86
Annemarie van Klaarbergen (357765) Msc. Marketing Management Coach: Dr. Y. M. van Everdingen Rotterdam School of Management Date: June 20 th , 2016 Co-reader: Prof. Dr. L. M. Sloot Rijksuniversiteit Groningen Consumer Perceptions on Exclusive Retail Distribution The Influence of Exclusive New Products on Store Choice and Brand Choice

The Influence of Exclusive New Products on Store Choice ... · offering a differentiated and therefore more attractive assortment. Assortment is defined here ... Exclusive Pringles

  • Upload
    ngotruc

  • View
    218

  • Download
    0

Embed Size (px)

Citation preview

Annemarie van Klaarbergen (357765) Msc. Marketing Management

Coach: Dr. Y. M. van Everdingen

Rotterdam School of Management

Date: June 20th, 2016

Co-reader: Prof. Dr. L. M. Sloot Rijksuniversiteit Groningen

Consumer Perceptions on Exclusive Retail

Distribution

The Influence of Exclusive New Products on Store Choice and Brand Choice

Preface The copyright of the master thesis rests with the author. The author is responsible for its contents. RSM is only responsible for the educational coaching and cannot be held liable for the content. ____________________________________ ________________________________ Date Signature

Acknowledgements First of all, I would like to give a special thanks to my coach Dr. Y. M. van Everdingen and coreader Prof. Dr. L. M. Sloot for their perfect guidance during the thesis process. I am very grateful for the way in which you supported me: with great expertise and kindness. A special thank you also goes to my family, who have always supported and motivated me during my studies, and gave me so many opportunities for which I am endlessly grateful. Further, my many thanks goes out to the friends that helped me by filling out questionnaires for the purpose of this thesis. Lastly, I would like to thank my colleagues working at the research agency Blauw Research and the sampling survey agency SSI because they helped me so well during the data collection process.

3

Abstract Since strict store loyalty is a rarity nowadays, and as assortment composition is a key determinant of store choice (SC), retailers try to differentiate themselves via their assortments and manufacturers are focused on differentiating their brand portfolio’s from competitors. While it is still a new phenomenon, exclusively distributing national brands through one sales channel is increasingly used in the FMCG as a differentiation strategy. However, a gap in the literature on exclusive retail distribution exists because the consumer’s perspective on exclusive products, which in turn influences SC and brand choice (BC), has never been measured yet. To fill this existing gap in the academic literature, the research objective of this study is to investigate how exclusive product offerings influence consumers’ SC and BC, and how these relationships are being moderated by the introduction price (price premium vs. no price premium) and the product category (utilitarian vs. hedonic). Among others, the study also controlled for BE and SE. 299 Dutch shoppers between the age of 18 and 65 participated in an online experimental survey, in which both the exclusivity of the new product, the introduction price and the product category were manipulated. Although the study did not find evidence for the exclusive new product significantly influencing store choice and brand choice (BC), controlling for store equity leads to a significant increase in SC. Moreover, when controlling for BE, the main effect between exclusivity and BC significantly increases. Mean scores for BC are (non-significantly) higher for the exclusive products both when BE is low and high. The relationships between exclusivity and SC and BC were not moderated by the product category. The main effect of exclusivity on BC was neither moderated by the product category. The relationship between exclusivity and BC was moderated by price, however this relationship is positive which means that both in the test and control condition, a normal price is preferred above the premium price. Moreover, a significant interaction effect of the product category and the test and control conditions on BC was found: exclusivity of the new product causes a higher BC for utilitarian products than for hedonic products and showing a non-exclusive product leads to a lower BC for utilitarian products than for hedonic products. Despite rejecting several of the hypotheses, the study triggers future research in multiple ways. Keywords: Retailing, Assortment, Exclusive Distribution, Fast Moving Consumer Goods, Store Choice, Brand Choice

4

Table of Content Preface ...................................................................................................................................................... 2

Acknowledgements .................................................................................................................................. 2

Abstract .................................................................................................................................................... 3

Table of Content ...................................................................................................................................... 4

1. Introduction .......................................................................................................................................... 6

2. Literature Review ............................................................................................................................... 10 2.1 Exclusive Products ...................................................................................................................... 10 2.2 Advantages and Disadvantages for Retailers and Manufacturers ............................................... 10

2.2.1 Advantages and Disadvantages for Retailers ....................................................................... 10 2.2.2 Advantages and Disadvantages for Manufacturers .............................................................. 11

2.3 Research on Launching New Products Exclusively .................................................................... 12 2.4 Determinants of Store Choice ..................................................................................................... 14

2.4.1 Overview of Possible Influencing Factors ........................................................................... 14 2.4.2 Product Factors ..................................................................................................................... 15 2.4.3 Market Factors ..................................................................................................................... 17 2.4.4 Situational Factors ................................................................................................................ 18

2.5 Determinants of Brand Choice .................................................................................................... 18 2.5.1 Overview of Possible Influencing Factors ........................................................................... 18 2.5.2 Store-related Factors ............................................................................................................ 20 2.5.3 Psychological Factors ........................................................................................................... 23

2.6 The Influence of Exclusive Products on Store and Brand Choice .............................................. 24 2.6.1 Store Choice, Brand Choice and Exclusivity ....................................................................... 25 2.6.2 The Moderating Effects of Price Premium and the Hedonic Level of the Product ............. 26 2.6.3 Control Variables Store Equity and Brand Equity ............................................................... 28

3. Methods.............................................................................................................................................. 30 3.1 Overview Pretests ........................................................................................................................ 30 3.2 Product Category Selection ......................................................................................................... 30 3.3 First Pretest .................................................................................................................................. 31

3.3.1 Retailer Selection ................................................................................................................. 32 3.3.2 New Product Selection for Soft Drinks ................................................................................ 32 3.3.3 Main Manipulation ............................................................................................................... 33

3.4 Second Pretest ............................................................................................................................. 33 3.4.1 Final Retailer Selection ........................................................................................................ 34 3.4.2 New Scent Selection for Detergents .................................................................................... 34 3.4.3 Price Manipulation ............................................................................................................... 35 3.4.4 Folder Design ....................................................................................................................... 36

3.5 Third Pretest ................................................................................................................................ 36 3.5.1 New Product Selection for Detergents ................................................................................. 37 3.5.2 Main Manipulation ............................................................................................................... 37 3.5.3 Folder Design ....................................................................................................................... 38 3.5.4 Manipulation Check Hedonic Level .................................................................................... 38

3.6 Experimental Design and Procedure ........................................................................................... 38 3.6.1 Subjects and Experimental Design ....................................................................................... 38 3.6.2 Procedure .............................................................................................................................. 39

5

3.7 Measures ...................................................................................................................................... 40 3.7.1 Dependent Variables ............................................................................................................ 40 3.7.2 Control Variables ................................................................................................................. 41

4. Analysis.............................................................................................................................................. 43 4.1 Normality and Outliers ................................................................................................................ 43 4.2 Manipulation Checks ................................................................................................................... 43 4.3 Descriptive Statistics ................................................................................................................... 44 4.4 Main Effect of New Product Exclusivity on the Dependent Variables ....................................... 45 4.5 Main Effect of New Product Exclusivity while controlling for Brand Equity and Store Equity 46 4.6 Moderating and Interaction Effects of Price Premium and Product Category ............................ 49 4.7 Additional Analyses .................................................................................................................... 52 4.8 Hypothesis Results ...................................................................................................................... 53 4.9 Control Variables ......................................................................................................................... 53

4.9.1 Extra Analyses on the Control Variables ............................................................................. 55

5. Discussion and Conclusion ................................................................................................................ 56 5.1 Managerial Implications .............................................................................................................. 58 5.2 Limitations and Future Research ................................................................................................. 58 5.3 Conclusion ................................................................................................................................... 60

References .............................................................................................................................................. 62 Internet resources ............................................................................................................................... 74 References of Excusive Products Cases ............................................................................................ 74

Appendix A Definition and Overview of Variables ......................................................................... 76

Appendix B Pretest Findings New Product Selection Soft Drinks .................................................. 77

Appendix C Pretest Findings New Product Selection Detergents ................................................... 78

Appendix D Folder Design for Eight Conditions ............................................................................. 79

Appendix E Final Sample Demographics ........................................................................................ 83

Appendix F Age and Gender Division Across Conditions .............................................................. 83

Appendix G Sample Background Variables Across Conditions ...................................................... 84

6

1. Introduction In today’s grocery retailing environment, strict store loyalty is a rarity. This is due to the increasing multi-store patronage, which refers to the number of stores that consumers patronize (Maruyama & Wu, 2014; Gielens, Gijsbrechts, & Dekimpe, 2014). As a response, retailers try to differentiate themselves. Because assortment composition is a key determinant of store choice (SC) and retail success (Grewal, Krishnan, Levy, & Munger, 2010), differentiation is focused on assortments. Most retailers carry similar categories, so retailers mostly differentiate their assortment within the category. Another differentiation strategy is called exclusive distribution. From a manufacturers perspective, this can be defined as severely limiting the number of intermediaries handling a company’s goods or services. From a retailer’s perspective, this means that the retailer exclusively offers key national brands through its sales channel. The strategy aims to attract additional consumers to the store by offering a differentiated and therefore more attractive assortment. Assortment is defined here as "the number of different items in a merchandise category" (Levy & Weitz, 1995, p. 30).

Exclusive retail distribution is already popular in the toy and apparel industry (Gielens et al., 2014). Until recently, it was still rarely used in grocery markets (Sudhir & Rao, 2006). However, more and more manufacturers launch new products through exclusive sales channels. In 2008, “Blue Band Goede Start” was the first premium product that was exclusively sold at Albert Heijn (Unilever, 2015) in the Netherlands. From that time on, other manufacturers, shown in Table 1, followed.

Table 1 Examples of Exclusive Product Launches in the FMCG

Exclusive product Retailer Year Manufacturer Category Country Hertog Jan Ongekend Jan Linders 2015 AB InBev N.V. Beer NL Capri Sun Fruit Crush Albert Heijn 2015 Coca Cola Beverages NL Grolsch Gerijpte Herfstbok Jumbo 2015 SABMiller Beer NL Persil Washing Powder WalMart 2015 Henkel Detergents US Velours Noir Albert Heijn 2013 Mondelez Coffee NL Hands Off My Chocolate Albert Heijn 2013 Hands Off B.V. Snacks NL Beanies coffee Sainsbury's 2013 Beanies Coffee UK Alvons Duschdas DM 2013 Unilever Body Care D Sisi Fruitbommetje C1000 2012 Vrumona Beverages NL De Klok Beer C1000 2012 SABMiller Beer NL Smith Kronkelzoutjes Albert Heijn 2012 PepsiCo Snacks NL Steamed mackerel Albert Heijn 2012 John West Fish NL Exclusive Pringles Morrissons 2010 P&G Snacks UK Designer Diapers Target 2010 P&G Toiletries US Becel bread Albert Heijn 2009 Unilever Bakery NL Gijs regional products (e.g. bread, dairy, fruit, vegetables) Plus 2009

Regional craftsmen Multiple NL

Vaseline Cacao Butter Body Lotion Tesco 2008 Unilever Body Care UK Blue Band Goede Start Albert Heijn 2008 Unilever Bakery NL Exclusive Pringles Walmart 2003 P&G Snacks US

Sources can be found in the literature list

7

Exclusively offering new products through a single sales channel is a remarkable choice of manufacturers operating in the fast moving consumer goods (FMCG) industry. Traditionally, they prefer to have a distribution coverage that is as intensely as possible to be able to increase sales (Coughlan et al., 2001). The FMCG is characterized by high purchase volume and thus by short shelf-lives of the products. Many buyers do not like to put too much effort into a purchase (Gielens et al., 2014). If the willingness to search for a firm’s products is low and distribution is restricted, sales probably decrease (Heide, 1994; Andritsos & Tang, 2010). Because most purchase decisions are in-store decisions (Dreze, Hoch, & Purk, 1994) and as consumers switch stores often (Shukla & Babin, 2013), it is extremely important to generate store traffic. Van Everdingen et al (2011) explain that manufacturers prefer a high distribution coverage because they need to earn back money spent on R&D and advertising. Because many new products in the marketplace fail in their first year and as developing new products is costly (Gielens & Steenkamp, 2007), it is not often sufficient to sell a product in a single country only (Golder, 2000). Let alone launching a product through one sales channel. Nevertheless, manufacturers increasingly launch new products exclusively. This study assumes that this may be due to the fact that adding exclusive products to retailers’ assortments and manufacturers brand portfolio’s (or assortments) is a strategic action that increases consumers’ positive assortment perceptions.

As researchers observe declining consumer loyalty and increased store switching among consumers (Wright & Riebe, 2010), since the amount of exclusive product offerings is increasing, and since exclusive products make the retailers’ assortment more unique (Briesch, et al., 2009), it is interesting to investigate whether exclusive products contribute to a higher SC. The study is also interesting for manufacturers, who invest much money to optimize their product portfolio’s. Just like for retailers, exclusivity is assumed to be a strategic tool to differentiate their assortments. But just like for SC, there is no research on brand choice (BC) and exclusive products yet. It would therefore be interesting to investigate whether exclusivity can lead to a higher BC. Since exclusivity is a strategic tool to make assortments more attractive, and as assortments turn out to be an important determinant for both SC and BC, this study’s focus is on assortments.

Offering more items in a category is costly for both the retailer and consumer (Ailawadi & Keller, 2004). Yet much resources are spent on assortment variety. On that account, and since assortment composition is a key determinant of SC and retail success (Grewal et al., 2010), the assumption was made that retailers need to know what influence exclusive products have on consumers’ SC. Manufacturers need to know whether exclusive products are attractive to make their assortments more attractive so they can use it as a strategic tool. Store uniqueness caused by a unique assortment and brand uniqueness caused by a unique brand portfolio are hereby the focus. In practice, new product launches are often combined with price discounts to increase both SC and BC. Because exclusivity can be a source of attraction in itself, it is assumed that a price discount is not needed. However, it would be interesting to investigate whether a price premium can be asked for exclusive new products without leading to a decreased likelihood to visit the store or to buy the brand. Furthermore, SC and BC in the context of exclusivity are likely to be different for various product categories because of the different benefits that they provide. Hedonic products, that deliver experiential consumption (e.g. pleasure and experience) are supposed to be the opposite of utilitarian goods, that provide functional and instrumental benefits (Dhar & Wertenbroch, 2000; Batra & Ahtola,

8

1990). Because consumers have a greater need for variety in hedonic categories (Van Trijp et al., 1996), it is assumed that they would value exclusivity more. Therefore, it is expected that the hedonic level of the product moderates the relationship between exclusiveness and store and BC. Because exclusive products contribute to both the retailers’ and the manufacturers’ positioning and assortment strategy, it would be a logical choice to investigate these strategic-aspect related moderating variables. Hence, the objective of the proposed study is to investigate how exclusive product offerings in the FMCG influence consumers’ SC and BC and how introduction price and the hedonic level of the product moderates these relationships. To achieve the research objective, the following research question should be answered:

How do exclusive product offerings influence consumers’ (a) store choice and (b) brand choice and how are these relationships moderated by the introduction price and the hedonic

level of the exclusive new product? Managerial Relevance. Manufacturers and retailers have other interests concerning shelf space. Manufacturers strive for sales and profit maximization of their products, while retailers want to maximize category sales and profits, regardless of brand identity (Dreze et al., 1994). Despite different interests of retailers and suppliers, this study is of managerial importance for both parties. Retailers need to know whether exclusive products impact store perception and therefore the willingness to visit the store more often (Gielens et al., 2014). In turn, this would make consumers more loyal and thus increase sales and profits in the long run. Manufacturers need to know whether consumers’ BC increases due to increased store and brand loyalty because of exclusive products. Exclusive offerings are most advantageous for retailers. They can enjoy benefits such as a higher retail price (Briesch et al., 2009; Hermalin & Katz, 2013), higher store traffic that could lead to spill-over effects when consumers also buy products from other categories, differentiation from competition because of a unique assortment and eventually new customers. Moreover, because the product is only available at one single store, price competition (that could drive prices down) between different retailers carrying the brand is avoided (Cheng, 2008; Andritsos & Tang, 2010). Overall, exclusivity is assumed to increase revenues and profits for retailers (Gielens et al., 2014). The situation for manufacturers is more complicated. On the one hand, exclusivity leads to more control of the exclusive product (Frazier & Lassar, 1996). On the other hand, investing into the relationship with the retailer is costly and has to be traded-off against investing into a broader market coverage (Frazier, 1999). However, if exclusiveness would lead to higher BC, the manufacturer would have more advantages too which makes this study important for them.

The outcomes of this research may give both retail and brand managers insight into whether exclusive product launches in the FMCG are generally a good idea. As exclusive products are part of the retailers’ assortment, this is of utmost importance since shelf space is one of the scarcest resources that retailers have (Kahn, 1999). Manufacturers invest a lot of money into launching new products exclusively, which makes the study relevant for them as well.

9

Academic Contribution. So far, some research has been done to exclusive channels (e.g. Gielens et al., 2014; Andritsos & Tang, 2010) and the implications of exclusive partnerships for both the manufacturer and the retailer (e.g. O'Brien & Shaffer, 1997; Marx & Shaffer, 2007). Only one study takes into consideration the consumer’s perspective on assortments that carry exclusive products (Van Everdingen & Ten Berge, 2016). This study makes four theoretical contributions. First, it is the first study within the academic marketing literature that sheds light on the consumers’ perspective on exclusive products. The study focuses thereby on the strategic aspects of exclusive products: both the dependent variables and the moderators can be used by manufacturers and retailers to make strategic managerial decisions. By using product category as moderator, the study is generalizable to more product categories which increases its strategic relevance. Second, the study enhances both retailers’ and manufacturers’ understanding of SC and BC in the context of exclusive new products (which is special because manufacturers and retailers often have conflicting interests). Therefore, this study makes an academic contribution to the SC and BC literature as well. Third, the author made an overview of possible definitions of the concepts SC and BC. A vast amount of studies address the topics SC, BC and purchase intention, but only a few pay attention to the determinants of SC and BC. Therefore, the author made an overview of possible influences on SC and BC focusing on the assortment, as this study investigates the impact of exclusivity as an assortment differentiator. This relationship between the assortment and SC and BC is not new, but the overviewing literature review about these concepts and the context of exclusivity is. Further, the study investigates the implications of the introduction price and the product category in which the exclusive new product falls, which have never been investigated before.

10

2. Literature Review This section provides a theoretical background about retail exclusivity. First, several definitions used in the literature for exclusivity are discussed, followed by the definition that is used in this study. Then, the advantages and disadvantages for retailers and suppliers of selling products exclusively are outlined, which is important to understand the strategic aspects of exclusivity. Thereafter, a short literature review about retail exclusivity is given, that results in the identification of the literature gap and the introduction of the conceptual model for this study. Further, the literature about the dependent variables and the two moderators is discussed. Like this, the reader is guided to the end of the section where the research hypotheses are introduced.

2.1 Exclusive Products To date, the term ‘exclusivity’ often appears in the economics, legal and marketing literature, while it describes different concepts. Most obviously, exclusivity refers to exquisite products that are valuable and rare (Joy et al., 2014). This explanation is often used in studies related to art and clothing industries (e.g. Sorescu et al., 2011). Secondly, some studies have used the term exclusivity to describe exclusive promotions (Grewal et al., 2011) or exclusive price discounts, customized to individual consumers (Barone & Roy, 2010). Thirdly, exclusivity can refer to exclusive distribution agreements between retailers and suppliers. Several distribution options exist, that vary from exclusive through selective to intensive distribution (Frazier, 1999). Exclusivity in this context refers to the ‘maximum degree of limitation’ (Fein & Anderson, 1997). Chang (1992) further differentiates between ‘downward’ and ‘upward’ exclusive dealing. The former is also called ‘territory exclusivity’ in the marketing literature, which is the degree to which the amount of parties operating in a geographic market is limited by the supplier (Fein and Anderson, 1997; Nair, Tikoo & Liu, 2009). The latter, also called brand exclusivity or exclusive retail distribution, is the exclusive right to distribute key national brands to consumers through a single sales channel (Gielens et al., 2014; O’Brien & Shaffer, 1997; Fein & Anderson, 1997). The latter definition is leading in this study, where the introduction of new exclusive products in grocery retailing is the central topic. A new product can be a new flavor, package or variant of an existing brand. For the purpose of this study, it should be remarked that retail exclusivity does not involve private label products, but branded products that are only available at one retailer.

2.2 Advantages and Disadvantages for Retailers and Manufacturers As the literature to date about exclusive products points out, introducing products through a single sales channel can have both advantages and disadvantages for retailers and manufacturers, which are outlined in this chapter. Retailers seem to have especially advantages. Findings about the supplier’s position are mixed. 2.2.1 Advantages and Disadvantages for Retailers Reviewing the literature about exclusive products especially reveals advantages for retailers. As the retailer is the only one that sells the exclusive product, and since assortments that carry exclusive products are more unique (Briesch et al., 2009), store traffic increases (Andritsos & Tang, 2010) which in turn leads to spill-over effects: consumers that visit the store now probably also buy products from other categories (Cachon & Kök, 2007). In case of an exclusive arrangement, price competition with other retailers is avoided. Hermalin & Katz (2013) show that both retailers and suppliers can earn higher profits on exclusive products.

11

Contrarily, findings by Gielens et al (2014) suggest that suppliers gain higher sales in a non-exclusive arrangement, while exclusivity is appealing for the retailer. This is in line with O’Brien & Shaffer’s (1997) study, who show that suppliers strive for an equilibrium without exclusivity, because that leads to higher profits. On the contrary, retailers prefer an equilibrium with an exclusive arrangement. A disadvantage is that the retail relies on only one manufacturer’s supply. On the other hand, the manufacturer assures that the retailer is supplied with the exclusive new product (O'Brien & Shaffer, 1997). In the long run, this can lead to a reliable relationship with the manufacturer. Peres & Van den Bulte (2014) show that there is another disadvantage that has consequences for both the retailer and the manufacturer: the firm that sells exclusive products cannot take advantage from positive word-of-mouth spillover effects from customers of rival firms. 2.2.2 Advantages and Disadvantages for Manufacturers In contrast to all the retailers’ advantages, the manufacturer’s position is more ambivalent. On the one hand, once a dealing arrangement is closed, shelf space is guaranteed (van Everdingen, 2015). O’Brien & Shaffer (1997) argue that in an exclusive arrangement, the manufacturer is assured of a reliable demand or supply at a given price. Moreover, if needed, the manufacturer can encourage the retailer to pay more attention to the launch of the exclusive product which is beneficial for the manufacturer’s product launch and its brand name. Frazier & Lassar (1996) say that distribution to a single retailer enhances the manufacturers’ control over transshipment, stocking levels and the pricing strategy compared to intensive distribution via many retailers. Further, in a non-exclusive setting, both consumers and manufacturers benefit from a lower retail price due to price competition among retailers. Andritsos & Tang (2010) showed that in an exclusive arrangement, when retail competition is reduced, the manufacturer can always set a higher wholesale price. The authors also show that the supplier has more control over the customer’s experience in the store. Finally, Fein and Anderson (1997) argue that an exclusive arrangement can serve as a signal for commitment to the retailer. But there are also disadvantages. Although extra store traffic is generated because of the exclusivity of the new product, there is a potential for lost sales opportunities because the manufacturer restricts its new product exposure to only one retailer (Cai, Dai, & Zhou, 2012). Now the success of the new product fully depends on this retailer. Frazier (1999) argues that investing into a long term relationship with such a retailer is costly (e.g. communication, channel personnel, specialized investments) so time is wasted if the new product fails. According to van Everdingen (2015), connections with other retailers might get harmed if they notice the partnership with the manufacturer. In the future, potential retailers with which the manufacturer can work together might avoid future cooperation. Table 2 gives an overview of the advantages and disadvantages of exclusive product offerings in a retail setting for both retailers and suppliers.

