Price .000 .000 Marketing Effort - .014 .017 Product variety - .001 .001 Relative brand differentiation .000 .001 Relative brand relevance - .000 .000 Relative brand esteem - .000 .001 Relative brand knowledge - .003 .003 Quarter2_dummy - .009 .026 Quarter3_dummy - .022 .028 Quarter4_dummy .039 .035 Brand dummies (brand2 - brand37) a Instrument Variables: Brand knowledge engagement .047 *** .005 Objective product evaluation .754 *** .071 Constant .108 .202 Notes: *** indicates p - value < .001; ** indicates p - value < .05; * indicates p - value < .10. N=34,493 . R - Square=82.33%. a F or the sake of parsimony, we do not report param eter estimates of brand dummies Individually . The results indicate that, except for o ne brand, all brand dummies are 80 Table 1.6 : T he Effect of Experiential Design on Purchase Conversion Model I Threshold Not Exceeded Model II Threshold Exceeded (Functional + Aesthetic Design > 2) Model II Threshold Exceeded (Functional + Aesthetic Design > 2) Coefficient Robust Std. Err. t - value Coefficient Robust Std. Err. t - value Coefficient Robust Std. Err. t - value Experiential design - .273 .214 - 1.28 .373 ** ** .058 6.39 .429 *** .056 7.72 Experiential design_square - .015 ** .007 - 2.04 Experiential design_cubic * .001 ** .000 2.06 Functional design .138 *** .030 4.61 .085 *** .030 2.83 .083 ** .030 2.76 Aesthetic design .023 * .012 1.95 . 034 *** .015 2.33 .034 ** .015 2.35 MKT communication ambiguity .048 .108 .44 .060 .191 .30 .058 .198 .29 Customer i magery subjectivity 1.445 * .850 1.70 - 1.213 **** .228 - 5.31 - 1.221 *** .228 - 5.35 Age .003 * .002 1.69 .004 * .003 1.32 .004 .003 1.33 Gender .030 .030 .99 .059 ** .032 1.87 .059 * .032 1.84 Marital Status - .003 .018 - .17 - .022 ** .013 - 1.66 - .021 .013 - 1.36 Income - .000 .001 - .20 .002 * ** .001 3.17 .002 ** .001 3.21 Price - .000 .000 - 1.19 - .000 * .000 - 1.47 - .000 .000 - 1.48 Marketing effort .141 *** .038 3.76 .071 * *** .020 3.48 .072 *** .020 3.50 Product variety - .003 * .002 - 1.83 - .004 ** .001 - 2.61 - .004 ** .001 - 2.65 Relative brand differentiation .002 .001 1.40 - .004 * .003 - .004 .003 - 1.57 Relative brand relevance - .001 .001 - 1.07 .000 .001 .000 .001 - .29 Relative brand esteem .002 .002 .80 .002 .002 .002 .002 .69 Relative brand knowledge - .001 .003 - .40 .010 ** .006 .010 * .006 1.73 Quarter2_dummy .012 .043 .28 - .041 .036 - 1.08 - .039 .036 - 1.08 Quarter3_dummy - .036 .045 - .80 - .040 * .024 - 1.61 - .038 .024 - 1.61 Quarter4_dummy .032 .050 .65 - .116 *** .045 - 2.65 - .119 ** .045 - 2.65 Brand dummies a Residual for endog eneity correction .429 * .219 1.96 - .265 * *** .056 - 4.77 - .264 *** .003 - 4.77 Constant - 2.105 *** .268 - 7.84 - .919 * 1.090 - .84 - .934 1.093 - .85 Insigma2: Brand ID .006 .007 .82 - .002 .009 - .25 - .002 .008 - .23 Log Pseudolikelihood= - 6,613.966 Sample size=23,719 Iteration=8 Log Pseudolikelihood= - 5,442.976 Sample size=10,774 Iteration=7 Log Pseudolikelihood= - 5,441.442 Sample size=10,774 Iteration=7 Notes: *** indicates p - value < .001; ** indicates p - value < .05; * indicates p - value < .10. a F or the sake of parsimony, we do not report parameter estimates of brand dummies individually. 81 Table 1.7 : The Effect of Experiential Design on Customer - Brand Relationship Cultivation Outcome Model of Purchase Conversion Group Outcome Model of Non - Purchase Conversion Group Treatment Model Coefficient Robust Std. Err. t - value Coefficient Robust Std. Err. t - value Coefficient Robust Std. Err. t - value Purchase Conversion a ATET (Purchase Converted vs Not) .123 ** .052 2.39 (.123 ** .050 2.48) Experiential design .158 ** * .003 6.99 .198 *** .019 10.31 .065 *** .011 6.01 Functional design .020 .058 .34 - .151 * .082 - 1.85 .123 *** .021 5.90 Aesthetic design - .045 .053 - .86 - .027 .050 - .33 .018 .022 .82 MKT communication ambiguity .123 ** .050 2.44 .076 .032 1.44 - .280 *** .022 - 12.84 Customer imagery subjectivity .056 .035 1.59 - .068 ** .002 - 2.10 - .014 .018 - .76 Age - .000 .003 - .13 - .001 .061 - .55 .011 *** .001 8.43 Gender .149 * .079 1.88 .371 * ** .036 6.21 .218 *** .034 6.32 Marital Status - .035 .045 - .77 - .016 .001 - .44 - .053 ** .020 - 2.63 Income .002 .001 - .92 .001 .001 1.21 .001 ** .000 2.61 Price .000 ** .000 2.90 - .000 .000 1.22 - .000 *** .000 - 6.55 Marketing effort - .336 ** .114 - 2.94 .069 * .041 1.70 .254 *** .032 7.90 Product variety - .003 ** .001 - 2.30 - .000 .001 - .26 .003 ** * .001 4.25 Relative brand differentiation .005 * .003 1.87 .006 ** .002 2.85 .004 ** .001 3.20 Relative brand relevance .000 .002 - .20 - .001 .001 - .91 .000 .001 - 1.57 Relative brand esteem .006 .004 1.48 - .002 .003 - .54 - .004 ** .002 .25 Relative brand knowledge .006 .008 .79 .012 ** .005 2.14 - .002 .003 - 2.43 Constant 7.406 *** .287 25.84 7.287 ** * .296 24.61 - 1.934 *** .129 - 15.02 Notes: *** indicates p - value < .001; ** indicates p - value < .05; * indicates p - value < .10. a The first line reports bootstrapping estimates (1000 draws). 