Online Purchase Intentions for Luxury and Non-Luxury Brands
in the Fashion Industry
Michael Tanteles - 11186623
University of Amsterdam - Faculty of Economics and Business
MSc in Business Administration – Marketing Track Supervisor: drs. Antoon Meulemans
Statement of Originality
This document is written by Michael Tanteles who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original
and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
TABLE OF CONTENTS Table of Contents Table of Contents ... 3 List of Tables ... 5 List of Figures ... 6 Abstract ... 7 1. Introduction ... 8 2. Literature Review ... 12 2.1 Promotion ... 12 2.2 Additional Costs ... 18 2.3 Brand Luxury ... 23 2.4 Research Model ... 35 3. Method ... 37 3.1 Fashion Industry ... 37 3.2 Pre-test ... 38
3.2.1 Data Collection Procedure ... 38
3.2.2 Sample ... 39
3.2.3 Measurement of Variables ... 39
3.2.4 Statistical Procedure ... 40
3.3 Survey ... 46
TABLE OF CONTENTS 3.3.2 Sample ... 47 3.3.3 Measurement of Variables ... 51 3.3.4 Normality Check ... 51 4. Results ... 53 4.1 Pre-Test ... 53 4.2 Hypotheses Testing ... 58 5. Discussion ... 77
5.1 Discussion and conclusions of results ... 77
5.2 Theoretical Implications ... 80 5.3 Managerial Implications ... 81 5.4 Limitations ... 81 5.5 Future Research ... 82 References ... 83 Appendix ... 105
LIST OF TABLES List of Tables
Table 1 Review of Brand Luxury Dimensions ... 34
Table 2 Cronbach’s Alpha scores – Pre-test ... 42
Table 3 Means, Standard Deviations, Correlations and Reliabilities per Brand – Pre-test ... 43
Table 4 Normality Check – Skewness & Kurtosis – Pre-test ... 45
Table 5 Summary of demographic profile ... 48
Table 6 Summary of demographic profile per brand ... 49
Table 7 Summary of online purchase profile per brand ... 50
Table 8 Cronbach’s Alpha scores – Pre-test ... 51
Table 9 Normality Check – Skewness & Kurtosis ... 52
Table 10 T-test – Paired Sample test ... 54
Table 11 Mean scores per brand ... 55
Table 12 Rankings for dimensions and traits ... 56
Table 13 Mean scores per brand after traits’ ranking ... 57
Table 14 Means scores per brand after dimensions’ ranking ... 58
Table 15 Summary of Descriptive Data ... 60
Table 16 Mauchly’s Test of Sphericity ... 61
Table 17 Means, Standard Deviations and Correlations ... 64
Table 18 Between-Subjects Effects ... 71
Table 19 Within-Subjects Effects ... 72
Table 20 Within-Subjects Contrasts ... 72
LIST OF FIGURES List of Figures
Figure 1 Brand Equity Levels ... 26
Figure 2 Brand Knowledge ... 27
Figure 3 Brand Personality ... 27
Figure 4 Research Model ... 36
Figure 5 Sampling framework ... 46
Figure 6 Three-factor experiment structure ... 59
Figure 7 Price Promotion estimated marginal means ... 62
Figure 8 Additional Costs estimated marginal means ... 63
Figure 9 Brand Luxury estimated marginal means ... 63
Figure 10 Brand Luxury × Price Promotion – Price Promotion × Brand Luxury ... 66
Figure 11 Brand Luxury × Additional Costs – Additional Costs × Brand Luxury ... 67
Figure 12 Price Promotion × Additional Costs – Additional Costs × Price Promotion ... 69
Figure 13 Low Brand Luxury: Price Promotion × Additional Costs ... 73
Figure 14 Medium-Low Brand Luxury: Price Promotion × Additional Costs ... 73
Figure 15 Medium Brand Luxury: Price Promotion × Additional Costs ... 74
Figure 16 Medium-High Brand Luxury: Price Promotion × Additional Costs ... 74
ABSTRACT Abstract
The study examines the moderating effect of brand luxury in online purchase intentions, under different promotional strategies. A pre-test was conducted to establish a hierarchical distinction on a 5-tier luxury scale among five brands. Following the pre-test, a survey was administered to a non-probability sampling in 23 countries. The empirical results showed that online promotion strategies must be adjusted according to the luxury level of a brand in order to increase its value, performance, and positioning compared to rivals. Theoretical and managerial implications are discussed.
Keywords: online purchase intentions, brand luxury, price promotion, shipping costs, fashion
INTRODUCTION 1. Introduction
The creation, delivery, and consumption methods of goods and services are undergoing a complete revolution (Okonkwo, 2010: 11; Achrol and Kotler, 2012: 50). Over the last two decades, traditional retailers have suffered from the emergence of electronic commerce
(e-commerce) (Arnold and Reynolds, 2003: 77; Wessel and Christensen, 2012: 7). This has induced them to also engage in digital practices and ultimately offer consumers a holistic omnichannel experience (Gallo and McAlister, 2003 in Kim, Kim, and Kandampully, 2009: 1189; Nirmala and Dewi, 2011: 67; Rigby, 2011: 5), which in turn has resulted in the ascent of online retailing (Hahn and Pisani, 2016).
Forrester Research forecasts that online retail sales in Europe will annually grow by 12 percent, reaching €233.9 billion in 2018 (Beeson, Evans, and Causey, 2014), while the respective percent for the U.S. is 11 percent (Wu, 2015). Especially for the fashion e-commerce, only in the U.S., it is expected that total sales will climb from $52.2 in 2014 to $86.4 billion in 2018
(eMarketer, 2015). These facts lead to the assumption that the study of online consumer behavior is indispensable in order to profitably manage online stores (Puccinelli, Goodstein, Grewal, Price, Raghubir, and Stewart, 2009: 15).
Consensus is evident in the literature about the benefits of electronic shopping compared to physical shopping. Such benefits include the ability to choose among a number of alternatives, convenience and accessibility, information bounty, and time-efficiency (Breitenbach and Van Doren,1998: 573; Brynjolfsson and Smith, 2000: 568; Häubl and Trifts, 2000: 17; Rosen and Howard, 2000: 80; Wolfinbarger and Gilly, 2001: 35; Brown, Pope, and Voges, 2003: 1667; Park and Kim, 2003: 16; Young Kim and Kim, 2004: 885; Gounaris, Dimitriadis, and Stathakopoulos, 2010: 142; Rigby, 2011: 5).
INTRODUCTION Although commerce offers key values to consumers as well as financial prosperity to e-tailers (electronic ree-tailers), the latter should be aware of its potential backfire effects due to consumers’ low switching costs and information abundance (Reibstein, 2002: 465). Therefore, how can fashion e-tailers unleash the full potential of their online shopping channels, and convert customers to members or even ambassadors (Mossinkoff, 2015: 9)? Which factors motivate consumers to engage in online fashion shopping? Considering that “the question challenging today’s entrepreneur is not whether to have a website but how to become a winner in internet competition” (Chen and Wells, 1999: 36), the answers to the above questions can have a positive impact on a brand’s online marketing strategy (Wolfinbarger and Gilly, 2001: 34).
