Master Thesis
Exclusivity of Assortment
A quantitative investigation of the motivation of manufacturers to engage in exclusive assortment agreements with retailers.
University of Groningen
MANAGEMENT SUMMARY
Three important trends are shaping the retail landscape at the moment. First, consolidation led to a power shift from manufacturers to increasingly large retailers. Second, new (online) players are making the market more transparent which increased competition through price convergence and led to margin erosion. Third, there is a higher need for differentiation from retailers. Assortment is the most important factor for consumer store choice behavior, which makes it attractive for retailers to differentiate with assortment. However, since the share of private label products can only be raised up to a certain threshold until potential consumer revenue is lost, exclusive assortment from national brand manufacturers can become very attractive for retailers as an alternative to differentiate themselves. The motivation for retailers to request exclusive assortment is clear since it supports their differentiation efforts and it is generally profitable. However, the motivation for manufacturers to provide this exclusive assortment demanded further investigation. This research analyzed quantitative data gathered by Sloot & van Everdingen (2015) in order to extract generalizable findings about the influencing factors for the manufacturers’ willingness to provide exclusive assortment to a retailer. One third of manufacturers indicated to be (very) likely to offer exclusive products. Statistically significant findings were that the manufacturers’ willingness to engage in exclusive assortment agreements is positively influenced by (1) the market share of the retailer, (2) the fair-share performance of the retailer, (3) the uniqueness of the retailer image, (4) the relationship to the retailer (positivity and importance of the retailer with regard to sales), (5) and the private label share in the product group. However, manufacturers were negatively influenced in their decision to provide exclusive assortment by (1) the positivity (strength) of the own brand image and by (2) the anticipated negative reactions of retailers that were excluded from such an exclusive agreement.
TABLE OF CONTENT
INTRODUCTION ... 4
PROBLEM STATEMENT ... 6
LITERATURE REVIEW ... 7
Relationship between Manufacturers and Retailers ... 7
Dutch Food Retail ... 9
Dutch Food Manufacturing Industry ... 11
The Role & Importance of Assortment ... 12
Private Label Brands ... 13
Exclusivity Agreements from Different Perspectives ... 15
Exclusivity from the Retailers’ Perspective ... 16
Exclusivity from the Manufacturers’ Perspective ... 17
METHODOLOGY ... 19
Research Method ... 19
Data Collection & Participants ... 19
Variables used in the Survey ... 21
Hypotheses ... 24
Conceptual Model ... 31
Plan of Analysis ... 31
DATA ANALYSIS ... 32
Coding of Variables ... 32
Data Screening and Cleaning ... 32
Descriptive Statistics ... 33
Factor Analysis of Dependent Variable ... 35
Factor Analyses of Independent Variables ... 35
Multiple Regression: Analyzing the Effects of the Independent Variables ... 36
Hypotheses Results ... 40
DISCUSSION ... 45
MANAGERIAL IMPLICATIONS ... 48
DIRECTIONS FOR FUTURE RESEARCH ... 51
LIMITATIONS ... 53
REFERENCES ... 54
APPENDICES ... 61
Appendix 1 - Coding of the Variables ... 61
Appendix 2 - Extensive Factor Analysis of Dependent Variable ... 63
Appendix 3 - Extensive Factor Analyses of Independent Variables ... 65
INTRODUCTION
Three important trends are shaping food retailing at the moment. First, consolidation led to a power shift from manufacturers to retailers. Second, the market got more transparent which increased competition and led to margin erosion. Third, there is a higher need for differentiation. Creating a unique assortment might be one way for retailers to cope with those trends. There is an emerging phenomenon in retailing driven by those trends: retailers demanding exclusive assortment from manufacturers.
Concentration in the retail market increases already since years (Messinger & Narasimhan, 1995). The majority of national food retail markets in (northern) Europe is highly concentrated (five-firm concentration ratio between 60-75%; Wrigley, 2002) and there is an overall trend towards more concentration on the national level for food retailers in the EU (European Commission, 2014). This is mainly motivated by the retailers wanting to utilize synergies and to improve their negotiation power (Wrigley, 2002). Generally, this has led to a power shift from suppliers to retailers (PwC, 2007). This is linked to retailers being able to extract a higher proportion of the channels profits due to their higher bargaining power (Iyer & Villas-Boas, 2003). Also, it was found that those powerful retailers approach manufacturers and force them to e.g. renegotiate terms and conditions (Sloot & Verhoef, 2008).
New (online) players in food retail will increase competition even further and squeeze prices and margins more than ever before (KPMG, 2016). This can be explained by the fact that the Internet empowers consumers by improving their market knowledge, which in turn contributes towards price convergence (Pires, Stanton & Rita, 2006). Increasing market and price transparency means that retailers have to find others ways to attract customers. Offering products that are attracting consumers and that are not available at competitors is one of the main challenges for food retailers (Simonson,
1999).