82 Table 2.1 : Representative Studies on Customer Interactions and Value Creation Studies Empirical C - C Interactions C - F Interactions Value Creation Proposition Schmitt, Skiera, and Van den Bulte (2011) Incentive referral Direct effect on prospective customer base and market performance; Indirect effect by enhancing marketing effort effectiveness. Garnefeld et al. (2013) Incentive referral Yang et al. (2012) WOM referral Villanueva, Yoo, and Hanssens (2008) WOM referral Trusov, Bucklin and Pauwels (2009) WOM referral Baker, Donthu, and Kumar (2016) General WOM Bansal and Voyer (2000) General WOM Berger, Sorensen, and Rasmussen (2010) General WOM East, Hammond, and Lomax (2008) General WOM Rosario et al. (2016) General WOM Adjei, Noble, and Noble (2010) Brand community Thompson and Shiha (2008) Brand community Carlson, Suter, and Brown (2008) Brand community Schau, Muniz, and Arnould (2009) Brand community Brand community Stage1: Improve customer knowledge and adaptive capability of resource allocation; Stage 2: Employ the improvements in stage one to achieve superior market performance. Kumar et al. (2010) Social influence and referral Customer knowledge development Joshi and Sharma (2004) Customer knowledge development Vargo and Lusch, (2004) Value co - creation & co - developing Prahalad and Ramaswamy (2004) Value co - creation & co - developing Lengnick - Hall (1996) Customer participation Fang (2008) Customer participation Fang, Palmertier, and Evans (2008) Customer participation Arnold, Fang, and Palmatier (2011) Customer reaction Day (2011) Customer reaction 83 Table 2.2 : Study I. Descriptive Statistics across 33 Brands Series Mean S.D. Min Max Customer B ase 5.295 6.275 .156 31.227 Market Share 3.151 4.132 .004 18.860 Change in Adv. Intensity - .066 1.075 - 10.552 7.936 Customer Profile: Male 2.388 2.946 .023 15.225 Customer Profile: Married 3.349 3.349 .033 20.187 84 Table 2.3 : Study I. Empirical Findings for Brands with Significant Customer - Relationships Assets Brand ID VAR Estimates Impulse Response Function IRF Cumulative Impulse Response Function CIRF R - Square (Share) R - Square (Customer Base) Coefficient a (S.E.) Coefficient (S.E.) Estimate (S.E.) Brands with Positive Customer - Relationships Assets: Brand 7 53.07% 88.63% .108 (.043) .011 (.005) .026 (.011) Brand 8 43.67% 91.42% .115 (.040) .015 (.005) .035 (.014) Brand 9 35.64% 81.80% .343 (.197) .096 (.056) .220 (.125) Brand 11 28.96% 81.24% .213 (.074) .032 (.011) .041 (.017) Brand 18 22.07% 75.01% .209 (.085) .021 (.016) .046 (.025) Brand 22 46.91% 86.62% .188 (.070) .025 (.010) .009 (.006) Brand 31 28.60% 87.97% .031 (.014) .000 (.001) .003 (.003) Brands with Negative Customer - Relationships Assets: Brand 15 49.35% 79.45% - .330 (.080) - .080 (.020) - .168 (.051) Brand 25 32.69% 90.52% - .133 (.076) - .015 (.009) - .005 (.005) Brand 26 42.41% 79.40% - .052 (.029) - .013 (.007) - .039 (.023) Brand 27 71.96% 46.24% - .060 (.023) - .021 (.010) - .088 (.039) - .040 (.023) Brand 28 20.47% 33.27% - .055 (.031) - .017 (.010) - .028 (.017) Brand 37 40.50% 87.98% - .067 (.022) .002 (.002) - .007 (.004) Notes: a T hese parameter estimates represent whether and how much lagged customer base significantly lead s to market share. 85 Table 2.4 : Study II. Descriptive Statistics Acquisition Sample Retention Sample Number of respondents 8,192 3,086 Number of purchases 6,705 2,633 Number of observations 83,184 32,567 Mean SD Mean SD Customer Profiles Purchase probability 9.04% .050 8.85% .053 Age 42.635 14.000 46.524 14.302 Gender (male) 45.80% 46.92% Income <$25,000 6.65% 4.96% $25,001 - $50,000 16.72% 15.07% $50,001 - $75,000 16.53% 16.53% $75,001 - $100,000 19.07% 20.67% $100,001 - $125,000 11.72% 13.35% $125,001 - $150,000 8.72% 8.39% $150,001 - $175,000 5.27% 5.35% $175,001 - $200,000 3.67% 3.76% $200,001 - $225000 2.10% 1.69% $225,001 - $250,000 1.40% 1.13% >$250,000 3.37% 2.43% Consideration Set Size 6.848 2.809 6.288 2.781 Product - Level Descriptive Statistics Interactive search (dummy) 1.72% 2.18% Objective search (range 1 - 3) .042 .281 .049 .314 Marketing rearch (range 1 - 3) .046 .034 .055 .318 Price 34,060.33 19,700.59 31,960.44 13,925.09 Product variety 26.832 26.832 35.583 25.017 Test drive (dummy) 1.20% 1.61% 86 Table 2.5a : Study II. Model Results for Acquisition and Retention (Copulas Endogeneity Correction) Acquisition Retention Coefficient Robust Std. Err. t - value Coefficient Robust Std. Err. t - value Hypotheses Testing: Customer relationships asset (CRA) 128.843 *** 21.413 6.02 136.651 *** 38.355 3.56 Consideration set size - .040 * * .019 - 2.13 - .078 *** .021 - 3.64 CRA x Consideration set size .074 ** .036 2.04 - .032 .064 - .50 Interactive information search .734 *** .043 16.94 .641 ** * .070 9.15 CRA x Interactive information search 3.155 ** 1.000 3.15 4.481 ** 1.862 2.41 Objective information search .259 ** * .023 11.02 .184 ** * .037 4.99 CRA x Objective information search 2.048 ** .633 3.23 - 1.733 1.085 - 1.60 MKT reach .210 ** * .021 9.92 .188 ** * .043 5.49 CRA x MKT reach - 1.012 .743 - 1.36 1.482 * .879 1.69 Controls: Price .000 .000 1.13 .000 .000 1.02 Product variety .007 * ** .000 28.46 .010 *** .000 25.33 Test drive .565 *** .052 10.82 1.200 *** .074 16.11 Age - .001 ** .003 - 2.53 .001 ** .000 2.05 Gender - .016 ** .008 - 2.11 .003 .016 .24 Income - .000 .000 - .51 - .001 ** .000 - 2.49 Month2_dummy .008 .009 .08 - .020 .016 - 1.22 Month3_dummy .008 .009 - .91 - .020 .016 - 1.25 Brand dummies a Consideration set size _Endog1 .015 .055 .28 .118 * .063 1.87 Constant - 1.191 * ** .138 - 8.61 - 1.167 *** . 