E-shoppers’ characteristics and online purchase intentions have been widely studied since the beginning of the 21st century (Harris and Dennis, 2008: 102), presumably due to the
differences between offline and electronic shoppers (Wolfinbarger and Gilly, 2001: 35; Cho, Kang, and Cheon, 2006: 262) as well as due to the latter’s heterogeneity (Reibstein, 2002: 473). Meanwhile, the distinction of fashion as an important online shopping product category (Kim et al., 2009: 1190) has afforded many academicians and practitioners to turn their attention to the factors that affect consumers’ purchase behavior. These factors include price (Reibstein, 2002: 473; Rosen and Howard, 2000: 72; Kwon and Noh, 2010: 345), which encompasses promotional activities (Dawson and Kim, 2010: 240), low-price guarantee and money back guarantee tactics (McWilliams and Gerstner, 2006: 106); e-service quality, which includes information
availability, aesthetics, ease of use, online completeness/ transaction capability and entertainment (Park and Kim, 2003: 23; Shankar, Smith, and Rangaswamy, 2003: 73; Van der Heijden,
Verhagen, and Creemers, 2003: 45; Young Kim and Kim, 2004: 892; Ha and Stoel, 2009: 570;
Kim and Damhorst, 2009: 14; Jones and Kim, 2010: 634; Hsin Chang and Wang 2011: 352;
INTRODUCTION
Kim, Choi, and Lee, 2015: 396); customization features (Lee, Damhorst, Campbell, Loker, and Parsons, 2011: 326; Kang and Kim, 2012: 99; Aichner and Coletti, 2013: 30); referrals and e-word-of-mouth (Gupta and Harris, 2010: 1048; Prendergast, Ko, and Siu Yin, 2010: 703; Guo, Wang, and Leskovec, 2011: 166; Fan and Miao, 2012: 178; Jalilvand and Samiei, 2012: 471); shipping costs (Lewis, Singh, and Fay, 2003: 31; Young Kim and Kim, 2004: 892; Lewis, 2006: 21; Lewis, Singh, and Fay, 2006: 62); visual product presentation (Park, Stoel, and Lennon, 2008: 84; Won Jeong, Fiore, Niehm, and Lorenz, 2009: 119; Lee, Κim, and Fiore, 2010: 149;
Okonkwo, 2010: 219; Shim and Lee, 2011: 955; McCormick and Livett, 2012: 34; Song and Kim, 2012: 352); virtual try-on technology (Kim and Forsythe, 2008: 56; Merle, Senecal, and St-Onge, 2012: 57; Yuan, Khan, Farbiz, Yao, Niswar, and Foo, 2013: 1967); security and privacy risks (Belanger, Hiller, and Smith, 2002: 266; McKnight, Choudhury, and Kacmar, 2002: 317;
Kim, Ferrin, and Rao, 2008: 556; Ha and Stoel, 2009: 570; McCole, Ramsey, and Williams 2010: 1022); as well as age and gender (Young Kim and Kim, 2004: 892; Hansen and Møller Jensen, 2009: 1168; Kwon and Noh, 2010: 346; Law and Ng, 2016: 16).
Yet, limited empirical evidence exists as to the impact of the aforementioned factors and brand luxury on online purchase behavior. And, although the effect of brand equity and its components (see Figures 1 and 2) on online fashion purchases is prominently featured in the literature (Lee and Huddleston, 2006: 23; Park and Lennon, 2009: 156; Jones and Kim, 2010: 634; Aghekyan-Simonian et al., 2012: 325), there is a lack of understanding on how different levels of brand luxury can affect patronage. Hence, this study aims to fill this void by exploring the combined effect of promotional activities and additional costs — shipping and return costs — on online purchase intentions through the moderating effect of brand luxury, and specifically of fashion brands that demonstrate different levels of luxuriousness.
INTRODUCTION The remainder of the study is organized as follows. The Literature Review section
reviews academic and management papers that will support the formation of the research hypotheses. It specifically delves deeper into the online promotion literature by presenting the pros and cons of discount policies, their impact on consumer purchase behavior and the different types that exist, with particular reference to promotional codes. It also sheds light to the available shipping fees schemes and their implications for customer acquisition and retention, and for conversion rates accordingly. It also briefly discusses the concepts of brand equity and brand personality, and reviews the brand luxury indexes, with particular focus on the luxury brand personality upon which the research of this study was built. The section ends with a research model which depicts the formed hypotheses.
The Method section introduces the research methodology applied, by elaborating on the data collection procedure, the sample characteristics, the measurement of variables and the statistical procedure used for both the pre-test and the survey. The Results section presents the empirical findings that will lead to the support or rejection of the formed hypotheses.
The study concludes with the Discussion section and casual inferences about consumers’ online purchase intentions for brands with different levels of luxuriousness under various
promotional settings. The results reveal that there are no one-size-fits-all online marketing strategies, as they have to be customized based on the luxury level of the brand. The findings are useful for marketing managers in enhancing their digital marketing efforts and in pursuing strategies that will connect their brand with potential and current customers, both intended to maximize conversion, engagement, and loyalty. The findings both support and refute existing literature, thus addressing theoretical implications for the relevant literature. The study ends with the limitations of the research and areas for future research.
LITERATURE REVIEW 2. Literature Review
2.1 Promotion
The massive expansion of internet use has afforded the development of “customer-responsive promotions” (Chatterjee and McGinnis, 2010: 13), while pricing and technology innovations have enabled the successful targeting of customers in the online environment (Grewal, Ailawadi, Gauri, Hall, Kopalle and Robertson, 2011: S43), which in turn can improve the competitiveness of a brand (Lai and Vinh, 2013: 16). Given that promotion is “an element of price” (Lewis, 2006: 15) and an intrinsic part of marketing strategy (Monroe, 2003: 478), its perceived value is relatively high, especially for price-conscious customers (Nirmala and Dewi, 2011: 69; Workman and Cho, 2012: 270). Therefore, it would be utopic if e-tailers expected that consumers would instantly buy a product without previously searching for price promotions (Close and Kukar-Kinney, 2010: 987). Literature has widely recognized the usefulness of promotion strategies and has focused on both their positive and negative aspects.
Although some luxury brands have been reluctant to discount policies (Okonkwo, 2010: 249), many e-tailers capitalize on patrons’ price-sensitivity (Reibestein, 2002: 473) by embracing promotional pricing to attract them (Park and Lennon, 2009: 150; Jung and Lee, 2010: 23). Price promotions have been proved to drive sales (Brohan, 1999; Dawson and Kim, 2010: 242; Kwon and Noh, 2010:345, Zoellner and Schaefers, 2015: 528), increase market share (Grewal et al., 2011: S43), and preserve loyalty among existing customers (Hsieh, Chiu, and Chiang, 2005: 80).
Previous research has confirmed that customers highly value online price promotions, which positively affects store image (Park and Lennon, 2009: 156), customer satisfaction (Wang and Huarng, 2002: 634; Oliver and Shor, 2003: 129; Honea and Dahl, 2005: 550; Dawson and Kim, 2010: 240), and customer loyalty (Bagozzi, 1998 in Park and Lennon, 2009: 150). In
LITERATURE REVIEW contrast, Jin and Park (2006: 208) did not find any positive relationship between price promotion and customer satisfaction; nonetheless, they aver that the former builds consumer trust.
Marketers view customization as the primary ingredient for a successful promotion strategy (Zhang and Wedel, 2009: 204). Technological advancements have enabled marketers to target individual market segments (Chen, Monroe, and Lou, 1998: 354; Zhang and
Krishnamurthi, 2004: 561; Zhang and Wedel, 2009: 204; Grewal et al., 2011: S44), aiming at higher conversion rates and customer loyalty (Zhang and Wedel, 2009: 191). The idea behind customized promotions is that e-tailers can offer price discounts only when customers actually appreciate them and there are heightened chances of utilizing them (Zhang and Wedel, 2009: 205). Indeed, Zhang and Krishnamurthi (2004: 575) posit that timing should be taken into account when implementing a customized promotion campaign. Studies suggest that tailored price promotions monopolize consumers’ attention (Zhenxiang and Lijie, 2011: 198;
Hanafizadeh and Behboudi, 2012: 27) and act as an incentive for them to complete the
transaction (Honea and Dahl, 2005: 544), especially for those with low willingness to pay (Shor and Oliver, 2006: 437). This phenomenon is evident in both offline (Chatterjee and McGinnis, 2010: 17; Dholakia, 2000: 956; Youn and Faber, 2000) and online transactions (Koufaris, Kambil, and LaBarbera, 2002: 117; Darke and Dahl, 2003: 336; Dawson and Kim, 2010: 240).
Over and above, Grewal et al. (2011: S48) contend that customized promotions are positively perceived by customers, mainly because they “create a sense of exclusiveness” (Chen et al., 1998: 366), which could explain why those type of promotions are more lucrative than mass-distributed ones (Zhang and Wedel, 2009: 205). What is more, online promotions that target existing customers – loyalty promotions – are more effective than those targeting potential customers – competitive promotions (Bawa and Shoemaker, 1987b: 376; Zhang and Wedel, 2009: 204).