12

Table 2 Advantages and Disadvantages of Exclusive Product Offerings Advantages Disadvantages Retailer • Higher retailer price

(Briesch et al., 2009; Hermalin & Katz, 2013) • Unique assortment (Briesch et al., 2009) • More store traffic (Andritsos & Tang, 2010) • Spill-over effect (Cachon & Kök, 2007)

Manufacturer • Guaranteed shelf space (van Everdingen, 2015) • More control (Frazier & Lassar, 1996) • Product launch receives more attention

(O'Brien & Shaffer, 1997) • Higher wholesale price (Andritsos & Tang, 2010) • Signal for commitment (Fein & Anderson, 1997)

• Costly relationship investment (Frazier, 1999)

• Lost sales (Cai et al., 2012) • Damaged relationships with other

retailers (van Everdingen, 2015)

Both • Reliable supply/demand (O'Brien & Shaffer, 1997)

• Dependance on a single party (van Everdingen et al., 2011)

• Word-of-mouth from other distributors (Peres & Van den Bulte, 2014)

Since the overview of advantages and disadvantages of exclusivity in general has been given, in the following chapter this thesis deep dives into the existing literature on launching new products exclusively and further illustrates the study’s academic literature gap. 2.3 Research on Launching New Products Exclusively Several streams of literature have examined exclusive partnerships instead of retail exclusivity, such as the economics, marketing and legal literature. Studies are rarely empirical. The economics literature focuses on why exclusive partnerships exist and developed numerous analytical explanations for this (e.g. Hermalin & Katz, 2013; O’Brien & Shaffer, 1997). Most marketing literature describe exclusive partnerships as a marketing strategy (e.g. Sorescu et al., 2011) or explains possible advantages and disadvantages (e.g. Peres & Van den Bulte, 2014; Nair et al., 2009). Legal literature especially explains how exclusive partnerships affect marketplace functioning (e.g. Vickers & Waterson, 1991). However, not much research exists on exclusivity in grocery retailing, which is the main focus of this study. In this chapter, an overview of the studies that do touch this subject is given and serves as a guide to the conceptual model. First, Gielens et al (2014) constructed a model to assess the gains and losses for the manufacturer and retailer when a manufacturer exclusively launches a new brand line. In their model, sales gains resulting from (i) an exclusive distribution setting and (ii) an intensive distribution scenario including all leading retailers are compared to a pre-introduction scenario. Next to the finding that only retailers benefit from higher sales due to exclusivity because the manufacturer suffers from foregone sales, the authors’ results suggest that extra feature support for the exclusive product could make the exclusive deal beneficial for the supplier. Secondly, Peres & Van den Bulte (2014) focused on social and competitive elements to determine whether or not to sell exclusive products. They found that firms holding exclusive products don’t benefit from positive word-of-mouth of consumers from rival firms and therefore risk decreasing sales. The presence of loyal customers, opinion leaders that are loyal customers and low price-sensitive customers increase the firms’ advantages of not

13

offering exclusive products. Lastly, Van Everdingen & Ten Berge (2016) shed light on the consumer’s perspective of an assortment that contains exclusive products. This assortment was found to be more unique, more varied and more attractive. To the author’s knowledge, empirical studies other than this study do not exist yet. Table 3 gives an overview of the existing studies that exist on exclusive products in a retail setting. There exists a gap in the academic literature caused by the lack of empirical research and the lack of investigation into SC and BC, while it is clear that assortment is an important determinant of SC (Pan & Zinkhan, 2006) and BC (Berger et al., 2007). Table 3 Overview of Previous Exclusive Product Studies in Retailing

Study Data/analysis Dependent variable(s) (DVs)

Independent variable(s) (IVs)

Key findings

Gielens, Gijsbrechts, & Dekimpe (2014)

Household spending data in 53 retail categories from Gfk Benelux, analyzed by a proposed unified framework

Gains and losses for the retailer and manufacturer in several categories

Brand introduction under exclusive vs. intensive distribution

Intensive distribution causes higher sales for the supplier than an exclusive arrangement. Contrarily, exclusivity is appealing for the retailer. Extra feature support can make an exclusive product also profitable for the supplier

Peres & Van den Bulte (2014)

Both computer simulations and a game-theoretic model were used to analyze 900 customer responses to the introduction of a new product into a market Profitability Exclusivity

Firms carrying exclusive products miss out sales because of word-of-mouth from customers of rival firms. This is magnified by the presence of loyal customers, opinion leaders that are loyal customers and low price-sensitive customers

Van Everdingen & Ten Berge (2016)

Online experimental study in which consumer responses to an assortment with different facings and different shelf positions were measured

Perceived attractiveness, variety and uniqueness of the assortment

Exclusivity of new product

Both the perceived assortment’s attractiveness, variety and uniqueness increase due to exclusivity. The relationship between the DV and IV is not moderated by the number of facings, but it moderated by the shelf position (higher shelf position gives a higher score on all DV’s)

Van Klaarbergen (2016)

Online experimental study in which consumer responses to a promotion folder with an (exclusive) new product with different introduction prices (price premium vs. no price premium) and product categories (hedonic vs. utilitarian) were measured

Store choice and brand choice

Exclusivity of new product See chapter 5

14

By adding exclusivity to the assortment as an extra differentiator, Van Everdingen & Ten Berge (2016) made a contribution to the assortment literature. This study aims to fill the literature gap on assortments in several ways. First, it introduces a complete new framework with focus on exclusivity with the strategic variables SC and BC as dependent variables. Contrary to Van Everdingen & Ten Berge’s study, it also offers a more strategic view on exclusivity by introducing strategic moderators. This is interesting because next to the strategic independent variable, the moderators are about strategic aspects too. Further, it extends the assortment literature by adding the relationship between assortments and SC/BC, where exclusivity can be seen as an assortment differentiator. Where SC is especially important for retailers and where BC is interesting to measure for manufacturers, the study is relevant for both parties. Most often, because retailers and manufactures have conflicting interests, studies are only relevant for one party. The research is also interesting for empirical objectives, because next to Van Everdingen & Ten Berge’s study no empirical research exists on this topic in the academic literature yet.

In the following two chapters, a literature review about the dependent variables is given, in which the focus is on assortments, because exclusivity as strategic tool can change the consumers’ perception of the assortment and therefore is assumed to increase SC (chapter 2.4) and BC (chapter 2.5).

2.4 Determinants of Store Choice Since there is no research to the effect of exclusive products on SC, it is important to give a broad overview of possible influences on this variable. The focus is especially on the assortment, because exclusive products influence SC via the assortment. While assortments are only one of the determinants of SC, other possible influences on SC have been taken into consideration either as control variables or as moderators. In Figure 1, a framework is presented in which the influences on SC are outlined.

2.4.1 Overview of Possible Influencing Factors Many studies have investigated what determinates store choice (intentions or behavior) towards a retailer. Some of these studies examine idiosyncratic relationships specified on certain shopping modes (such as internet experience as a determinant of online shopping intention, Brengman & Geuens, 2002), which is out of the realm of the present study. Other studies are of a more empirical nature. Since SC is measured as store choice intention (SCI, see chapter 3.7.1), the concept is defined here as a consumer’s likelihood to patronize a retailer (Pan & Zinkhan, 2006). The terms SCI and store patronage intentions are both used in the marketing literature to refer to the same concept. The term SC is however used in this study to refer to SCI, to prevent confusion. Studies investigating SC predictors have used a variety of research methods and contexts, which have resulted in opposing evidence about relationship directions and strengths. Pan & Zinkhan (2006) have given a quantitative summary of empirical works on SC from the retail literature. They selected eleven antecedents on SC, put them into a formal meta-analysis and found that assortment had the strongest correlation with SC. Other predictors of SC – ranked on strength of correlation - are found to be service, quality, store atmosphere, low prices, location, fast checkout, convenient opening hours, friendly salespeople and convenient parking facilities. These antecedents are presented in Figure 1, where they are categorized

15

into product factors, market factors and situational factors. Thereafter, their relationship with SC will be discussed. Figure 1 Determinants of Store Choice

2.4.2 Product Factors In this section, the product factors ‘assortment’, ‘quality’ and ‘price’ will be discussed. Assortment. Retailers nowadays operate in an area of hyper competition. One way of differentiating themselves is to respond to heterogeneous consumer tastes by varying in the breath (number of brands) and depth (number of stock keeping units, SKUs) of their assortments (Dhar, Hock, & Kumar, 2001). Next to the fact that this variation is a response to a high demanding consumer, offering a varied assortment also makes the shopping task easier because this allows consumers to make comparisons with other products at the same store. As such, perceived costs such as effort and travelling time can be diminished (Dellaert, Arentze, & Bierlaire, 1998). ‘Product selection’ or ‘assortment’ has the highest average correlation with SC and therefore seems to be the most important driver of SC (Pan & Zinkhan, 2006). In other words, if consumers have a positive assortment perception, their SC seems to rise. However, findings about about this subject in the literature are mixed. In general, two research streams describe a different effect of assortment perceptions on SC. The first research stream states that retail patronage positively relates to assortment size, but inversely relates to the distance to the consumer’s home (e.g. Huff, 1964; Brown, 1989). This theory believed that consumers prefer stores with large assortments. First, because a large assortment increases the likelihood that consumers find the product that exactly matches their needs (Baumol & Ide, 1956). Secondly, because a large assortment can fulfill needs from consumers that like variety seeking (McAlister & Pessemier, 1982; Kahn, 1995) and thirdly because variety may increase consumers’ experience of utility (Kahn, Moore, & Glazer,

Product factors (2.4.2)

Assortment

Quality

Low price

Market factors (2.4.3)

Service

Store atmosphere

Convenience (location, opening hours,

parking)

Fast checkout

Friendliness of salespeople

Situational factors (2.4.4)

Shopping task

16

1987). And more authors emphasize the fact that assortment variety is a key driver of consumer satisfaction, store image and SC (Hoch, Bradlow & Wansink, 1999; Ailawadi & Keller, 2004; Fox, Montgomery & Lodisch, 2004). Briesch et al (2009) explain that the importance of assortment variety is a logical outcome of today’s retail formulas since most supermarkets carry the same product categories. Hence, assortments determine almost the entire variation in product selection across stores. The authors also investigated the effect of different forms of assortment variety on SC. They found that assortment breath positively influences SC and concluded that SKUs per brand, the amount of SKUs that are unique to that store (with which private label brands are meant) and sizes per brand, negatively affect SC.

Contrarily, other authors suggest that SC is determined by the consumer’s perception on assortment variety (Kahn et al., 1987), which can be influenced by many different factors, such as the presence or absence of a favorite product (Broniarczyk, Hoyer, & McAllister, 1998), the assortments’ arrangement (Hoch et al., 1999), the amount of good alternatives (Kahn & Lehmann, 1991) and the category space. Dreze et al (1994) did not find increased grocery sales due to assortment size. When 10% of less popular SKUs were deleted from the assortment and more shelf space was dedicated to popular items, sales even went up by 4%. This is in line with findings from Boatwright & Nunes (2001), who state that grocery retailers are over-assorting their stores. Assortment size negatively affected category sales in their study, where an internet grocer was used for measurement. The mixed findings in the literature on assortment perception and SC might be the result of different definitions of the assortment, which makes it important to specify exactly what is meant by this term.

There are more conflicting results about the importance of assortments. While Arnold, Oum & Tigert (1983) found that shoppers ranked assortments third as a SC criterion behind location and low prices, more recent research where household-level market basket data was used, found that assortment changes are more important in SC decisions than price (Briesch et al., 2009). This is in line with research from Stassen, Mittelstaedt & Mittelstadt (1999), who show that differentiated assortments are as important as variables like price and location as a SC predictor, if not more important.

Quality. The traditional SC literature generally agrees that perceived merchandise quality positively relates to SC (Darley & Lim, 1993; Jacoby & Mazursky, 1985; Olshavsky, 1985). This should be true because product quality partly determines the retailer’s reputation. According to Darden & Schwinghammer (1985), it also works the other way around: the authors found that the store type where the product is bought influences the consumers’ perceived product quality. More recent studies found that merchandise quality positively influences the perceived merchandise value, which in turn influences the consumer’s SC (Grewal, Baker, Levy, & Voss, 2003; Baker, Parasuraman, Grewal, & Voss, 2002).

Low Price. Argued from a consumer perspective, price is ‘what is given up or sacrificed to obtain a product’ (Zeithaml, 1988, p. 10). A common thought is that low prices lead to more retail purchases (Tigert, 1983; Rockney & Rinne, 1986). However, findings about the relationship between price and product quality perceptions are mixed. Some authors say that price positively affects perceived quality (Rao & Monroe, 1988; Dodds, Monroe & Grewal, 1991). Consumers may therefore choose a retailer and buy high-priced products because those products are expected to have an enhanced quality (Tellis & Gaeth, 1990). Pan & Zinkhan

17

(2006) however found the general price level in a store to be negatively related to SC. This can be explained by findings from Dodds et al (1991), who show that the price-quality relationship weakens if cues such as store and brand name are available. Price negatively affects the perceived product value and willingness to buy.

Exclusive products can be used by retailers as a strategic tool: they can attract customers to their stores. Since price is an instrument that can be used for strategic objectives too, it is an interesting variable to take into consideration in this study. Price potentially moderates the relationship between exclusive products and SC/BC: as price is a quality cue, a higher price of an exclusive product impacts the consumers’ willingness to buy the product. Price therefore potentially changes both the nature and strength of the relationship between exclusiveness and SC/BC.

2.4.3 Market Factors Market factors are determinants that retailers compete on. Reviewing the literature, these factors turn out to be service, friendliness of salespeople, store atmosphere, convenience and fast checkout. They will be discussed in this paragraph. Service and Friendliness of Salespeople. Research showed that service quality is directly linked to SC (e.g. Sirohi, McLaughlin, & Wittink, 1998; Zeithaml, Berry, & Parasuraman, 1996). This is in line with findings from Baker et al (2002), who found a direct significant link between interpersonal service quality (e.g. in-store treatment by salespeople, personal attention, prompt service) and store patronage intentions. Friendliness of salespeople has also been found to be a determinant of SC (Pan & Zinkhan, 2006). More traditional literature explains this by the fact that consumers visit stores for social experience (Tauber, 1972) or to diminish feelings of loneliness or boredom (Rubenstein & Shaver, 1980). Store Atmosphere. Store atmospherics are about physical aspects of the store, such as store layout and methods of displaying merchandise. Although many studies focus on the store atmosphere and the related store choice intentions of shopping centers (e.g. Finn & Louviere, 1990; Gentry & Burns, 1977), other research focus on smaller retail stores and found that store atmosphere negative influences SC (e.g. Grewal et al 2003; Baker, Grewal, & Levy, 1992; Van Kenhove & Desumaux, 1997). Baker et al (2002) found that design perceptions such as attractive facilities and organized merchandise in a card-and-gift store influenced SC via an enhanced merchandise value perception. An unpleasant perceived atmosphere in this store negatively influenced SC. Convenience (location, opening hours, parking). The consumer perception on convenience (e.g. location, opening hours, parking facilities) positively relates to their satisfaction about service. Moreover, this relationship has been shown to be moderated by the consumer’s invested time and effort (Berry, Seiders, & Grewal, 2002), which can be influenced by the store. These findings correspond with conclusions from a more traditional study by Bellenger, Robertson, & Greenberg (1977), who showed that an easier accessible shopping center highly correlates with its selection. Thelen & Woodside (1997) also discuss the importance of convenience. They asked Austrian households to identify several store attributes of supermarkets. In the study, store attributes were shown to respondents and supermarkets that automatically popped up in the consumer’s mind were recorded. Associations with supermarkets are useful in the purchase or consumption situation. Three out of four respondents said that their primary store was closest to their residence. More than 8 out of 10

18

respondents said that their primary store had the most convenient opening hours. 65% of the consumers’ primary stores were the easiest to reach by car. Fast Checkout. Grewal et al (2003) investigated how waiting time influenced consumers’ SC at a jewelry store. According to them, waiting time negatively affects SC. Despite the fact that interaction with salespeople is necessary to be able to even try or touch the jewelry, the situation at the jewelry can be compared to waiting at the checkout in a supermarket. The principle of waiting as negative experience is the same. Other research has also shown that if consumers expect to be forced to spend more time in a store due to a long waiting time, they may avoid entering that store (e.g. Hul, Dube, & Chebat, 1997). 2.4.4 Situational Factors Many aspects of the situation in which a consumer is, can influence his or her behavior. One of the aspects that was mentioned most often in the literature is the consumer’s shopping task. Shopping Task. The SC decision is also influenced by the shopping task, which is the set of goals formed by the shopper to resolve the needs derived out of a specific situation (Marshall, 1993). The literature refers to these kind of situational conditions and the arising shopping strategies as ‘task definitions’. Van Kenhove, De Wulf, & Van Waterschoot (1999) studied SC for do-it-yourself (DIY) products and discovered that the SC decision varied for different tasks like urgent purchase, regular purchase, large quantities, difficult job and get ideas. Whereas the task ‘difficult job’ refers to special material needed for DIY products, the other tasks would be appropriate in a grocery context too. Thelen & Woodside (1997) concluded that the task type determines the consumers’ primary SC. Their research shows that consumers have specific goal-derived store categories in their long-term memory, which are used when consumers have to decide to go to which store. In the following chapter, the determinants of brand choice (BC) will be discussed so that it becomes more clear which relationship this concept could have with exclusivity. 2.5 Determinants of Brand Choice Just like for SC, there is no research on BC and exclusive products yet. To gain more insight into the topic, a literature review about possible influencing factors on BC (intention) and purchase intention is therefore given in this chapter. Again, the focus is on the assortment because exclusive products as a strategic tool are assumed to influence BC via the assortment. Because assortments are only one determinant of BC, other possible determinants have been taken into consideration as control variables or as moderators in this study. An overview of these factors, which seem to overlap partly with those of SC, is given in Figure 2. 2.5.1 Overview of Possible Influencing Factors A brand is defined as ‘a name, term, sign, symbol or design, or a combination of them which is intended to identify the goods and services of one seller or a group of sellers and to differentiate them from those of competitors (Kotler, 1997, p. 443). Brands are thus used for differentiation, which explains why manufacturers in todays’ competitive environment focus so much on building their brands. SC is a cognitive process that shares many similarities with BC. Sinha & Banerjee (2004) explain that the only difference between the two terms is the spatial dimension: SC depends very much on location (Fotheringham, 1988), whereas BC does not. BC is more influenced by mental processes and information search, which will be discussed in chapter 2.5.2. The

19

variable is defined here as the likelihood that a consumer buys products from a certain brand. Because no research exists on BC and exclusive products, the literature review about this construct focuses on its determinants. This may be helpful in understanding how exclusive products can influence BC. Defining Brand Choice. BC in the marketing literature has often been measured using sales data (e.g. Pauwels, Hanssens, & Siddarth, 2002) or electronic scanner data (e.g. Guandagni & Little, 1983) as input for a variety of models. In these sort of studies, BC is defined as the actual ‘customer choice among product alternatives in a product category’ (e.g. Guandagni & Little, 1983). The focus lies on the action of purchasing a brand. Others defined BC as the ‘willingness to buy’ a product (e.g. Dodds et al, 1991; Dodds & Monroe, 1985), also called ‘purchase intention’. BC differs from purchase intention in the sense that purchase intention focuses on the likelihood of purchasing one specific product instead of a product from a certain brand. In this study, both terms have been combined so that BC is defined here as ‘the likelihood that a consumer buys products from a certain brand (based on Dodds et al., 1991). This concept gives insight into how an exclusive new product influences the consumer’s perception of the whole brand. Because it is an empirical study, no sales data was taken into consideration but the focus is on the intention to buy a brand. Determinants of Brand Choice. To get a broad overview of possible determinants of BC, the author discusses the literature about both BC (intention) and purchase intention. Currently, to the author’s knowledge, the literature does not provide a meta-analysis on BC. BCI even has not been measured as a dependent variable before. Therefore, the author reviewed the marketing literature on purchase intention and brand choice (intention) herself resulting in the identification of determinants of BC that are most frequently mentioned. These possible determinants of BC can be divided into store-related and psychological factors, which are presented in Figure 2. The concepts ‘brand familiarity’ and ‘information search & product categories’ have a rather pale color, because they are indirect influencers of BC. However, they should be explained to understand the mental processes in the consumer’s mind when making BC decisions.

20

Figure 2 Determinants of Brand Choice

2.5.2 Store-related Factors Factors that influence the buyer’s BC and that are related to the store, are called store-related factors in this study. Next to the assortment, products’ prices and brand promotions, there is a more psychological process that is related to the store’s assortment and that has to be explained to further understand the consumer’s buying process. Therefore, next to the determinants of BC, brand familiarity, information search and the moderator ‘product category’ will be discussed in this section as well. Assortment. As explained in chapter 2.4.2, a common thought in the marketing literature is that consumers prefer larger assortments (Chernev, 2003). This is in line with economic theories stating that larger assortments potentially better match consumers’ preferences (e.g. Lancaster, 1990). How about the brand’s assortment, or in other words, the product portfolio? According to Berger, Draganska & Simonson (2007), variety in a brand’s product portfolio is positively related to purchase likelihood. This can be explained by the fact that variety serves as a quality cue, for instance because manufacturers that offer more products should have category expertise. So in case a consumer has no prior brand preference, when this consumer has to choose between two brands and one of those brands offers more variety, it is likely that this brand will be chosen. This is in contrast with research about the variety of a store’s assortment. On the one hand, Iyengar & Lepper (2000) found that when choosing from a more varied assortment, the decision-making process was enjoyed more. However, people also experienced difficulty in choosing, felt greater frustration and were less likely to

Store-related factors (2.5.2)

Assortment

Brand Familiarity

Information Search & Product Categories

Price

Brand Promotions

Psychological factors(2.5.3)

Brand Credibility

Brand and Store Name

Attitude towards the brand

Psychological Proximity

21

purchase products. And more studies argue that a large assortment negatively influences purchase intentions because assortment variety could lead to uncertainty and choice overload, and thus decrease the brand’s purchase likelihood (Greenleaf & Lehmann, 1995; Tversky & Shafir, 1992; Iyengar & Lepper, 2000). When consumers know exactly what they want (Hoch et al., 1999), more variety could lead to negative assortment perceptions that may negatively impact BC. Other studies found that more variety is not always needed. The authors saw aggregate sales rise when less popular items were eliminated from the assortment (Dreze et al., 1994; Boatwright & Nunes, 2001). The fact that the literature on BC describes both the positive and negative impact of assortment variety, makes it interesting to investigate the effect of exclusive products on consumers’ assortment perceptions.

Brand Familiarity. It is favorable for a brand to be familiar, because brand familiarity leads to BC in the end. Brand familiarity has been shown to positively influence brand attitude (e.g. Laroche, Kim, & Zhou, 1996), which in turn influences BC. This can be explained by the fact that people like stimuli (such as brands) more if they are repeatedly exposed to it (Zajonc & Markus, 1982). Moreover, when a consumer is familiar with a brand, information search automatically focuses on the familiar brand alternatives (Avery et al., 2007). To understand the role of brand familiarity, it should first be explained how the consumer decision making process of choosing a brand works (Narayana & Markin, 1975). All brands available in the market are called the total set. The consumer only knows a subset of these brands, mostly those that are available in the shelf, called the awareness set. The assortment is therefore a determining factor again. Then a critical phase follows: the consumer decides which brands meet his or her initial buying criteria and thus starts to seek for information. The consideration set is the subset of brands that consumers actually consider when making a BC (Murthi & Rao, 2012). After elaboration about the brands in the consideration set, the consumer makes a final choice from the choice set, the subset of brands that have remained interesting after information search and elaboration.

Information Search and Product Categories. How much effort a consumer puts into information search and elaboration about the purchase depends on how involved the consumer is with the product category (Voss, Spangenberg, & Grohmann, 2003). Two types of product categories exist, based on the reason why people buy products: hedonic goods deliver sensations, fun, pleasure and experience at the moment of consumption. Flowers, ice cream and chocolate fall into this category. This product category can be considered as the opposite of products with utilitarian benefits (hereafter called ‘utilitarian products’), which are necessities bought primarily because of instrumental, effectivity, functional reasons. Examples are detergents, margarine and toilet paper (Khan, Dhar, & Wertenbroch, 2005). Research suggests that different independent components of products can be either hedonic or utilitarian, implying that product categories are relatively hedonic or utilitarian in nature (Batra & Ahtola, 1990). Following Dhar & Wertenbroch (2000), it should be noted that in this study hedonic product categories refer to categories that are superior on the hedonic dimension and utilitarian product categories refer to categories that are superior on the utilitarian dimension. Some products such as shampoo fulfill both hedonic (nice smell) and utilitarian (the need to wash hair) benefits.

Voss et al (2003) found that attitudes towards hedonic products are formed by affective involvement, which means that the consumer expends emotional energy in the purchase

22

decision. Attitudes towards utilitarian products are formed by cognitive involvement, meaning that consumers find it interesting to think about the purchase. Because attitudes towards both product categories are formed in a different way, the product category may also affect the consumers’ view about exclusive products in that category. Therefore, the hedonic level of the product is assumed to be a moderator of the relationship between exclusiveness and SC and BC1. Attitudes towards the category can be also measured as ‘product category involvement’ (Mittal, 1989), which has been taken into consideration as control variable. Since different brand familiarities for products have been proved to influence BC as well, this variable will be included as control variable as well (see chapter 2.6 for the conceptual model).

Price. The impact of price on BC has to do with reference pricing, which is the allowance to compare prices among. Reference pricing is related to prospect theory, which states that consumers don’t merely judge the absolute price, but rely on evaluation against reference prices for their decision making (Kahneman & Tversky, 1979; Russell, 1986; Bolton & Shankar, 2003, van Oest, 2013). Consumers’ evaluation against reference prices implies that consumers experience a loss (gain) if they pay more (less) than the reference price. Current shelf prices are called ‘external reference prices’. The internal reference price is formed by past experiences and exists in the consumers’ mind (van Oest, 2013).