160 - 7.31 Model Fit Log Pseudolikelihood = - 20,587.404 Wald chi2(45) = 5,315.55 Sample size = 83, 086 Iteration=53 Avg o bs per group=10.4 Log Pseudolikelihood = - 7,294.854 Wald chi2(43) = 2,703.74 Sample size = 31,373 Iteration=42 Avg o bs per group=10.4 Notes: *** indicates p - value < .001; ** indicates p - value < .05; * indicates p - value < .10. a for the sake of parsimony, we do not report parameter estimates of brand dummies individually. 87 Table 2.5b : Study II. Results Validation (Modified Control Function Approach for Endogeneity Correction) Acquisition Retention Coefficient Robust Std. Err. t - value Coefficient Robust Std. Err. t - value Hypotheses Testing: Customer relationships asset (CRA) 129.005 *** 21.383 6.03 136.467 *** 38.339 3.56 Consideration set size - .040 ** * .002 - 22.38 - .041 *** .003 - 14.29 CRA x Consideration set size .079 * * .038 2.07 - .033 .065 - .50 Interactive information search .731 ** * .043 16.84 .654 ** * .071 9.27 CRA x Interactive information search 3.157 ** .998 3.16 4.415 ** 1.854 2.38 Objective information search .259 ** * .023 11.02 .184 ** * .037 4.98 CRA x Objective information search 2.038 * * .630 3.24 - 1.715 1.081 - 1.59 MKT Reach .213 *** .021 10.00 .185 ** * .035 5.35 CRA x MKT Reach - 1.025 .738 - 1.39 1.505 * .876 1.72 Controls: Price .000 .000 1.07 .000 .000 1.08 Product variety .007 * ** .000 28.49 .010 ** * .000 25.41 Test drive .572 * ** .052 10.92 1.191 *** .075 15.86 Age - .001 ** .003 - 2.61 .001 ** .000 2.14 Gender - .016 * * .008 - 2.12 .003 .014 .24 Income - .000 * .000 - .53 - .001 ** .000 - 2.49 Month2_dummy - .000 .009 - .04 - .019 .016 - 1.19 Month3_dummy .008 .009 .385 - .020 .016 - 1.22 Brand dummies a Consideration set size_Endog2 .126 *** .017 7.62 .080 * * .027 2.93 Constant - 1.192 *** .048 - 26.65 - 1.394 ** * .090 - 15.46 Model Fit Log Pseudolikelihood = - 20,591.736 Wald chi2(45) = 5,213.80 Sample size = 83 ,184 Iteration=53 Avg Obs per group=10.4 Log Pseudolikelihood = - 7297.351 Wald chi2(43) = 2,641.28 Sample size = 31,386 Iteration=54 Avg Obs per group=10.4 Notes: *** indicates p - value < .001; ** indicates p - value < .05; * indicates p - value < .10. a F or the sake of parsimony, we do not report parameter estimation of brand dummies individually. 88 Appendix B: Figures Figure 1.1 : Conceptual Framework Notes : The solid lines represent the hypothesized relationships, whereas the dashed lines indicate that experiential design is influenced by functional & aesthetic design, marketing communication ambiguity, and customer imagery subjectivity. The relations hips described by H1b, H2a, H2b, H3a, and H3b are proposed under the condition that functional & aesthetic design achieves the threshold to reveal the positive impact of experiential d esign. 89 Figure 1.2 : The Illustration of Experiential Design and Its Cont ingencies Notes: Functional & aesthetic design indicates dominant effect on purchase conversion, when the level is low (e.g. the sum of functional and aesthetic design is below 2). When the threshold is achieved ( e .g., the sum of functional and aesthetic design is greater than 2), the effect of experienti al design is positive and much stronger (i . e . prominent) than that of functional and aesthetic design; experiential design exhibits an inverse S - Shaped function form. The effect of experiential design reaches the highest when the level of marketing communi cation ambiguity is higher and the level of customer imagery subjectivity is moderate. 90 Figure 1.3a : The Marginal Effects of Experiential Design at Levels of Marketing Communication Ambiguity 0 0.101 0.105 0.109 0.112 0.116 0 1 2 3 4 5 6 The effect of experiential design is not significant (p - value >.05) Marketing Communication Ambiguity The Marginal Effect (Size) of Experiential Design 91 Figure 1.3b : The Marginal Effects of Experiential Design at Levels of Customer Imagery Subjectivity 0.084 0.103 0.107 0.095 0.071 0.042 0.020 0 0.5 1 1.5 2 2.5 3 Customer Imagery Subjectivity The Marginal Effect (Size) of Experiential Design 92 Figure 2.1 : Conceptual Model of Customer - Relationships Asset 93 Figure 2.2 : Modeling Procedure of Study I 94 Figure 2.3a : Study I. Cumulative Impulse Response Functions for Brands with Positive Customer - Relationships Assets 95 Figure 2.3a ( ) 96 Figure 2.3b : Study I. Cumulative Impulse Response Functions for Brands with Negative Customer - Relationships Assets 97 Figure 2.3b ( ) 98 REFERENCES 99 REFERENCES Adjei, Mavis T, Stephanie M Noble, and Charles H Noble (2010), "The Influence of C2C Communications in Online Brand Communities on Customer Purchase Behavior," Journal of the Academy of Marketing Science, 38 (5), 634 - 53. Ahearne, Michael, Chitrabhan B Bhattacharya, and Thomas Gruen (2005), "Antecedents and Consequences of Customer - Company Identification: Expanding the Role of Relationship Marketing," Journal of Applied Psychology, 90 (3), 574. Algesheimer, René, Utpal M Dholakia, and Andreas Herrmann (2005), "The Social Influence of Brand Community: Evi dence from European Car Clubs," Journal of Marketing, 69 (3), 19 - 34. Anderson, Eric T and Duncan I Simester (2004), "Long - Run Effects of Promotion Depth on New versus Established Customers: Three Field Studies," Marketing Science, 23 (1), 4 - 