LITERATURE REVIEW Recent marketing literature has identified different types of promotions, including “buy one-get one free deal, coupons, percentage off when spend a certain limit, free gift with purchase, free shipping or shipping discount, ability to return online purchase in a physical store, contests or sweepstakes, membership discounts” (Dawson and Kim, 2010: 236), and invitation-only (Grewal et al., 2011: S47). Previous research has shown that consumers value promotion efforts differently as they do not process promotion communication uniformly (Suri, Swaminathan, and Monroe, 2004: 82), which could explain why not all sales promotion techniques have a positive impact on purchase satisfaction (Teck Weng and Cyril de Run, 2013: 86). Seeing that most online marketing efforts adopt the promotion code selling policy in order to boost sales (Chatterjee and McGinnis, 2010: 13), this study will focus exclusively on this method.
Coupons have been widely used at promotional campaigns since the 90s (Blattberg and Neslin, 1990: 6). The proliferation of internet-based economy has digitized coupons (Machlis, 1998 in Suri et al., 2004: 75) and has evolved them into the form of “promotion codes” (Oliver and Shor, 2003: 121). This development has afforded them greater popularity (Bawa, Srinivasan, and Srivastava, 1997: 517; Blundo, Cimato, and De Bonis, 2005: 117) and efficiency (Jung and Lee, 2010: 25), both time- and cost-wise (Carmody, 2001: 50, 56). Indeed, consumers prefer online coupons vis- à-vis offline ones when it comes to high-involvement purchases (Suri et al., 2004: 83), such as fashion products (Yoh, Damhorst, Sapp, and Laczniak, 2003: 1097).
Therefore, it is no coincidence that the redemption rates of online coupons are higher, which could also be attributed to the online coupons’ customization features (Fortin, 2000: 530; Jung and Lee, 2010: 33). Bawa and Shoemaker (1987a: 110) have argued that consumers’
characteristics can also indicate the predisposition towards coupons. More specifically, they argue that coupon users are well-educated and upper-income, they live in urban areas and they are not brand loyal, a view also shared by Chiou-Wei and Inman (2008: 305).
LITERATURE REVIEW Several definitions exist in the literature for coupons: “certificates that give buyers a saving when they purchase specified products” (Kotler and Armstrong, 2012: 483); “a certificate that entitles a consumer to some sort of incentive to buy a product or service” (Jung and Lee, 2010: 24); or just “documents that entitle the holder to a special price” (Monroe, 2003: 481). As stated by Shor and Oliver (2006: 424), coupons “ permit a retailer to price discriminate by selling the same good or service to different consumers at varying prices”. From here onwards, online coupons and promotional codes will share the same meaning.
The basic benefit types of coupons are discount and free, with the latter scoring higher in redemption rates compared to the former (Jung and Lee, 2010: 3). The discount coupons are offered under two framing messages; as a monetary reduction to the original price or as a percentage discount (Della Bitta, Monroe, and McGinnis, 1981: 425; Grewal et al., 2011: S48). Empirical evidence suggests that percentage discounts lead to higher purchase intentions
compared to monetary reductions (Chen et al., 1998: 366), while the former is more efficient for low-priced products and the latter for durable and high-priced products (above $2.000) (Chen et al., 1998: 365; Grewal et al., 2011: S48).
Further, sufficient evidence exists to support that the higher the face value of the coupon (worth of discount or percentage-off) the higher the redemption rates (Bawa and Shoemaker, 1987b: 376; Leone and Srinivasan, 1996: 284; Suri et al., 2004: 83; Jung and Lee, 2010: 33), especially for durable goods (Chiou-Wei and Inman, 2008: 305). Nevertheless, there is a recommended threshold for percentage discounts, 20 or 50 percent and higher, meaning that beyond these thresholds consumers will start doubting the credibility of the discount (Jung and Lee, 2010: 35), or underrate it (DelVecchio, Krishnan, and Smith, 2007: 167). A respective threshold exists for monetary discounts too (Jung and Lee, 2010: 35), but their perceived value is contingent upon relative criteria and not absolute monetary terms (Grewal and Marmorstein,
LITERATURE REVIEW
1994: 459). In other words, consumers’ satisfaction drawn from a monetary discount depends on both the discount amount and the initial price of the product, whereas for a percentage discount the percentage itself is enough to compare its attractiveness with other products (Chen et al., 1998: 355). Suri et al. (2004: 83) have added another element on how customers evaluate
promotions and this is their motivation to process the available information. The authors aver that online coupons are more thoroughly processed by less motivated consumers, who usually are more price sensitive. However, this view contradicts Bawa and Shoemaker (1987b: 376) and Zhang and Wedel’s (2009: 204) findings that customized promotions are more effective, which by default are addressed to motivated consumers.
Going back to customized promotions, empirical evidence suggests that in contrast to mass-distributed coupon promotions (price promotions), customized coupon promotions are more profitable, highly perceived by consumers, and can enhance purchase intentions (Chen et al., 1998: 365; Zhang and Wedel, 2009: 204; Chatterjee and McGinnis, 2010: 18). Chatterjee and McGinnis (2010: 13) contend that price promotions have a negative effect on brand equity and on consumers' purchasing behaviors, a line of reasoning which contrasts Suri et al.’s assertion that price promotions can improve purchase intentions (2004: 83).
Nonetheless, online coupons are not without their drawbacks. Oliver and Shor (2003: 129) maintain that the absence of promotional codes can lead to dissatisfaction and then to incomplete transactions, while Leonard and Ariely (2006: 65) have empirically proved that promotional codes have more chances to be redeemed at the early stages of the shopping process. Darke and Dahl (2003: 336) have also found that promotional reward imbalance can yield
negative effects on non-targeted customers. Despite their limitations, major e-tailers use online coupons for the marketing of their products (Suri et al., 2004: 74) and thanks to the recent
LITERATURE REVIEW economic turmoil coupons have emerged as one of the most valuable promotion tools (Jung and Lee, 2010: 23).
As can be expected, the positive effects of price promotion on purchase decisions are also evident in the online fashion shopping (Rhee, 2006); they may account for product evaluations (Raghubir, 2004: 185) and even serve as value determinants (Park and Lennon, 2009: 157). According to Grace and O’ Cass (2005: 113), controlled communications, such as promotions, have positive impact on brand attitude and can influence consumers’ brand choice (Sun, Neslin, and Srinivasan, 2003: 401; Alvarez Alvarez and Vázquez Casielles, 2005: 68), including fashion brands (Park, Kim, Funches, and Foxx, 2012: 1588). Considering that customers tend to abandon their online shopping carts in await of a promotional code or a better deal, promotion codes are an advantageous opportunity for e-tailers to boost their conversion rates (Kukar-Kinney and Close, 2010: 249). Therefore, it is essential to examine consumers’ behavior towards coupons so that e-tailers can implement the corresponding online marketing strategies and ultimately increase their redemption and conversion rates (Bawa et al., 1997: 517; Jung and Lee, 2010: 23). Based on the above, the following hypothesis is suggested:
H1: Price promotion positively affects online purchase intentions1.
LITERATURE REVIEW 2.2 Additional Costs
Shopping cart abandonment exists since the birth of e-commerce (Magill, 2005: 18). Although one would expect that abandonment rates would have slowed down thanks to information technology’s breakneck advancements and to the accelerating power of digital technologies, the trend has been increasing the past decade. Abandonment rates in 2003 were recorded as high as 67 percent (Lueker, 2003: 16) and 74.3 percent in Q1 2016, thus justifying that e-tailers are still struggling to increase their conversion rates (eMarketer, 2016).
A plethora of reasons have been identified for this trend, with the majority of consumers indicating the additional costs appeared in the final step of the online purchasing process
(Campanelli, 2002; Kukar-Kinney and Close, 2010: 249; Prieto, 2016; Quillfeldt, 2016). From a managerial perspective, high abandonment rates pose a threat to a company’s revenues and accordingly to its existence (Lewis et al., 2003: 4; Moore and Mathews, 2006: 74; Egeln and Joseph, 2012: 1; Meola, 2016). For the purpose of this study, the term additional costs will be limited to shipping and/or return costs.