Grewal, Monroe & Krishnan (1998b) used the context of price comparison advertising and found that perceived product value2 positively influences purchase intention. Grewal, Krishnan, Baker & Borin (1998a) found that this perceived product value is positively influenced by the brand’s quality, which in turn influences purchase intention. However, consumers have scare resources and can not always buy the most expensive product with the highest quality. They thus need to make trade-offs. It would therefore be interesting to see how exclusive products influence BC when price is taken into consideration as a moderating factor.

Brand Promotions. Guadagni & Little (1983) were the first authors that modeled BC as the utility of a brand and described it as a function of its display and feature (which are both promotional variables), loyalty variables and price. Many authors followed their example and found promotional variables to be predictors of brand utility and thus of BC (e.g. Gupta, 1988; Bucklin & Lattin, 1991). Inman, McAlister, & Hoyer (1990) found that even though promotion material does not communicate price reductions, it does function as a cue for a temporary reduced price and thus increases BC. Namely, brand promotions help to define the consideration set by excluding non-promoted brands (Fader & McAlister, 1990). With a promotion, search costs are reduced and the promoted brand is more likely to become the final BC.

1 To be able to suggest a negative or positive relationship of the moderator, the variable name ‘hedonic level’ was chosen. Later on, the variable will be also referred to as ‘product category’ with the levels ‘utilitarian’ (relative low hedonic level) and ‘hedonic’ (relative high hedonic level). 2 Product value is based on price comparisons with internal reference prices.

23

2.5.3 Psychological Factors The last determinants of BC that will be discussed have a more psychological nature: brand credibility, brand/store name, attitude towards the brand and the psychological proximity to the brand are reviewed in this chapter. Brand Credibility. The consideration set is influenced by brand credibility as well, which has two components: brand trustworthiness and expertise (Erdem & Swait, 2004; Swait & Erdem, 2007). In case of credible brands, consumers expect lower costs due to lower information costs. Moreover, the perceived risk is smaller for credible brands. This accounts for a broad range of products, including grocery products (juice and hair shampoo were included in Erdem & Swait’s study, 2004).

Favorable Brand and Store Name. In the supermarket, consumers mainly use heuristic search strategies to make decisions. These are quick rules of thumb or shortcuts such as recommendations, past experiences or brand names (Avery et al., 2007) that influence BC in the end. Dodds et al (1991) studied how price, brand name and store name influence consumers’ product quality perceptions, value perceptions and purchase intention. They conclude that favorable brand and store names cause enhanced quality and value perceptions, and therefore positively influences purchase intention. A favorable brand name in their study had a high perception of quality and is familiar and knowledgeable among consumers. A favorable store name also had a high perception of quality and met the criteria of being perceived as a store with high quality and high store satisfaction. This is in line with research conducted by Jacoby, Szybillo, & Busato-Schach (1977). They found that prior to making purchase decisions, consumers used less information on the package when brand name was available as information cue. In contrast, Dodds & Monroe (1985) did not find a significant positive effect between brand name and purchase intention. Attitude towards the brand. Brand attitude, which is an ‘individual’s internal evaluation of the brand’ (Mitchell & Olson, 1981, p. 318) seems to be positively related with the intention to buy a particular brand (e.g. Spears & Singh, 2004; Laroche et al., 1996). A negative attitude towards a brand is negatively related with BC (Laroche & Brisoux, 1989). This means that a favorable attitude towards a brand is linked to the behavioral intention to purchase the brand. Attitude are enduring and are formed by e.g. advertisements, emotional feelings and past experiences, which are differently interpreted by individuals. Psychological Proximity. Consumers often buy brands because they are attracted by the brand’s image. Some brands stand for status or power and individuals may be triggered by these brands because its purchase enhances the consumer’s status in turn. Psychological proximity, which means ‘matching the brand profile with consumer self-perception’ turned out to be important to determine store BC (Baltas, 1997), which is therefore also assumed to be important for ‘normal’ BC. The literature review in the previous chapters, combined with the reason for this study, has been the basis of the conceptual model that will be presented in Figure 3 in chapter 2.6.

24

Price Premium (M1)

Store Choice (DV1)

Brand Choice (DV2)

Hedonic Level (M2)

Exclusivity of New Product

(IV)

H1 (+)

H2 (+)

H3a (-) H3b (-)

H4a (+) H4b (+)

H5a

H5b

Demographic Control Variables

Age

Gender

Nationality

Behavioral Control Variables

Product Category Involvement

Brand Familiarity

H6

Buying Frequency

2.6 The Influence of Exclusive Products on Store and Brand Choice Prior research doesn’t provide insight into how the consumer’s perspective in terms of SC and BC are affected by a new exclusive product. This study aims to answer the question how exclusivity affects these variables and how these relationships are moderated by the introduction price (chapter 2.4.2 ‘low price’ and 2.5.2 ‘price’) and the hedonic level of the new product (see chapter 2.5.2 ‘information search and product categories’). Some of the control variables that were added to the framework have been explained in chapter 2.4 and 2.5. The control variables brand equity and store equity will be discussed in chapter 2.6.3 and other control variables are discussed in chapter 3.7.2. The conceptual model for the study is presented in Figure 3:

Figure 3 Conceptual Model

Extra Control Variables

Brand Equity Store Equity

25

2.6.1 Store Choice, Brand Choice and Exclusivity Consumers respond to retailers’ marketing activities in terms of evaluations, preferences and eventually behavior, such as SC and BC (Hartman & Spiro, 2005). It is this behavior that is the outcome of all the retailers’ effort and that causes sales eventually. SC and BC are therefore important variables to measure for both the retailer and the manufacturer. This section explains why it is expected that retail exclusivity leads to a higher SC and BC.

Store Choice. Because assortment is a key determinant of SC (Briesch et al., 2009), it is not surprising that retailers try to differentiate themselves through their assortments in order to gain more store traffic. Traditionally, retail differentiation was generated by offering private label (PL) products or store brands. Some studies found that PLs help generating store differentiation, loyalty and store traffic (Corstjens & Lal, 2000; Ailawadi, Scott, & Gedenk, 2001). Baltas, Argouslidis, & Skarmeas (2010) found that store brand proneness, which reflects the overall likelihood of purchasing PLs (Richardson, Jain, & Dick, 1996), diminishes the amount of stores that are patronized. However, Ailawadi, Pauwels & Steenkamp (2008) found that if retailers push their PLs too much, eventually this leads to negative returns. The authors explain that this is due to an inverted U-shaped effect between the amount of private label products offered (PL share) and share of wallet3. Consumers that buy no PLs at all are not loyal to the store. Consumers that buy some PLs from a chain probably build some loyalty and people that buy a lot of PLs are more focused on savings than on the store itself and are thus not loyal either. This is a huge problem because non-loyal consumers switch stores often. One consequence of the increased store-switching in the FMCG (Shukla & Babin, 2013), is that retailers now obtain exclusive agreements with suppliers as an alternative way to differentiate their assortments (Gielens et al., 2014). Unilever’s customer development director Erik Bras confirmed that innovative ways of value creation such as exclusive products are needed in a sector where sales are under pressure (Bras, 2003). Just like PLs have always done, exclusive products could attract consumers to the store as well. Since their availability is small, their quality is high and as assortment is an important SC indicator, combining this with the fact that exclusive products increase the assortment’s attractiveness (Van Everdingen & Ten Berge, 2016), it is expected that exclusive products lead to SC. Since exclusivity is still a growing phenomenon in grocery retailing and as the amount of exclusive products is still small, the effect of the amount of exclusive products in the shelves will not be taken into consideration in this study. H1: An exclusive new product has a positive effect on store choice.

Brand Choice. BC is likely to be influenced by exclusive products as well. The assumed relationship between exclusivity and BC follows a same logic as the one for the relationship between exclusivity and SC. A varied product portfolio positively relates to purchase likelihood (Berger et al., 2007) because variety serves as a quality cue. However, this accounts for consumers who have no prior brand preference. As explained in chapter 2.5.2, two contrasting beliefs theories regarding BC and assortment variety are found. One stream found assortment variety to make the decision process more enjoyable and a second stream

3 Share of wallet refers to the euros spent on PLs as a percentage of a consumer’s total spending on supermarket products.

26

found that variety leads to choice overload and therefore to a decreasing purchase likelihood. Despite these findings, the author adopts the first mentioned theory and expects a more varied brand portfolio due to an exclusive new product to increase the overall brand liking and therefore the BC. H2: An exclusive new product has a positive effect on brand choice.

2.6.2 The Moderating Effects of Price Premium and the Hedonic Level of the Product Price, brand name and store name are three external cues that influence product quality and value perceptions, and therefore willingness to buy (Zeithaml, 1988). Rao & Monroe (1988) found that both price and brand name have a moderating effect on consumers’ quality perceptions. Hence, it can be expected that price has a moderating effect on SC and BC. Because there is no existing research on how product related factors influence the perception of exclusive products and assortments yet, it would be interesting to investigate how these factors impact SC and BC. Furthermore, because consumers have a different view on purchase decisions depending on the product group, it can be expected that the hedonic level of the product moderates the relationship between exclusivity and SC and BC as well.

Price Premium. Price discounts usually cause increasing category sales and generate store traffic (Putsis & Dhar, 1999). However, one might argue that a price promotion is not needed if a new product is exclusively introduced at one retailer, because the exclusiveness is an incentive in itself to buy the new product. Furthermore, a new exclusive product doesn’t compete on price with other retailers because other retailers don’t sell that exclusive product (Cheng, 2008; Andritsos & Tang, 2010). Therefore, it is assumed that a price promotion such as a discount is not needed per se. Where retailers often use both skimming strategies (prices start high and slowly drop over time) and penetration strategies (using a low introduction price to serve the mass market) as pricing strategies for new products (Kotler & Keller, 2012), it would be interesting to see whether retailers can ask a price premium for exclusive new products. Such a pricing strategy would offer long term price premiums and therefore increasing sales which is of interest for both the retailer and the manufacturer. In general, consumers prefer low prices. High prices lead to a lower willingness to buy because it increases immediate costs (Dodds et al, 1991; Walters & Rinne, 1986). But in contrast to economic theory, consumers often do not buy the lowest priced product in the category either (Ding, Ross, & Rao, 2010). One behavioral explanation for this is that consumers infer quality information from prices (Lichtenstein & Burton, 1989; Monroe & Krishnan, 1985; Lichtenstein, Bloch, & Black, 1988). Cooper (1969) argues that consumers get suspicious of a product’s quality if the price is too much below what is considered as an acceptable price. An unacceptable high price leads to the inference that an offer has little or no net perceived value (Dodds et al., 1991). The price-quality relationship is found to be even stronger for non-durable goods (Lichtenstein & Burton, 1989) and therefore probably accounts for products in grocery retailing as well. Yet, whether price is used as a quality indicator depends on access to other diagnostic information (Rao & Monroe, 1988). Dodds et al (1991) found that the price-quality

27

relationship weakens if other extrinsic cues such as store and brand name are present. Because both store and brand name will be present during the manipulation, it is expected that absence of a price premium for the exclusive product leads to smaller differences in SC and BC for a new non-exclusive product. Therefore, the following hypothesis was set: H3: A higher introduction price due to a price premium compared to a ‘normal’

introduction price will have a smaller effect on the relationship between exclusiveness on (a) store choice and (b) brand choice for a new exclusive product than for a new non-exclusive product.

Hedonic Level of the Product. People make purchase decisions in different ways depending on the product category that the product falls in. Avery et al (2007) suggest that consumers who purchase utilitarian goods mainly focus on rational arguments such as price when making a purchase decision. Their finding is in line with research showing that utilitarian products can better engage in sales promotions when that enhances efficiency while shopping (Chandon, Wansink, & Laurent, 2000). Another study suggests that consumers prefer to expend other combinations of time, effort and money depending on the product category. During an experiment conducted by Okada (2005), particpants got the offer to travel to another store to buy the same camera for $50 less. People buying a relatively more hedonic camera were willing to spend more time to travel to another store than people that bought a relatively more utilitarian camera and got the same offer. Okada (2005) concludes that people thus rather ‘pay’ in time for hedonic goods and prefer to pay more money instead of time for utilitarian goods. The results show that hedonic vs. utilitarian product categories lead to different consumer responses because of the nature of the product categories and the benefits that they deliver. Because this accounts for ‘normal’ products, it is assumed that this is also true for exclusive products. If people rather spend more time than money on hedonic vs. utilitarian (exclusive) products, it can be expected that people who buy hedonic grocery goods would be more willing to spend time to travel to the store that offers exclusive products. Thus, a new exclusive hedonic product can be a determinant of SC. On the contrary, consumers would not put extra effort into buying (exclusive) utilitarian goods. Furthermore, Van Trijp et al (1996) discovered that consumers have a smaller need for variety for utiliarian goods than in hedonic categories. Because exclusive products cause more variety in an assortment it is expected that exclusive hedonic products can increase the SC. BC is also likely to increase when an assortment contains exclusive hedonic products, because the brand liking will increase due to the increased variety of the brand portfolio.

Sloot, Verhoef, & Franses (2005) invested consumer out-of-stock (OOS) responses in a grocery setting for hedonic vs. utilitarian products. According to their research, consumers buying high-equity hedonic brands, compared to high-equity utilitarian brands, were more likely to switch to another item by that brand in case of OOS. They thus did not postpone their purchase, probably because people buy hedonic goods based on emotional motives and therefore will be disappointed if they are OOS (Dhar & Wertenbroch, 2000). The research showed that the hedonic level of products negatively affected consumer brand switching. It

28

can therefore be assumed that consumers will be more brand loyal to hedonic (exclusive) products because they provide more emotional value, and therefore this will lead to SC. Following this rationale, it can also be assumed that hedonic (exclusive) products make a whole brand more exclusive and unique and therefore improve the emotional value provided by the product.

Based on SC drivers in the Indian market, Sinha & Banerjee (2004) already suggested product category to be a moderator of SC drivers on store patronage loyalty. In line with their suggestion for future research and because of the other reasons meantioned above, the following hypothesis is proposed:

H4: The hedonic level of a product positively moderates the relationship between (a) exclusivity and store choice and (b) exclusivity and brand choice.

2.6.3 Control Variables Store Equity and Brand Equity The constructs store equity (SE) and brand equity (BE) will be taken into consideration as control variables as they seem to influence consumer responses to (new) products. They will be discussed in this section.

Store Equity. SE is defined as ‘the differential effect of store knowledge on customer response to the marketing activities of the store’ (Hartman and Spiro, 2005, pp. 1113). SE consists of three components. First, the differential effect: SE should be measured relative to consumer responses to marketing activities of other supermarkets. Second, store knowledge, which refers to all associations that consumers have when they think about the store name. The third component is the consumer response to marketing activities, such as consumer evaluations, behavior (such as SC and BC) and preferences. Because SE is related to consumer behavior, it should be taken into consideration as a control variable when measuring how consumers respond to new (and exclusive) product offerings. Responses might be biased if it is not measured. Therefore, the following hypothesis was made:

H5: A high perceived store equity (vs. low perceived store equity) leads to higher (a) store choice and (b) brand choice when an exclusive new product is offered

Brand Equity. Brands can be either high- or low-equity brands (Chandon, Wansink, & Laurent, 2000). Keller (1993) explains that a brand has a high customer-based BE when consumers react more favorably to a product when the brand is shown than when it is not shown’. Low-equity brands are mainly bought because of their lower price while high-equity brands are valued more because they provide more benefits (Chandon et al., 2000). In addition, high-equity brands have a higher perceived quality, brand awareness and brand preference. A common thought in the marketing literature is that both the hedonic level and the BE of the product determine how consumers respond to marketing stimuli (e.g. Batra & Ahtola, 1990; Chandon et al., 2000; Dhar & Wertenbroch, 2000). To illustrate: consumers who intend to buy coffee will not buy a brand that they do not value, only because it is in promotion. The product category also plays an important role in consumer spending: a consumer would be more prone to buy a hedonic product in promotion when they do not need it, than a utilitarian product. Since high-equity brands enjoy a higher brand preference and because brand preferences can be very strong for certain brands, BE may influence the liking

29

of the exclusive brand and therefore the SC and BC. Thus, it is important to control for BE.

If a high perceived BE indeed increases the SC and BC, retailers can use BE as a criterion to decide whether to include an exclusive product in their assortment or not. Manufacturers could use this information to decide on which brand to offer exclusively. Second, it would be interesting to investigate whether the effect is higher for the exclusive new product than for the new product. Therefore, the following hypothesis was formed:

H6: A high perceived brand equity vs. a low perceived brand equity of the exclusive new product leads to a higher (a) store choice and (b) brand choice

30

3. Methods This section discusses the methods for the research. In this chapter, an overview of the three pretests used to determine the manipulations will be given first, followed by an explanation of the stimuli selection based on outcomes of these pretests. At the end of the chapter, the measures for the dependent variables and control variables will be discussed. It should be noted that after the three pretests, an experimental online survey was conducted using a 2x2x2 between-subjects design. The data was collected between April 25 and May 3, 2016. To answer the research question, respondents were asked to answer several questions about their opinion on the supermarket Dirk and the brand Robijn (representing the utilitarian product categories) or Fanta (representing the hedonic product categories). Thereafter, they were exposed to a two-folder page of Dirk, after which the same questions were asked again. This information is crucial to understand the following sections. 3.1 Overview Pretests Three pretests were conducted to determine the best way to manipulate the independent variable, the moderators and the dependent variables. In the following sections, findings from the three pretests will be used to explain what consequences the outcomes of the pretests have had for the retailer selection, the product category selection, the new product selection, the folder design, the main manipulation and the manipulation for the moderators. In Table 4 an overview of the pretests’ objectives and the participating respondents is shown: Table 4 Overview Pretests

Pretest Overall Objective Respondents Nationalities

1 To determine the manipulation of the independent variable

N = 20 Nfemale = 10, Nmale = 10

90% Dutch; 10% other European (2 nationalities)

2 To determine the manipulation of the dependent variables (1) and the price manipulation (2)

N = 30 Nfemale = 15, Nmale = 15

63% Dutch; 17% other European (3 nationalities); 20% non-European (4 nationalities)

3 To test whether the final manipulation is carried out correctly (1) and to determine the retailer selection (2)

N = 24 Nfemale = 10, Nmale = 14

54% Dutch; 43% other European (4 nationalities); 3% non-European (1 nationality)

3.2 Product Category Selection No pretest was used for the product category selection. However, the selection process should be explained first to understand the selection of the other stimuli explained in this chapter. A suitable product category has been selected from among product categories in which one or more products have been launched exclusively. An overview of these exclusive products can be found in Table 1. Below the selection process for both the hedonic and utilitarian product category has been clarified.

Utilitarian Product Category. The selection criteria are as follows. First, the utilitarian products, which are bread, dairy, detergents, fish, fruit, toiletries and vegetables, were selected from the categories in Table 1. Secondly, the product should be relevant to both men

31

and women (Gardner, 1970). If not, likeliness of trial and brand familiarity will vary too much. Here, toiletries were eliminated. Third, as the (exclusive) product is new, it should be easy to come up with realistic and attractive new scents or flavors for the product category. In this phase, fish, fruit, and vegetables were eliminated. Finally, the product category should carry a large amount of international brands. The reason for this is that the sample for the pretests consisted of people living in Europe and not in one specific country only. If this criterion is not met, familiarity could differ between nationalities. For all the reasons above, the detergents category was selected as utilitarian product category.

Hedonic Product Category. The same selection criteria were used for the selection of a hedonic product category. First, all categories with hedonic features were selected: beer, body lotion, candy, chocolate, crisps, coffee, shower gel and soft drinks are all categories that bring pleasure at the time of consumption. In the second selection phase, beer, chocolate, shower gel and body lotion were eliminated because preferences for these products differ depending on gender. In the third selection phase, no categories were eliminated because for both crisps, coffee and soft drinks it is easy to come up with new flavors. In the last selection phase, crisps and coffee were ruled out because those categories do not carry many international brands. As a consequence, the soft drink category was selected. Soft drinks are defined here as carbonated lemonade. Please find an overview of the selection criteria for the product category selection in Figure 4.

Figure 4 Product Category Selection

The selected categories have additional advantages. First, they are dominated by premium brands, not by private label brands. Showing a folder without private label brands is therefore realistic. Second, both detergents and soft drinks are weekly promoted in supermarkets’ folders. This indicates that they are important categories for retailers because they attract customers to the store. Moreover, this indicates that these categories are important for consumers as well. It should therefore be easy to find buyers of detergents or soft drinks to fill out the online experimental questionnaire. 3.3 First Pretest Focus of the first pretest was on setting the manipulation of the independent variable. In addition, some general questions about exclusivity were asked, which were used to formulate questions in the other pretests and the final questionnaire. Based on this pretest, the initial

32

retailer was selected following the retail selection procedure. Furthermore, as well for the utilitarian product category (detergents) as for the hedonic product category (soft drinks), a new product was presented in the Dirk folder. In the first pretest, the soft drink brand was selected. Finally, the manipulation for the independent variable was determined. 3.3.1 Retailer Selection To be able to measure SC, a retailer has been selected during the first pretest. A fictitious retailer is not an option, because SC is influenced by many factors about which respondents need to have an overall impression beforehand to answer questions about the construct. In general, a distinction can be made between two different retailer types: the mainly service-oriented (e.g. Albert Heijn, Jumbo, PLUS) and mainly price-oriented (e.g. Aldi, Lidl, Dirk) retailers. Service-oriented retailers tend to offer more exclusive products, as can be seen in Table 1. Results from the first pretest on a 7-point Likert scale [1=very unfamiliar, 7=very familiar] indicated that Albert Heijn (M=6.60, SD=0.68) is the most familiar among both Dutch and international people, probably because of its high market share (35.1%). Moreover, this retailer is most likely to introduce exclusive products because of its willingness to offer an exclusive product variant (M=5.82), measured on a 7-point Likert scale [1=not at all, 7=extremely willing]. To compare, Jumbo (5.03), Plus (3.82) and Aldi (2.94) had the highest score after Albert Heijn (Van Everdingen & Sloot, 2015). Therefore, Albert Heijn was initially selected as a retailer. 3.3.2 New Product Selection for Soft Drinks New product introductions in the FMCG are often new packages or new flavors (Gielens et al., 2014). Inventing a new detergent and soft drink in a similar way can be easily done by selecting an existing brand for each category and adding a new scent or flavor to it. Because it is more likely that brands with a higher familiarity and experience generate trial (Gielens & Steenkamp, 2007), brands with a high familiarity were selected for the new product introduction. In the first pretest, brand familiarity of international soft drink brands sold in the Netherlands was measured by asking respondents how familiar and experienced they were with them on a 7-point Likert scale [1=not at all, 7=extremely familiar/experienced] (Kent & Allen, 1994). Further, they were asked how likely they were to purchase those brands on a 7-point Likert scale. Kent & Allen also measured ‘knowledgeability’ as a brand familiarity scale component. However, it was assumed that a consumer is knowledgeable about a brand if he is familiar with the brand, which made the author choose to only measure ‘familiarity’ and ‘experience’ as scale components. Familiarity is ‘the overall awareness of the brand and understanding of what kind of products the brand represents’ (Jacobson & Lane, 1995). ‘Experience’ was measured because a familiar brand is not necessarily liked and therefore consumed (Jacobson & Lane, 1995). The term refers to the amount of times that people have consumed the brand (Kent & Allen, 1994). Results from the first pretest showing that higher familiarity scores indeed show a higher likelihood of trial, are presented in Appendix B. In selecting the new soft drink flavor, respondents were shown nine non-existing soft drink fruit flavors in the local market, inspired by existing fruit flavors for juices and ice cream. Results can be found in Appendix B as well.

The first pretest showed the highest familiarity (M=6.60) and experience (M=5.75) score for Coca Cola. However, likelihood of trail was lower for this brand (M=3.80). Fanta had the second highest familiarity (M=6.45), experience (M=5.40) and likelihood of trial (M=4.45) score. Spa fruit had the highest likelihood of trail (M=4.70) score, but received lower

33

familiarity (M=5.85) and experience (M=5.05) scores. Therefore, Fanta was chosen as the exclusive brand. Average flavor preferences (see Appendix B) turn out to be highest for Red fruit and Wild berry (MRed fruit=5.50, MWild berry=5.15), however female (MRed fruit=6.00, MWild

berry=5.90) and male preferences (MRed fruit=5.00, MWild berry=4.00) are large for both flavors. Although male averages (M=3.99) are lower than female averages (M=4.86) in general, differences are smaller for both Raspberry (MMale=4.70, MFemale=5.30) and Blackberry (MMale=4.60, MFemale=5.10), who have the third and fourth highest average score (MRaspberry=5.00, MBlackberry=4.85). Hence it was decided to combine raspberry and blackberry into the new flavor Wild Berry. Because Fanta already sells the purple colored drink Fanta Cassis and respondents should not be confused with both drinks, the color for the new Fanta Wild Berry was chosen to be dark red. Moreover, the label on the bottle of the new Fanta Wild Berry shows a raspberry and blackberry, which is referred to in the text in the folder. It was decided to show a 1.5-liter bottle in the folder, because this size is most common in the assortment of Dirk and after reviewing several Dirk folders it can be said that 1.5-liter soft drink bottles are most often promoted.