20. Argyriou, E vmorfia (2012), "Consumer Intentions to Revisit Online Retailers: A Mental Imagery Account," Psychology & Marketing, 29 (1), 25 - 35. Armelini, Guillermo and Julián Villanueva (2011), "Adding Social Media to the Marketing Mix," IESE Insight, 9 (9), 29 - 36. Ar nold, Todd J, Eric Er Fang, and Robert W Palmatier (2011), "The Effects of Customer Performance," Journal of the Academy of Marketing Science, 39 (2), 234 - 51. Nilüfer Z and Luca Cian (2014), "Show Me the Product, Show Me the Model: Effect of Picture Type on Attitudes toward Advertising," Journal of Consumer Psychology, 24 (4), 506 - 19. (2016), "The Effect of Electronic Word o f Mouth on Sales: A Meta - Analytic Review of Platform, Product, and Metric Factors," Journal of Marketing Research, 53 (3), 297 - 318. Baker, Andrew M, Naveen Donthu, and V Kumar (2016), "Investigating How Word - of - Mouth Conversations about Brands Influence Purchase and Retransmission Intentions," Journal of Marketing Research, 53 (2), 225 - 39. Bansal, Harvir S and Peter A Voyer (2000), "Word - of - Mouth Processes Within a Services Purchase Decision Context," Journal of Servi ce Research, 3 (2), 166 - 77. Bartel, Caroline A (2001), "Social Comparisons in Boundary - Spanning Work: Effects of Community Outreach on Members' Organizational Identity and Identification," Administrative Science Quarterly, 46 (3), 379 - 413. 100 Bergami, Massimo and Richard P Bagozzi (2000), "Self Categorization, Affective Commitment and Group Self Esteem as Distinct Aspects of Social Identity in the Organization," British Journal of Social Psychology, 39 (4), 555 - 77. Berger, Jonah, Alan T Sorensen, and Scott J R asmussen (2010), "Positive Effects of Negative Publicity: When Negative Reviews Increase Sales," Marketing Science, 29 (5), 815 - 27. Bhattacharya, Chitrabhan B and Sankar Sen (2003), "Consumer - Company Identification: A Relationships with Companies," Journal of Marketing, 67 (2), 76 - 88. Bloch, Peter H (2011), "Product Design and Marketing: Reflections after Fifteen Years," Journal of Product Innovation Management, 28 (3), 378 - 80. ---- (1995), "Seeking the Ideal Form: Pro duct Design and Consumer Response," Journal of Marketing, 16 - 29. Bodapati, Anand V and Aimee Drolet (2005), "A Hybrid Choice Model That Uses Actual and Ordered Attribute Value Information," Journal of Marketing Research, 42 (3), 256 - 65. rnd H Schmitt, and Shi Zhang (2014), "Experiential Product Attributes and Preferences for New Products: The Role of Processing Fluency," Journal of Business Research, 67 (11), 2291 - 98. Branco, Fernando, Monic Sun, and J Miguel Villas - Boas (2015), "Too Much Information? Information Provision and Search Costs," Marketing Science, 35 (4), 605 - 18. Bronnenberg, Bart J., Vijay Mahajan, and Wilfried R. Vanhonacker (2000), "The Emergence of Market Structure in New Repeat - Purchase Categories: The Interplay of Market Share and Retailer Distribution," Journal of Marketing Research, 37(1), 16 - 31. Burnkrant, Robert E and H Rao Unnava (1995), "Effects of Self - Referencing on Persuasion," Journal of Consumer Research, 22 (1), 17 - 26. Carlson, Brad D, Tracy A Suter, and Tom J Brown (2008), "Social versus Psychological Brand Community: The Role of Psychological Sense of Brand Community," Journal of Business Research, 61 (4), 284 - 91. Cavaliere, Giuseppe and Giorgio Tassinari (2001), "Advertising Effect on Primary Demand: A Coint egration Approach," International Journal of Advertising, 20 (3), 319 - 39. Chitturi, Ravindra, Rajagopal Raghunathan, and Vijay Mahajan (2008), "Delight by Design: The Role of Hedonic versus Utilitarian Benefits," Journal of Marketing, 72 (3), 48 - 63. ---- (2007), "Form Versus Function: How the Intensities of Specific Emotions Evoked in Functional versus Hedonic Trade - Offs Mediate Product Preferences," Journal of Mark eting Research, 44 (4), 702 - 14. 101 Choi, Jeonghye, David R Bell, and Leonard M Lodish (2010), " Traditional and IS - Enabled Customer Acquisition for an Internet Retailer: Why New Buyer Acquisition Varies over Geographies and by Method," Knowledge@ Wharton. Datta, Hannes, Bram Foubert, and Harald J Van Heerde (2015), "The Challenge of Retaining Custome rs Acquired with Free Trials," Journal of Marketing Research, 52 (2), 217 - 34. Day, George S (2011), "Closing the Marketing Capabilities Gap," Journal of Marketing, 75 (4), 183 - 95. Deighton, John, Caroline M Henderson, and Scott A Neslin (1994), "The Effect s of Advertising on Brand Switching and Repeat Purchasing," Journal of Marketing Research, 28 - 43. Dekimpe, Marnik G, Dominique M Hanssens, Vincent R Nijs, and Jan Benedict EM Steenkamp (2005), "Measuring Short - and Long - Run Promotional Effectiveness on Sc anner Data Using Persistence Modelling," Applied Stochastic Models in Business and Industry, 21 (4 5), 409 - 16. Desmet, Pieter and Paul Hekkert (2007), "Framework of Product Experience," International Journal of Design, 1 (1). Drolet, Aimee, Itamar Simonson , and Amos Tversky (2000), "Indifference Curves That Travel with the Choice Set," Marketing Letters, 11 (3), 199 - 209. East, Robert, Kathy Hammond, and Wendy Lomax (2008), "Measuring The Impact of Positive and Negative Word of Mouth on Brand Purchase Probab ility," International Journal of Research In Marketing, 