Shipping and return fees have been found to impede patrons to buy online (Rosen and Howard, 2000: 80; Lueker, 2003:16; Magill, 2005: 18), therefore it is no coincidence that much focus has been directed on their implications for the online strategy of an e-tailer (Campanelli, 2002; Lewis, 2006: 21). A survey conducted by Global Millenia Marketing showed that 69 percent of respondents abandoned their shopping carts because the shipping cost was too high and unknown until the checkout page (Lueker, 2003: 16). In addition, a report from Cambridge unveiled that 57 percent of e-shoppers did not finalize their order because they were unwilling to incur the shipping costs (Magill, 2005: 18). Recent studies, though, yielded mixed results: the least promising findings estimate the respective rate at approximately the same levels – 61 and 56 percent – (Baymard Institute, 2016; Quillfeldt, 2016), while the most optimistic ones show a drop
LITERATURE REVIEW of the respective percentage to 35.7 (Meola, 2016). At the turn of the century, many companies, even the disruptive ones that are currently leaping ahead in the e-commerce race, such as Amazon.com, had struggled to identify the sweet spot between online consumer behavior and shipping costs (Hellweg, 2002) and this was reflected in “the level of experimentation occurring
online” (Lewis et al., 2006: 52).
Yet, where do these costs come from and what are the implications for consumers’ online purchase behavior? In the e-commerce environment, the spatial distance between the retailers’ physical shop and customers entails additional costs related to the assembly and/or shipping of the product (Rosen and Howard, 2000: 80; Lewis et al., 2003: 2; Yao and Zhang, 2012: 368). Basically, shipping fees “increase the sacrifice asked of the consumer without changing the utility of the products received”, thus introducing “an element of nonlinear pricing on otherwise
straightforward transactions” (Lewis, 2006: 15). The cost of this sacrifice, though, is usually so high that e-tailers have to pass it onto consumers (Lewis, 2006: 13). That said, how to allocate these costs and ultimately what shipping-fee scheme to adopt is a core marketing strategy decision that e-tailers have to make (Lewis et al., 2003: 2; Lewis et al., 2006: 51; Li and
Dinlersoz, 2012: 276). The gravity of this decision is such that it can have major consequences on customer acquisition and retention (Lewis, 2006: 21; Lewis et al., 2006: 52), on profitability (Pyke, Johnson, and Desmond, 2001: 26; Lewis et al., 2006: 62), and even on the company’s competitive edge over incumbents (Dinlersoz and Li, 2006: 408; Kukar-Kinney and Close, 2010: 249; Yao and Zhang, 2012: 368). Nonetheless, little attention has been given by academicians and researchers to how e-tailers should price shipping and return costs (Lewis et al., 2006: 51;
Yao and Zhang, 2012: 369).
According to the relevant literature, there are three main shipping fee schemes: free
LITERATURE REVIEW
368), cost splitting between the e-tailer and the patron (Clay, Krishnan, Wolff, and Fernandes, 2002: 365), and profit-shipping (Yao and Zhang, 2012: 368) in which e-tailers earn money from charging a higher fee than their actual costs (Schindler, Morrin, and Bechwati, 2005: 44). Cost splitting includes graduated fees that vary according to the order size threshold, either monetary or weight-related, which means higher (lower) shipping fees below (above) a specific threshold (Lewis et al., 2003: 4), free large in which e-tailers provide incentives for larger orders by waiving shipping fees above a value threshold (Lewis, 2006: 21), and flat rate which is a standard fee rate (Lewis et al., 2003: 2). This study will test the relative effectiveness of free shipping and flat rate.
Shipping fees are an integral part of price (Lewis, 2006: 15) and sometimes a costly one (Prieto, 2016). This cost can make consumers susceptible to them (Lewis et al., 2006: 51), even more than the product’s actual price (Smith and Brynjolfsson, 2001: 549). Academic research has shown that shipping fees influence both orders’ volume and size (Lewis et al., 2003: 31; Lewis, 2006: 21; Lewis et al., 2006: 52). When shipping fees increase as the shopping cart fills up, the order size decreases because patrons end up removing items from the cart (Lewis, 2006: 21). On the contrary, Morwitz, Greenleaf, and Johnson (1998: 460) found that customers buying for the first time from an e-tailer tend to overlook additional costs. This phenomenon, however, is not evident among existing customers, as they are already familiar with the surcharges. Hence, experience has a negative effect on the ignorance of additional costs.
Yet, how can e-tailers tackle this Gordian knot and induce patrons to not alter their buying behavior based on the shipping costs? Sales promotions that encourage larger orders positively affect customers’ online purchases (Lewis et al., 2003: 31), which surprisingly enough account for a large percentage of the total daily sales (Lewis, 2006: 21). Unfortunately, though, these types of promotions do not manage to increase order volumes (Lewis et al., 2006: 62). In the
LITERATURE REVIEW meantime, the introduction of a flat shipping rate that would allow patrons to shop unlimitedly, without troubling themselves with variable shipping rates, could be a constructive choice for e-tailers (Kukar-Kinney and Close, 2010: 249). Another option that was tested in the past by Amazon.com is shipping fee waiver for orders that exceed a pre-defined maximum credit (Courogen, 2002).
The fact, though, that three out of five online customers report that the decision on which e-tailer to patronize was contingent upon the availability of free shipping (Evans, 2009) makes free shipping or free return shipping undeniably a drastic solution for increasing conversion rates (Lisanti 1999 in Lewis et al., 2003: 3; Lewis et al., 2003: 31; Lewis et al., 2006: 62; Close and Kukar-Kinney, 2010: 987; Dawson and Kim, 2010: 242; Kukar-Kinney and Close, 2010: 249). Although free shipping is a prominent promotion tool (Chatterjee and McGinnis, 2010: 14), it leads to “smaller orders” (Lewis et al., 2003: 31), and the boosted conversion rates will not counterbalance the lost revenues from shipping fees (Lewis et al., 2006: 52).
Customer acquisition and retention are also affected by the e-tailer’s shipping fee policy. Specifically, graduated fees positively influence both new and/or potential patrons’ conversion rates and total order value, while free shipping is the most efficient method to attract new customers. Per contra, existing customers mostly favor flat shipping fee rates (Lewis, 2006: 21).
All in all, Lewis et al. (2006: 52) assert that patrons’ attitude towards free shipping is inconsistent – their study showed that a free-shipping policy increased conversions rates between 10 and 35 percent. On the contrary, Kim and Niehm (2009: 230) aver that free shipping is an exceptional promotion tool in terms of perceived value, which according to Chatterjee and McGinnis (2010:18) can positively influence purchase intentions. In consonance with the above arguments, the following hypothesis is proposed:
LITERATURE REVIEW H2: Additional costs negatively affect online purchase intentions2.
Drawing from Dawnson and Kim’s (2010: 236) contention that free shipping and
shipping discounts can count as a price promotion tool, as well as from H1 and H2, the following hypothesis is proposed:
H3: The relationship between price promotion and additional costs positively affects online purchase intentions2.
LITERATURE REVIEW 2.3 Brand Luxury
With an annual growth rate between 10 and 15 percent (Silverstein and Fiske, 2003: 51), the luxury market dominates the list of the fastest-paced industries worldwide (Fionda and Moore, 2009: 347) and contributes to the strengthening of the industrialized economies’ GDP (Silverstein and Fiske, 2003: 51). Drivers of this growth have principally been socioeconomic, demographic (Christodoulides, Michaelidou, and Li, 2009: 396), and technologic (Atwal and Williams, 2009: 339; Economist, 2012). These include the increase of available income, the ever increasing purchasing power of women (Silverstein and Fiske, 2003: 52; Roche, Silverstein, and Charpilo, 2008: 11; Fionda and Moore, 2009: 347; Truong, McColl, and Kitchen, 2009: 375), the higher education levels (Silverstein and Fiske, 2003: 53), the rise of a segment craving for luxury brands (Fionda and Moore, 2009: 347; Okonkwo, 2010: 266), the decrease in production costs (Truong et al., 2009: 375), and the evolution of technology and mass production (Atwal and Williams, 2009: 339).
In light of this progression, “the luxury market has transformed from its traditional conspicuous consumption model to a new experiential luxury sensibility marked by a change in the way consumers define luxury” (Wiedmann, Hennigs, and Siebels, 2007: 1). It hence became imperative for academicians and researchers to discover the legitimate reasons why consumers buy luxury goods, how they define luxury, and how the perceived luxury value of a product or a brand influences their purchasing behavior (Wiedmann et al., 2007: 1). This study is
predominantly concerned with delving deeper into how luxury perceptions impact online purchase intentions.