3.3.3 Main Manipulation The independent variable has two levels, namely ‘new brand’ and ‘exclusive new brand’ which were manipulated by showing different tags in the 2-folder page that was shown to the respondents during the experiment. In the first pretest respondents were shown a tag in a promotion of a new Spa Fruit flavor ‘Pear and Orange’, saying ‘new’. All respondents understood that this term referred to a new flavor. Then, it was examined whether respondents understood the term exclusivity by showing them two different sorts of tags in the same promotion: ‘new and exclusively available at Albert Heijn’ and ‘new – only available at Albert Heijn’. All respondents understood that this tag referred to exclusive distribution and a new product. One respondent remarked that the term ‘exclusive’ sounded more luxury, therefore the tag ‘new – exclusively available at (name of the store)’ was used for the other pretests and during the final manipulation. Yet, aided awareness of exclusivity in a grocery setting was low (50%) and when respondents were asked to give examples of exclusive products, only one respondent could give a correct example. Some mentioned private label products or could not give any examples. Therefore, more questions about the adjusted tag were asked during the third pretest. 3.4 Second Pretest The second pretest, in which respondents were divided into two groups, had multiple objectives. First, to investigate whether there were differences in store and brand image between groups of respondents. To achieve the first objective, group 1 saw a promotion for the new Fanta Wild Berry with the tag ‘new’ and group 2 saw the same promotion with the tag ‘new – exclusively available at Albert Heijn’. Hereafter, store image was measured for Albert Heijn and brand image was measured for Fanta. Based on these results, the retailer selection was adjusted (see section 3.4.1). Secondly, as it was decided to investigate the moderator ‘product category’, brand liking was measured for detergents and Robijn Sweet Jasmine was chosen as new product for the detergents category. The third objective was to determine how the moderator ‘price’ had to be manipulated. To reach the second objective, both groups saw the same promotion again, but now the tag for both groups was ‘new – exclusively available at Albert Heijn’. Moreover, group 1 saw the price ‘€2.16’ (which is a price premium of 20% on top of the normal price for Fanta) and group 2 saw the price ‘€2.34’ (which is a 30% price premium). After showing the promotion, questions were asked about

34

the quality of the product, the height of the price and willingness to buy the new and exclusive Fanta Wild Berry. Based on the outcomes of the first two pretests, and based on the outcome of the second pretest that store image (Mnew=5.62, Mnew and exclusive=5.49) and brand image (Mnew=5.33, Mnew and exclusive=5.11)4 did not seem to differ between group 1 and 2, a folder design was made for the supermarket Dirk. Finally, the folder design was set up. 3.4.1 Final Retailer Selection Although Albert Heijn was initially selected as a retailer due to outcomes of the first pretest, results from the second pretest revealed that Albert Heijn already had a high store image as well for the group that saw the tag ‘new’ (M=5.62, SD=1.01) as for the group that saw the tag ‘new – exclusively available at Albert Heijn’ (M=5.49, SD=2.90). Store image was measured on a 7-point Likert scale [1=strongly disagree, 7=strongly agree] with 3 items (attractive, distinctive and high quality service). Store image is even lower for the exclusive product in group 2, which is probably due to the low purchase frequency for this group: 75% buys soft drinks less than once per month vs. 29% in group 1. Because exclusivity does not seem to increase Albert Heijn’s store image and as improving the high score would be hard, it was decided to choose a price-oriented retailer. Those with the highest market shares in the Netherlands are Aldi (7.3%), Lidl (10.0%) and Dirk (3.0%). Aldi and Lidl do not sell much premium brands, but Dirk does and therefore it was selected as retailer (Foodpersonality, 2015). It is interesting to see whether exclusivity can improve the SC (which is linked to store image) for this price-oriented retailer especially because their willingness to offer exclusive products is not as high as for service-oriented retailers (Van Everdingen & Sloot, 2015). Because of Dirk’s lower familiarity (M=3.50, SD=2.04) it was decided to provide some information and pictures of Dirk during the introduction of the experiment. 3.4.2 New Scent Selection for Detergents Appendix C shows the outcomes of the brand familiarity, brand experience and brand trial scores for detergents that were measured during the second pretest. These scores have been measured in the same way as in the first pretest (see section 3.3.2). In pretest 2, respondents were shown eight non-existing detergent scents in the local market which were inspired by existing scents for air fresheners and cleaning products. Liking of these flavors and scents is shown in Appendix C as well. Introducing a new product with a flavor or scent that is liked increases the chance of acceptance of the new product (Gielens & Steenkamp, 2007), therefore the flavor and scent with the highest liking was selected. For detergents, it was decided to use liquid detergents for colored laundry as this is the largest product group within the detergents category. Most respondents will therefore sometimes buy this product group. The second pretest proves Robijn to be the brand with the highest scores on both familiarity (M=6.17), experience (M=5.03) and likelihood of trial (M=5.10). To compare, average familiarity (M=3.42), experience (M=3.20) and likelihood of trial (M=4.58) scores are lower. Robijn was thus selected as detergent brand. Scent preferences are highest for Soft Linen (M=5.10), White Lilies (M=4.93) and Sweet Jasmine (M=4.90). However, male and female preferences differ for these scents. The smallest differences can be found for Sweet Jasmine (MMale=4.80, MFemale=5.00), which made the author select Robijn Sweet Jasmine as a new product representing the utilitarian category. Robijn offers both concentrated and normal

4 Store and brand image were measured on a three-item 7-point Likert scale. The items for both scales were ‘attractive’, ‘distinct from other brands’ and ‘high quality’.

35

liquid detergents for colored laundry (both for 20 washes). Therefore, it was decided to measure what package respondents preferred during the third pretest (see section 3.5.1).

Since consumer brand preferences for the soft drink category are very strong, answers to SC and BC might be biased. The same accounts for the brand Robijn, which is a very familiar brand. Hence, BE and product category involvement were included as control variables. SE was taken into consideration as control variable, because store preferences may bias consumer responses to store evaluations (and thus SC).

3.4.3 Price Manipulation The moderator ‘price premium’ has two levels: ‘price premium’ vs. ‘no price premium’. During the second pretest, respondents were first asked whether they could give a price indication of a normal 1.5-liter bottle of Fanta. Only one out of three respondents knew that a normal 1.5-liter bottle of Fanta costs between €1.70 and €1.90 (a normal 1.5-liter bottle of Fanta costs €1.80). The mean for all respondents was €1.91. Thus, most respondents are not very price-conscious for a Fanta soft drink. Then, participants in group 1 were shown a promotion for the new Fanta Wild Berry without a price indication. Group 2 was shown a promotion for the new and exclusive Fanta Wild berry, also without a price tag. The mean price they were willing to pay for Fanta Wild Berry was higher than €1.80 (Mgroup1 ‘new’=2.04, Mgroup2 ‘new&exclusive’=1.92). The high average amount that the first group is willing to pay is probably due to the fact that their buying frequency for soft drinks is higher. Group 1 also seems to have a more positive attitude towards the Fanta brand. On a 7-point Likert scale they rate the Fanta Wild Berry quality higher (M=5.40) than group 2 (M=5.07). These differences were controlled for in the final experiment by controlling for BE. As a higher price does not immediately lead to a higher perceived quality, it was presumed that asking a lower price premium will not lower the perceived quality of the product. Looking at the average prices that both groups are willing to pay for the new Fanta Wild Berry, a price premium of 10% seems to be the best option so far. Thereafter, it was examined how respondents that were randomly allocated to two groups reacted to a price premium of 20% (€2.16 – group 1) or 30% (€2.34 – group 2) for the new and exclusive Fanta Wild Berry. Both groups think the price is too high, looking at the willingness to buy the product for the indicated price (Mgroup1=3.71, Mgroup2=3.00) on a 7-point scale [1=strongly disagree, 7=strongly agree]. Both groups do like the product, because after they were being told that a normal 1.5- liter Fanta bottle costs €1.80, their willingness to buy the new Fanta Wild Berry for that price is higher (Mgroup1=5.71, Mgroup2=4.83), measured on the same 7-point Likert scale. Moreover, there is a discrepancy between what people say and what they intend to decide in the end: when asking whether respondents were willing to pay more for a new and exclusive soft drink, some participants agreed (Mgroup1=4.07, Mgroup2=3.53). These results also indicate a lower price premium than 20% would be better. Hence it was decided to use a price premium of 10% for both Fanta and Robijn in pretest 3 and in the final questionnaire. To assure that the price manipulation is executed well, respondents were asked how they perceived the price of the new (exclusive) product in the final questionnaire as a manipulation check. Authors that investigated price-quality relationships mainly used reference prices (Darke & Chung, 2005; Lichtenstein & Bearden, 1989) to communicate price discounts. Research also found that reference prices influence BC at the retail level (Kumar & Leone, 1988). By showing another brand in the same category in the lower side of the left page of the folder, Dirk helps consumers to set an external reference price. Based on the reference pricing

36

theory (see chapter 2.5.2, section ‘price’), respondents are assumed to evaluate the product with a price premium as ‘high priced’. As Dirk would never communicate that they ask a price premium for a new product, during the final experiment, respondents were not made aware of the fact that the price they saw was ‘normal’ or a ‘premium price’ compared to the other conditions. 3.4.4 Folder Design During the experiment, each respondent was shown two pages of a Dirk folder. In Appendix D, an overview of the different conditions and the accompanying folders that were shown in each condition is given. All 2-page folder versions only differed on stimuli. The folder was designed with the objective to make it as realistic as possible. Normally, Dirk shows the same kind of products on a folder page. Therefore, folder 1-4 had another setting than folder 5-8. First, in folder 1-4 the (exclusive) new brand was Robijn and in folder 5-8 this was Fanta. Second, the products on the right folder page were products that are normally shown on the same page as the (exclusive) new brand. Promotions from the Dirk folder of week 15 were used to fill this right page. The brand on the lower side of the left page was used as a reference price for the (exclusive) new product on the upper side of the left page. When a price premium was asked, this product can be used to compare the price of the (exclusive) new product to. For detergents, Fleuril Brilliant Color is shown on the lower side of the left page. The second pretest showed that both its familiarity (MFleuril=3.73, MRobijn=6.17) and experience (MFleuril=2.50, MRobijn=5.03) is lower than for Robijn, which makes the new Robijn Sweet Jasmine extra attractive. Further, showing Fleuril next to Robijn is realistic as Dirk usually also promotes brand with a high familiarity next to those with a lower familiarity. The normal price of a bottle of Fleuril Brilliant Color is €4.99, which is the same as the average price of a bottle of a similar size of Robijn. For soft drinks, a 1.5-liter bottle of Dr. Pepper is shown on the lower side of the left page for the following reasons. First, because the Fanta Wild Berry is a 1.5-liter bottle too, second because the bottle has almost the same price (€1.74) as a Fanta 1.5-liter bottle (which costs €1.80). Both Fanta and Dr. Pepper were priced €1.80 in the final questionnaire. The second pretest showed that both Dr. Pepper’s familiarity (MDr. Pepper=4.40, MFanta=6.45) and experience (MDr. Pepper=2.90, MFanta=5.40) is lower than for Fanta, which makes the Fanta promotion extra attractive. Again, promoting a high-familiarity brand next to a lower-familiarity brand is realistic. To make the folder not too complicated for the respondents, both products on the left page had no discounted price. Now it can be assumed that there will be no confusion of the price for the products. Further, the promotion for the new (exclusive) Fanta Wild Berry and Robijn Sweet Jasmine is the most eye catching because of its size and the yellow tag ‘new’ or ‘new - exclusively available at Dirk. It is realistic that Dirk would put extra effort into promoting a new (and exclusive) product to boost sales, because the product is not familiar yet. 3.5 Third Pretest Based on Wetzel (1977), the third pretest investigated whether the final manipulation, based on the output of the first two pretests, would be carried out correctly. Moreover, the detergents package for Robijn Sweet Jasmine was chosen. Further, it was tested whether the folder for the supermarket Dirk would lead to SC and BC (measured in the same way as in the final experiment, see chapter 3.7.1). The new product Fanta Wild Berry was used to test BC. Respondents were randomly distributed across two scenarios: group 1 saw the new Fanta

37

Wild Berry promotion (€1.80) and the tag ‘new – exclusively available at Dirk’, while group 2 saw the same promotion and price with the tag ‘new’. Hereafter, all respondents saw a folder page with the new and exclusive Robijn Sweet Jasmine (€1.80) to test the manipulation for the utilitarian product category. Finally, the hedonic level manipulation check was done to assure that this manipulation check would work in the final experiment. 3.5.1 New Product Selection for Detergents As Robijn sells both concentrated and non-concentrated or normal detergent, it was measured what package respondents preferred when buying detergents. Results show that 17% preferred normal detergents vs. 13% for concentrated detergents. 54% had no opinion or buys the one that is on promotion and 17% never buys (liquid) detergents. Therefore, it was decided to choose the normal version for the new product introduction. Robijn currently offers a pink and blue version of its normal detergent. After a small field research in the supermarket, an orange version of liquid detergents for colored laundry seemed not to exist yet while other remarkable colors suiting the category such as purple (Omo), green (Persil), red (Fleuril) and blue (Persil) did. Therefore, it was chosen to make the new Robijn Sweet Jasmine detergent orange. The green label and cap match the orange colored detergent. It was decided to show a bottle of Robijn for 20 washes, because this size is similar to Robijn’s current liquid detergent bottles.

3.5.2 Main Manipulation During the third pretest, respondents were asked again what the label ‘new - exclusively available at Dirk’ meant to them after they saw the Robijn folder (priced €4.99). All respondents (n=24) understood the manipulation, but some said that it may seem that the offer itself was only available at Dirk. Therefore, it was decided to further explain the term in the final survey to the groups that saw a folder with an exclusive product and to keep the folder page on screen while answering the questions. Further, to emphasize the product’s exclusiveness, it was told that there were currently around 12 exclusive products in the Dutch supermarkets. During the first part of the third pretest, it was also measured which soft drink promotion was most appealing to the respondents. The majority (83%, n=10) of group 1 (new&exclusive) answered ‘Fanta’. Reasons were the special taste, Fanta’s quality, the bigger size of the promotion and its exclusivity. The two respondents that liked Dr. Pepper more explained that they did not like to try a new flavor. The same percentage (83%, n=10) of group 2 (new) liked the new Fanta promotion more. Reasons were the liking of the brand and curiosity to the new flavor. Participants liking the Dr. Pepper promotion more explained that they trusted the Dr. Pepper taste better. Color of the Tags. The tags ‘new’ and ‘new - exclusively available at Dirk’ needed to have an appropriate color that attracts attention. Dirk normally uses yellow tags with black letters for price promotion information in its folder. To make the folder realistic, a yellow tag would thus be most appropriate. Results from the third pretest indicated that it was obvious for group 1 that the new Fanta Wild Berry was new and exclusive (M=5.17) and for group 2 that it was new (M=4.17) on a 7-point Likert scale [1=not obvious at all, 7=very obvious]. For the new Robijn Sweet Jasmine, measured on the same scale, exclusivity was even more obvious (M=5.96). To emphasize the newness and exclusiveness of the product even more, it was therefore decided to point the respondents to the fact that the folder showed a promotion of a new (exclusive) product for a certain price after showing the 2-page folder.

38

3.5.3 Folder Design Results from the third pretest indicated that the designed folder was seen as realistic (Mgroup1

‘new&exclusive’=5.67, Mgroup2 ‘new’=5.33). In addition, as well in the first as in the second group, 83% thinks the Fanta promotion is most attractive. Reasons are: Fanta’s better quality and its familiarity, the size of the Fanta promotion and the exclusiveness/newness of the product. Reasons for finding the Dr. Pepper promotion more attractive are: its familiarity and its taste. In the same pretest, 88% of the respondents indicated that they liked the Robijn promotion more than the Fleuril promotion, mainly because of Robijn’s familiarity, the detergent’s color and scent, its exclusiveness/newness, the size of the promotion. Respondents that found Fleuril more attractive liked the Fleuril scent more than Sweet Jasmine or find an existing scent more convenient. One respondent mentioned that the orange detergent did not look good at all. Unfortunately, not every consumer may thus like the new scent, however most respondents do. Based on the findings described above, the folder design in itself was not adjusted compared to the third pretest.

3.5.4 Manipulation Check Hedonic Level To reduce time and effort, the mean hedonic level of both detergents and soft drinks was measured during the third pretest (Perdue & Summers, 1986). All respondents (n=24) first evaluated the hedonic level of detergents, and thereafter evaluated the hedonic level of soft drinks. Five items were measured on a 7-point semantic differential scale (Voss et al, 2003). The scores for the five items [fun, exciting, delightful, thrilling, enjoyable] were averaged and recoded so that 1=low hedonic level, 7=high hedonic level. A low hedonic level refers to a utilitarian product category. A high hedonic level refers to a hedonic product category. The same accounts for the measurement of the hedonic level in the main experiment. As all participants in the third pretest answered the manipulation check for both product categories, a one-sample t-test was performed to compare the mean scores for the hedonic level. The test indicated that the manipulation was executed successfully as the mean score of the hedonic level for detergents [M=3.12, SD=1.50] was significantly lower than for the soft drinks [M=5.21, SD=.65; t(23)=-6.852, p=.000].

3.6 Experimental Design and Procedure Based on the information gathered in the three pretests, a final experimental survey was designed and filled out by participants. This section successively describes the subjects, the experimental design and the procedure that was followed during the experiment.

3.6.1 Subjects and Experimental Design 299 respondents participated in the online experimental survey. Participants were approached by the survey sampling agency SSI to fill out the survey on a personal computer or tablet. As Dirk is a supermarket in the western part of the Netherlands, the sample consisted of Dutch shoppers living in this part of the Netherlands. Finally, since this group is doing groceries it would be most interesting to examine their potential shopping behavior. To assure a good representation of Dutch shoppers in the western part of the Netherlands, ages were between 18 and 65 years. First, 9 responses were removed because of incompleteness. Non-Buyers. At the beginning of the survey, consumers were asked how often they bought soft drinks or liquid detergents for colored laundry (depending on the group that they were assigned to before they entered the questionnaire). 7 non-buyers, which are people that never buy soft drinks or liquid detergents for colored laundry, were screened out the questionnaire because they are not part of Dirk’s target group. Moreover, non-buyers will visit supermarkets

39

less often and therefore answering questions about the store would be difficult. Thus, next to the removed participants due to incompleteness, the non-buyers were removed from the sample. Main Manipulation Check. Following Perdue & Summers (1986) the main manipulation check was included in the final experiment because it was adjusted after the third pretest, in which four questions were asked to be able to formulate the final manipulation check. First, respondents read an introduction text about the 2-page folder of Dirk they were going to see. After having seen the folder, the control conditions (group 1, 2, 5 and 6)5 were asked where the new Robijn Sweet Jasmine/Fanta Wild Berry was available [options: only at Dirk (1); at all supermarkets (2)]. The test conditions (group 3, 4, 7 and 8) were asked what was said in the folder about the new Robijn Sweet Jasmine/Fanta Wild Berry [options: A free product is included in the package deal when buying a bottle of Robijn Sweet Jasmine/Fanta Wild Berry (1); Robijn Sweet Jasmine/Fanta Wild Berry is a new scent/taste (2)]. The folder stayed on screen while answering the manipulation check questions. It was decided to remove 44 respondents (which is 15.5% of 283) were removed from the sample due to misunderstanding of the main manipulation. Sample Demographics. An overview of the final sample characteristics can be found in Appendix E. It was decided not to remove the respondents that misperceived the hedonic level of the product and the price manipulation check, because groups would be too small then. Most respondents would have been removed because they thought the ‘normal’ price was too high (6) or way too high (7). The final sample contained 239 shoppers, of which 30% is male and 70% is female. All respondents reside in the western part of the Netherlands and have at least a lower primary education degree. 7.1% has a lower primary education degree, 44.3% has followed average education, 10.5% holds a degree in overall higher or pre-university education, 22.6% followed higher education and 15.5% has a degree in scientific education. Figure 6 in Appendix G shows at which supermarkets respondents normally do groceries. On the individual level, consumers’ perspective on service is influenced by the stores that they normally visit. Most people visit Albert Heijn (75.7%), Lidl (49.4%) and Dirk (46.9%), which means that the sample in general does not specifically prefer a service- or price-oriented retailer. To be able to compare the eight conditions among, in each condition both age and gender were correctly represented so that each group is a correct representation of shoppers in the western part of the Netherlands. As not each respondent understood the main manipulation correctly and since some respondents were removed due to incompleteness or because they were screened out, the sampling survey agency approached new respondents that fit the sample demographics criteria for those specific conditions afterwards. Now, each condition satisfies the shopper criteria for age and gender, which means that male and female respondents were evenly spread over the age categories. Details can be found in Appendix F. The mean age in the sample was 42.11 years. 3.6.2 Procedure Respondents were approached via the survey sampling agency SSI, since this agency has access to a large and differentiated pool of respondents within a short amount of time. The data was collected between April 25 and May 3, 2016. The experimental online survey was completed on personal computers or tablets at different locations (e.g. at home). The

5 For the group numbers, see Table 9

40

experimental survey was selected as research design since it enables to control for influencing variables from the environment. Although some studies that measure SC or store-related variables use videos to simulate a store’s environment, it is better to measure consumer perception while controlling for stimuli from the store’s environment. The experiment now controlled for the shopping task and for the convenience factors for all stores nearby the respondent’s residence, which will be explained below. During the experiment, respondents were randomly assigned to one of the eight conditions. There were four test and four control conditions. As can be seen in the conceptual model (see Figure 3), both price (price premium of 10% vs. no price premium) and the product category (utilitarian vs. hedonic) were manipulated. Thus, a 2x2x2 between-subjects design was used. First, respondents were given a short introduction about Dirk, illustrated by three pictures. They were explained that Dirk is price-oriented and that the supermarket sells less premium brands than e.g. Albert Heijn or Jumbo. Then, participants were asked to imagine they were going to do groceries and needed detergents or soft drinks (dependent on the condition they were assigned to). To control for convenience, consumers were asked to imagine that there were several supermarkets nearby their residence. The convenience factors were thus equal for all supermarkets. Now, SC, SE, BC and BE were measured. It was decided to measure these variables before and after showing the folder to be able to precisely track differences. Participants in the test conditions were thereafter given a short introduction about exclusive products in the Netherlands. After a short introduction about the 2-page folder, strengthening the manipulations because it draws attention to the price of the (exclusive) new product, it was shown. The tag for the test cells was ‘new – exclusively available at Dirk’, and ‘new’ for the control cells. The new (and exclusive) Robijn Sweet Jasmine was promoted in the folder pages for the conditions seeing the ‘utilitarian’ product. The new Fanta Wild Berry was shown to the respondents in the conditions for the hedonic category (see Appendix D). Immediately after this, the main manipulation check and price manipulation check were placed (based on Wetzel, 1977). The folder pages stayed on screen while asking these questions. Then, the same questions about Dirk and Fanta or Robijn were asked while the folder stayed on screen, followed by a purchase intention and some normal shopping behavior questions. Finally, the background questions were asked. 3.7 Measures In this section, an overview of the measurement instruments for the dependent variables and control variables is given. Cronbach’s alpha’s were calculated based on the final data. 3.7.1 Dependent Variables This study is focused on SC, which is measured as store choice intention, since the study is not a real-life experiment but an online questionnaire. Store choice behavior as such is hard to measure in an experiment. Namely, the actual visiting of a store depends on external variables such as availability of time and transport, current promotions and current needs for products. Although SC is often measured in the marketing literature, SCI has rarely been taken into consideration as dependent variable in the marketing literature. A study conducted by Grewal et al (2003) is one of the scarce studies that does measure this concept as a dependent variable. Hence, the multi-item scale that they used was based on Dodds et al (1991) and was selected to measure SCI in the present study. BC is referred to as the act of actually purchasing a product, which is often measured by analyzing sales and electronic scanner data (e.g. Guandagni & Little, 1983). However, this study is focused on the intention of purchasing a product from a certain brand as that is measurable in an online experiment. BCI as such has, to the author’s knowledge, not been measured as a dependent variable before. The scale used

41

for BCI was therefore based on a ‘willingness to buy’ (a product) scale, which was adjusted so that it measures the intention to buy a product from a brand. A renowned study in the marketing literature from Dodds et al (1991), where ‘willingness to buy’ is the dependent variable, formed the basis for the BCI scale. The scales’ reliability was tested using the Cronbach’s alpha. Both scales have a Cronbach’s alpha above the critical mark of 0.70. BCI had the highest Cronbach’s alpha (0.95). In Table 5 an overview of the dependent variables and the items that they consist of is given. Both variables were measured on a 7-point Likert scale, as all the Likert scales were in the study. Because SCI and BCI are both three-item variables, scores for the three items were averaged. The average scores thus represent the dependent variables. Further, all scales that were originally developed in English, were translated into Dutch for the final experiment following usual procedures for translation (Craig & Douglas, 2005).