25 (3), 215 - 24. Enders, Walter (2008 ), Applied Econometric Time Series: John Wiley & Sons. Epstein, Seymour (1994), "Integration of the Cognitive and the Psychodynamic Unconscious," American Psycholog ist, 49 (8), 709. Evangelidis, Ioannis and Jonathan Levav (2013), "Prominence versus Dominance: How Relationships between Alternatives Drive Decision Strategy and Choice," Journal of Marketing Research, 50 (6), 753 - 66. Fang, Eric (2008), "Customer Particip ation and the Trade - Off between New Product Innovativeness and Speed to Market," Journal of Marketing, 72 (4), 90 - 104. Fang, Eric, Robert W Palmatier, and Kenneth R Evans (2008), "Influence of Customer Participation on Creating and Sharing of New Product V alue," Journal of the Academy of Marketing Science, 36 (3), 322 - 36. Garnefeld, Ina, Andreas Eggert, Sabrina V Helm, and Stephen S Tax (2013), "Growing Existing Customers' Revenue Streams through Customer Referral Programs," Journal of Marketing, 77 (4), 17 - 32. 102 Gilly, Mary C, John L Graham, Mary Finley Wolfinbarger, and Laura J Yale (1998), "A Dyadic Study of Interpersonal Information Search," Journal of the Academy of Marketing Science, 26 (2), 83 - 100. Gupta, Sunil, Donald R Lehmann, and Jennifer Ames Stuar t (2004), "Valuing Customers," Journal of Marketing Research, 41 (1), 7 - 18. Haenlein, Michael (2013), "Social Interactions in Customer Churn Decisions: The Impact of Relationship Directionality," International Journal of Research in Marketing, 30 (3), 236 - 48. Hagtvedt, Henrik (2011), "The Impact of Incomplete Typeface Logos on Perceptions of the Firm," Journal of Marketing, 75 (4), 86 - 93. He, Yi, Qimei Chen, Leona Tam, and Ruby P Lee (2016), "Managing Sub - Branding Affect Transfer: The Role of Consideration Set Size and Brand Loyalty," Marketing Letters, 27 (1), 103 - 13. Higgins, E Tory (2006), "Value from Hedonic Experience and Engagement," Psychological Review, 113 (3), 439 - 60 . Higgins, E Tory and Abigail A Scholer (2009), "Engaging the Consumer: The Science and Art of the Value Creation Process," Journal of Consumer Psychology, 19 (2), 100 - 14. Hoch, Stephen J (2002), "Product Experience Is Seductive," Journal of Consumer Research, 29 (3), 448 - 54. Homburg, Christian, Martin Schwemmle, and Christina Kuehnl (20 15), "New Product Design: Concept, Measurement, and Consequences," Journal of Marketing, 79 (3), 41 - 56. Hu, Ye, Rex Yuxing Du, and Sina Damangir (2014), "Decomposing the Impact of Advertising: Augmenting Sales with Online Search Data," Journal of Marketing Research, 51 (3), 300 - 19. Huston, Ted L, Sylvia Niehuis, and Shanna E Smith (2001), "The Early Marital Roots of Conjugal Distress and Divorce," Current Directions in Psychological Science, 10 (4), 116 - 19. Iyengar, Raghuram, Kamel Jedidi, Skander Essegaier , and Peter J Danaher (2011), "The Impact of Tariff Structure on Customer Retention, Usage, and Profitability of Access Services," Marketing Science, 30 (5), 820 - 36. Jiang, Yuwei, Gerald J Gorn, Maria Galli, and Amitava Chattopadhyay (2015), "Does Your Com pany Have the Right Logo? How and Why Circular - and Angular - Logo Shapes Influence Brand Attribute Judgments," Journal of Consumer Research, 42 (5), 709 - 26. Jindal, Rupinder P, Kumar R Sarangee, Raj Echambadi, and Sangwon Lee (2016), "Designed to Succeed: Dimensions of Product Design and Their Impact on Market Share," Journal of Marketing, 80 (4), 72 - 89. 103 Joshi, Amit and Dominique M Hanssens (2004), "Advertising Spending and Market Capitalization," MSI Report (04 - 110). Joshi, Ashwin W and Sanjay Sharma (2004 ), "Customer Knowledge Development: Antecedents and Impact on New Pr oduct Performance," Journal of M arketing, 68 (4), 47 - 59. Jones, Michael A., Kristy E. Reynolds, David L. Mothersbaugh, and Sharon E. Beatty (2007), "The Positive and Negative Effects of Sw itching Costs on Relational Outcomes," Journal of Service Research, 9(4), 335 - 355. Keller, Kevin Lane (1993), "Conceptualizing, Measuring, and Managing Customer - Based Brand Equity," Journal of Marketing, 57 ( 1 ), 1 - 22. Keller, Kevin Lane and Donald R Lehman n (2003), "How Do Brands Create Value?" Marketing Management, 12 (3), 26 - 31 . Keller, Kevin Lane, Tony Apéria, and Mats Georgson (2008), Strategic Brand Management: A European Perspective: Pearson Education. Keller, Punam Anand and Ann L Mcgill (1994), "Dif ferences in The Relative Influence of Product Attributes under Alternative Processing Conditions: Attribute Importance versus Attribute Ease Of Imagability," Journal of Consumer Psychology, 3 (1), 29 - 49. Kim, Ho and Dominique M Hanssens (2017), "Advertisin g and Word - of - Mouth Effects on Pre - Launch Consumer Interest and Initial Sales of Experience Products," Journal of Interactive Marketing, 37 (2017) , 57 - 74. Kumar, V and Denish Shah (2009), "Expanding the Role of Marketing: From Customer Equity to Market Capi talization," Journal of Marketing, 73 (6), 119 - 36. ---- (2015), Handbook of Research on Customer Equity In Mark eting: Edward Elgar Publishing. Kumar, V and Morris George (2007), "Measuring and Maximizing Customer Equity: A Critical Analysis," Journal of th e Academy of Marketing Science, 35 (2), 157 - 71. Lampert, Shlomo I and Larry J Rosenberg (1975), "Word of Mouth Activity as Information Search: A Reappraisal," Journal of the Academy of Marketing Science, 3 (3 - 4), 337 - 54. Landwehr, Jan R, Aparna A Labroo, a nd Andreas Herrmann (2011), "Gut Liking for The Ordinary: Incorporating Design Fluency Improves Automobile Sales Forecasts," Marketing Science, 30 (3), 416 - 29. Landwehr, Jan R, Daniel Wentzel, and Andreas Herrmann (2013), "Product Design for the Long Run: Consumer Responses to Typical and Atypical Designs at Different Stages of Exposure," Journal of Marketing, 77 (5), 92 - 107. Lapersonne, Eric, Gilles Laurent, and Jean - Jacques Le Goff (1995), "Consideration Sets of Size One: An Empirical Investigation of Aut omobile Purchases," International Journal of Research in Marketing, 12 (1), 55 - 66. 