“The first luxury brands consisted of silverware, glassware and china made industrially in France and England” (Nueno and Quelch, 1998: 62). But what is a luxury brand? Luxury is a
LITERATURE REVIEW concept hard to define (Kapferer, 1997: 251; Kapferer, 1998: 44; Cornell, 2002: 47) and
unfortunately there is no consensus in the literature regarding what makes a brand luxurious (Christodoulides et al., 2009: 397; Heine, 2012: 10). “The word luxury derives from the Latin word luxus, which according to the Latin Oxford Dictionary signifies ‘soft or extravagant living, (over-)indulgence’ and ‘umptuousness, luxuriousness, opulence’ ”(Dubois, Czellar, and Laurent, 2005: 115). In brief, luxury brands are those offered in limited quantities (Bearden and Etzel, 1982: 184;Pantzalis 1995: 145; Cornell, 2002: 47), their price and quality are inseparable and the highest in the market (McKinsey, 1990 in Wiedmann et al., 2007: 2; Kapferer, 1997: 252; Dubois and Czellar, 2002: 31), and their consumption conveys value to both users and society (Solomon, 1983: 327; Elliott, 1997: 287; Kapferer 1997: 253; Wiedmann et al., 2007: 3; Sung, Choi, Ahn, and Song, 2015: 121), thus evoking both psychological and utility benefits (Dubois and
Duquesne, 1993b: 41; Vigneron and Johnson, 2004: 486; Kim and Johnson, 2015: 430). A rather contradictory approach holds that “luxury brands are those whose ratio of functional utility to price is low while the ratio of intangible and situational utility to price is high” (Nueno and Quelch, 1998: 62). Although luxury is a key differentiator between brands (Allérès 1991 in Wiedmann et al., 2007: 1) and specifically between luxury and non-luxury brands (Vigneron and Johnson, 2004: 486), Veblen (1899 in Christodoulides et al., 2009: 396) argued that this distinction should be based on the socio-economic environment regardless of the product’s appearance and intrinsic attributes. In the same line of thoughts, luxury has been identified as a culturally (Kemp, 1998: 604; Christodoulides et al., 2009: 403) and socially relative term (Heine, 2012: 9).
Within the luxury brands category, a further distinction was introduced by Vigneron and Johnson (2004: 486) between those with upper and low range of luxury, in a sense that not all luxury brands are equally luxurious. Nueno and Quelch (1998: 62) have determined several
LITERATURE REVIEW characteristics of luxury brands among them being premium quality, craftsmanship heritage, style and design, limited supply, the brand’s universal reputation, country-of-origin quality
associations, and product uniqueness. Many of these characteristics have been corroborated by Kapferer (1998: 47) and Dubois, Laurent, and Czellar (2001: 8), with the former also adding magic and the latter superfluousness. Luxury brands can be found in four distinct categories:
fashion, perfumes and cosmetics, wine and spirits, and watches and jewelry (Jackson, 2004 in Fionda and Moore, 2009: 348).
With reference to the hierarchy of luxury goods, three levels have been identified in terms of their accessibility. These are from the top to the bottom: inaccessible luxury, affordable only by the upper class, intermediate luxury, affordable by the professional and socioeconomic class, and accessible luxury, which is affordable by the middle class aiming at elevated social status (Allérès, 1990 in Sung et al. 2015:130; Christodoulides et al., 2009: 397). Conversely, Silverstein and Fiske (2003: 50) made a somewhat different distinction: accessible superpremium, very high-priced products that are affordable by the middle class, old-luxury brand extensions, cheaper lines of high-end brands, and mass prestige or “masstige” products whose price value is lower than superpremium and old-luxury.
The recent rise of new luxury brands and the fostering competition among established firms have resulted in marketers struggling to diversify the functional attributes of their brands (Sung et al., 2015: 130). Kim and Johnson (2015: 430) aver that if firms want to survive, they should “manage customers’ perceptions or valuation of a brand to assist customers in their ability to differentiate brands and to make their brand distinct among competitors”. Brand personality is claimed to work towards shaping consumers’ preferences (Zhang, 2007 in Kang and Sharma, 2012: 325; Kang and Sharma 2012: 324), by contributing to the creation of consumer brand
LITERATURE REVIEW relationships (Biel, 1993 in Kang and Sharma, 2012: 324), as well as brand trust and brand loyalty (Wysong, Munch, and Kleiser, 2004 in Kang and Sharma, 2012: 325).
Brand personality is defined as the symbolic meanings of a brand and specifically “the set of human characteristics associated with a brand” that differentiates it from its competitors (Aaker, 1997: 347; Keller, 2013: 333). These symbolic attributes are part of the brand awareness, which is one of the key elements of brand knowledge (Keller, 1993: 8, 12) and accordingly of brand equity (Lemon, Rust, and Zeithaml, 2001: 22) (Figure 1). Brand equity was defined by Farquhar (1989: 24) as “the “added value” with which a given brand endows a product”. In a later model, Keller (1998: 94; 2013: 548) (Figure 2) identified brand personality as part of the non-product related attributes of a brand. All-in-all, brand personality is a strong brand dimension (Grace and O'Cass, 2002: 106) associated with the values that consumers attach to brands (Keller, 1993: 4) and with how they in turn attach them to their own personality and to the way they perceive the personalities of the people within their social environment (Solomon, 1983: 326). The brand personality scale by Aaker (1997: 352) includes five dimensions: sincerity, excitement,
competence, sophistication, and ruggedness (Figure 3).
Figure 1 Brand Equity Levels
Brand Equity Attitude toward the Brand Corporate Ethics Brand Knowledge Lemon et al., 2001: 22
LITERATURE REVIEW Figure 2 Brand Knowledge
Figure 3 Brand Personality
Attributes Uniqueness of brand associations Attitudes Price User and usage imagery Brand Personality Brand Awareness Brand Image Brand Awareness Brand Knowledge Types of brand associations Brand Recognition Brand Recall Favorability of brand associations Strength of brand associations Non-product related Product related Benefits Functional Experiential Symbolic Brand Personality Competence Sophistication Excitement Ruggedness Sincerity Reliable Intelligent Successful Upper class Charming Daring Spirited Imaginative Up-to-date Outdoorsy Tough Down-to-earth Honest Wholesome Cheerful Keller, 2013: 549 Aaker, 1997: 352
LITERATURE REVIEW As stated by Dubois and Duquesne (1993a: 115), if marketers wish to grasp a market segment, they first need to comprehend its relationship with the available brands. This can be even more imperative if we take into account that consumers’ purchase behavior is shaped by the available information on the brands (Shugan, 1980: 109; Alba and Hutchinson, 1987: 437). In the case of the luxury market, a scale that would measure consumers’ perceptions of luxury would be appropriate (Vigneron and Johnson, 2004: 484). The need for the study of luxury brands was identified by Andrus et al. (1986: 5) in the late 80’s and although Vigneron and Johnson (2004: 484) have argued that the measurement of a brand’s perceived luxuriousness has not been empirically investigated, several other attempts have been made in the past towards indexing brand luxury (Dubois and Laurent, 1994; Kapferer, 1998: 46; Vigneron and Johnson, 2004: 499;
Wiedmann et al., 2007: 4; Christodoulides et al., 2009: 401; Kim and Johnson, 2015: 439; Sung et al., 2015: 129). The most important findings are presented below in a chronological order.
At the initial attempt, Dubois and Laurent (1994) found that consumers’ attitudes are influenced by how they generally perceive the luxury world and how they think they fit in it. A few years later, Kapferer (1997: 251) sought to interpret the luxury brands’ symbols and how they differ from the basic and up-market brands. Although he did not build a luxury index, he set the conceptual groundworks of luxury: “luxury is enlightening […]3 Luxury is thus both creation and the vital source of inspiration […] Luxury items provide extra pleasure and flatter all senses at once […] Luxury brands are just perpetuating and exemplifying the signs and attitudes of former aristocracy” (Kapferer, 1997: 253). On a later study, Kapferer (1998: 46) aimed to ascertain how consumers appreciate luxury by identifying five dimensions: beauty, excellence,
magic, uniqueness and tradition and savoir-faire. An analogous study from Dubois et al. (2001:
LITERATURE REVIEW
39) revealed six dimensions: quality, time incorporation, hedonism, scarcity, high price, and
superfluousness; however both attempts were criticized as incomplete (Vigneron and Johnson, 2004: 485).