Table 5 Dependent Variables

Construct Item [Scale: strongly disagree (1) – (7) strongly agree] Based on Cronbach’s alphab

Store Choice Intention (SCI)

1. It is likely that I would shop at Dirk 2. I would be willing to buy merchandise at Dirk 3. I would be willing to recommend Dirk to my friends

Grewal et al (2003); Dodds et al (1991)

0.94

Brand Choice Intention (BCI)

1. It is likely that I would purchase a (product) from (brand)a 2. The probability that I would consider buying (product) from (brand) is very high 3. I am willing to buy a (product) from (brand)

Dodds et al (1991)

0.95

a ‘(product)’ refers either to detergents or soft drinks. ‘(brand)’ refers either to Fanta or Robijn b calculated for SC_after and BC_after, because these variables were used during the data analysis (see chapter 4.3) 3.7.2 Control Variables Scales for the control variables (see Table 6 and 7) were designed using existing scales from relevant literature. First, the multi-item perceived BE scale was based on Sloot & Verhoef (2008), who in turn based the scale on Batra & Ahtola (1990). BE and SE can be seen as related concepts, since both scales measure equity of a brand or store. Hartman & Spiro (2005) introduced SE as a parallel of BE based on Keller (1993). To make the BE more similar to the SE scale, it was decided to add the third item ‘uniqueness’ to the scale. The second control variable scale, the SE scale, was based on Hartman & Spiro (2005). Third, the product category involvement scale was based on Mittal (1995). The reliability of these scales was found to be good for SE (α =.89) and product category involvement (α=.81) because their Cronbach’s alpha’s were higher than the threshold of 0.70. BE scored lower than the threshold, but according to Murphy and Davidshofer (1988) it is not unacceptable so no items were removed. As well brand familiarity, buying frequency, the shopping task, convenience (location, opening hours, parking facilities at the store, fast checkout), as the demographic control variables were included as control variables in the experimental setting. Brand familiarity was determined in the first pretest (as described in chapter 3.6.2) and buying frequency was used as a screen-out criteria. In chapter 3.6.2 it was also explained how the experiment controlled for convenience. The demographic control variables were gender, age and nationality.

42

Table 6 Multi-Item Control Variables

Construct Item Based on Cronbach’s alpha

Brand Equity (BE)a

1. Compared to other (product) brands, the price of the (product)a offered by the brand is very high 2. Compared to other (product) brands, the quality of the products offered by the brand is very high 3. (Products) from the brand are unique compared to other (product) brands 4. My preference for the brand is very high compared to other (product) brands [strongly disagree (1) – (7) strongly agree]

Sloot & Verhoef (2008), Batra & Ahtola (1990)

0.69

Store Equity (SE)b

1. Compared to other retailers/supermarkets, Dirk has a good reputation 2. Dirk is a unique retail formula/supermarket, compared to other supermarkets 3. Compared to other retail formulas/supermarkets, Dirk is known as a high quality retailer 4. Compared to other retailers/supermarkets, Dirk is innovative 5. When Dirk is located nearby my residence, I strongly prefer to shop at Dirk [strongly disagree (1) – (7) strongly agree]

Hartman & Spiro (2005)

0.89

Product Category Involvement

1. In selecting from the many types of (product) available in the market, I care a great deal about which one I buy 2. It is important to me to make the right choice of (product) 3. In making my selection of (product), I am very concerned about the outcome of my choice [strongly disagree (1) – (7) strongly agree]

Mittal (1989), Mittal (1995)

0.81

Brand familiarity

1. How familiar are you with the brand?b 2. How experienced are you with the brand?b

[not at all familiar/experienced (1) – (7) extremely familiar/experienced]

Kent & Allen (1994)

n/a

a ‘(product)’ refers either to detergents or soft drinks b brands shown for both detergents and soft drinks can be found in Appendix B and C c αcalculated based on BE_after, which is the brand equity measured after the folder was shown

d αcalculated based on SE_after, which is the store equity measured after the folder was shown

Table 7 Other Control Variables

Construct Item Based on Cronbach’s alpha

Buying Frequencya

How often do you buy (product)? [strongly disagree (1) – (7) strongly agree]

n/a

a ‘(product)’ refers either to soft drinks or detergents

43

4. Analysis The main outcomes of the experimental survey will be outlined in this chapter. First, preliminary assumptions and the manipulation checks of the final experiment will be discussed. Then, the main effects will be tested without and with taking into consideration the control variables brand equity and store equity. Then, the moderating effects are discussed. Finally, an overview of the control variables is presented and some extra analyses are performed. It should be noted that all conducted significance tests are two-tailed. 4.1 Normality and Outliers Data can often be approached as if it has a normal ‘bell-shaped’ curve. If this normality assumption is not met, results may be not trustworthy (e.g. Thode, 2002). Therefore, normality assumptions were tested for the dependent variables. As both SC and BC were measured before and after the participants saw the folder, these four measured were assessed by performing the Kolmogorov-Smirnov test. The assumption of normality was violated for the test conditions, control conditions and for all eight conditions a few cases (3 of 32) failed the test at p-values of 0.1. Following Stevens (1996), visual inspection of histograms and box plots was done for additional inspection. Distributions of the dependent variables that failed the assumption were highly skewed and had long tails. Several potential outliers (>1.5 IQRs from box edge) were detected. There were only very few extreme outliers (>3 IQRs from box edge). Removing outliers is not always desirable because it changes the original data (Osborne & Waters, 2002). Glass & Stanley (1970) noted that skewness in a univariate distribution has a minor effect on the test robustness if the number of observations is equal in each cell, which is the case in this study. Therefore, no outliers were removed. 4.2 Manipulation Checks Manipulation checks can be executed during the pretest or during the main experiment. Buying frequency was used as a filter question during the main experiment. To reduce time and effort, and to use its outcome for the main experiment, the manipulation check for brand familiarity was done during the pretests (Perdue & Summers, 1986) so that is is now assumed that both Fanta and Robijn are high familiarity brands. Normality and outliers were evaluated for the perceived hedonic level and price manipulation, which were done during the main experiment. Only the perceived hedonic level for the Fanta test group (condition 5-8) was normally distributed, for the other test groups, control groups and eight conditions together the normality assumption was violated. Inspecting the box plots, some potential outliers were detected. However, for the same reason as mentioned in chapter 4.1, no outliers were removed. Both manipulations do not have a significant linear relationship with each other [t(1)=11.064, p=.294]. Manipulation Check Hedonic Level. The manipulation check for the hedonic level was measured on a three point 7-item semantic differential scale [fun, exciting, enjoyable; low hedonic level (1) to (7) high hedonic level]. An independent samples t-test was conducted to compare the manipulation check scores between the utilitarian (detergents) and hedonic product category (soft drinks). The test proved that the mean score of the hedonic level for detergents [M=4.42, SD=1.21] was significantly lower than for soft drinks [M=4.96, SD=1.11; t(237)=-3.569, p=0.000]. The manipulation check was thus executed successfully. Manipulation Check Price. During the main experiment, respondents were asked what they thought about the price of the new (and exclusive) Robijn (condition 1-4) or Fanta (condition 5-8) compared to other detergents/soft drinks. The price manipulation check was measured on

44

a 7-point Likert scale [1=way too low, 7=way too high]. Again, an independent samples t-test was run to compare the mean scores for the ‘normal price’ [M=4.71, SD=.91] and ‘premium price’ group [M=4.99, SD=1.03]. The means differed significantly [t(237)=-2.274; p=0.024] and the manipulation was thus effective. Differences Between Eight Conditions. Two one-way ANOVAs were conducted to assess differences between the eight conditions on the manipulation checks. There is a weak but significant difference between the eight conditions on the hedonic level manipulation check [F(7)=2.174, p=.037]. The LSD post-hoc test reveals that significant differences can be found only among the detergents conditions (1-4) and the soft drink conditions (condition 5-8). There was no significant difference between the eight conditions on the price manipulation check [F(7)=.993, p=.437].

4.3 Descriptive Statistics This section shows some important descriptive statistics of the experiment. As shown in the tables in chapter 3.7, the Cronbach’s alpha’s for the used scales in the experiment were found to be high enough to be reliable. In Table 8 average scores for the most important variables (which were originally measured on multiple-item scales) are shown. It should be noted that both SC and BC were measured as SCI and BCI, as explained in chapter 3.7.1. Scores on the multiple-item scales SCI, BCI, SE and BE were averaged. Both SCI, BCI, SE and BE were measured before and after the moment that respondents were shown the two-page folder so that differences among could be compared. Both measurements and the differences are given in the table below. In the table, Robijn represents the detergents and therefore utilitarian categories in the supermarket. Fanta represents soft drinks and thus the hedonic categories. To make the numbers easy to read, it was decided to only show standard deviations for the purchase intention of the new product. Table 8 Descriptives of the Final Experiment Test cells

(“New – exclusively available at Dirk”) Control Cells

(“New”) Normal

Robijn Normal Fanta

Premium Robijn

Premium Fanta

Normal Robijn

Normal Fanta

Premium Robijn

Premium Fanta

SCI_before 4.92 5.00 4.77 4.52 4.74 4.89 5.10 3.87 SCI_after 3.93 4.37 4.53 4.11 4.17 4.54 4.43 3.82 Difference -0.99 -0.63 -0.24 -0.41 -0.57 -0.35 -0.67 -0.05 BCI_before 4.57 4.31 4.37 3.18 4.31 4.17 4.42 4.66 BCI_after 3.99 4.04 4.04 3.07 3.80 3.66 4.02 3.96 Difference -0.58 -0.27 -0.33 -0.11 -0.51 -0.51 -0.40 -0.70 SE_before 4.75 4.53 4.58 4.32 4.53 4.71 4.67 4.30 SE_after 4.61 4.47 4.71 4.21 4.37 4.71 4.57 4.21 Difference -0.14 -0.06 0.13 -0.11 -0.16 0.00 -0.10 -0.09 BE_before 4.63 4.54 4.50 4.05 4.68 4.64 4.56 4.42 BE_after 4.67 4.58 4.60 4.04 4.62 4.50 4.56 4.44 Difference 0.04 0.04 0.10 -0.01 -0.06 -0.14 0.00 0.02 Purchase Intention New Product (sd)

4.10 (1.47)

3.63 (1.90)

4.13 (1.50)

3.24 (1.68)

3.90 (1.63)

4.27 (1.78)

3.63 (1.50)

3.68 (1.76)

All variables were measured on a seven-point (1-7) scale

45

As differences for SCI, BCI, SE and BE before and after showing the two-page folder are small and in most cases even negative, it was decided to only use the average scores measured after the participants saw the two-page folder. A possible reason for these small differences is that respondents are not sensitive to the folder promotions which could be due to the low purchase intention of some of the respondents. Another reason could be that respondents were bored when filling out the same scales again. Other statistics about background variables are shown in Appendix G. To further explain what manipulations were used for the different conditions, Table 9 gives a textual summary of the manipulations used. Numbers of the conditions are shown in this table as well, which are corresponding with the numbers for the conditions in Appendix D. In the following sections, these condition numbers will be referred to. Table 9 Overview of the Manipulations Across Conditions Test cells

(“New – exclusively available at Dirk”) Control Cells

(“New”) Normal

Robijna Normal Fanta

Premium Robijn

Premium Fanta

Normal Robijn

Normal Fanta

Premium Robijn

Premium Fanta

Condition 1 2 5 6 3 4 7 8

New product Product Category Price

Exclusive Utilitarian

Normal

Exclusive Hedonic Normal

Exclusive Utilitarian Premium

Exclusive Hedonic Premium

New Utilitarian

Normal

New Hedonic Normal

New Utiliarian Premium

New Hedonic Premium

4.4 Main Effect of New Product Exclusivity on the Dependent Variables H1 and H2 predict that exclusiveness positively affects SC and BC. These hypotheses will be tested. SC and BC were measured as SCI and BCI. Since there were only small differences between SCI and BCI before and after showing the folder, it was decided to only take into consideration the dependent variables after showing the folder (see Table 10). The dependent variables are only weakly correlated (.37) and the correlation is far below the threshold of 0.7 (Leeflang et al., 2000). Running a linear regression shows a variance inflation factor (VIF) of 1.00, which is below the threshold of 6 and thus it indicates that there are no multicollinearity issues. Table 10 Means, SD and Pearson Correlation Coefficients (n=239) Dependent Variable Mean SD 1 2

SCI_after 4.24 1.44 -- BCI_after 3.83 1.48 .37*** -- +p<0.10, * p<0.05, ** p<0.01, *** p<0.001 Assumptions for normality, homogeneity of variances, linearity, and univariate and multivariate outliers were tested. Both dependent variables are normally distributed [K-SSCI_after=.125, p=.000; K-SBCI_after=.166, p=0.000], Levene’s test shows that variances of both dependent variables are equal (pSCI_after=.862, pBCI_after=.570) and there is a significant linear relationship. Some potential outliers were detected; however, they were not removed for the same reason as described in chapter 4.1. Hypothesis 1 and 2. Table 11 shows the results of two independent samples t-tests, testing whether the test condition (‘new – exclusively available at Dirk’) differs significantly from

46

the control condition (‘new’). SCI turns out to be the same for the test (M=4.24) and control condition (M=4.24). H1 is thus not supported [t(237)=.004, p=.997]. It can be said that overall, adding one exclusive new product to the retailer’s assortment does not impact SCI. This might be due to the fact that only one exclusive product does not make the whole assortment more attractive. Further, exclusivity is only one aspect of the assortment, while e.g. the presence or absence of a favorite product (Broniarczyk, Hoyer, & McAllister, 1998) or the assortments’ arrangement (Hoch et al., 1999) and the amount of good alternatives (Kahn & Lehmann, 1991) also impact the consumer’s perception on assortment variety. H2 neither is supported. The mean score for BCI is even lower for the test condition (M=3.79) than for the control condition (M=3.86), but differences are not significant [t(237)=.356, p=.722]. The lower score in the test condition could be due to the lower buying frequency in the test condition (M=4.24) compared to the control condition (M=4.58). Buying frequency was measured as a background variable. Though this difference could be true in a real life situation as well, a higher buying frequency logically points towards a higher attraction to the promotion in the control condition. Other reasons could be that adding only one exclusive product to a brand portfolio does not impact the consumer’s perspective as such, or the respondents are loyal to other brands and are thus not interesting in the promotion of the exclusive new product. In the sample, Coca Cola (31.4%) and Fanta (25.9%) are i.a. bought most often. For detergents, Robijn (25.9%) is bought most often i.a., followed by Ariel (19.2%). See Figure 8 in Appendix G. In Table 13 the results for the moderating effects of price and product category have been summarized. There, the main effects of exclusivity on SC [F(1, 238)=.734, p=.393] and BC [F(1,238)=.997, p=.319] are given while controlling for product category involvement. The results indicate that there are no significant main effects either when this control variable is included. Table 11 Independent Samples T-tests of the Main Effects Variable

Condition t-value Sig.

Test (SD)

Control (SD)

SCI_after 4.24 (1.52)

4.24 (1.36)

.004 .997

BCI_after 3.79 (1.52)

3.86 (1.45)

.356 .722

The dependent variables were measured on a seven-point (1-7) scale +p<0.10, * p<0.05, ** p<0.01, *** p<0.001 4.5 Main Effect of New Product Exclusivity while controlling for Brand Equity and Store Equity The hypotheses 5 and 6 investigate what effect the control variables brand equity (BE) and store equity (SE) have on the main effects. Both control variables were measured before and after the participants were shown the two-page folder. As there were hardly any differences between the two measurements, only the control variables that were measured after the folder was shown (BE_after and SE_after) are taken into consideration (see chapter 4.3). In this way, respondents have taken into consideration the folder pages in their answers. Most ANOVA assumptions were already tested and explained in the former chapter. Besides, Table 12 shows that there are no variables violating the assumption of correlation between the control variables, neither the assumption of linearity between the control variables and the dependent variables. The high correlation between SE and SCI did not give a multicollinearity

47

issue as the VIF score is lower than 6 (1.180). Both BE and SE were dummies when conducting the analyses, where a high BE is equal to a mean score on the BE items >4. A low score is equal to <4. The same split was used for SE. The number 4 was chosen as it is the upper limit of the mean of the 7-point Likert scale. In the table below both BE and SE are presented as mean scores. Table 12 Means, SD and Pearson Correlation Coefficients (n=239) Variable Mean SD 1 2 3 4 SCI_after 4.24 1.44 -- BCI_after 3.83 1.48 .37*** -- BE_after 4.50 .92 .28*** .68*** -- SE_after 4.48 1.13 .75*** .24*** .29*** -- +p<0.10, * p<0.05, ** p<0.01, *** p<0.001 Distribution of the respondents over the control variables: NBE_low=89, NBE_high=150, NSE_low=97, NSE_high=142 Hypothesis 5 and 6. H5 and H6 were tested using two ANCOVA’s where BE, SE and the control variable ‘product category involvement’ were added as covariates. Table 13 shows that while controlling for SE, the main effect on SCI [F(1, 238)=.432, p=.516] is still non-significant (compared to the main effects in Table 11). Also for BCI, the main effect is still non-significant [F(1, 238)=.009, p=.926]. However, hypothesis 5a is proven to be true as controlling for SE significantly influences the main effect. A low perceived SE leads to a lower SCI (M=3.30) than a high perceived SE (M=4.88), which is a logical outcome as SE and SCI are highly correlated (r=.75). It should be noted that the main effect did not change significantly when controlling for SE, meaning that exclusivity does not improve SE such that SCI is significantly improved. Interestingly, however, investigating differences between means for SE6 with two independent samples t-tests reveals that both in the test and control condition a high (vs. low) SE relates to a significantly higher SCI [ttest(117)=-6.438, p=.000; tcontrol(118)=-7.577, p=.000]. People that have a positive attitude towards the store, will thus be more likely to go to that store when it sells an exclusive or new non-exclusive product. The reason for this could be that people that rate a store as high-equity, will see an offer of a new (and exclusive) product as a conformation of their liking of that store and act upon that. Because the effect is also true for new products, exclusivity (vs. non-exclusivity) does thus not have a higher impact on SCI when SE is high. Hypothesis 5b is proven to be incorrect [F(1, 238)=.536, p=.465], which means that SCI does not improve significantly when BE is perceived as high. A well-perceived BE does thus not influence people’s intention to visit the store. In chapter 7 an extra analysis will be done to see the impact of exclusivity on BE as a dependent variable. Hypothesis 6a is not true [F(1,238)=1.937, p=.165]. A higher SE does thus not significantly influence consumers’ SCI when they perceive the brand as attractive (see Table 13). However, an independent samples t-test shows that the mean BC for a low perceived BE (M=2.90) is significantly lower than when BE is perceived as high [M=4.38, t(237)=-2.812, p=.005]. In chapter 7 an extra analysis will be done to test the impact of exclusivity on BCI. In Table 14 the mean scores are presented.

6 Mean SCI in test condition: SE_low=3.24, SE_high=4.84 vs. control condition: SE_low=3.35, SE_high=4.92. The higher mean scores in the control condition might be due to the higher buying frequency in this condition (M=4.58) compared to the test condition (M=4.24).

48

Although the main effect of exclusivity on BCI has not become significant as a consequence of controlling for BE, hypothesis 6b is true [F(1, 238)=146.801, p=.000] as BE significantly influences the main effect. Adding an exclusive or non-exclusive product with a high perceived BE to a manufacturer’s brand portfolio, will thus contribute to a consumer’s likelihood to purchase that specific brand. Two independent samples t-tests between the means for low and high BE7 reveals that BCI for high-equity brands is significantly higher for both new and exclusive products [ttest(117)=-6.035, p=.000; tcontrol(118)=-5.974, p=.000]. Interestingly, both for a low and high BE scores for exclusive products are higher. Although the effects between the test and control condition are not significant [tBE_low(87)=-.135, p=.893; tBE_high(148)=-.255, p=.799], there seems to be a small effect of exclusivity since the mean scores on BCI are higher in the test condition. Logically speaking, a high perceived BE leads to a higher BCI as the liking of the brand is high. This could be due to the fact that when consumers rate a brand as high-equity, and when seeing a promotion of a new or exclusive product of this brand, their perceptions of the brand are confirmed which in turn leads to buying the brand. Table 13 ANCOVA’s of the Main Effects Controlling for SE and BE

Dependent Variables

SCI_after BCI_after Test variables F

(Sig.) F

(Sig.) Main Effects: Exclusivity vs. non-exclusivity .432

(.516) .009

(.926)

Control variables

Store Equity (SE_after) 249.570*** (.000)

1.937 (.165)

Brand Equity (BE_after) .536 (.465)

146.801*** (.000)

Product Category Involvement 5.520* (.020)

1.390 (.240)

Variables were measured on a seven-point (1-7) scale +p<0.10, * p<0.05, ** p<0.01, *** p<0.001

7 Mean BCI in test condition: BE_low=2.91, BE_high=4.41 vs. control condition: BE_low=2.88, BE_high=4.35.

49

Table 14 SCI and BCI Means for Low vs. High SE and BE

Dependent Variables

SCI_after BCI_after Test Variable Mean Mean

SE Low

3.30 (1.21)

3.51 (1.35)

High 4.88 (1.22)

4.05 (1.54)

BE Low 3.82 (1.42)

2.90 (1.26)

High 4.48 (1.40)

4.38 (1.32)

Variables were measured on a seven-point (1-7) scale

4.6 Moderating and Interaction Effects of Price Premium and Product Category In this chapter, the moderating effects of the introduction price and the product category on the relationship between the independent variable and the dependent variables will be tested. In section 4.3, most assumptions for ANOVA have already been tested. These account also for the ANCOVA that was executed to test hypothesis 3a, 3b, 4a and 4b. Table 15 shows that there is no correlation between the covariates (which are the moderating variables) and the assumption of linear relationships between the covariates and the dependent variables is true. Because the correlations are low, no multicollinearity was perceived. During the tests that were conducted to analyze the moderating effects, both price premium (0=no price premium, 1=price premium) and product category (0=utilitarian/Robijn, 1=hedonic/Fanta) were added as dummy variables. Table 15 Means, SD and Pearson Correlation Coefficients (n=239) Variable Mean SD 1 2 3 4 SCI_after 4.24 1.44 -- BCI_after 3.83 1.48 .37*** -- Price premiuma .50 .501 -.02 -.09 -- Product categoryb .50 .501 -.01 -.03 -.004 -- +p<0.10, * p<0.05, ** p<0.01, *** p<0.001 a Dummy variable (0=no price premium, 1=price premium) b Dummy variable (0=utilitarian/Robijn, 1=hedonic/Fanta) Moderating Effects. To investigate the effect of price premium as a moderating variable, two two-way full factorial univariate ANCOVAs with the control variable ‘product category involvement’ were conducted8. Results are given in Table 16.

8 see Table 20 for an overview of the mean scores per condition for ‘product category involvement’

50

Table 16 ANCOVA’s Testing Moderating Effects of Price and Product Category

Dependent Variables

SCI_after BCI_after Test variables F

(Sig.) F

(Sig.) Main Effects: Exclusivity vs. non-exclusivity .734

(.393) .997

(.319)

Price: premium vs. normal .309 (.579)

2.909+ (.089)

Product category: hedonic vs. utilitarian

.005 (.943)

.224 (.636)

Interaction effects:

Exclusivity vs. Price .023 (.879)

1.785 (.183)

Exclusivity vs. Product Category

.194 (.660)

6.078* (.014)

Control variable Product Category Involvement 33.621***

(.000) 17.891***

(.000) Variables were measured on a seven-point (1-7) scale +p<0.10, * p<0.05, ** p<0.01, *** p<0.001

Price Premium – Hypothesis 3a and 3b. Though the mean scores on the dependent variables differ between the test and control conditions (see Table 17), there was found no significant effect of price premium on SCI [F(1, 238)=.309, p=.579]. Neither an interaction effect of price premium and the main effect on SCI was found. For BCI however, a marginally significant moderating effect of price premium on BCI was found [F(1, 238)=2.909, p=.089]. However, H3a states that a price premium will have a smaller effect on BCI comparing the test and control condition than for the normal price. This is not true, as the difference between the test and control condition for the normal price is smaller (M difference=4.02-3.92=.1) than for the price premium group (M difference=3.56-3.81=-.25). Moreover, BCI in the test condition for the normal price (M=4.02) is higher than when a premium price is asked (M=3.56). The introduction price thus moderates the relationship between exclusivity and BCI, but not in a negative way. No interaction effect of the moderator price premium and the main effect was found. Thus, neither H3a nor H3b was supported (see Table 16). Although there was no significant interaction effect of price premium on SCI, interestingly, its mean score is lower in the control condition (M=4.17) than in the test condition (M=4.24) when a price premium is asked. Although an independent samples t-test shows that this difference is non-significant [t(117)=.278, p=.782], the difference between a normal (M=4.23) and premium price (M=4.24) in the test condition is negligible, while the differences between the normal (M=4.31) and premium price (M=4.17) in the control situation are larger. Although an independent samples t-test shows that this difference is non-significant [t(118)=.602, p=.602], both observations point towards the fact that consumers are willing to pay more for an exclusive new product and exclusivity makes a price premium less important.