104 Lapersonne, Eric, Gilles Laurent, and Jean - Jacques Le Goff (1995), "Consideration Sets of Size One: An Empirical Investigation of Automobile Purchases," International Journa l of Research in Marketing, 12 (1), 55 - 66. Lemon, Katherine N, Roland T Rust, and Valarie A Zeithaml (2001), "What Drives Customer Equity," Marketing Management, 10 (1), 20. Lengnick - Hall, Cynthia A (1996), "Customer Contributions to Quality: A Different V iew of the Customer - Oriented Firm," Academy of Management Review, 21 (3), 791 - 824. Lewis, Michael (2006), "Customer Acquisition Promotions and Customer Asset Value," Journal of Marketing Research, 43 (2), 195 - 203. ---- (2005), "Incorporating Strategic Cons umer Behavior into Customer Valuation," Journal of Marketing, 69 (4), 230 - 38. ---- (2006), "The Effect of Shipping Fees on Customer Acquisition, Customer Retention, and Purchase Quantities," Journal of Retailing, 8 2 (1), 13 - 23. Libai, Barak, Eitan Muller, and Renana Peres (2013), "Decomposing the Value of Word - of - Mouth Seeding Programs: Acceleration versus Expansion," Journal of Marketing Research, 50 (2), 161 - 76. Liu, Yan, Krista J Li, Haipeng Chen, and Subramanian Balachander (2017), "The Effects of Produ - Mix Effectiveness: The Role of Segment Prototypicality and Brand Consistency," Journal of Marketing, 81 (1), 83 - 102. Lojacono, Gabriella and Gianfranco Zaccai (2004), "The Evolution of the Design - Inspired Enter prise," MIT Sloan Management Review, 45 (3), 75. Lowe, Michael L and Kelly L Haws (2017), "Sounds Big: The Effects of Acoustic Pitch on Product Perceptions," Journal of Marketing Research, 54 (2), 331 - 46. Luo, Lan, PK Kannan, and Brian T Ratchford (2008), "Incorporating Subjective Characteristics in Product Design and Evaluations," Journal of Marketing Research, 45 (2), 182 - 94. Macinnis, Deborah J and Linda L Price (1987), "The Role of Imagery in Information Processing: Review and Extensions," Journal of Co nsumer Research, 13 (4), 473 - 91. McCarthy, Daniel, Peter Fader, and Bruce Hardie (2016), "V (CLV): Examining Variance in Models of Customer Lifetime Value." https://www.mckinsey.com/business - functions/marketing - and - sales/our - insights/the - consumer - decision - journey McKinsey & Company ( 20 https://www.mckinsey.com/business - functions/marketing - and - sales/our - insights/th e - new - consumer - decision - journey 105 https://www.mckinsey.com/business - functions/marketing - and - sales/our - insights/building - a - design - driven - culture Miller, Darryl W and Michael Stoica (2004), "Comparing th e Effects of a Photograph versus Artistic Renditions of A Beach Scene in a Direct - Response Print Ad for a Caribbean Resort Island: A Mental Imagery Perspective," Journal of Vacation Marketing, 10 (1), 11 - 21. Miller, Elizabeth G and Barbara E Kahn (2005), " Shades of Meaning: The Effect of Color and Flavor Names on Consumer Choice," Journal of Consumer Research, 32 (1), 86 - 92. Mishra, Abhishek (2016), "Attribute - Based Design Perceptions and Consumer - Brand Relationship: Role of User Expertise," Journal of Busi ness Research, 69 (12), 5983 - 92. Mourali , Mehdi and Zhiyong Yang (2013), "The Dual Role of Power in Resisting Social Influence," Journal of Consumer Research, 40 (3), 539 - 54. Muniz, Albert M and Thomas C O'guinn (2001), "Brand Community," Journal of Consum er Research, 27 (4), 412 - 32. Nam, Sungjoon, Puneet Manchanda, and Pradeep K Chintagunta (2010), "The Effect of Signal Quality and Contiguous Word of Mouth on Customer Acquisition for a Video - On - Demand Service," Marketing Science, 29 (4), 690 - 700. Newman, C hristopher L, Elizabeth Howlett, and Scot Burton (2015), "Effects of Objective and Evaluative Front - of - Package Cues on Food Evaluation and Choice: The Moderating Influence of Comparative and Noncomparative Processing Contexts," Journal of Consumer Research , 42 (5), 749 - 66. Nitzan, Irit and Barak Libai (2011), "Social Effects on Customer Retention," Journal of Marketing, 75 (6), 24 - 38. Noseworthy, Theodore J and Remi Trudel (2011), "Looks Interesting, But What Does It Do? Evaluation of Incongruent Product Fo rm Depends on Positioning," Journal of Marketing Research, 48 (6), 1008 - 19. Nyer, Prashanth U and Mahesh Gopinath (2005), "Effects of Complaining versus Negative Word of Mouth on Subsequent Changes in Satisfaction: The Role of Public Commitment," Psycholog y & Marketing, 22 (12), 937 - 53. Oakley, James L, Adam Duhachek, Subramanian Balachander, and S Sriram (2007), "Order of Entry and the Moderating Role of Comparison Brands in Brand Extension Evaluation," Journal of Consumer Research, 34 (5), 706 - 12. Olsen, Svein Ottar (2002), "Comparative Evaluation and the Relationship between Quality, Satisfaction, and Repurchase Lo yalty," Journal of the Academy o f Marketing Science, 30 (3), 240 - 49. 