A few years later, Vigneron and Johnson (2004: 500) developed a scale to measure the degree of luxury among luxury brands. Their brand luxury index (BLI) included five perceived values, which were classified to personal and non-personal oriented perceptions. The former category contained hedonic and extended self factors, and the latter conspicuous, uniqueness, and
quality factors (Vigneron and Johnson, 2004: 488). According to the authors (2004: 500), the scale can be utilized to measure consumers’ perceived luxuriousness of specific brands or products vis-à-vis Dubois and Laurent’s (1994) scale that measures brand luxury as a concept. Nonetheless, this BLI model was majorly criticized for drawing responses from business students instead from actual luxury product users (Christodoulides et al., 2009: 397).
Another attempt was made by Wiedmann et al. (2007: 5) who developed a framework consisting of four main value dimensions along with their antecedent constructs. These dimensions were financial, functional, individual, and social values. However, the framework must be empirically tested before utilized to measure consumers’ attitude towards luxury (Wiedmann et al., 2007: 9).
Further, Christodoulides et al. (2009: 396) pursued to enhance Vigneron and Johnson’s BLI index by validating the existing scale with responses from luxury brand consumers.
Nonetheless, according to the authors (2009: 402), their findings were questionable and required further empirical investigation. Notwithstanding the limitations, one of the main contributions of the paper is the finding that brand luxury perceptions vary across cultures.
More recent, Kim and Johnson (2015: 439) followed the example of Christodoulides et al. and added to Vigneron and Johnson’s model. The improved BLI preserved three dimensions
LITERATURE REVIEW from the previous model – quality, extended self, and hedonism – and featured two new ones –
accessibility, which shared items from conspicuousness and uniqueness of the former BLI, as
well as tradition. However, the authors argued that their findings were inadequate for use as further validity and reliability was required (Kim and Johnson, 2015: 442).
Finally, Sung et al. (2015: 129) examined how consumers interpret luxury brands by focusing on the dimensions of luxury brand personality. As reported by the authors, the brand personality dimensions are valuable for both researchers and marketers to effectively “understand consumers who express themselves through purchasing and using luxury brands” (2015: 122). They further posit that although much attention had been devoted to Aaker’s brand personality dimensions, these do not effectively capture “the special symbolic and social meanings embodied by luxury brands” (2015: 130). Therefore, they extended Aaker’s brand personality scale (1997: 352), which was developed for basic brands, in order to examine the luxury brand personality (2015: 125). Their results revealed six brand personality dimensions and 36 personality traits:
excitement, sincerity, sophistication, professionalism, attractiveness, and materialism (see Table 1). The first three dimensions carry the same meaning with those identified by Aaker (1997: 352), that is for non-luxury brands; yet the remaining three are pertinent only in luxury brands (Sung et al., 2015: 129), and can be used to identify the symbolic meanings attached to them (2015: 130). A similar approach has been employed by Su and Tong (2015), although the scale corresponds to sportswear only. Table 1 presents a review of the BLIs.
It could be somehow argued that the brand luxury dimensions embody the psychological benefits that accrue from the purchase and utilization of luxury brands, as mentioned by Dubois and Duquesne (1993b: 41), Vigneron and Johnson (2004: 486), and Kim and Johnson (2015: 430). The last aver that the psychological benefits derived from the consumption of luxury products are the main differentiator from non-luxury products (2004: 486), while Chattalas and
LITERATURE REVIEW Shukla (2015: 51) argue that “the social, personal and functional value perceptions play key roles in consumers’ luxury purchase intentions”. Similarly, Salehzadeh and Pool (2016: 4) confirmed that these values can induce higher purchase intentions for luxury products vis-à-vis non-luxury ones. These lead to the following set of hypotheses:
H4: Brand luxury positively affects online purchase intentions4.
H5a: High brand luxury levels induce higher online purchase intentions. H5b: Medium brand luxury levels induce mediocre online purchase intentions. H5c: Low brand luxury levels induce lower online purchase intentions4.
Although, it is evident from the literature that price promotions can positively influence fashion purchase intentions and that (customized) percentage discount coupons are the most effective ones, little is known on how much influence promotion codes exert on purchase intentions towards fashion brands with various levels of brand luxury. Nevertheless, since redemption rates of durable goods are positively influenced by the increase of the face value of the coupon (Bawa and Shoemaker, 1987b: 376; Leone and Srinivasan, 1996: 284; Suri et al., 2004: 83; Chiou-Wei and Inman, 2008: 305; Jung and Lee, 2010: 33) and online coupon users are more price sensitive (Suri et al., 2004: 83), the following set of hypotheses is suggested:
H6a: Brand luxury moderates the relationship between promotion and online purchase
intentions, so that this relationship is weaker for higher levels of brand luxury.4
LITERATURE REVIEW H6b: Brand luxury moderates the relationship between promotion and online purchase
intentions, so that this relationship is neither strong nor weak for medium levels of brand luxury.
H6c: Brand luxury moderates the relationship between promotion and online purchase
intentions, so that this relationship is stronger for lower levels of brand luxury5.
Turning to additional costs, how would patrons respond to different shipping schemes in the fashion industry, and particularly to brands whose luxury level varies? Unfortunately, literature has not shed light on it yet and this study aims to fill this gap. Drawing on the findings of Lewis et al. (2006: 52) and Chatterjee and McGinnis (2010:18), the following set of
hypotheses is suggested:
H7a: Brand luxury moderates the relationship between additional costs and online purchase intentions, so that this relationship is stronger for higher levels of brand luxury.
H7b: Brand luxury moderates the relationship between additional costs and online purchase intentions, so that this relationship is neither strong nor weak for medium levels of brand luxury.
H7c: Brand luxury moderates the relationship between additional costs and online purchase intentions, so that this relationship is weaker for lower levels of brand luxury5.
LITERATURE REVIEW Based on H3, H6a-c, and H7a-c, we conclude to the following set of hypotheses:
H8a: Brand luxury moderates the relationship among price promotion, additional costs,
and online purchase intentions, so that this relationship is weaker for higher levels of brand luxury.
H8b: Brand luxury moderates the relationship among price promotion, additional costs,
and online purchase intentions, so that this relationship is neither strong nor weak for medium levels of brand luxury.
H8c: Brand luxury moderates the relationship among price promotion, additional costs,
and online purchase intentions, so that this relationship is stronger for lower levels of brand luxury.6
.
LITERATURE REVIEW Table 1 Review of Brand Luxury Dimensions
Kapferer, 1999
Dubois et al., 2001
Vigneron and Johnson, 2004
Wiedmann et al., 2007 Christodoulides et al., 2015
Kim et al., 2015
Sung et al., 2015 Dimensions Dimensions Dimensions Traits Dimensions Traits Dimensions Dimensions Dimensions Traits
Beauty Quality Conspicuousness Conspicuous, Elitist, Extremely expensive, For wealthy Financial value
Price Conspicuousness Accessibility Excitement Energetic, Exciting, Adventurous, Fun, Daring, Outgoing, Cool, Colorful Excellence Time incorporation Uniqueness Very exclusive, Precious, Rare, Unique Functional value Usability, Quantity, Uniqueness
Uniqueness Tradition Sincerity Laidback, Simple, Gentle, Family-oriented,
Down-to-earth, Sensitive, Thoughtful, Warm
Magic Hedonism Quality Crafted,
Luxurious, Best quality, Sophisticated, Superior Individual value Self-identity, Hedonic, Materialistic
Quality Quality Sophistication Upper-class, Wealthy, Status conscious, Stylish,
Sophisticated Uniqueness Scarcity Hedonism Exquisite,
Glamorous, Stunning
Social value Conspicuousness, Prestige
Hedonism Hedonism Professionalism Mature, Professional, Intelligent, Reliable,
Refined Tradition,
Savoir-faire
High price Extended Self Leading, Very powerful, Rewarding,
Successful
Extended Self Extended self Attractiveness Beautiful, Attractive, Good-looking, Artistic, Gorgeous
Superfluousness Materialism Selfish,
Materialistic, Struck-up, Pretentious, Showy
LITERATURE REVIEW 2.4 Research Model
The conceptual framework articulates how various factors influence online purchase intentions. Albeit the existence of various aspects that affect online purchase intentions, this study focuses on assessing the impact of price promotions, additional costs, and brand luxury. The literature review revealed four hypotheses and four sets of hypotheses.