To differentiate, retailers can utilize different factors (Pan & Zinkhan, 2006) like
influential factor for retail store choice (Pan & Zinkhan, 2006) and is more important than retail prices (Briesch, Chintagunta & Fox, 2009). Also, it was found that, in global comparison, assortment is relatively more important for consumers in Europe (Nielsen, 2015). Since the need for differentiation intensifies and assortment is the most important factor for retailers, they might demand more exclusive manufacturer products (PwC, 2007). In a UK survey, more than one third of brand manufacturers reported that they have been approached by retailers to provide exclusive assortment (Gielens, Gijsbrechts & Dekimpe, 2014).
Exclusive assortment could include:
- Brand exclusivity (a brand for only one retailer) - Size exclusivity (a special size of a certain product)
- Exclusive editions/variants (exclusive variants of one brand) - Time-bound exclusivity (e.g. six months after launch)
The leading German food retailer EDEKA utilizes exclusivity agreements to potentially achieve competitive advantages (EDEKA, 2013). Examples for exclusives at EDEKA are (1) KNORR “Weltgerichte (world-dishes)” by Unilever, (2) specially developed muesli “Mühlenklassiker” by Kölln and (3) “Henderson London Dry Gin” by Henderson.
manufacturers the outcome was less positive. Only if margins are renegotiated, the overall effect for both can be positive.
Furthermore, it should be taken into account that every new product or line extension increases complexity for any organization and that the additional results of adding SKU should contribute positively to the manufacturers success (Gottfredson & Aspinall, 2005). Additionally, suppliers are becoming more eager to simplify trade terms (Strategy&, 2015) and manufacturers like Unilever and Nestlé are constantly trying to reduce complexity by cutting down on the number of SKU (Financial Times, 2013). Also Procter & Gamble rationalized product lines in order to cut down on cost (Gourville & Soman, 2005). Exclusive agreements might be contrary to these attempts to reduce complexity.
PROBLEM STATEMENT
There is little research in the field of assortment exclusivity and specifically there is not much insight yet into the motivation of manufacturers to provide exclusive assortment. Existing research focused on exploring the profitability of such agreements (Gielens, Gijsbrechts & Dekimpe, 2014) or qualitatively exploring possible motivations for manufacturers and retailers (Valkenburg, 2014). There has not been an approach yet to quantify insights about the motivations of manufacturers to provide exclusive assortment. The following research aims at closing this gap by researching the underlying motivations of manufacturers on the willingness to engage in exclusive assortment agreements with retailers. Thus, the research question is:
“What are the influencing factors for the willingness of manufacturers to engage in exclusive assortment agreements with retailers?”
Sub-Questions being explored in this research cover the following topics:
LITERATURE REVIEW
Relationship between Manufacturers and Retailers
The relationship between food manufacturers and retailers has dramatically changed in the past. Traditionally, manufacturers were dictating the rules of engagement in this relation. However, channel power in this relationship has shifted from the manufacturers to the retailers (Geylani, Dukes, Srinivasan, 2007; Chambolle & Villas-Boas, 2015). There are two main drivers of this development: (1) a consolidation of retailers and (2) an increasing focus of retailers on private labels. The first main factor driving this development is an ongoing process of consolidation between large retailers over the last two decades (Messinger & Narasimhan, 1995; Grewal, Krishnan, Levy & Munger, 2006), mainly caused by mergers and acquisitions (Goodman & Dion, 2001). This led to large retail groups (Chambolle & Villas-Boas, 2015), especially in western Europe. Food retailers were trying to utilize synergies and obtain advantages regarding their buyer power (Wrigley, 2002). European buying organizations are taking this even a step further by combining the buying power of several large international retailers (Gruijters, 2016). Those large retailers utilize their channel power in the relationship with manufacturers to purchase products at lower prices and pressurize the manufacturers’ margins (Geylani, Dukes & Srinivasan, 2007; Chambolle & Villas-Boas, 2015). They are able to do this since they can achieve more attractive trade agreements and conditions with the manufacturers (Inderst & Shaffner, 2007) based on their size in terms of revenue and goods sold. This is linked to retailers being able to extract a higher proportion of the channels profits due to their improved bargaining power (Iyer & Villas-Boas, 2003). The more powerful a retailer and the more customers are being served, the higher are the chances that threatening behavior from the focal retailer towards a manufacturer will be observed (Fein & Anderson, 1997).
margins on those private label products (Meza & Sudhir, 2010). Higher margins are typically realized by saved costs on R&D, marketing, and overhead. The increase in buying power, however, seems to be stronger against niche brands compared to national mass market brands (Meza & Sudhir, 2010). This would imply that larger manufacturers enjoy a better position (i.e. are less threatened) against large retailers than relatively smaller manufacturers. When having adequate private label brands, food retailers are also more likely to delist national brand products to motivate manufacturers to compete harder for their patronage (Inderst & Shaffner, 2007). Also with regard to (the threat) delisting of products, large manufacturers that are selling important brands for the retailer are in an advantageous position (Bundeskartellamt, 2014). The amount of national brand manufacturers with a size big enough to resist those threats can be considered as rather small (6% in Germany).