106 Osinga, Ernst C, Peter SH Leeflang, and Jaap E Wieringa (2010), "Early Mar keting Matters: A Time - Varying Parameter Approach to Persistence Modeling," Journal of Marketing Research, 47 (1), 173 - 85. Palmatier, Robert W, Mark B Houston, Rajiv P Dant, and Dhruv Grewal (2013), "Relationship Velocity: Toward a Theory of Relationship D ynamics," Journal of Marketing, 77 (1), 13 - 30. Palmatier, Robert W, Rajiv P Dant, and Dhruv Grewal (2007), "A Comparative Longitudinal Analysis of Theoretical Perspectives of Interorganizational Relationship Performance," Journal of Marketing, 71 (4), 172 - 94. Park, Sungho and Sachin Gupta (2012), "Handling Endogenous Regressors by Joint Estimation Using Copulas," Marketing Science, 31 (4), 567 - 86. Patrick, Vanessa M and Henrik Hagtvedt (2011), "Aesthetic Incongruity Resolution," Journal of Marketing Researc h, 48 (2), 393 - 402. Pauwels, Koen, Dominique M Hanssens, and Sivaramakrishnan Siddarth (2002), "The Long - Term Effects of Price Promotions on Category Incidence, Brand Choice, and Purchase Quantity," Journal of Marketing Research, 39 (4), 421 - 39. Pauwels, K oen, Jorge Silva - Risso, Shuba Srinivasan, and Dominique Hanssens (2003), "The Long - Term Impact of New - Product Introductions and Promotions on Financial Performance and Firm Value." Tuck S chool of Business Working Paper, ( 03 - 05 ). Pauwels, Koen, Jorge Silva - Risso, Shuba Srinivasan, and Dominique M Hanssens (2004), "New Products, Sales Promotions, and Firm Value: The Case of the Automobile Industry," Journal of Marketing, 68 (4), 142 - 56. Peppers, Don and Martha Rogers (2005), Return on Customer: Creating Maximum Value from Your Scarcest Resource: Currency. Peracchio, Laura A and Joan Meyers - Levy (1994), "How Ambiguous Cropped Objects in Ad Photos Can Affect Product Evaluations," Journal of Consumer Research, 21 (1), 190 - 204. Petersen, J Andrew and V Kumar (2009), "Are Product Returns a Necessary Evil? Antecedents and Consequences," Journal of Marketing, 73 (3), 35 - 51. Petrin, Amil and Kenneth Train (2010), "A Control Function Approach to Endogeneity in Consumer Choice Models," Journal of Marketing Research , 47 (1), 3 - 13. Petrova, Petia K and Robert B Cialdini (2005), "Fluency of Consumption Imagery and the Backfire Effects of Imagery Appeals," Journal of Consumer Research, 32 (3), 442 - 52. Prahalad, Coimbatore K and Venkat Ramaswamy (2004), "Co - Creation Expe riences: The Next Practice in Value Creation," Journal of Interactive Marketing, 18 (3), 5 - 14. Ramani, Girish and Vipin Kumar (2008), "Interaction Orientation and Firm Performance," Journal of Marketing, 72 (1), 27 - 45. 107 Rao, Hayagreeva, Gerald F Davis, and Andrew Ward (2000), "Embeddedness, Social Identity and Mobility: Why Firms Leave the NASDAQ and Join The New York Stock Exchange," Administrative Science Quarterly, 45 (2), 268 - 92. Reinartz, Werner, Jacquelyn S Thomas, and Viswanathan Kumar (2005), "Balanc ing Acquisition and Retention Resources to Maximize Customer Profitability," Journal of Marketing, 69 (1), 63 - 79. Richardson, Alan (1969), Mental Imagery. London: Routledge & Kegan Paul. Rindova, Violina P and Antoaneta P Petkova (2007), "When Is a New Thi ng a Good Thing? Technological Change, Product Form Design, and Perceptions of Value for Product Innovations," Organization Science, 18 (2), 217 - 32. Rust, Roland T, Katherine N Lemon, and Valarie A Zeithaml (2004), "Return on Marketing: Using Customer Equi ty to Focus Marketing Strategy," Journal of Marketing, 68 (1), 109 - 27. Schau, Hope Jensen, Albert M Muñiz Jr, and Eric J Arnould (2009), "How Brand Community Practices Create Value," Journal of Marketing, 73 (5), 30 - 51. Schlosser, Ann E (2003), "Experienci ng Products in the Virtual World: The Role of Goal and Imagery an Influencing Attitudes versus Purchase Intentions," Journal of Consumer Research, 30 (2), 184 - 98. Schmitt, Philipp, Bernd Skiera, and Christophe Van den Bulte (2011), "Referral Programs and C ustomer Value," Journal of Marketing, 75 (1), 46 - 59. Schulze, Christian, Bernd Skiera, and Thorsten Wiesel (2012), "Linking Customer and Financial Metrics to Shareholder Value: The Leverage Effect in Customer - Based Valuation," Journal of Marketing, 76 (2), 17 - 32. Schwarz, Norbert (2004), "Metacognitive Experiences in Consumer Judgment and Decision Making," Journal of Consumer Psychology, 14 (4), 332 - 48. Schweidel, David A, Young - Hoon Park, and Zainab Jamal (2014), "A Multiactivity Latent Attrition Model for Customer Base Analysis," Marketing Science, 33 (2), 273 - 86. Shah, Denish, Vipin Kumar, and Kihyun Hannah Kim (2014), "Managing Customer Profits: The Power of Habits," Journal of Marketing Research, 51 (6), 726 - 41. Shin, Jiwoong, K Sudhir, and Dae - Hee Yoon Cost - Based Pricing," Management Science, 58 (5), 932 - 47. Shiv, Baba and Joel Huber (2000), "The Impact of Anticipating Satisfaction on Consumer Choice," Journal of Consumer Research, 27 (2), 202 - 16. Skiera, Bern d, Manuel Bermes, and Lutz Horn (2011), "Customer Equity Sustainability Ratio: A New Metric For Assessing a Firm's Future Orientation," Journal of Marketing, 75 (3), 118 - 31. 