51

Moreover, although H3b was not true, BCI is (non-significantly) higher in the test condition (M=4.02) than in the control condition (M=3.92) for a normal price [t(117)=-.377, p=.707], which points towards the fact that exclusivity does have some effect. Future research should deep-dive into these figures (see chapter 5.2). Table 17 SCI and BCI Means for the Moderators Price and Product Category Dependent Variable

SCI_after BCI_after

Variable Level Test Condition

Control Condition

Test Condition

Control Condition

Price Normal 4.23 (1.54)

4.31 (1.26)

4.02 (1.49)

3.92 (1.44)

Premium 4.24 (1.52)

4.17 (1.46)

3.56 (1.53)

3.81 (1.47)

Product Category

Utilitarian 4.15 (1.54)

4.36 (1.29)

4.02 (1.44)

3.73 (1.30)

Hedonic 4.33 (1.51)

4.12 (1.43)

3.56 (1.57)

3.99 (1.58)

Variables were measured on a seven-point (1-7) scale. Standard deviations are given in parentheses. Product category – Hypothesis 4a and 4b. In Table 16, the results for the two two-way full factorial univariate ANCOVAs for the moderator ‘product category’ are presented as well. Hypothesis 4a was not supported, since there is no significant moderating effect on SCI [F(1, 238)=.005, p=.943]. There was neither a significant interaction effect of exclusivity and the product category. Although the product category did not seem to moderate BCI either [F(1, 238)=.224, p=.636], which means that H4b is not supported, there was found a weak but significant (on the .05 level) interaction effect of exclusivity and the product category on BCI [F(1, 238)=6.078, p=.014]. The mean scores for the test and control conditions are given in Table 17. Looking at Table 16 and 17, the opposite (but non-significant) interaction effect can be found for the product category and the main manipulation on SCI: SCI is higher for the control group (M=4.36) than for the exclusive product (M=4.15) in case of utilitarian products, while exclusive hedonic products lead to a higher SCI (M=4.33) than its new version (M=4.12). The higher score for hedonic products in the test condition might be due to the fact that soft drinks are bought more often and therefore it is a reason to visit the store for this particular promotion. It is interesting that the interaction effect of the product category for SCI is the opposite of the one for BCI. In Figure 5, the interaction effect of the product category and the main effect on BCI is graphically presented. Interestingly, participants in the control group that saw the Fanta folder have a higher BCI (M=3.99, SD=1.58) than those in the test group (M=3.56, SD=1.57). The opposite accounts for the ‘Robijn’ group: participants in the test group had a higher BCI (M=4.02, SD=1.44) than the control group (M=3.73, SD=1.30). For hedonic products, exclusivity thus seems not to work as BCI is lower than for the non-exclusive new Fanta. In contrast, the exclusive new utilitarian product causes a higher BCI than its non-exclusive version. Since utilitarian products have less hedonic features, people might not expect these

52

products to be exclusive which could trigger them to purchase the brand. Another reason could be that in general Robijn Sweet Jasmine was found to be a more attractive offer than Fanta Wild Berry. Figure 5 Interaction Effect of Product Category on Brand Choice

4.7 Additional Analyses As discussed in chapter 4.5, both H5b and H6a were not true. In this section, some extra analyses will be conducted regarding these hypotheses. As there seems to be some kind of relationship between exclusivity and the control variables BE_after and SE_after, these control variables were taken into consideration as dependent variables. Brand Equity as Dependent Variable. Hypothesis 5b, stating that a higher perceived BE would show a higher SCI compared to a low BE of the (exclusive) new product, was rejected. However, performing an independent samples t-test revealed that the mean scores on SCI for a low perceived BE (M=3.82) and high perceived BE (M=4.48) differed significantly [t(237)=-3.509, p=.001]. It would be interesting to see what direct influence exclusivity has on BE as a construct. If there is an effect, it would be interesting to investigate whether BE influences BCI in turn, so that BE could maybe be seen as a mediating effect on BCI. Another independent samples t-test shows however that the BE score in the test [M=4.47, SD=.92] and control situation [M=4.53, SD=.93] does not differ significantly [t(237)=.454, p=.650]. Adding one exclusive product to the brand portfolio does thus not directly influence BE of the whole brand. It should be noted that only one product was added to the brand portfolio. In chapter 5.2 some suggestions will be done for future research regarding this issue. BE is a construct that is hard to measure as it covers many concepts of the brand. To illustrate, Keller (1993), Sloot & Verhoef (2008), Aaker (1991) and Agarwal & Rao (1996) all define and measure BE in different ways. As the concept covers so many aspects, a possible reason for the non-difference could be that adding one exclusive new product to the brand portfolio was not enough to impact the broad construct. The same differences between means were found for the scores regarding H6a. This hypothesis neither seemed to be true, however, since the means for a low perceived SE (M=3.51) and high perceived SE (M=4.05) differ significantly on a 5% significance level as

3,3

3,4

3,5

3,6

3,7

3,8

3,9

4

4,1

New & Exclusive (test) New (control)

Mea

n B

rand

Cho

ice

Condition

Utilitarian

Hedonic

53

well [t(237)=-2.812, p=.005]. Thus, it would be interesting to see what the relationship is between exclusivity and SE as a dependent variable. If there is a perceived effect, SE might be a sort of mediating effect of exclusivity on SCI. However, an independent samples t-test shows that the slightly higher mean score on SE for an exclusive new product [M=4.50, SD=1.08] and new product [M=4.46, SD=1.18] did not differ significantly [t(237)=-.267, p=.789]. Adding one new exclusive product to the store’s assortment does thus not impact the store’s equity. The term SE appears less often in the literature and no consistent way of measuring this concept has been given yet. SE is, just as the BE construct, a broad term which could explain the non-difference in SE when adding only one exclusive product to the assortment. 4.8 Hypothesis Results An overview of the hypotheses results is given in Table 18. Table 18 Hypotheses Results Hypothesis Result Independent

Variable Dependent Variable

Moderator Control Variable

F- or t-value

1 Rejected Exclusivity of New product

SCI .004

2 Rejected BCI .356 3a Rejected SCI Price premium .309

3b Rejecteda BCI Price premium 2.909+

4a Rejected SCI Product Category .005 4b Rejected BCI Product Category .224 5a Accepted SCI Store Equity 249.570*** 5b Rejected SCI Brand Equity .536 6a Rejected BCI Store Equity 1.937 6b Accepted BCI Brand Equity 146.801***

+p<0.10, * p<0.05, ** p<0.01, *** p<0.001 a The effect is marginally positively significant, while H3b is that the effect has a negative relationship, which makes the hypothesis false.

4.9 Control Variables Differences between the means of the control variables among the eight conditions were investigated by one-way between groups multivariate ANOVAs (Table 20). Just like for the SCI_after and BCI_after scales, the three-item product category involvement scale was averaged. The Kolmogorov-Smirnov test showed that the assumption of normality was not violated for ‘product category involvement’ (K-S=.098, p=.000). Some potential outliers were detected; however, they were not removed for the same reason as described in chapter 4.1. As can be seen in Table 19, almost no multicollinearity among the control variables and between the control variables and the dependent variables were found (most correlations were >0.7). For SE and SCI, there might to be a correlation issue. However, VIF is 1.180 which is lower than the threshold of 6, which indicates that there is no multicollinearity.

54

Table 19 Means, SD and Pearson Correlation Coefficients (n=239) Variable Mean SD 1 2 3 4 5 SCI_after 4.24 1.44 -- BCI_after 3.83 1.48 .37*** -- BE_after 4.50 .92 .28*** .68*** -- SE_after 4.48 1.13 .75*** .24*** .29*** --

-- Product Category Involvement 4.81 1.06 .35*** .23*** .41*** .35***

Variables were measured on a seven-point (1-7) scale +p<0.10, * p<0.05, ** p<0.01, *** p<0.001 No significant differences were found, which means that the control variables did not significantly influenced the results (see Table 20). Table 20 ANOVA Control Variables

Test Cells (“new – exclusively available at

Dirk”)

Control Cells (“new”)

Condition Normal Robijn

Normal Fanta

Premium Robijn

Premium Fanta

Normal Robijn

Normal Fanta

Premium Robijn

Premium Fanta

F (Sig.)

BE_after 4.67 (1.03)

4.60 (.96)

4.58 (.75)

4.04 (.81)

4.62 (.83)

4.56 (.95)

4.50 (.75)

4.44 (1.16)

1.351 (.227)

SE_after 4.61 (1.06)

4.71 (.91)

4.47 (1.99)

4.21 (1.32)

4.37 (1.12)

4.57 (1.16)

4.71 (1.14)

4.21 (1.29)

.971 (.453)

Product Category Involvementa

4.71 (1.25)

4.94 (1.00)

4.99 (.95)

5.21 (.89)

4.91 (1.05)

4.52 (1.31)

4.71 (.85)

4.51 (1.09)

1.565 (.147)

Ageb 3.23 (1.38)

3.13 (1.48)

3.27 (1.41)

3.31 (1.42)

3.28 (1.39)

3.23 (1.38)

3.20 (1.40)

3.16 (1.42)

.361e

(1.000)

Genderc .70 (.47)

.70 (.47)

.70 (.47)

.72 (.45)

.72 (.45)

.70 (.47)

.70 (.47)

.68 (.48)

.228f

(1.000)

Nationalityd n/a n/a n/a n/a n/a n/a n/a n/a -- +p<0.10, * p<0.05, ** p<0.01, *** p<0.001. Standard deviations are given in parentheses. a Measured on a seven-point (1-7) scale b Age categories: 18-24 (1), 25-34 (2), 35-44 (3), 45-54 (4), 55-65 (5) c Dummy variable with male (0) and female (1) d All respondents in the sample had the Dutch nationality e The Kruskal-Wallis test was used to determine the difference between the eight conditions f A Pearson Chi-square test was used to determine the difference between the eight conditions

55

4.9.1 Extra Analyses on the Control Variables Inspired by the mean scores presented in Table 20, some extra analysis on the control variables were conducted. Main Effects. An independent samples t-test shows that the mean scores on SE in the test condition (M=4.50, SD=1.08) and control condition (M=4.46, SD=1.18) did not differ significantly [t(237)=-.267, p=.789]. SE did thus not rise due to exclusivity. A similar test concludes the same for BE: the mean BE score in the test condition (M=4.47, SD=.92) is not significantly different from the control condition (M=4.53, SD=.93), meaning that adding one exclusive new product to the assortment does not lead to an increased BE. Store Equity. In the test condition, an independent samples t-test with a filter on the exclusive cases and Fanta-conditions shows that the perceived SE for Fanta is marginally significantly higher (M=4.71) when a normal price is asked than when a premium price is asked [M=4.21, t(57)=1.695, p=.096]. In contrast, in the control condition, SE for Robijn is higher (M=4.71) when a premium price is asked compared to a normal price (M=4.37). Since the latter difference is non-significant [t(57)=-1.161, p=.250] and since it can be found in the control condition, it is less relevant in the context of exclusivity. A possible explanation for the difference between Fanta and Robijn could be that Fanta is bought more often than Robijn, so that consumers are less willing to pay more for the Fanta promotion. Taking into consideration SE, it would thus not be recommended to ask a price premium for exclusive hedonic products.

56

5. Discussion and Conclusion Where exclusive distribution of key national brands is already a common phenomenon in the toy and apparel industry (Gielens et al., 2014), this strategy is rather new within the FMCG. Because of the multi-store patronage of today’s consumers and because the assortment is a key determinant of retail success (Grewal et al., 2010), retailers try to differentiate their assortments and manufacturers try to differentiate their brand portfolio’s with the help of this strategy. This study investigates whether both retailers and manufacturers can benefit from exclusive retail distribution by examining how an exclusive vs. non-exclusive new product promotion presented in a Dirk-folder influences store choice (SC) and brand choice (BC).

Main Effect on SC. However, the hypothesis that SC is significantly influenced by exclusivity is not supported. Pan & Zinkhan (2006) found that the assortment is the most important determinant of SC. Many authors suggest that a larger assortment increases its variety and therefore increases SC (e.g. Hoch, Bradlow & Wansink, 1999; Ailawadi & Keller, 2004; Fox, Montgomery & Lodisch, 2004). Others did not find higher sales due to assortment size (e.g. Dreze et al., 1994) or state that retailers make the mistake of ‘over-assorting’ stores (Boatwright & Nunes, 2001). The finding seems to be in line with the last research stream, since SC did not increase due to the promotional offer of the exclusive new product. Another reason could be that exclusivity is only one factor that could impact assortment variety, while e.g. the presence or absence of a favorite product (Broniarczyk, Hoyer, & McAllister, 1998) or the assortments’ arrangement (Hoch et al., 1999) and the amount of good alternatives (Kahn & Lehmann, 1991) also impact the consumer’s perception on assortment variety.

Main Effect on BC. Secondly, BC was not significantly higher due to exclusivity either. Berger, Draganska & Simonson (2007) concluded that a more varied product portfolio of a brand is positively related to purchase intention. Contrarily, Iyengar & Lepper (2000) explain that a more varied assortment in their study led to decreased purchase intention because consumers expierenced difficulty in choosing. In this study, adding one exclusive product to the brand portfolio was no significant trigger for increased BC. Reasons could be choice overload (Greenleaf & Lehmann, 1995) or loyalty to other brands9. As respondents mentioned in the third pretest, another reason could be resistance to trying new products.

Controlling for Store Equity. When controlling for store equity (SE), the main effects of exclusivity on BC and SC are still non-significant. Adding one exclusive product with a high perceived BE to the assortment does thus not lead to a significant increase in SC and BC compared to new products. However, interestingly, the main effect of exclusivity on SC does increase to some extent: people that perceive the store’s equity as high have a significant higher SCI (M=4.88) than when they perceive the store’s equity as low (M=3.30). Research suggests that high-equity brands are perceived as more valuable due to the extra benefits that they provide (Chandon et al., 2000), which can be an explanation for the impact that they have on SCI. Another explanation could be that people that rate a store as high-equity, will see an offer of a new (and exclusive) product as a conformation of their liking of that store and act upon that. In contrast, controlling for SE does not significantly impact the effect of exclusivity on BC.

9 25.9% in the sample normally buys Fanta i.a; accidentally 25.9% also buys Robijn i.a. See Figure 8 in Appendix G.

57

Controlling for Brand Equity. Controlling for brand equity (BE) did not lead to a significant main effect of exclusivity on BC and SC either, but the main effect was improved significantly. An interesting finding is that both for a low and high BE, BC was higher when an exclusive product was shown in the two-page folder. Looking at this result, exclusivity might therefore be used to differentiate from other high-equity brands. As the main effect is still non-significant, it cannot be said that BE is the only tool that should be taken into consideration when introducing exclusive products. It could however improve BCI. It is a common thought in the marketing literature that BE determines how consumers respond to marketing stimuli (e.g. Batra & Ahtola, 1990; Chandon et al., 2000; Dhar & Wertenbroch, 2000), which is in line with this finding. Another reason could be that consumers’ perceptions of the brand are confirmed when seeing a promotion of an exclusive or new offer of a product from a brand that they rate as high-equity, leading to buying this brand.

Price. Findings in the literature about the role of price on SC are mixed. For instance, Pan & Zinkhan (2006) found that the store’s general price level negatively relates to SC. There is other evidence that price serves as a quality cue (Tellis & Gaeth, 1990), which led to the assumption that price could influence the strength of the relationship between the (non-)exclusive product and SC and BC. A moderating effect of the introduction price on the consumer’s SC has not been found, suggesting that the relationship between exclusivity and SC did not change when a normal vs. premium price was asked for the exclusive new product. Means for SC were even almost similar for the exclusive new product when a normal vs. premium price was asked. It would thus not matter for the SC to increase the price of the exclusive product with 10%, which is interesting for retailers. Since all the other factors were held constant, the difference should be due to exclusivity. Although not significant, exclusivity does seem to have some effect. To compare, a premium price led to a lower SC compared to a normal price for the new product. Looking at the mean scores on the control variable SE, the perceived SE for the exclusive Fanta offer is marginally significantly higher when a normal price is asked than when a premium price is asked. A possible explanation for the fact that there is no difference for Robijn could be that Fanta is bought more often than Robijn, so that consumers are less willing to pay more for the Fanta promotion.

For BC, a positive moderating effect of price on exclusivity has been found, which means that the nature and strength of the relationship changed when a price premium vs. a normal price was asked. The hypothesis that a price premium for exclusive products would have a smaller effect on BC than for new products was not true, as BC was highest for the normal price for an exclusive product. In case of a premium price, BC is highest for the new product, which is in line with the fact that consumers prefer low prices in general (Dodds et al, 1991; Walters & Rinne, 1986). For the purpose of BC, it would thus not be recommended to ask a price premium for exclusive products.

Product Category. No evidence is found for a moderating effect of the product category (utilitarian vs. hedonic) on SC and BC either, which means that the product category does not significantly influence the relationship between exclusivity and SC or BC. Interestingly, the study finds a significant interaction effect of the product category and the test and control conditions on BC: exclusivity of the new product causes a higher BC for utilitarian products than for hedonic products and showing a non-exclusive product leads to a lower BC for utilitarian products than for hedonic products. This means that for hedonic products,

58

exclusivity does not seem helpful while for utilitarian products, it does. The finding is contrary to existing research stating that consumers have a smaller need for variety for utilitarian products (Van Trijp et al., 1996). A possible explanation for this phenomenon is that people might not expect to see an exclusive utilitarian product, so that it has extra impact if such a utilitarian product is introduced compared to exclusive hedonic products. Another explanation could be that the Robijn Sweet Jasmine promotion was found to be a more attractive offer in the two-page folder compared to the Fanta Wild Berry promotion. For SC, the opposite (non-significant) interaction effect was found. Here, the exclusive new hedonic product causes a higher SC than the exclusive new utilitarian product. For the new product, the opposite was true. The higher SC for the hedonic exclusive product compared to the utilitarian product might be due to the fact that soft drinks are bought more often and therefore it is a reason to visit the store for this particular promotion10. It should be noted that because only one product in the study represented the utilitarian category and one product represented the hedonic category, generalizing the results to all utilitarian and hedonic product categories should be done carefully (see chapter 5.2 on how external validity could be improved).

5.1 Managerial Implications This research has implications for both brand managers and retailers. The research is important for academics as they could do follow-up research based on the study. Based on the research, it can be concluded that offering one exclusive new product to the assortment does not increase SC and BC directly for a price-oriented retailer. However, there are indications that exclusivity can provide some extra advantages to retailers and manufacturers. First, the control variable ‘product category involvement’ significantly influenced the main effects of exclusivity on both SC and BC. Although not significant, the extent to which a consumer is involved with the category influences its perception on the introduction of new products. When introducing exclusive new products, it is recommended to target consumers that have a high product category involvement. For instance, magazines targeting food lovers could be used to advertise an exclusive new soft drink and magazines targeting mothers could show promotions for an exclusive detergent.

Further, the findings suggest retailers to invest into SE in general, because a higher SE leads to higher SC for both exclusive and new products. This would strengthen their position at manufacturers. For manufacturers, it is recommended to invest into BE in general as higher BE causes a higher BC. This is true for both new and exclusive products. However, the mean scores were some higher (not significantly) for exclusive products, so this should be a topic for future research. Introducing (for manufacturers) or accepting (for retailers) an exclusive product with a high BE seems to significantly increase BC, which is therefore recommended.

Based on this study, it is needed to do additional research into the influence of pricing on exclusivity (see chapter 5.2). Because no significant differences were found on SC for a normal vs. a premium price, it is possible to ask small (10%) price premiums for exclusive products from a retail perspective. Taking into consideration the control variable SE, it would however not be recommended to ask a price premium for exclusive hedonic products that are

10 Mean buying frequency for soft drinks (in the soft drink conditions) is 5.23 vs. 3.58 for detergents (in the detergent conditions).

59

bought often. Futher, since price significantly moderates the relationship between exclusivity and BC in the sense that a normal price causes a higher BC than when a price premium is asked, it would not be recommended for manufacturers to ask a higher price. Based on this study, a recommendation for manufacturers would be to make their brands more attractive by adding exclusive products to utilitarian product categories. This seems to have a larger effect on BC than when hedonic products are made exclusive. For retailers, it would be interesting to focus on adding hedonic products to their assortments. However, it should be noted that no direct effect of adding one new exclusive product to the assortment was found on SC and BC, which means that the results should be interpreted carefully and that future research is needed. The findings of this research contribute to the existing literature on retail exclusivity and more suggestions for research to follow on this study will be given in chapter 5.2.

5.2 Limitations and Future Research There are several limitations of the study that could be triggers for future research. Moreover, some topics trigger future research as there seem to be relationships with exclusivity. Retail formulas. First, the study is limited to one country only as the Dutch supermarket Dirk was chosen as retailer. Moreover, the sample consisted of Dutch shoppers. The results can therefore not be generalized to other countries that may have other supermarket formulas. It would be interesting to conduct research in other countries’ supermarkets too and, for instance, compare them among. Consumer perceptions might be different in different countries as well. Moreover, the study was conducted for only one retail formula, namely for Dirk, which is a price-oriented retailer implying that consumers might have other expectations of the store and the assortment than when shopping at a service-oriented retailer. It would be interesting to investigate perceptions between price-oriented and service-oriented retailers, or even between supermarkets and department stores. Consumers might expect that supermarkets have smaller assortments than department stores. Supermarket size might play a roll as well. Small supermarkets have a more limited assortment than larger ones, implying that exclusive products might therefore be more applicable for certain retail formulas. The shopping task might influence SC and BC as well, so future research could take different shopping tasks into consideration as moderator or control variable. Exclusive Products. Another limitation of the research is that it measures the impact of adding only one new exclusive product to the retailers’ assortment. It was decided to introduce only one exclusive product in the folder, because it would not be realistic to introduce more exclusive products simultaneously in the weekly folder. As the consumers’ perception of the assortment is influenced by multiple factors, such a small amount of exclusive products has a smaller effect on consumer intentions than more exclusive products might have. Future research might therefore extend this study by investigating consumer response to assortments with more exclusive products across multiple categories. A suggestion would be to measure perceptions at multiple points in time. Longitudinal studies can be insightful to tackle wrong interpretations due to time constraints of the variables (Andrews, 1988). That is another limitation of the study: SC and BC were measured based on only one folder, while perception of a store and brand might change over time if more exclusive products would be added to the assortment.

60

Strategic Variables. Since consumer perception on exclusive retail distribution in the FMCG is a relative new field of research, and as it concerns a strategic tool for retailers and manufacturers, it would be interesting to investigate the influence of other strategic variables such as the timing of introduction or the newness of the exclusive product (Gielens et al., 2014 investigated the influence of the product’s newness on its acceptance). Moreover, the influence of other pricing strategies (such as introduction discounts or seasonal promotions) on the consumer’s perception of exclusive products could be investigated as well. Price. The outcomes of this research suggest that retailers are able to ask a price premium for exclusive products without risking a decreasing SC. However, asking a premium price for an exclusive product seems to decrease BC. Price thus seems to have a larger effect on BC than on SC. Since exclusivity does lead to a higher BC when a normal price is asked, future research should try to investigate how come that SC does decrease and BC does. It would be interesting to see what effect price has on brands, BC and on stores and SC in general. Differences might become larger if a price premium of e.g. 20% would be asked. Product Category. Further, this study shows an interaction effect between the product category and BC. The hedonic levels of both detergents and soft drinks differed significantly, however, given the fact that an interaction effect was found, it would be interesting to extend this study by investigating other product categories that differ more on their hedonic levels. This would improve external validity, can help retailers to decide which products to offer exclusively and helps manufacturers decide for which brands exclusive new products should be introduced. Moreover, this interaction effect for BC shows that utilitarian product should be made exclusive, while the (non-significant) interaction effect for SC shows that the hedonic product should be made exclusive as SC is higher in that case. Since groups were small (n=30) and since differences between means are small, future research could investigate why people would visit the store if an exclusive hedonic product is offered and why a utilitarian product seems to improve BC. More understanding of product categories and their influence on SC and BC is needed to understand these differences. If more exclusive products would be added to the folder, this might also increase the SC and BC, which is a suggestion for a future experiment as well. Brand Equity and Store Equity. Both SE and BE are intangible assets and multidimensional, complex constructs that are hard to measure and thus it is hard to investigate how to improve these variables. Despite growing interest in the concepts, no consistent way of measuring BE and SE in the literature can be found. For instance, Keller (1993), Sloot & Verhoef (2008), Aaker (1991) and Agarwal & Rao (1996) all define and measure BE in different ways. The term SE appears less often in the literature and no consistent way of measuring this concept has been given either. Therefore, and as both SE and BE seem to have a link with exclusivity, measuring and improving SE and BE could be a topic for future research.