Dutch Food Retail
The majority of national food retail markets in north-western Europe is highly concentrated (five-firm concentration ratio between 60-75%; Wrigley, 2002) and there is an overall trend towards more concentration on the national level for food retailers in the EU (European Commission, 2014). This also holds for the Dutch food retail market where the five largest retailers have a combined market share of 80% (see figure 1). After a difficult period following the last financial crisis, the Dutch food retail market seems to have recovered. The first four months of 2016 showed a revenue growth for Dutch supermarkets of 2.6% (to 11.21 bn€) compared to a revenue of 10.93 bn€ in Q1 2015 (GfK, 2016a).
Figure 1: Overview of Dutch food retail market (distrifood.nl, 2016)
The Dutch food retail industry is increasingly focusing on satisfying the consumers’ needs with regard to food instead of focusing on traditional concepts. New (online) competitors like Picnic and HelloFresh are entering the market and offer a new dimension of convenience to the customers. The borders between retail channels are increasingly blurring (Rabobank, 2015): Picnic offers free of charge delivery of groceries with products at the same price level as normal supermarkets. Traditional food retailers are also starting to react to those new consumer needs by differentiating their formats. Albert Heijn has introduced “AH to go” in order to attract customers that demand a high level of convenience at central locations like train stations or city centers where traditional retail formats were not preset yet. Jumbo has introduced its Foodmarkt: combining a traditional supermarket concept with new customer experience elements (Rabobank, 2015). Online shopping is a large opportunity as well as a huge threat at the same time for traditional food retailers in the Netherlands. 20% of Dutch people said that they ordered groceries online in the last 6 months compared to 10% in 2014 (GfK, 2016b). However, putting those recent developments aside, there are more fundamental developments that are driving the behavior of Dutch food retailers.
Buying groups have increased the power of food retailers over the manufacturers. This can be explained by increased productivity and decreased costs (Geyskens, Gielens & Wuyts, 2015) which are linked to economies of scale when greater amounts of goods are ordered. Additionally, private label brands are gaining importance for the retailers, which enables them to threaten manufacturers with substituting their products with own brands (Inderst & Shaffner, 2007). However, since there seems to be a maximum threshold for private label share for retailers (Ailawadi, Pauwels & Steenkamp, 2008), they have to consider other ways to differentiate with their assortment.
Dutch Food Manufacturing Industry
The Dutch food manufacturing industry is relatively heterogenic with approximately 5.500 companies (FNLI, 2016). This includes many small- and medium sized (regional) food manufacturers but also large multinationals like Heineken and Unilever. Four of the 30 largest food manufacturers of the world have their origin in the Netherlands. Those national brand manufacturers invest strongly in marketing and sales to build up their brands. However, both small and large companies are confronted with an ever increasing power of the food retailers through an ongoing process of concentration. Additionally, manufacturers are confronted with discounters (e.g. Aldi, Lidl) that follow a tough negotiation strategy and operate with extremely sharp margins, which further threatens the manufacturers (Chambolle & Villas-Boas, 2015). The amount of sales that is being realized by the large food retailers, however, makes cooperating with food retailers a necessity (Goodman & Dion, 2001; Geylani, Dukes & Srinivasan, 2007). Only 6% of national food brands are considered to be a “must-have” for the retailer and are thus less affected by delisting threats from large retailers. Those products are too important for the food retailers, since they are expected by the consumers to be available in the store (DIW, 2011).
The Role & Importance of Assortment
To understand the reasoning of retailers behind requests and demands for exclusive assortment from manufacturers, it is key to recognize what role assortment plays for the retailer and how important it is to attract customers.
Consumers’ retail store choice is influenced by a set of diverse factors like location, price, assortment, service, and shopping environment (Arnold, Oum, Tigert, 1983). Thus, retailers can utilize various levers (like assortment) to positively influence retail store patronage (i.e. store choice and frequency of visits) (Pan & Zinkhan, 2006). Assortment can be defined as the number of categories offered (breadth) and the number of stock-keeping units within a category (depth) (Simonson, 1999; Pan & Zinkhan, 2006; Broniarczyk, 2008). Consumers appreciate variety in assortment since they are more likely to find what they desire (Hoch, Bradlow, Wansink, 1999). Retailers can utilize assortment to increase purchase intentions (Simonson, 1999) and to drive store satisfaction and positively influence store choice (Hoch, Bradlow, Wansink, 1999). Often retailers include many products in their stores since they try to minimize shopping costs (in terms of effort & actual monetary costs) for the consumer and to facilitate one-stop-shopping (Hamilton & Richards, 2009). However, retailers are limited by the amount of money to invest in assortment and by the physical space available (Mantrala, Levy, Kahn, Fox, Gaidarev, Dankworth & Shah, 2009). In a meta-analysis of the existing literature about retail patronage behavior of consumers, Pan & Zinkhan (2006) found that a wide selection of products has the highest correlation with store choice (amongst factors like service, prices, and location). It has been confirmed by further research that assortment is even more influential for consumer store choice than retail prices, which were for a long time considered as the most important factor (Briesch, Chintagunta & Fox, 2009).