Three hypotheses refer to the direct effect of price promotion, additional costs and brand luxury respectively on purchase intentions. The remaining hypothesis deals with the direct effect of the relationship between price promotion and additional costs on purchase intentions.
Turning to the set of hypotheses, the first refers to the indirect relationship between price promotions and online purchase intentions, which is moderated by the perceived brand luxury; and the second one deals with the indirect relationship between additional costs and online purchase intentions, which is moderated by brand luxury. To put it more simply, price promotions and additional costs act as the independent variable that indirectly influences the dependent variable, i.e online purchase intentions, while brand luxury has a moderating effect on the relationship between the variables. The third set of hypotheses refers to the direct relationship of perceived brand luxury with the online purchase intentions and the last one refers to the
indirect effect of the relationship between price promotion and additional costs on purchase intentions, which is moderated by brand luxury.
The expected relationship between price promotions, additional costs and brand luxury on online purchase intentions is depicted into the conceptual diagram below (Figure 4).
LITERATURE REVIEW Figure 4 Research Model
Price Promotion Online Purchase Intentions Brand Luxury Additional Costs H4 H1 H2 H6a – H6b – H6c H7a – H7b – H7c H5a – H5b – H5c H4 H8a - H8b - H8c
METHOD 3. Method
The following section introduces the preliminary part of the empirical research. First and foremost, the study’s targeted industry is presented and the reasons that led to its selection. Next, details about the pre-test and the chosen fashion brands will be given, including the data
collection procedure, the characteristics of the sample, and the measure development. Following this, the questionnaire’s variables and corresponding reliabilities are explained. The section concludes with the statistical approach that was adopted for the examination of the relationships between the variables (see section 2.4).
3.1 Fashion Industry
Among the four categories where luxury brands are evident, fashion accounts for the 42 percent of total sales (Fionda and Moore, 2009: 348). This percentage is not surprising,
especially if we take into account that fashion has become an important segment of large e-commerce companies, such as Amazon (Hahn and Pisani, 2016). Fashion industry is the most complicated and pricey of the four, mainly because of the need to “maintain a consistent brand image while providing a diverse and contemporary product assortment” (Fionda and Moore, 2009: 348), with costs and competition constantly rising. Therefore, firms must learn to become agile and to understand consumers’ needs in order to target the proper market segment and to make sound marketing mix decisions (Keller, 1993: 1). Although fashion industry has embraced digital technology (Okonkwo, 2010: 16), many luxury fashion brands are reluctant to go online fearing that if they do their products will lose the sense of exclusivity (Rowley, 2009: 365;
Cecilio, 2015; Rogerson, 2015). Considering that e-commerce has become an integral part of marketing strategy (Close and Kukar-Kinney, 2010: 986) and that still many luxury companies
METHOD cannot leverage internet’s true potential (Okonkwo, 2010: 21), there is an ever increasing need for becoming familiar with online consumer behavior and its relationship with shopping
orientations and socio-demographics. Since little empirical evidence exists with reference to “the creation and maintenance of luxury fashion brands” (Fionda and Moore, 2009: 360), this study aims to provide suggestions to luxury fashion brands on how to enhance their online marketing strategy, and specifically on how to leverage their promotional mix in order to increase
conversion rates.
3.2 Pre-test
3.2.1 Data Collection Procedure
The primitive step to examine the suggested hypotheses is to identify luxury and non-luxury fashion brands through a pre-test. Forasmuch as there are many fashion brands that belong to both categories, five brands were randomly selected. The purpose of the pre-test is to examine the perceived level of brand luxury and specifically to establish a hierarchical
distinction on a 5-tier luxury scale – low, medium-low, medium, medium-high, and high – for the following unisex brands: Diesel, Gap, Gucci, H&M, and Replay. Each respondent completed the rating for all brands – within subject design experiment (Field, 2009: 15) – and the order in which the brands were presented was systematically rotated to avoid possible order biases. Considering that the perceived level of luxury will yield the results that this study aims for, the gravity of this pre-test is self-evident.
METHOD 3.2.2 Sample
The data were collected from a convenience sample. A total of 32 respondents completed the pre-test out of 40, who started filling out the questionnaire (response rate 80%). Among the participants, 15 were male (Mage = 33.3, SD = 7.8, age range 23-46 years) of whom 9 (Vp = 60%) had completed a master’s degree. 17 respondents were female (Mage = 29.2, SD = 7.1, age range 22-46) of whom 10 (Vp = 58.8%) had completed a master’s degree. Taking all 32
respondents together, males represented 46,9 percent of the total respondents and females 53.1 percent respectively (Mage = 31.2, SD = 7.6, age range 22-46). A majority of 59.4 percent had completed a master’s degree and 28.1 percent a bachelor’s degree. The remaining 12.5 percent held doctoral, associate, or college degrees.
3.2.3 Measurement of Variables
In pursuance of high reliability and internal validity levels, the Luxury Brand Index scale of Sung et al. (2015: 126) was utilized (Cronbach’s α = .942). The measure consists of six
dimensions and 36 traits, which are indicative for perceived brand luxury. The respondents were asked to rate how descriptive these traits are for each and every brand by using a single answer Likert scale from a scale of 1 (not at all descriptive) to 7 (extremely descriptive). Finally, participants were asked to rank the importance of these traits by using a drag and drop rank order, with the most preferred item on top and vice versa.
METHOD 3.2.4 Statistical Procedure
An online survey was administered the period November 18th – 21st in order to obtain the necessary data. Eight (8) cases were recorded with missing values and were excluded on a listwise basis (Field, 2009: 654). Mean substitution could not be adopted as thanks to the force response that was applied to all survey questions the missing values corresponded to more than 60 percent of the total survey questions. Pairwise deletion was also not preferred to avoid bias into estimates and inconsistency in results. No recoding was deemed necessary, as no counter-indicative items were available.
Following the listwise deletion, a reliability analysis was performed to examine the findings’ consistency (Saunders, Lewis, and Thornhill, 2009: 156). Although the scale was already validated by Sung et al. (2015: 126) and has a Cronbach’s Alpha of .942, reliability analyses were run for excitement, sincerity, sophistication, professionalism, attractiveness, and materialism. This was done to establish a good correlation between the items and the scales’ score (.942), as well as to check if any of the items would have a negative impact on reliability if they were deleted (Field, 2009: 678). All six dimensions for all five brands have a Cronbach’s alpha > .7, which indicates high level of reliability (see Table 2).
The correlation analysis yielded a correlation matrix for all the trait combinations per brand and showed in total 59 significant positive correlations and 10 significant negative ones (see Table 3). A factor analysis, and specifically a confirmatory factor analysis, was going to be executed in order to verify scale construction and to evaluate its goodness, however, it was not deemed necessary as it was already implemented by Sung et al. (2015: 128).