5.3 Conclusion Since strict store loyalty is a rarity nowadays, and as assortment composition is a key determinant of store choice (SC in short, see Grewal et al., 2010), retailers try to differentiate themselves via their assortments and manufacturers are focused on differentiating their brand portfolio’s. While it is still a new phenomenon, exclusive retail distribution is increasingly used in the FMCG as a differentiation strategy. To fill the existing gap in the academic literature, it was investigated how exclusive product offerings influence consumers’ SC and brand choice (BC). As exclusivity is a strategic phenomenon, the product category (utilitarian

61

vs. hedonic) and the introduction price (price premium vs. no price premium) were taken into consideration as moderating variables. Control variables were, among others, brand equity (BE) and store equity (SE). Next to the exclusivity of the new product, the product category and price of the exclusive new product were manipulated. Although the study did not find evidence for the exclusive new product significantly influencing store choice and brand choice (BC), controlling for store equity leads to a significant increase in SC. Moreover, when controlling for BE, the main effect between exclusivity and BC significantly increases. Mean scores for BC are (non-significantly) higher for the exclusive products both when BE is low and high. The relationships between exclusivity and SC and BC were not moderated by the product category. The main effect of exclusivity on BC was neither moderated by the product category. The relationship between exclusivity and BC was moderated by price, however this relationship is positive which means that both in the test and control condition, a normal price is preferred above the premium price. Moreover, a significant interaction effect of the product category and the test and control conditions on BC was found: exclusivity of the new product causes a higher BC for utilitarian products than for hedonic products and showing a non-exclusive product leads to a lower BC for utilitarian products than for hedonic products. Despite rejecting several of the hypotheses, the study triggers future research in multiple ways.

62

References D. Aaker (1996). In D. Aaker, Building strong brands. New York: Free Press. Aaker, D. A (1991), Managing Brand Equity, New York: Free Press. Ailawadi, K., & Keller, K. (2004). Understanding retail branding: conceptual insights and research priorities. Journal of Retailing , 80, 331-342. Ailawadi, K., Scott, A., & Gedenk, K. (2001). Pursuing the Value-Conscious Consumer: Store Brands Versus National Brand Promotions. Journal of Marketing , 65 (1), 71-89. Ailawadi, K., Pauwels, K., & Steenkamp, J. (2008). Private-label use and store loyalty. Journal of Marketing , 72 (6), 19-30. Agarwal, M. K., Rao, V. (1996), "An empirical comparison of consumer based measures of brand equity", Marketing Letters, 7 (3), 237-247. Alba, J., & Hutchinson, J. (1987). Dimensions of Consumer Expertise. Journal of Consumer Research , 13, 411-454. Andritsos, D., & Tang, C. (2010). Launching new products through exclusive sales channels. European Journal of Operational Research, 204 (2), 366-375. Andrews, J. (1988). Motivation, Ability, and Opportunity to Process Information: Conceptual and Experimental Manipulation Issues. Advances in Consumer Research , 15 (1), 219-225. Arnold, S., Oum, T., & Tigert, D. (1983). Determinant Attributes in Retail Patronage: Seasonal, Temporal, Regional and International Comparisons. Journal of Marketing Research, 20, 149-157. Avery, J., Kozinets, R. V., Mittal, B., Raghubir, P., & Woodside, A. G. (2007). Consumer Behavior: Human Pursuit Of Happiness In The World Of Goods. Cincinnati: Open Mentis. J. Bain (1956), Barriers to New Competition. Cambridge: Harvard University Press. Baker, J., Grewal, D., & Parasuraman, A. (1994). The Influence of Store Environment on Quality Inferences and Store Image. Journal of the Academy of Marketing Science , 22 (4), 328-339. Baker, J., Parasuraman, A., Grewal, D., & Voss, G. (2002). The Influence of Multiple Store Environement Cues on Perceived Merchandise Value and Patronage Intentions. Journal of Marketing , 55, 120-141. Baltas, G. (1997). Determinants of store brand choice: a behavioral analysis. Journal of Product & Brand Management , 6 (5), 315-324. Baltas, G., Argouslidis, P., & Skarmeas, D. (2010). The Role of Customer Factors in Multiple Store Patronage: A Cost–Benefit Approach. Journal of Retailing , 1, 37-50.

63

Barone, M., & Roy, T. (2010). Does Exclusivity Always Pay Off? Exclusive Price Promotions and Consumer Response. Journal of Marketing , 74 (2), 121-132. Batra, R., & Ahtola, O. (1990). Measuring the Hedonic and Utilitarian Sources of Consumer Attitudes. Marketing Letters, 2, 159-170. Baumol, W., & Ide, E. (1956). Variety in Retailing. Management Science, 3 (1), 93-101. Bellenger, D., Robertson, D., & Greenberg, B. (1977). Shopping center patronage motives. Journal of Retailing, 53 (2), 29-38. Berger, J., Draganska, M., & Simonson, I. (2007). The Influence of Product Variety on Brand Perception and Choice. Marketing Science, 26 (4), 460-472. Berry, L., Seiders, K., & Grewal, D. (2002). Understanding service convenience. Journal of Marketing, 66, 1-17. Boatwright, P., & Nunes, J. (2001). Reducing assortment: An attribute-based approach. Journal of Marketing, 65, 50-63. Bolton, R., & Shankar, V. (2003). An empirically derived taxonomy of retailer pricing and promotion strategies. Journal of Retailing, 79, 213-224. Boulding, W., Lee, E., & Staelin, R. (1994). Mastering the Mix: Do Advertising, Promotion, and Sales Force Activities Lead to Differentiation? Journal of Marketing Research, 31, 159-172. Bras, E. (2003, December 1). Unilever wil vaker exclusief met retailers samenwerken. (Marketingonline, Interview) Brengman, M., & Geuens, M. (2002). Profiling Internet users based on their propensity to adopt online shopping. Asia Pacific Advances in Consumer Research, 5, 5-36. Briesch, R., Chintagunta, P., & Fox, E. (2009). How does assortment affect grocery store choice? Journal of Marketing Research, 46 (2), 176-189. Broniarczyk, S., Hoyer, W., & McAllister, L. (1998). Consumers’ Perceptions of the Assortment Offered in a Grocery Category: The Impact of Item Reduction. Journal of Marketing Research, 35, 166-176. Brown, S. (1989). Retail Location Theory: The Legacy of Harold Hotelling. Journal of Retailing, 65 (4), 450-470. Bucklin, R., & Lattin, J. (1991). A Two State Model of Purchase Incidence and Brand Choice. Marketing Science, Winter (1), 24-39. Cachon, G., & Kök, A. (2007). Category management and coordination in retail assortment planning in the presence of basket shopping consumers. Management Science, 53 (6), 934-951.

64

Cai, G., Dai, Y., & Zhou, S. (2012). Exclusive Channels and Revenue Sharing in a Complementary Goods Market. Marketing Science, 31 (1), 172-187. Campbell, M., & Keller, K. (2003). Brand Familiarity and Advertising Repetition Effects. Journal of Consumer Research, 30, 292-304. Chakravarti, A., & Janiszewski, C. (2004). The Influence of Generic Advertising on Brand Preferences. Journal of Consumer Research, 30, 487-502. Chandon, P., Wansink, B., & Laurent, G. (2000). A benefit congruency framework of sales promotion effectiveness. Journal of marketing, 64 (4), 65-81. Chang, M. (1992). Exclusive dealing contracts in a successive duopoly with side payments. Southern Economic Journal, 59 (2), 180-193. Chen, S., Monroe, K., & Lou, Y. (1998). The Effects of Framing Price Promotion Messages on Consumers' Perceptios and Purchase Intentions. Journal of Retailing, 74 (3), 353-372. Cheng, A. (2008). US retailers hope licensing brands will help traffic, profits. Dow Jones Newswires. Chernev, A. (2003). When More Is Less and Less Is More: The Role of Ideal Point Availability and Assortment in Consumer Choice. Journal of Consumer Research, 30 (2), 170-183. Choi, J. (1998). Brand extension as informational leverage. Review of Economic Studies, 65 (225), 655−669. Collins-Dodd, C., & Lindley, T. (2003). Store brands and retail differentiation: the influence of store image and store brand attitude on store own brand perceptions. Journal of Retailing and Consumer Services, 10, 345-352. Cooper, P. (1969). The Begrudging Index and the Subjective Value of Money. In B. Taylor, & G. Wills, Pricing Strategy (pp. 122-131). London: Staples Press, Ltd. Corstjens, M., & Lal, R. (2000). Building Store Loyalty Through Store Brands. Journal of Marketing Research, 37 (August), 281-291. Coughlan, A. T., Stern, L. W., Anderson, E., & El-Ansary, A. I. (2001). In Marketing Channels (Sixth edition). Upper Saddle River: Prentice Hall. C. S. Craig, & S. P. Douglas (2005). International Marketing Research 3rd edition. John Wiley & Sons Ltd. Darden, W., & Schwinghammer, J. (1985). The influence of social characteristics on perceived quality in Patronage Choice Behavior. In J. Jacoby, & C. Olson, Perceived Quality: How Consumers View New Stores and Merchandise (pp. 161-172). Lexington, MA: D. C. Heath and Company.

65

Darke, P., & Chung, C. (2005). Effects of pricing and promotion on consumer perceptions: it depends on how you frame it. Journal of Retailing, 81 (1), 35-47. Darley, W., & Lim, J. (1993). Store-Choice Behavior for Pre-Owned Merchandise. Journal of Business Research, 27, 17-31. Dellaert, B., Arentze, T., & Bierlaire, M. (1998). Investigating consumers’ tendency to combine multiple shopping purposes and destinations. Journal of Marketing Research , 3 (2), 177-188. Dhar, R., & Wertenbroch, K. (2000). Consumer choice between hedonic and utilitarian goods. Journal of Marketing Research, 37 (1), 60-71. Dhar, S., Hock, S., & Kumar, N. (2001). Effective category management depends on the role of the category. Journal of Retailing, 77 (2), 165-184. Ding, M., Ross, W., & Rao, V. (2010). Price as an Indicator of Quality: Implications for Utility and Demand Functions. Journal of Retailing, 86 (1), 69-84. Dodds, W., & Monroe, K. (1985). The effect of brand and price information on subjective product evaluations. Advances in consumer research, 12 (1), 85-90. Dodds, W., Monroe, K., & Grewal, D. (1991). The effects of price, brand, and store information on buyers’ product evaluations. 28, 307-319. Dreze, X., Hoch, S., & Purk, M. (1994). Shelf Management and Space Elasticity. Journal of Retailing , 70 (4), 301-326. Erdem, T., & Swait, J. (2004). Brand Credibility, Brand Consideration and Choice. Journal of Consumer Research, 31 (June), 191-198. Erdem, T., Zhao, Y., & Valenzuela, A. (2004). Performance of store brands: a cross-country analysis of consumer store-brand preferences, perceptions, and risk. Journal of Marketing Research , 41 (1), 86-100. Fader, P., & McAlister, L. (1990). An Elimination by Aspects Model of Consumer Response to Promotion Calibrated on UPC Scanner Data. Journal of Marketing Research, 27 (August), 322-332. Fein, A., & Anderson, E. (1997). Patterns of Credible Commitments: Territory and Brand Selectivity in Industrial Distribution Channels. Journal of Marketing, 61 (2), 19-34. Finn, A., & Louviere, J. (1990). Shopping center patronage models: fashioning a consideration set segmentation solution. Journal of Business Research, 21 (3), 159-175. Fotheringham, A. (1988). Note-Consumer Store Choice and Choice Set Definition. Marketing Science, 7 (3), 299-310. Fox, E., Montgomery, A., & Lodish, L. (2004). Consumer Shopping and Spending Across Retail Formats. Journal of Business, 77 (2), S25-S60.

66

Frazier, G. (1999). Organizing and managing channels of distribution. Journal of the Academy of Marketing Science, 27 (2), 226-240. Frazier, G., & Lassar, W. (1996). Determinants of distribution intensity. Journal of Marketing, 60 (4), 39-51. Gardner, D. (1970). An Experimental Investigation of the Price-Quality Relationship. Journal of Retailing, 46 (3), 25-41. Gentry, J., & Burns, A. (1977). How ‘important’ are evaluative criteria in shopping center patronage? Journal of Retailing, 53 (4), 73-86, 94. Gielens, K., & Steenkamp, J. (2007). Drivers of consumer acceptance of new packaged goods: An investigation across products and countries. International Journal of Research in Marketing , 24, 97-111. Gielens, K., Gijsbrechts, E., & Dekimpe, M. (2014). Gains and losses of exclusivity in grocery retailing. International Journal of Research in Marketing, 31, 239-252. Glass, G. & Stanley, J. (1970). Statistical methods for education and psychology. Englewood Cliffs, NJ: Prentice-Hall. Golder, P. N. (2000). Insights from senior executives about innovation in international markets?. Journal of Product Innovation Management, 17 (5), 326. Greenleaf, E., & Lehmann, D. (1995). Reasons for substantial delay in consumer decision making. Journal of Consumer Research, 22, 186-199. Grewal, D., & Marmorstein, H. (1994). Market Price Variation, Perceived Price Variation and Consumers' Price Search Decisions for Durable Goods. Journal of Consumer Research, 21, 452-460. Grewal, D., Ailawadi, K., Gauri, D., Hall, K., Kopalle, P., & Robertson, J. (2011). Innovations in Retail Pricing and Promotions. Journal of Retailing, 87S (1), S43-S52. Grewal, D., Baker, J., Levy, M., & Voss, G. (2003). The effects of wait expectations and store atmosphere evaluations on patronage intentions in service-intensive retail stores. Journal of Retailing, 79, 259-268. Grewal, D., Krishnan, R., Baker, J., & Borin, N. (1998a). The Effect of Store Name, Brand Name and Price Discounts on Consumer's Evaluations and Purchase Intentions. Journal of Retailing, 74 (3), 331-352. Grewal, D., Krishnan, R., Levy, M., & Munger, J. (2010). Retail success and key drivers. In Retailing in the 21st century: Current and future trends (pp. 15-30). Heidelberg: Springer. Grewal, D., Monroe, K., & Krishnan, R. (1998b). The effects of price-comparison advertising on buyers' perceptions of acquisition value, transaction value, and behavioral intentions. The Journal of Marketing (April), 46-59.

67

Guandagni, P., & Little, J. (1983). A logit model of brand choice calibrated on scanner data. Marketing Science, 2 (3), 203-238. Gupta, S. (1988). Impact of Sales Promotion on When, What, and How Much to Buy. Journal of Marketing Research, 25, 342-355. Hartman, K., & Spiro, R. (2005). Recapturing store image in customer-based store equity: a construct conceptualization. Journal of Business Research, 58, 1112-1120. Heath, T., Chatterjee, S., & France, K. (1995). Heath, Timothy B., Subimal Chatterjee, and Karen Russo France. (1995). "Mental Accounting and Change in Price: The Frame Dependence of Preference Dependence," . Journal of Consumer Research, 22, 90-97. Heide, J. (1994). Managing the distribution system. In The AMA management handbook (pp. 3-26). N. Capon. Hermalin, B., & Katz, M. (2013). Product Differentiation through Exclusivity: Is there a One-Market-Power-Rent Theorem? Journal of Economics and Management Strategy, 22 (1), 1-27. Hildebrandt, L. (1988). Store image and the prediction of performance in retailing. Journal of Business Research, 17 (1), 91-100. Hoch, S., Bradlow, E., & Wansink, B. (1999). The Variety of an Assortment. Marketing Science, 18 (4), 527-546. Hoyer, W., & Brown, S. (1990). Effects of brand awareness on choice for a common, repeat-purchase product. Journal of Consumer Research, 17 (2), 141-148. Huff, D. (1964, February 28). Redifining and Estimating a Trading Area. Journal of Marketing, 34-38. Hul, M., Dube, L., & Chebat, J. (1997). The impact of music on consumers' reactions to waiting for services. Journal of Retailing, 71 (1), 87-104. Inman, J., McAlister, L., & Hoyer, W. (1990). Promotion Signal: Proxy for a Price Cut. Journal of Consumer Research, 17, 74-81. Iyengar, S., & Lepper, M. (2000). When choice is demotivating: Can one desire too much of a good thing? Journal of Personality and Social Psychology, 79, 995-1006. Jacobson, R., & Lane, V. (1995). Stock Market Reactions to Brand Extension Announcements: The Effects of Brand Attitude and Familiarity,”. Journal of Marketing, 59 (January), 63–77. Jacoby, J., & Mazursky, D. (1984). Linking brand and retailer images: do the potential risks outweigh the potential benefits? Journal of Retailing 60 (2), 105–122. Journal of Retailin , 60 (2), 105-122.

68

Jacoby, J., & Mazursky, D. (1985). The impact of linking brand and retailer images on perceptions of quality. In J. Jacoby, & C. Olson, Perceived Quality: How Consumers View New Stores and Merchandise (pp. 155-159). Lexington, MA: D. C. Heath and Company. Jacoby, J., Szybillo, J., & Busato-Schach, J. (1977). Information Acquisition Behavior in Brand Choice Situations. Journal of Consumer Research, 3 (4), 209-216. Joy, A., Wang, J., Chan, T., Sherry Jr., J., & Cui, G. (2014). M(Art)Worlds: Consumer Perceptions of How Luxury Brand Stores Become Art Institutions. Journal of Retailing, 90 (3), 347-364. Kahn, B. (1995). Consumer Variey-Seeking Among Goods and Services: An Integrative Review. Journal of Retailing and Consumer Services, 2 (3), 139-148. Kahn, B. (1999). Introduction to the special issue: assortment planning. Journal of Retailing , 75 (3), 289-293. Kahneman, D., & Tversky, A. (1979). Prospect Theory: An Analysis of Decision Under Risk. Econometrica, 47, 263-291. Kahn, B., & Lehmann, D. (1991). Modeling Choice Among Assortments. Journal of Retailing, 67 (3), 274-299. Kahn, B., Moore, W., & Glazer, R. (1987). Experiments in Constrained Choice. Journal of Consumer Research, 14 (June), 96-113. Keller, K. (1993). Conceptualizing, Measuring and Managing Customer-Based Brand Equity. Journal of Marketing, 57 (1), 1-22. Kent, R., & Allen, C. (1994). Competitive Interference Effects in Consumer Memory for Advertising: The Role of Brand Familiarity. 58, 97-105. Khan, U., Dhar, R., & Wertenbroch, K. (2005). A behavioral decision theory perspective on hedonic and utilitarian choice. In S. Ratneshwar, & D. G. Mick, Inside consumption: Frontier of research on consumer motives, goals and desires (pp. 144-165). Kirmani, A., Sood, S., & Bridges, S. (1999). The Ownership Effect in Consumer Responses to Brand Line Stretches. Journal of Marketing, 63, 88-101. Kotler, P. (1997). Marketing Management. Englewood Cliffs: Prentice-Hall. Kotler, P. & Keller, K. (2012), Marketing Management (Vol. 14th Edition, pp. 412-413). Edinburgh Gate, England: Pearson Education Limited. Krishna, A., Briesch, R., Lehmann, D., & Yuan, H. (2002). A meta-analysis of the impact of price presentation on perceived savings. Journal of Retailing, 78, 101-118. Kumar, V., & Leone, R. (1988). Measuring the effect of retail promotion son brand and store substitution. Journal of Marketing Research, 25 (2), 178-185.

69

Kunkel, J., & Berry, L. (1968). A Behavioral Conception of Retail Image. Journal of Marketing, 32 (4), 21-27. Lancaster, K. (1990). The economics of product variety: A survey. Marketing science, 9, 189-206. Laroche, M., & Brisoux, J. (1989, September). Incorporating Competition into Consumer Behavior Models: The Case of the Attittude-Intention Relationship. Journal of Economic Psychology , 343-362. Laroche, M., Kim, C., & Zhou, L. (1996). Brand familiarity and confidence as determinants of purchase intention: An empirical test in a multiple brand context. Journal of Business Research, 37 (2), 115-120. P. S. Leeflang, D. R. Wittink, M. Wedel, & P. A. Naert. (2000). Building Models for Marketing Decisions. Dordrecht, The Netherlands: Kluwer Acaemic Publishers. Levy, M., & Weitz, B. (1995). In Retailing management (p. 30). Chicago: Irwin. Lichtenstein, D., & Bearden, W. (1989). Contextual influences on perceptions of merchant-supplied reference prices. Journal of Consumer Research, 16, 55-66. Lichtenstein, D., & Burton, S. (1989). The Relationship Between Perceived and Objective Price–Quality. Journal of Marketing Research, 26, 429-443. Lichtenstein, D., Bloch, P., & Black, W. (1988). Correlates of Price Acceptability. Journal of Consumer Research, 15 (2), 243-252. Marshall, D. (1993, July). ppropriate Meal Occasions: Understanding Conventions and Exploring Situational Influences on Food Choice. The international Review of Retail, Distribution and Consumer Research, 261-301. Maruyama, M., & Wu, L. (2014). Multiple store patronage: The effects of store characteristics. Journal of Retailing and Consumer Services, 21 (4), 601-609. Marx, L., & Shaffer, G. (2007). Upfront payments and exclusion in downstream markets. Journal of Economics, 38 (3), 823-843. Mazursky, D., & Jacoby, J. (1985). Forming Impressions of Merhandise and Service Quality. In J. Jacoby, & J. Olson, Perceived Quality (pp. 139-154). Lexington: Lexington Books. McAlister, L., & Pessemier, E. (1982). Variety Seeking Behavior: An Interdisciplinary Review. Journal of Consumer Research, 9 (3), 311-322. Mitchell, A., & Olson, J. (1981). Are Product Beliefs the Only Mediator of Advertising Effect on Brand Attitude. Journal of Marketing Research, 18 (August), 318-332. Mittal, B. (1995). A comparative analysis of four scales of consumer involvement. Psychology & Marketing, 12 (7), 663-682.

70

Mittal, B. (1989). Measuring Purchase-decision involvement. Psychology & Marketing, 6 (2), 147-162. Monroe, K., & Krishnan, R. (1985). The Effect of Price on Subjective Product Expectations. In J. Jacoby, & J. Olson, Perceived Quality: How Consumers View Stores and Merchandise (pp. 209-239). Lexington: Lexington. Murthi, B., & Rao, R. (2012). Price Awareness and Consumers' Use of Deals in Brand Choice. Journal of Retailing, 1, 34-46. Murphy, K & Davidshofer, C. (1988) Psychological Testing: Principles and Applications. Englewood Cliffs, NJ: Prentice-Hall. Myer, J. (1967). Determinants of private brand attitude. Journal of Marketing Research , 4 (1), 73-81. Nair, S., Tikoo, S., & Liu, S. (2009). Valuing Exclusivity from Encroachment in Franchising. Journal of Retailing, 85 (2), 206-210. Narayana, C., & Markin, R. (1975). Consumer behavior and product performance: An alternative conceptualization. The Journal of Marketing, 1-6. Netemeyer, R., & Krishan, B. (2004). Developing and validating measures of facets of customer-based brand equity. Journal of Business Research 57, 209–224. Journal of Business Research, 57, 209-224. Nevin, J., & Houston, M. (1980). Images as a Component of Attractiveness to Intra-Urban Shopping Areas. Journal of Retailing , 56 (Spring), 77-93. O'Brien, D., & Shaffer, G. (1997). Nonlinear supply contracts, exclusive dealing and equilibrium market foreclosure. Journal of Economics and Management Strategy , 6 (4), 755-785. Okada, E. (2005). Justification effects on consumer choice of hedonic and utilitarian goods. Journal of marketing research, 42 (1), 42-53. Olshavsky, R. (1985). Perceived quality in consumer decision making: an integrated theoretical perspective. In J. Jacoby, & C. Olson, Perceived Quality: How Consumers View Stores and Merchandise (pp. 3-29). Lexington, MA: D. C. Heath and Company. Osborne, J., & Waters, E. (2002). Four assumptions of multiple regression that researchers should always test. Practical assessment, research & evaluation, 8 (2), 1-9. Pan, Y., & Zinkhan, G. (2006). Determinants of retail patronage: A meta-analytical perspective. Journal of Retailing, 82 (3), 229-243. Pauwels, K., Hanssens, D., & Siddarth, S. (2002). The Long-Term Effects of Price Promotions on Category Incidence, Brand Choice, and Purchase Quantity. Journal of Marketing Research, 39 (4), 421-439.