the likelihood that consumers find desired products (Gourville & Soman, 2005). However, purchase intention (conversion) was found to be greater for smaller choice sets (Iyengar & Lepper, 2000; Chernev & Hamilton, 2009). An increase in the breadth and depth of an assortment can have positive effects on sales but there were also harmful effects if the assortment is too large (Dhar, Hoch & Kumar, 2001). Large assortments can mentally stimulate consumers but too much choice is associated with a high cognitive load for the consumers, leading to regretting the choice or even avoiding the choice (Broniarczyk, 2008).
Thus, a key challenge for food retailers is offering a unique assortment that attracts consumers and that is not available at competing retailers (Simonson, 1999). This can lead to a stronger differentiation and prevent price erosion. Products that emerge to be popular are quickly introduced by competing retailers that want “a piece of the pie” as well. Offering a distinct and unique assortment is vital for retailers since many customers are not only loyal to one retail store but patronize multiple ones and divide their spending over more than one food retailer (Baltas, Argouslidis & Skarmeas, 2010).
An established method for retailers to differentiate themselves is to provide exclusive assortment in the form of private label brands that are only available at the focal retailer.
Private Label Brands
a trade-off between PL and national brands. PL products are only profitable as an addition to national brands if a significant amount of consumers buys the national brand (Corstjens & Lal, 2000). Furthermore, it was found that retailers cannot increase the share of PL indefinitely. There is an inverted u-relationship between the private label share and share-of-wallet (Ailawadi, Pauwels & Steenkamp, 2008), which implies that retailers can only raise the PL share to a certain threshold until they start to lose potential consumer spending. This shows the complementary role of national brands and PL brands (Corstjens & Lal, 2000).
Exclusivity Agreements from Different Perspectives
Legal & Power Perspective
Exclusivity agreements are vertical contracts between a supplier and a retailer that restrict one or both parties to engage in transactions with other third parties (Matouschek & Ramezzana, 2007). Exclusivity has often been investigated in academic literature in the form of exclusive dealing: a manufacturer delivers its products to a retailer and prohibits the dealer to sell (certain) products of direct competitors (Bernheim & Whinston, 1998; O’Brien & Shaffer, 1998). Whereas in exclusive dealing the restrictions lay upon the retailer to not sell products competing to the manufacturers’, in exclusive distribution, the manufacturer is constrained by only selling its products to a certain retailer (Moner-Colonques, 2006). Raising vertical constraints emphasizes that the main concern for exclusivity is the control that the manufacturer and retailer can have over each other respectively (Iglesias, Trespalacios & Vazquez, 2000). From the perspective of channel power between manufacturers and retailers, both parties strive to be as independent as possible from the other to avoid being exposed to abuses of influence of the dominating party (Fein & Anderson, 1997).
Relationship Perspective
been shown to improve the economic outcomes of the relationship (Rodríguez, Agudo & Gutiérrez, 2006).
Product Perspective
Exclusive distribution may improve the perceived image of the brand in terms of exclusivity (Iglesias, Trespalacios & Vazquez, 2000). There might also be advantages due to perceived product scarcity by consumers if an exclusive product is only available at a certain retailer. Scarcity implies that if a resource becomes limited, it increases in value (Cialdini, 1993). It could be a signal to (uninformed) consumers for superior product quality (Stock & Balachander, 2005). Furthermore, it was found that with exclusivity, higher retail prices can be realized since the product is not widely distributed (Andritsos & Tang, 2010). However, there is also a downside from the product perspective. Every new product or line extension increases complexity and costs for organizations. Added SKU should contribute positively to the manufacturers success (Gottfredson & Aspinall, 2005). Many manufacturers are rationalizing product lines in order to cut down on cost (Gourville & Soman, 2005). Thus, exclusive agreements might improve the image of the brand in terms of perceived value, but also might be conflicting with attempts to reduce complexity within the organization.
Exclusivity from the Retailers’ Perspective
since it is an inverted u-relation (Ailawadi, Pauwels & Steenkamp, 2008). This is why exclusivity is a viable alternative to using private label brands to differentiate.
Retailers value initiatives introduced by manufacturers - especially when those initiatives do not only have an economic perspective but are more seen as an investment in the relationship (Rodríguez, Agudo & Gutiérrez, 2006) - which applies for exclusivity agreements (Matouschek & Ramezzana, 2007) since those can also be aimed at being an improvement for the relation to the focal retailer. However, for exclusive assortment agreements, the initiative is mostly coming from the retailer. This becomes obvious when summing up the advantages of those for retailers.