108 Slotegraaf, Rebecca J and Koen Pauwels (2008), "The Impact of Brand Equity and Inn ovation on the Long - Term Effectiveness of Promotions," Journal of Marketing Research, 45 (3), 293 - 306. Slotegraaf, Rebecca J, Christine Moorman, and J Jeffrey Inman (2003), "The Role of Firm Resources in Returns To Market Deployment," Journal of Marketing Research, 40 (3), 295 - 309. Srinivasan, Shuba and Dominique M Hanssens (2009), "Marketing and Firm Value: Metrics, Methods, Findings, and Future Directions," Journal of Marketing Research, 46 (3), 293 - 312. Srivastava, Rajendra K, Tasadduq A Shervani, and Li am Fahey (1998), "Market - Based Assets and Shareholder Value: A Framework for Analysis," Journal of Marketing, 62(1), 2 - 18. Stahl, Florian, Mark Heitmann, Donald R Lehmann, and Scott A Neslin (2012), "The Impact of Brand Equity on Customer Acquisition, Rete ntion, and Profit Margin," Journal of Marketing, 76 (4), 44 - 63. Thompson, Scott A and Rajiv K Sinha (2008), "Brand Communities and New Product Adoption: The Influence and Limits of Oppositional Loyalty," Journal of Marketing, 72 (6), 65 - 80. Tian, Kelly Tep per, William O Bearden, and Gary L Hunter (2001), "Consumers' Need for Uniqueness: Scale Development and Validation," Journal of Consumer Research, 28 (1), 50 - 66. Trusov, Michael, Randolph E Bucklin, and Koen Pauwels (2009), "Effects of Word - of - Mouth versu s Traditional Marketing: Findings from an Internet Social Networking Site," Journal of Marketing, 73 (5), 90 - 102. Tversky, Amos, Shmuel Sattath, and Paul Slovic (1988), "Contingent Weighting in Judgment And Choice," Psychological Review, 95 (3), 371. Vargo, Stephen L and Robert F Lusch (2004), "Evolving to a New Dominant Logic for Marketing," Journal of Marketing, 68 (1), 1 - 17. Venkatesan, Rajkumar and Vita Kumar (2004), "A Customer Lifetime Value Framework for Customer Selection and Resource Allocatio n Strategy," Journal of Marketing, 68 (4), 106 - 25. Verhoef, Peter C (2003), "Understanding the Effect of Customer Relationship Management Efforts on Customer Retention and Customer Share Development," Journal of Marketing, 67 (4), 30 - 45. Verhoef, Peter C a nd Bas Donkers (2005), "The Effect of Acquisition Channels on Customer Loyalty and Cross - Buying," Journal of Interactive Marketing, 19 (2), 31 - 43. 109 Villanueva, Julian, Shijin Yoo, and Dominique M Hanssens (2008), "The Impact of Marketing - Induced versus Word - of - Mouth Customer Acquisition on Customer Equity Growth," Journal of Marketing Research, 45 (1), 48 - 59. Vogel, Verena, Heiner Evanschitzky, and Balasubramani Ramaseshan (2008), "Customer Equity Drivers and Future Sales," Journal of Marketing, 72 (6), 98 - 1 08. Voss, Glenn B and Zannie Giraud Voss (2008), "Competitive Density and the Customer Acquisition Retention Trade - Off," Journal of Marketing, 72 (6), 3 - 18. Walters, Gabrielle, Beverley Sparks, and Carmel Herington (2007), "The Effectiveness of Print Adver tising Stimuli in Evoking Elaborate Consumption Visions for Potential Travelers," Journal of Travel Research, 46 (1), 24 - 34. Wangenheim, Florian V and Tomás Bayón (2007), "The Chain from Customer Satisfaction via Word - of - Mouth Referrals to New Customer Acq uisition," Journal of the Academy of Marketing Science, 35 (2), 233 - 49. Wiesel, Thorsten, Bernd Skiera, and Julian Villanueva (2008), "Customer Equity: An Integral Part of Financial Reporting," Journal of Marketing, 72 (2), 1 - 14. Williams, Richard (2009), "Using Heterogeneous Choice Models to Compare Logit and Probit Coefficients Across Groups," Sociological Methods & Research, 37 (4), 531 - 59. Wooldridge, Jeffrey M. 2007 http://www.nber.org/wne/lect_6_controlfuncs.pdf Wu, Freeman, Adriana Samper, Andrea C M Pretty to Use! When and How Enhanced Product Aesthetics Discourage Usage and Lower Consumption Enjoyment," Journal of Consumer Research, 44 (3), 651 - 72. Wu, Jianan and Arvind Rangaswamy (2003), "A Fuzzy Set Model of Search and Consideration with an Application to an Online Market," Marketing Science, 22 (3), 411 - 34. Yang, Sha, Mantian Hu, Russell S Winer, Henry Assael, and Xiaohong Chen (2012), "An Empirical Study of Word - of - Mouth Generation and Consumption," Marketing Science, 31 (6), 952 - 63. Yoo, Jungmin and Minjeong Kim (2014), "The Effects of Online Product Presentation on Consumer Responses: A Mental Imagery Perspective," Journal of Business Research, 67 (11), 2464 - 72. Zeithaml, Valarie A, Ruth N Bolton, John Deighton, Timothy L Keiningham, Katherine N Lemon, and J Andrew Petersen (2006), "Forward - Looking Focus: Can Firms Have Adaptive Foresight?," Journal of Service Research, 9 (2), 168 - 83. Zhang, Jonathan Z, George F Watson Iv, Robert W Palmatier, and Ra jiv P Dant (2016), "Dynamic Relationship Marketing," Journal of Marketing, 80 (5), 53 - 75. 110 Zhang, Kaifu, Theodoros Evgeniou, V Padmanabhan, and Emile Richard (2012), "Content Contributor Management and Network Effects in a UGC Environment," Marketing Scienc e, 31 (3), 433 - 47. Zhang, Shoutong Thomas (2016), "Firm Valuation from Customer Equity: When Does It Work and When Does It Fail?" International Journal of Researc h in Marketing, 33 (4), 966 - 70.