71

Perdue, B., & Summers, J. (1986). Checking the success of manipulations in marketing experiments. Journal of Marketing Research, 317-326. Peres, R., & Van den Bulte, C. (2014). When to take or forgo new product exclusivity: Balancing protection from competition against word-of-mouth spillover. Journal of Marketing, 78, 83-100. Porter, M. (1976). Intrabrand Choice, Media Mix, and Market Performance. American Economic Review, 66, 398-406. Putsis, W., & Dhar, R. (1999). Category Expenditure, Promotion, and Competitive Market Interactions: Can Promotions Really Expand the Pie? Working paper. Rao, A., & Monroe, K. (1988). The Moderating Effect of Prior Knowledge on Cue Utilization in Product Evaluations. Journal of Consumer Research, 15, 253-264. Richardson, P., Jain, A., & Dick, A. (1996). Household Store Brand Proneness: A Framework. Journal of Retailing, 72 (2), 159-185. Rao, A., & Monroe, K. (1988). The Moderating Effect of Prior Knowledge on Cue Utilization in Product Evaluations. Journal of Consumer Research, 15, 253-264. Rockney, W., & Rinne, H. (1986). n empirical investigation into the impact of price promotions on retail store performance. Journal of Retailing, 62 (Fall), 237-266. Rubenstein, C., & Shaver, P. (1980). Loneliness in two northeastern cities. In The Anatomy of Loneliness. New York: International Universities Press. Russell, W. (1986). A Reference Price Model of Brand Choice for Frequently Purchased Products. Journal of Consumer Research, 13, 250-256. Schiffman, L., Dash, J., & Dillon, W. (1977). The contribution of store image characteristics to store-type choice. Journal of Retailing, 53. Shukla, P., & Babin, B. (2013). Effects of consumer psychographics and store characteristics in influencing shopping value and store swithcing. Journal of Consumer Behavior, 12, 194-203. Sinha, P., & Banerjee, A. (2004). Store choice behavior in an evolving market. International Journal of Retail & Distribution Management, 32 (10), 482-494. Sirohi, N., McLaughlin, E., & Wittink, D. (1998). A model of consumer perceptions and store loyalty intentions for a supermarket retailer. Journal of Retailing, 74 (2), 223-245. Sloot, L., & Verhoef, P. (2008). The impact of Brand Delisting on Store Switching and Brand Switching Intentions. Journal of Retailing, 84 (3), 281-296. Sloot, L., Verhoef, P., & Franses, P. (2005). The impact of brand equity and the hedonic level of products on consumer stock-out reactions. Journal of Retailing, 81 (1), 2005.

72

Sorescu, A., Frambach, R., Singh, J., Rangaswamy, A., & Bridges, C. (2011). Innovations in Retail Business Models. Journal of Retailing, 87S (1), S3-S16. Steenkamp, J., & Wedel, M. (1991). Segmenting Retail Markets on Store Image Using a Consumer-Based Mathodology. Journal of Retailing, 67 (3), 300-320. Stevens, R. (1996). Applied multivariate statistics for the social sciences. Mahwah, NJ: Erlbaum. Spears, N., & Singh, S. (2004). Measuring attitude toward the brand and purchase intentions. Journal of Current Issues & Research in Advertising, 26 (2), 53-66. Stassen, R., Mittelstaedt, J., & Mittelstaedt, R. (1999). Assortment Overlap: Its Effect on Shopping Patterns in a Retail Market When the Distributions of Prices and Goods are Known. Journal of Retailing, 75 (3), 371-386. Stewart, D. (1992). Speculations on the Future of Advertising Research. Journal of Advertising, 21, 1-17. Sudhir, K., & Rao, V. (2006). Do Slotting Allowances Enhance Efficiency or Hinder Competition? Journal of Marketing Research, 43 (2), 153. Swait, J., & Erdem, T. (2007). Brand Effects On Choice And Choice Set Formation Under Uncertainty. Marketing Science, 26 (5), 679-697. Tauber, E. (1972). Why do people shop? Journal of Marketing, 36, 46-49. Teas, R., & Agarwal, S. (2000). The Effects of Extrinsic Product Cues on Consumer's Perceptions of Quality, Sacrifice, and Value. Journal of the Academy of Marketing Science, 28 (2), 278-290. Tellis, G., & Gaeth, G. (1990). Best value, price-seeking, and price aversion: the impact of information and learning on consumer choices. Journal of Marketing, 54 (2), 34-45. Van Everdingen, Y. & Ten Berge, S. (2016). Effects of Exclusive New Products on Consumers’ Assortment Perceptions, EMAC paper. Thelen, E., & Woodside, A. (1997). What evokes the brand or store? Consumer research on accessibility theory applied to modeling primary choice. International Journal of Research in Marketing, 14, 125-145. Thode, H. C. (2002), Testing for normality. CRC press. Tigert, D. (1983). Pushing the hot buttons for a successful retailing strategy. In R. Darden, & F. Lusch, Patronage Behavior and Retail Management. New York: North-Holland. Tversky, A., & Shafir, E. (1992). Choice under conflict: The dynamics of deferred decision. Psychological Science, 6, 358-361.

73

Van Everdingen, Y., & Sloot, L. (2015, April 10). Retail and Consumer Adoption of New Products. 30. Rotterdam: Rotterdam School of Management. Van Everdingen, Y., Sloot, L., Van Nierop, E., & Verhoef, P. (2011). Towards a Further Understanding of the Antecedents of Retailer New Product Adoption . Journal of Retailing , 87 (4), 579-597. Van Kenhove, P., & Desumaux, P. (1997, October). The relationship between emotional states and approach or avoidance responses in a retail environment. The International Review of Retail Distribution and Consumer Research , 351-368. Van Kenhove, P., De Wulf, K., & Van Waterschoot, W. (1999). The Impact of Task Definition on Store-Attribute Saliences and Store Choice. Journal of Retailing, 75 (1), 125-137. van Oest, R. (2013). Why are Consumers Less Loss Averse in Internal than External Reference Prices? Journal of Retailing, 89 (1), 62-71. Van Trijp, H., Hoyer, W., & Inman, J. (1996). Why switch? Product category-level explanations for true variety-seeking behavior. Journal of Marketing Research, 33 (3), 281-292. Verhagen, T., & Boter, J. (2005). The importance of website content in online purchasing across different types of products. Working Paper Research Memorandum, 10. Vickers, J., & Waterson, M. (1991). Vertical Relationships: An Introduction. The Journal of Industrial Economics , 39 (5), 445-450. Voss, K. E., Spangenberg, E. R., & Grohmann, B. (2003). Measuring the hedonic and utilitarian dimensions of consumer attitude. Journal of Marketing Research, 40 (3), 310-320. Walters, R., & Rinne, H. (1986). An empirical investigation into the impact of price promotions on retail store performance. Journal of Retailing (Fall), 237-266. Wagner, T., & Rudolph, T. (2010). Towards a hierarchical theory of shopping motivation. Journal of Retailing and Consumer Services, 17 (5), 415-429. Wetzel, C. (1977). Manipulation Checks: A Reply to Kidd. Representative Research in Social Psychology, 8 (2), 88-93. Wright, M., & Riebe, E. (2010). Double jeopardy in brand defection. European Journal of Marketing, 44 (6), 860–873. Yi, Y., & Jeon, H. (2003). Effects of Loyalty Programs on Value Perception, Program Loyalty, and Brand Loyalty. Journal of the Academy of Marketing Science, 31 (3), 229-240. Yu, F., Chavan, K., & Dowden, C. (2008). Selling the sizzle – The need for customer experience management and control in consumer electronics. Deloitte Consulting Newsletter.

74

Zajonc, R., & Markus, H. (1982, September). Affective and Cognitive Factors in Preferences. Journal of Consumer Research, 123-131. Zeihaml, V., Berry, L., & Parasuraman, A. (1996). The behavioral consequences of service quality. The Journal of Marketing, 31-46. Zeithaml, V. (1988). Consumer Perceptions of Price, Quality, and Value: A Means-End Model and Synthesis of Evidence. Journal of Marketing, 52 (3), 2-22. Zimmer, M., & Golden, L. (1988). Impressions of Retail Stores: A Content Analysis of Consumer Images. Journal of Retailing, 64 (Fall), 265-293. Internet Resources Online picture (front page), downloaded from http://nl.freepick.com/iconen-gratis/winkelmandje-cirkel_703346.htm on June 20th, 2016. Foodpersonality. (2016, January 25). Iri & Nielsen: de marktaandelen. Retrieved February 3, 2016, from www.foodpersonality.nl: http://www.foodpersonality.nl/categorieen/nieuws/11680-iri-nielsen-de-marktaandelen.html References of Exclusive Products Cases Aarnoudse, L. (2014, June 6). Hoe verover je een plekje in de shappen van Albert Heijn of Jumbo? Retrieved January 25, 2016, from NRCQ: http://www.nrcq.nl/2014/06/06/hoe-verover-je-een-plekje-in-de-schappen-van-albert-heijn-of-jumbo Beanies. (2013, August 7). The Grinder. Retrieved January 25, 2016, from Beanies the Flavour Company: http://www.beaniesflavourco.co.uk/blog/beanies-at-sainsburys/ Biernet.nl. (2015, September 22). Grolsch Gerijpte Herfstbok exclusief bij Jumbo. Retrieved February 3, 2016, from www.biernet.nl: http://www.biernet.nl/nieuws/grolsch-gerijpte-herfstbok-exclusief-bij-jumbo Coca-Cola Enterprises Nederland B.V. (2015, December). Drankportfolio - Merken - Caprisun. Retrieved February 19, 2016, from Coca Cola Nederland: http://www.cocacolanederland.nl/default.aspx?nodeID=330 Distrifood. (2009, September 10). AH en Unilever lanceren Becelbrood. Retrieved January 25, 2016, from Distrifood.nl: http://www.distrifood.nl/formules/nieuws/2009/9/ah-en-unilever-lanceren-becel-brood-10129762 Gijs. (2009). Gijs - Waar te koop. Retrieved February 19, 2016, from van-gijs.nl: http://www.van-gijs.nl/waar-te-koop/ Heineken. (2012, Juni). Notulen, Jaarverslagen en Personeelsbladen Heineken. Heineken NL Magazine. The Netherlands. Retrieved from Heineken.memorix.nl.

75

Innofood. (2012, October 23). Industrie zoekt heil in exclusieve producten voor specifieke supermarktketens. Retrieved January 25, 2016, from Innofood.org: http://www.innofood.org/nl/nieuws/9591/merkartikel-als-het-nieuwe-huismerk.html John West. (2012, June 14). John West in de media. Retrieved January 16, 2016, from John-west.nl: http://www.john-west.nl/john-west-in-de-media/artikel/john-west-introduceert-gestoomde-makreelfilet Levensmiddelenkrant. (2012, March 19). Grolsch introduceert De Klok exclusief bij C1000. Retrieved January 2016, 2016, from Levensmiddelenkrant.nl: http://www.levensmiddelenkrant.nl/nieuws/assortiment/grolsch-introduceert-de-klok-exclusief-bij-c1000 Marketing Tribune. (2013, October 7). Mondelez lanceert koffiemerk Velours Noir. Retrieved January 25, 2016, from http://www.marketingtribune.nl/food-en-retail/nieuws/2013/10/mondelez-international-lanceert-koffiemerk-velours-noir35_0/index.xml Rensen, E. (2015, November 19). Hertog Jan Ongekend exclusief bij Jan Linders. Retrieved February 3, 2016, from www.distrifood.nl: http://www.distrifood.nl/assortiment/nieuws/2015/11/hertog-jan-ongekend-exclusief-bij-jan-linders-10193650 Smit, P. (2013, December 11). Heet Nootje Exclusief bij Albert Heijn. Retrieved November 22, 2015, from Levensmiddelenkrant.nl: http://www.levensmiddelenkrant.nl/nieuws/handel/formules/heet-nootje-exclusief-bij-albert-heijn Unilever. (2015). Blue Band. Retrieved November 23, 2015, from http://www.unilever.nl/merken-in-actie/detail/Blue-Band/311775/ van Beek, M. (2014, September 4). Nieuw: Argentijnse Callia Alta wijnen exclusief bij PLUS supermarkten. Retrieved February 3, 2016, from Wijn.blog.nl: http://wijn.blog.nl/wijnstreken/2014/09/04/nieuw-argentijnse-callia-alta-wijnen-exclusief-bij-plus-supermarkten The Wall Street Journal. (2015, March 15). Wal-Mart to Sell Premium-Priced German Laundry Soap Persil. Retrieved April 25, 2016, from http://www.wsj.com/articles/wal-mart-to-sell-premium-priced-german-laundry-soap-persil-1426467859

76

Appendix A Definition and Overview of Variables

Variable Definition Measurement Instrument

Independent variable

Exclusive new product

A new product (e.g. new flavor or package) from a key national brand that is exclusively distributed to consumers through a single sales channel due to the retailer’s exclusive right for distribution (e.g. Gielens et al., 2014)

Dummy variable equal to 1 if the brand is presented as if it is exclusively distributed by the retailer

Dependent variables

Store choice (SC)

A consumer’s likelihood to patronize a retailer (Pan & Zinkhan, 2006)

Degree of store choice intention (SCI) on a three-statements scale consisting likelihood to visit, willingness to buy merchandise and willingness to recommend the store [strongly disagree (1) – (7) strongly agree]

Brand choice (BC)

The likelihood that a consumer buys products from a certain brand (based on Dodds et al., 1991)

Degree of brand choice intention (BCI) consisting of three statements measuring willingness to buy from brand [strongly disagree (1) – (7) strongly agree]

Moderating variables

Price premium Brand in folder was presented with a 10% markup on the normal price

Dummy variable equal to 1 if the brand was sold with a price premium

Hedonic level The degree to which a product provides hedonic benefits to the consumer i.e. experiential consumption, fun, excitement (Voss et al., 2003)

Dummy variable equal to 1 in case of a hedonic product (vs. utilitarian product). Soft drinks were identified as hedonic; detergents were identified as utilitarian in a pretest on a 7-point semantic differential scale consisting of not fun/fun, unexciting/exciting, not delightful/ delightful, not thrilling/thrilling, unenjoyable/enjoyable.

Control variablesa

Brand equity (BE)

The level at which consumers react more favorably to a product when the brand is identified than when it is not (Keller, 1993)

Dummy variable equal to 1 in case of a high brand equity (mean average score >4) on a four-item 7-point Likert scale [strongly disagree (1) – (7) strongly agree]

Store equity (SE)

The differential effect of store knowledge on customer response to the marketing activities of the store (Hartman & Spiro, 2005)

Dummy variable equal to 1 in case of a high store equity (mean average score >4) on a five-item 7-point Likert scale [strongly disagree (1) – (7) strongly agree]

Product Category Involvement

Degree to which a consumer cares about buying products from the category (Mittal, 1989)

Degree of Product Category Involvement on a three-item 7-point Likert scale [strongly disagree (1) – (7) strongly agree]. Items can be found in Table 6.

a measured during the experiment

77

Appendix B Pretest Findings New Product Selection Soft Drinks

Brand Average Familiarity Brand Average

Experience Brand Average

likelihood of triala

Coca Cola 6.60 Coca Cola 5.75 Coca Cola 3.80 Fanta 6.45 Fanta 5.40 Fanta 4.45 Pepsi 6.10 Pepsi 4.45 Pepsi 3.15 Seven Up 6.10 Seven Up 4.90 Seven Up 3.70 Lipton Ice Tea 5.85 Lipton Ice Tea 4.60 Lipton Ice Tea 3.90 Spa Fruit 5.85 Spa Fruit 5.05 Spa Fruit 4.70 Sisi 5.50 Sisi 4.10 Sisi 2.65 Sprite 5.40 Sprite 4.35 Sprite 3.10 Crystal Clear 5.30 Crystal Clear 4.00 Crystal Clear 3.20 Royal Club 5.00 Royal Club 3.65 Royal Club 2.75 Orangina 4.90 Orangina 3.35 Orangina 2.40 Rivella 4.65 Rivella 3.25 Rivella 2.10 Dr. Pepper 4.40 Dr. Pepper 2.90 Dr. Pepper 2.05

Holy soda 3.50 Holy soda 1.75 Holy soda 1.80

Average (n=20) 5.40 Average (n=20) 4.11 Average (n=20) 3.13 Scores were measured on a 7-point Likert scale [1=not at all familiar/experienced/likely, 7=very familiar/experienced/likely] a The question asked for likelihood of trial was: Imagine that the following soft drink brands will introduce a new flavor. How likely is it that you will buy a bottle from each of the brands below, assuming that the new flavor tastes good?

Flavor Average

Score (n=20)

Female Average (n=10)

Male Average (n=10)

Red fruit 5.50 6.00 5.00 Wild berry 5.15 5.90 4.00 Passion Fruit 5.10 5.60 4.60 Raspberry 5.00 5.30 4.70 Blackberry 4.85 5.10 4.60 Kiwi 4.00 3.90 4.10 Pomegranate 3.95 4.70 3.20 Pineapple 3.80 4.50 3.10 Banana 2.65 2.70 2.60 Average 4.44 4.86 3.99 Average Fanta Wild Berrya 5.13 5.58 4.58 Scales were measured on a 7-point Likert scale [1=don’t like at all, 7=like very much] a Red fruit, Wild Berry, Raspberry and Blackberry

78

Appendix C Pretest Findings New Product Selection Detergents

Brand Average Familiarity Brand Average

Experience Brand Average

likelihood of trial

Robijn 6.17 Robijn 5.03 Robijn 5.10 Ariel 5.93 Ariel 4.33 Ariel 4.33 Persil 5.87 Persil 3.43 Persil 4.03 Omo 5.60 Omo 2.97 Omo 3.30 Dreft 5.37 Dreft 4.80 Dreft 4.27 Witte Reus 5.20 Witte Reus 3.40 Witte Reus 3.63 Dash 4.37 Dash 2.43 Dash 2.77 Fleuril 3.73 Fleuril 2.50 Fleuril 2.80 Dixan 2.07 Dixan 1.67 Dixan 1.97 Le Chat 1.50 Le Chat 1.47 Le Chat 1.97

Average (n=30) 3.42 Average (n=30) 3.20 Average (n=30) 4.58 Scores were measured on a 7-point Likert scale where 1=not at all familiar/experienced/likely, 7=very familiar/experienced/likely a The question asked for likelihood of trial was: Imagine that the following detergent brands will introduce a new scent. How likely is it that you will buy a package from each of the brands below, assuming that the new flavor smells good?

Scent Average

Score (n=30)

Female Average (n=15)

Male Average (n=15)

Soft Linen 5.10 5.53 4.67 White Lilies 4.93 5.20 4.67 Sweet Jasmine 4.90 5.00 4.80 Satin & Rose 4.47 4.87 4.07 Red Roses 4.40 4.60 4.20 Lychee & Pink Tulips 4.00 4.87 3.13 Sparkling Orange 3.87 3.47 4.27 Aloe Vera 3.83 4.00 3.67 Average 3.55 3.75 3.35 Scores were measured on a 7-point Likert scale where 1=don’t like at all, 7=like very much

79

Appendix D Folder Design for Eight Conditions GROUP 1 – Exclusive new product; Price: €4.99; Brand: Robijn

GROUP 2 – Exclusive new product; Price: €5.49; Brand: Robijn

80

GROUP 3 – New product; Price: €4.99; Brand: Robijn

GROUP 4 – New product; Price: €5.49; Brand: Robijn

81

GROUP 5 – Exclusive new product; Price: €1.80; Brand: Fanta

GROUP 6 – Exclusive new product; Price: €1.98; Brand: Fanta

82

GROUP 7 – New product; Price: €1.80; Brand: Fanta

GROUP 8 – New product; Price: €1.98; Brand: Fanta

83

Appendix E Final Sample Demographics Test cells

(“New – exclusively available at Dirk”)

Control Cells (“New”)

Normal Robijn

Normal Fanta

Premium Robijn

Premium Fanta

Normal Robijn

Normal Fanta

Premium Robijn

Premium Fanta

Total

Surveys started 39 36 36 42 37 37 37 35 299

Surveys completeda

36 34 34 39 36 35 36 33 283

Correct main manipulationb

30 30 30 29 29 30 30 31 239

Product category manipulation cutc

27 30 25 29 27 30 28 30 226

Price manipulation cutd

21 23 24 28 22 23 28 30 169

a Partials were removed and non-buyers were screened out b Subjects that failed the main manipulation were removed c Subjects scoring >6 (utilitarian) and <2 (hedonic) on the hedonic level manipulation check were removed in this overview d Subjects scoring < 6 (normal) and <3 (premium) on the price manipulation were removed in this overview

Appendix F Age and Gender Division Across Conditions

a Mean age in the sample was 42.11 years.

Test cells (“New – exclusively available at Dirk”)

Control Cells (“New”)

Normal Robijna

Normal Fanta

Premium Robijn

Premium Fanta

Normal Robijn

Normal Fanta

Premium Robijn

Premium Fanta

Total

Age: 18-24 4 6 4 4 4 4 4 5 35 Age: 25-34 6 5 6 5 5 6 7 6 46 Age: 35-44 6 5 6 6 6 6 5 6 46 Age: 45-54 7 7 6 6 7 7 7 7 54 Age: 55-65a 7 7 8 8 7 7 7 7 58 Male 9 9 9 8 8 9 9 10 71 Female 21 21 21 21 21 21 21 21 168

Total 30 30 30 29 29 30 30 31 239

84

Appendix G Sample Background Variables Across Conditions Background Variables

Construct Item Based on Cronbach’s alpha

Purchase Intention

How likely is it that you would buy the new Robijn Sweet Jasmine/Fanta Wild Berry? [not likely at all (1) – (7) extremely likely]

Grewal et al (1998a)

n/a

Normal store At which stores do you normally do groceries?a

[Options: Albert Heijn, Jumbo, Dirk, Lidl, Aldi, Plus, Boni, Coop, Hoogvliet, Spar, Poiez, Local supermarket, Other].

n/a

Normal soft drinks

Which soft drink brands do you normally buy?a

[Options: Coca Cola, Fanta, Pepsi, Seven Up, Lipton Ice Tea, Spa Fruit, Sisi, Sprite, Crystal Clear, Royal Club, Orangina, Rivella, Dr. Pepper, Holy Soda, Private label brand, Other].

n/a

Normal detergents brands

Which detergents brands do you normally buy?a

[Options: Ariel, Persil, Robijn, Omo, Dreft, Witte Reus, Dash, Fleuril, Dixan, Le Chat, Private Lbael brand, Other].

n/a

Need for uniqueness

Please indicate how important the following store characteristics are for you when you engage in grocery shopping. 1. Luxurious, unique products 2. Exclusive products 3. Local products [not important at all (1) – (7) extremely important]

Wagner & Rudolph (2010) α=.96, Van Everdingen & Ten Berge (2016) α=.88

n/a

a Multiple answers possible

85

Background Statistics

Test cells (“New – exclusively available at

Dirk”)

Control Cells (“New”)

Normal Robijn

Normal Fanta

Premium Robijn

Premium Fanta

Normal Robijn

Normal Fanta

Premium Robijn

Premium Fanta

% that normally does (i.a.) groceries at Dirk within the condition

56.67% 46.67% 60.00% 41.38% 37.93% 40.00% 55.17% 38.71%

% that normally buys (i.a) Fanta within the condition

- 63.33% - 27.59% - 50.00% - 64.52%

% that normally buys (i.a) Robijn within the condition

30.65% - 25.81% - 24.19% - 19.35% -

Importance exclusive products when doing groceries

3.77 (1.65)

4.03 (1.25)

3.87 (1.63)

3.55 (1.55)

4.24 (1.24)

4.37 (1.47)

3.80 (1.42)

4.48 (1.43)

Importance luxorious/unique products when doing groceries

3.38 (1.58)

4.23 (1.36)

4.00 (1.36)

3.79 (1.47)

4.10 (1.23)

4.27 (1.34)

3.43 (1.17)

4.45 (1.61)

Importance local products when doing groceries

4.03 (1.43)

4.53 (1.31)

4.33 (1.27)

4.03 (1.43)

4.07 (1.36)

4.60 (1.13)

4.03 (1.35)

4.26 (1.59)

Standard deviations are given in parentheses

86

Figure 6 MC Question: “At which supermarkets do you normally do groceries?” 11

Figure 7 MC Question: “Which detergents brands do you normally buy?”

Figure 8 MC Question: “Which soft drink brands do you normally buy?”

11 Multiple Choice question.

0%

20%

40%

60%

80%

Albert Heijn Lidl Dirk Jumbo Aldi Hoogvliet Plus Coop Other

0%

10%

20%

30%

Robijn Ariel Home brand

Witte Reus

Fleuril Dash Persil Omo Dreft Other

0%

10%

20%

30%

40%

Coca Cola

Fanta Lipton Ice Tea

Home brand

Spa Fruit Pepsi Sprite Seven Up

Sisi Royal Club

Crystal Clear