It has been shown that in the context of (video-game) platforms and applications, an exclusive agreement gives the platform (retailer) the opportunity to differentiate itself (Hermalin & Katz, 2013). The video-on-demand platform Netflix uses exclusive content (e.g. the US-series “House of Cards” or “Narcos”) to differentiate itself from the many competitors on the video-on-demand market. This exclusive content is only available at Netflix and was often found to be a focal reason for consumers to sign up for a membership (TheWrap, 2014). Next to differentiation (Lang & Hunt, 2014), other advantages of exclusive assortment for the retailer are a higher retail prices (since the product is only available at the focal retailer) and higher store traffic by customers that are attracted by the exclusive product and that might purchase other products (spill-over effect) while being in the store (Andritsos & Tang, 2010; Gielens, Gijsbrechts & Dekimpe, 2014).
A first attempt by Gielens, Gijsbrechts & Dekimpe (2014) to quantify the results of exclusive assortment agreements between manufacturers and retailers showed a positive effect for the retailer.
Exclusivity from the Manufacturers’ Perspective
Since higher distribution of products increases market share (Reibstein & Farris, 1995), manufacturers are generally striving for intensive distribution in order to cover as much as possible of the market (Fein & Anderson, 1997). For manufacturers the main disadvantage of exclusivity agreements lies thus in the limited number of reachable customers (Peres & Van, 2014) since manufacturers can only access a smaller part of the market compared to a non-exclusivity situation (Andritsos & Tang, 2010). This implies that exclusivity agreements can lead to loss of sales opportunities for the manufacturer (Moner-Colonques, 2006; Gielens, Gijsbrechts & Dekimpe, 2014).
However, since the large retailers sell high volumes of products, it can still be attractive for the manufacturer to work together with them (Geylani, Dukes & Srinivasan, 2007). In a Dutch survey of food manufacturers, 94% of the suppliers stated that working together with (large) retailers is critical for long-term success and that they are willing to invest in collaborating with the retailer (EFMI, 2014).
Although initial research shows that quantitative results of exclusive assortment agreements are more positive for the retailer, there is a chance for manufacturers to nevertheless profit from those agreements. If both parties renegotiate margins specific to the exclusivity agreement in order to shift them gradually towards the manufacturer, it can be a win-win situation (Gielens, Gijsbrechts & Dekimpe, 2014) in which the manufacturer is compensated for lost sales opportunities (Andritsos & Tang, 2010).
Gielens, Gijsbrechts & Dekimpe (2014) list different types of exclusive assortment: - unique products lines (SKU range)
- exclusive SKU (size, packaging, variant)
- temporary exclusivity (time-limited exclusive launch)
METHODOLOGY Research Method
The aim of this research was to quantify insights about the willingness of national brand manufacturers to provide exclusive assortment for food retailers and about the influencing factors of this decision. This research is based on qualitative findings by Valkenburg (2014) and on the quantitative data collected by Sloot & van Everdingen (2015). Valkenburg (2014) investigated the motivations of Dutch manufacturers to engage in exclusive assortment agreements with retailers by conducting several in-depth interviews with manufacturers and retailers. Sloot & van Everdingen (2015) created a survey and collected data from Dutch manufacturers regarding the motivations to provide exclusivity. This data was prepared and analyzed in this research. Furthermore, it was initially intended to also collect data in Germany and replicate the survey conducted by Sloot & van Everdingen to compare the data between the Netherlands and Germany in order to explore potential country differences and similarities. However, that approach proved to be a challenge and was not successful; this will further be described in the next section.
Data Collection & Participants
The data about the motivations of Dutch manufacturers to engage in exclusivity agreements was collected by Sloot & van Everdingen (2015) with a quantitative survey. This survey was being distributed in 2015 to potential participants. 67 datasets were finally received from the respondents. Data was collected both via an internet-based as well as a printed survey. Targeted participants were chosen from small, medium and large companies mainly in the food manufacturing sector with only a small share of manufacturers being active in the non-food sector. Most of the targeted participants were decision makers from the field of marketing and sales with some participants being active in general management (director level).
Managers, Category Managers, Customer Directors, Marketing Directors, Trade Marketing Managers etc.) and ask them to participate in this study. More than 300 contact requests together with a short message explaining the context of the survey were sent to persons of interest. The replies, however, proved to be not as frequent as expected (less than 20% of potential participants replied) and a vast majority of the received answers were negative.
The following quotes of targeted participants are emphasizing this: § “… we use only standard sizes, no exclusive partnerships.”
§ “… information about our relationships with retailers will not be shared…” § “… we don’t have exclusive products, […] we find the topic legally
problematic…”
§ “… topic is not relevant.”
§ “… every customer is being treated equally, there are no customized products for any customer.”
Variables used in the Survey
This section will explain the concepts and variables that were used in the quantitative survey by Sloot & van Everdingen (2015). The variables will be explained briefly and the theoretical background on which the variables are based will be described together with the method of measuring the variable in the survey.
Hypotheses
After the variables and concepts measured in the dataset have been described, now the hypotheses will be established based on the variables and the literature review.