Before proceeding with the normality test, new variables as a function of existing ones were created by calculating the mean of all dimensions per brand (means and standard deviations
METHOD can be found at Table 3). Following this, descriptive statistics and particularly skewness and kurtosis normality tests were conducted. For Replay, all items have a skewness statistic between -0.6 and 0,4 and a kurtosis statistic between -0.9 and 0.2, which indicates that the distribution might be flatter. For Diesel, all items have a skewness statistic between –1.0 and -0.4 and a kurtosis statistic between -0.6 and 0.9, which indicates the possibility of a negatively skewed distribution. For Gap, all items have a skewness statistic between -0.7 and 0.9, but the kurtosis statistic ranges from 0.1 to 2.8, which indicates that the distribution is flatter than normal. For
Gucci, all items have a skewness statistic between 1.0 and 0.2 and a kurtosis statistic between
-1.1 and 0.9, which indicates the possibility of a negatively skewed distribution. Finally, for H&M all items have a skewness statistic between 0.1 and 0.7 and a kurtosis statistic between 1.2 and -0.8, which indicates the possibility of a positively skewed and flat distribution. Based on Table 4, it could be argued that most items tend towards normal distribution, except from Ga_SinTOT,
METHOD Table 2 Cronbach’s Alpha scores – Pre-test
N=32 Cronbach’s alpha N of items M SD
D_ExcTOT .915 8 5.24 1.06 D_SinTOT .914 8 4.13 1.29 D_SopTOT .894 5 4.81 1.32 D_ProTOT .938 5 4.70 1.43 D_AttTOT .904 5 5.26 1.14 D_MatTOT .955 5 4.59 1.46 Ga_ExcTOT .888 8 3.56 0.88 Ga_SinTOT .897 8 4.10 0.99 Ga_SopTOT .901 5 3.80 1.06 Ga_ProTOT .791 5 4.01 0.91 Ga_AttTOT .889 5 5.26 1.14 Ga_MatTOT .924 5 3.30 1.15 Gu_ExcTOT .960 8 4.39 1.71 Gu_SinTOT .940 8 3.76 1.72 Gu_SopTOT .890 5 5.97 1.13 Gu_ProTOT .933 5 5.53 1.32 Gu_AttTOT .972 5 5.37 1.68 Gu_MatTOT .895 5 5.53 1.26 HM_ExcTOT .952 8 3.42 1.47 HM_SinTOT .937 8 3.70 1.48 HM_SopTOT .861 5 2.45 1.12 HM_ProTOT .930 5 2.96 1.38 HM_AttTOT .957 5 3.39 1.58 HM_MatTOT .921 5 2.21 1.09 R_ExcTOT .926 8 4.59 0.97 R_SinTOT .927 8 3.59 1.07 R_ProTOT .898 5 4.48 1.06 R_SopTOT .813 5 4.00 1.02 R_AttTOT .848 5 4.83 0.93 R_MatTOT .878 5 4.03 1.18
METHOD Table 3 Means, Standard Deviations, Correlations and Reliabilities per Brand – Pre-test
Variables - Diesel M SD 1 2 3 4 5 6 7 8 9 Gender 1.53 0.51 - Age 31.16 7.60 -0.27 - Education 5.53 0.80 0.16 -0.21 - D_ExcTOT 5.24 1.06 0.13 -0.25 0.27 (.915) D_SinTOT 4.13 1.29 -0.22 -.568** 0.10 .585** (.914) D_SopTOT 4.81 1.32 -0.02 -.357* 0.03 .608** .633** (.894) D_ProTOT 4.70 1.43 -0.31 -0.31 0.09 .630** .715** .804** (.938) D_AttTOT 5.26 1.14 -0.09 -0.19 0.22 .667** .524** .640** .732** (.904) D_MatTOT 4.59 1.46 -0.07 -.560** 0.05 .514** .683** .775** .642** .354* (.955) Variables - Gap M SD 1 2 3 4 5 6 7 8 9 Gender 1.53 0.51 - Age 31.16 7.60 -0.27 - Education 5.53 0.80 0.16 -0.21 - Ga_ExcTOT 3.56 0.88 0.09 0.01 -0.19 (.888) Ga_SinTOT 4.10 0.99 0.24 0.02 -0.13 0.21 (.897) Ga_SopTOT 3.80 1.06 0.07 0.04 -0.31 .460** 0.21 (.901) Ga_ProTOT 4.01 0.91 0.15 0.13 -0.18 .380* .486** .825** (.791) Ga_AttTOT 5.26 1.14 -0.09 -0.19 0.22 0.00 0.14 -0.19 -0.16 (.889) Ga_MatTOT 3.30 1.15 -0.23 -0.28 -0.10 0.16 -0.19 .391* 0.31 -0.05 (.924) Variables - Gucci M SD 1 2 3 4 5 6 7 8 9 Gender 1.53 0.51 - Age 31.16 7.60 -0.27 - Education 5.53 0.80 0.16 -0.21 - Gu_ExcTOT 4.39 1.71 0.07 -.543** 0.24 (.960) Gu_SinTOT 3.76 1.72 -0.12 -.458** 0.16 .864** (.940) Gu_SopTOT 5.97 1.13 .414* -0.24 0.13 .508** .367* (.890) Gu_ProTOT 5.53 1.32 0.20 -0.26 0.17 .676** .587** .794** (.933) Gu_AttTOT 5.37 1.68 0.22 -0.19 0.07 .600** .595** .756** .897** (.972) Gu_ExcTOT 4.39 1.71 0.12 -.414* 0.12 .549** 0.35 .511** .603** .546** (.895)
METHOD Variables – H&M M SD 1 2 3 4 5 6 7 8 9 Gender 1.53 0.51 - Age 31.16 7.60 -0.27 - Education 5.53 0.80 0.16 -0.21 - HM_ExcTOT 3.42 1.47 0.31 0.27 -0.07 (.952) HM_SinTOT 3.70 1.48 0.23 0.27 -0.22 .784** (.937) HM_SopTOT 2.45 1.12 0.08 0.34 -0.27 .708** .656** (.861) HM_ProTOT 2.96 1.38 0.18 0.31 -0.24 .766** .826** .866** (.930) HM_AttTOT 3.39 1.58 0.30 0.22 -0.15 .801** .763** .810** .839** (.957) HM_MatTOT 2.21 1.09 -0.23 0.13 -.578** .364* 0.33 .626** .502** .404* (.921) Variables - Replay M SD 1 2 3 4 5 6 7 8 9 Gender 1.53 0.51 - Age 31.16 7.60 -0.27 - Education 5.53 0.80 0.16 -0.21 - R_ExcTOT 4.59 0.97 0,31 -0,04 0,11 (.926) R_SinTOT 3.59 1.07 -0,09 -.494** -0,13 0,22 (.927) R_SopTOT 4.48 1.06 0,07 -.424* -0,03 0,26 .560** (.898) R_ProTOT 4.00 1.02 -0,12 -.382* -0,03 0,3 .747** .804** (.813) R_AttTOT 4.83 0.93 0,11 -0,14 -0,06 .411* .562** .446* .604** (.848) R_MatTOT 4.03 1.18 -0,17 -.473** -0,05 0,02 .514** .700** .569** 0,18 (.878)
* Correlation is significant at the 0.05 level (2-tailed). ** Correlation is significant at the 0.01 level (2-tailed).
METHOD Table 4 Normality Check – Skewness & Kurtosis – Pre-test
N Skewness Kurtosis
Statistic Statistic Std. Error Statistic Std. Error
D_ExcTOT 32 -0.464 0.414 -0.564 0.809 D_SinTOT 32 -0.319 0.414 -0.640 0.809 D_SopTOT 32 -0.999 0.414 0.994 0.809 D_ProTOT 32 -0.622 0.414 -0.617 0.809 D_AttTOT 32 -0.766 0.414 0.011 0.809 D_MatTOT 32 -1.020 0.414 0.460 0.809 Ga_ExcTOT 32 0.153 0.414 -0.470 0.809 Ga_SinTOT 32 0.983 0.414 1.525 0.809 Ga_SopTOT 32 0.151 0.414 0.613 0.809 Ga_ProTOT 32 0.084 0.414 0.109 0.809 Ga_AttTOT 32 -0.766 0.414 0.011 0.809 Ga_MatTOT 32 0.556 0.414 2.838 0.809 Gu_ExcTOT 32 -0.064 0.414 -1.127 0.809 Gu_SinTOT 32 0.238 0.414 -1.170 0.809 Gu_SopTOT 32 -1.042 0.414 0.343 0.809 Gu_ProTOT 32 -0.594 0.414 -0.292 0.809 Gu_AttTOT 32 -0.930 0.414 -0.004 0.809 Gu_MatTOT 32 -1.050 0.414 0.986 0.809 HM_ExcTOT 32 0.146 0.414 -1.209 0.809 HM_SinTOT 32 -0.156 0.414 -0.960 0.809 HM_SopTOT 32 0.571 0.414 -0.868 0.809 HM_ProTOT 32 0.053 0.414 -0.985 0.809 HM_AttTOT 32 0.111 0.414 -1.069 0.809 HM_MatTOT 32 0.715 0.414 -0.941 0.809 R_ExcTOT 32 0.463 0.414 -0.561 0.809 R_SinTOT 32 -0.617 0.414 0.143 0.809 R_ProTOT 32 -0.384 0.414 -0.722 0.809 R_SopTOT 32 -0.209 0.414 -0.976 0.809 R_AttTOT 32 -0.104 0.414 -0.108 0.809 R_MatTOT 32 -0.663 0.414 0.247 0.809 Valid N (listwise) 32