Manufacturer Characteristics
Large retailers possess a high amount of channel power in the relationship with brand manufacturers (Geylani, Dukes & Srinivasan, 2007; Chambolle & Villas-Boas, 2015) and use this power to achieve more attractive trade agreements and conditions with the manufacturers (Inderst & Shaffner, 2007). The larger (in terms of revenue) and thus the more powerful the retailer, the more likely it is that this focal retailer shows threatening behavior towards a manufacturer (Fein & Anderson, 1997). However, some manufacturers are large enough to resist pressure by retailers since they offer brands that are expected in the stores by consumers (DIW, 2011). This implies that large and more powerful manufacturers with strong brands might be less willing to engage in exclusive assortment agreements since they are less receptive to retailers that utilize their power in the channel relationship. The following hypotheses are related to variables that are an indication for the power of the manufacturer in terms of e.g. market position, market share, and revenue.
H1: The yearly revenue of the manufacturer is negatively related to the willingness to
provide exclusive assortment.
H2: Brand manufacturers (type of manufacturer: main focus branded products) are
less willing to provide exclusive assortment.
H3: Market leaders (market position: first position) are less willing to provide
exclusive assortment.
H4: The higher the market share in the focal product group the less likely the
H5: The higher the market share of the focal brand the less willing the manufacturer
will be to provide exclusive assortment.
Manufacturers generally strive for a high distribution rate in order to cover as much of the market as possible (Fein & Anderson, 1997). For manufacturers the main disadvantage of exclusivity agreements is the limited number of customers that can be reached (Peres & Van, 2014). Assuming that manufacturers which manage products with a high distribution rate are also aiming for a high distribution rate of their products in the future, it is expected that a high distribution rate has a negative influence on the willingness to provide exclusive assortment. Contrary, manufacturers that already have a portfolio of niche brands that are not widely available might be more adapted to a low distribution rate and thus be more prone to accept an exclusive assortment agreement (which implies limited distribution).
H6: The distribution rate of the manufacturers’ products is negatively related to the
willingness to provide exclusive assortment.
Some large and powerful manufacturers might have very strong brands which enable them to resist threats and demands of retailers (DIW, 2011). Respondents were asked to rate the (perceived) brand image of the brand(s) from the manufacturer they are working at. A positive brand image can be an indicator for a strong brand. Manufacturers are aware that assortment is the most important factor for store choice behavior of consumers (Briesch, Chintagunta & Fox, 2009). Thus those manufacturers might utilize their strong brand image and be less likely to respond positively to requests for exclusive assortment since they know how important their brands are.
H7: A positive brand image of the manufacturers’ brand(s) has a negative influence
Retailer Characteristics
It is very attractive for manufacturers to work together with large retailers that sell a high amount of products (Geylani, Dukes & Srinivasan, 2007). Retailers with a high market share can be assumed to be able to sell large amounts of products for the focal manufacturer and thus be attractive for exclusive assortment. Thus:
H8: The market share of the retailer is positively related to the willingness of a
manufacturer to provide exclusive assortment.
The performance of the focal retailer from the manufacturers’ point-of-view was measured via the fair-share. When a retailer performs above the fair-share (share of retailer in product group relative to retailer overall market share), then this implies that the retailer is performing very well and this might be attractive for the manufacturer in terms of exclusivity. Even though exclusivity agreements can lead to loss of sales opportunities for the manufacturer (Moner-Colonques, 2006), it can still be very tempting for a manufacturer to cooperate with a strong retailer that is performing very well regarding the fair-share since a high sales volume can be realized in that specific product group.
H9: A high fair-share of the retailer is positively related to the manufacturer being
willing to provide exclusive assortment for the focal retailer.
For ‘price strategy of the retailer’ the respondents were asked to rate the pricing strategy of the retailer in the focal product group compared to other retailers. Higher prices set by a retailer in the focal product group could make it more interesting for manufacturers to engage in exclusive assortment agreements since those higher prices would allow it for both parties to renegotiate margins to eventually make such an agreement profitable for the retailer and the manufacturer (Gielens, Gijsbrechts & Dekimpe, 2014).
H10: The higher the retailers’ prices in the product group (price strategy of retailer),
The retailer image was measured with regard to quality of the retailer, uniqueness of retail formula, and experience with category management & new product introductions. It is likely that a manufacturer would prefer to partner up with a retailer that has a decent image and is experienced when deciding about the possibility to provide exclusive assortment or not. The retailer image was measured from the manufacturers’ point-of-view.
H11: The more positive the retailer image, the more likely it is that a manufacturer
will be willing to provide exclusive assortment.
Since the monetary benefits that manufacturers can gain from agreements about exclusive assortment are limited for manufacturers (unless margins are re-negotiated) (Gielens, Gijsbrechts & Dekimpe, 2014), it is assumed that manufacturers demand additional support from the focal retailer in case that exclusivity is granted. Thus, the more support is expected from the retailer in return for exclusive assortment, the more likely a manufacturer is assumed to be to be willing to provide exclusivity.
H12: A higher level of (expected) support is positively related to the willingness of a
manufacturer to provide exclusive assortment.
Retailers are confronted with the challenge to offer a unique assortment that attracts consumers and that is not available at competing retailers (Simonson, 1999). This was measured by the ‘role of product category at the retailer’ which indicated if the product category was a standard category for the retailer or used for differentiation. If it was a differentiation category, it would be more likely that a manufacturer will be eager to provide exclusive assortment since this could allow higher margins for both the retailer and the manufacturer (Gielens, Gijsbrechts & Dekimpe, 2014).
H13: If the product category of the manufacturer is used for differentiation (role of
product category: differentiation) at the focal retailer, it increases the willingness of
Product category characteristics
The three variables ‘competition intensity in product group’, ‘similar quality in product group’, and ‘large choice set in product group’ were aimed at investigating the competitiveness in the product group. A large number of similar products with an intense competition can be indicators that this category is commoditized (Koschate-Fischer, Cramer & Hoyer, 2014) which means that consumers view the products in this category as being equal. Those three variables are supposed to have a positive influence on the willingness of a manufacturer to provide an exclusive product for a retailer since exclusive assortment is an approach for within-category differentiation for the retailer (Briesch, Chintagunta & Fox, 2009) which can also be profitable for the manufacturer (Gielens, Gijsbrechts & Dekimpe, 2014).
H14: ‘Competition intensity in product group’, ‘similar quality in product group’,
and ‘large choice set in product group’ are positively related to the willingness of a manufacturer to provide exclusive assortment.
The more revenue and sales are achieved within a certain product group, the more important this product group is for both the manufacturer and the retailer. Large retailers that are able to sell a high volume of products are attractive for manufacturers to cooperate with (Goodman & Dion, 2001; Geylani, Dukes & Srinivasan, 2007). The category size is thus expected to have a positive influence on the willingness for exclusivity since manufacturers might be willing to “go the extra mile” (with exclusive assortment) to achieve their goals for this category. Also, the expected category growth can be an indicator for the attractiveness to provide exclusivity since future growth is expected to be a strong motivator for the manufacturers.
H15: The category size is positively related to the willingness to provide exclusivity.
H16: The expected category growth is positively related to the willingness to provide
Retailers can only raise the share of private label products in their stores to a certain threshold until they lose potential consumer spending (Ailawadi, Pauwels & Steenkamp, 2008). This emphasizes the complementary role of national brands and private label products (Corstjens & Lal, 2000). For retailers, exclusive assortment is an option to differentiate within a category if the opportunities regarding private labels are exhausted. The private label share within the product group thus might be an indicator for the likelihood that retailers ask for exclusive assortment. Manufacturers could then support these demands to support a retailer with differentiation.
H17: The higher the private label share in the product group, the higher the
willingness of a manufacturer to provide exclusive assortment.
Since margin re-negotiation is a feasible option to make exclusive assortment agreements profitable for the manufacturer (Gielens, Gijsbrechts & Dekimpe, 2014), a higher gross margin for the retailer gives more room for margin (re-)negotiations. Thus, a higher gross margin for the retailer is expected to have a positive influence on the willingness to provide exclusive assortment.
H18: Gross margin (for the retailer) is positively related to the willingness to provide
exclusive assortment.
Manufacturer & retailer relationship
Exclusivity agreements can be seen as a relational commitment to the partner (Fein & Anderson, 1997) and as an investment in the relationship (Matouschek & Ramezzana, 2007). Valkenburg (2014) found that manufacturers could both be willing to invest in a bad relationship to improve it or invest in a positive relationship to improve it even more. However, it was found that the positive motivation is slightly more relevant. Manufacturers are expected to be willing to invest in a positive relationship with exclusive assortment in order to improve this relationship (Valkenburg, 2014).
H19: The relationship (relationship quality) with the retailer is positively related to
Furthermore, manufacturers are dependent on large retailers to sell their products (Geylani, Dukes & Srinivasan, 2007), which is why a high dependency on the retailer can be an indicator for the manufacturers’ motivation to satisfy the need of the focal retailer by providing exclusive assortment. This would then not be seen as an intrinsic motivation of the manufacturer to engage in exclusivity but more as an extrinsic motivation in order to keep the (important) retailer happy and to ensure a positive relationship since the manufacturer is highly dependent on the sales realized by the focal retailer.
H20: Relationship dependence is positively related to the manufacturers’ willingness
to provide exclusive assortment.
Anticipated competitive reactions
In case a manufacturer agrees to provide a focal retailer exclusive assortment, it is expected that other (excluded) retailers feel either very negative about this since they were excluded or that other retailers would also like to get exclusive assortment (Sloot & van Everdingen, 2015). Both reactions could prevent manufacturers of providing exclusive assortment in the first place to the focal retailer. However, not only other retailers could feel excluded, but also other manufacturers could perceive an exclusive assortment agreement as a “provocation” (Sloot & van Everdingen, 2015). The decision to provide exclusive assortment is very strategic and thus the manufacturers are expected to weigh the benefits and the (anticipated) reactions of other players (retailers and manufacturers) carefully.
H21: A strong anticipated competitive reaction of excluded retailers is negatively
related to the willingness of the manufacturer to provide exclusive assortment.
H22: The stronger the anticipated reaction of other manufacturers, the lower the
Conceptual Model
Based on the previously described hypotheses, the following conceptual model is derived. The Independent Variables were grouped together into (1) Manufacturer characteristics, (2) Retailer characteristics, (3) Product group characteristics, (4) Manufacturer & Retailer relationship, and (5) Anticipated competitive reactions.
Figure 2: Conceptual Model Plan of Analysis
DATA ANALYSIS
The goal of this part was to prepare the data for analysis, give an overview of descriptive statistics, reduce the large number of variables in the data set into fewer underlying dimensions, and to create a regression model that allows a practical interpretation.
Coding of variables
In order to analyze the data with a regression analysis, it had to be ensured that the variables were coded correctly. Most of the variables within the dataset were measured by using a 7-point Likert scale. Those variables were not modified and were used directly for the regression analysis. However, some variables were measured nominal or ordinal (non-Likert scales). Those variables had to be checked and eventually re-coded to ensure (1) analyzability with the regression analysis and (2) interpretability. Generally, all variables were coded that an increase in the variable’s value indicates an increase in the actual value (e.g. increasing values from 1 - 5 for ‘yearly revenue’ imply an increasing revenue). The extensive overview of the variable coding can be found in Appendix 1. All other variables not mentioned in the coding table were measured on a 7-point Likert scale.
Data screening / cleaning
to engage in exclusivity), which was not available for that retailer. This fact alone resulted in 32 (n of online surveys) cases not being considered in further analysis by SPSS (listwise deletion) since the influence of the independent variables could not be researched without retailer specific data about the dependent variable. Since data was missing about one retailer for 32 cases, the number of cases that could be analyzed thus dropped from 536 to 504. Respondents filling in the same data for every variable were not an issue in this data set. Even though there were cases where respondents were filling in the variables in a very similar way, there was always a slight difference in ratings between the different retailers. Furthermore, some data had to be cleaned since some respondents did not comply with the stated input rules and entered information that was not appropriate or necessary. For example when asked for market share it was necessary to enter the information for a “35%” market share in the form of “35”, additional %-signs were deleted to allow analysis. Generally, text was deleted from questions where only numerical input was allowed. Furthermore, commas were replaced with points (“,” à “.”) to enable data processing by SPSS. Variables that were classified as string variables were changed into numeric variables when appropriate (e.g. the output for “market share” was classified as a string variable, however, it was changed into a numeric variable since this was more appropriate for further analysis).
Descriptive Statistics
Table 2 - Descriptive Statistics
total cases n = 536, all values in valid % or means (excl. missing values), 1 - 7-point Likert or numerical scales
Variable Result
Willingness to introduce a new exclusive brand. (%)
Not willing (1 - 3) 53.6
Neutral (4) 10.8
High willingness (5 - 7) 35.6
How long would you be willing to grant exclusivity? (%)
Shorter than 3 months 36.1
3 - 6 months 12.3
6 - 12 months 24.7
1 - 2 years 14.8
2 - 3 years 6.7
Longer than 3 years 5.4
Willingness to provide exclusive assortment...
... with regard to market share of the retailer (mean)*
1 - 9% 2.73
10 - 20% 4.52
> 20% 5.20
... with regard to the price strategy of the retailer (mean)
lower than average prices (1 - 2) 2.85
average prices (3 - 5) 3.55
higher than average prices (6 - 7) 2.88
... with regard to the fair-share of the retailer (mean)*
performs below fair-share (1 - 2) 2.35
performs around fair-share (3 - 5) 3.33
performs above fair share (6 - 7) 3.99
... with regard to the PL share in product group (mean)*
1 - 25% 2.71
26 - 50% 2.95
51 - 75% 3.92
76 - 100% 4.22
... with regard to the positivity of the relation (mean)*
relatively negative relation (1 - 2) 3.31
neutral relationship (3 - 5) 3.06
very positive relationship (6 - 7) 3.87
Factor Analysis of Dependent Variable
The questionnaire used multiple items to measure various variables. In order to explore if some of these items have common variances and potentially represent underlying dimensions, a Factor Analysis (FA) was conducted. The goal was (1) to establish a dependent variable that could be used for the multiple regression analysis, (2) to improve the parsimoniousness (keeping the regression model as simple as possible) and (3) to reduce multicollinearity, which can occur if variables are highly related. First the data was screened and then various assumptions (see e.g. James, 2008) were tested to ensure the factorability of the items. An extensive description of the four variables and of testing the assumptions can be found in Appendix 2.
Four variables measured the willingness of respondents to provide retailers with exclusive assortment. Those variables factored very well with a KMO measure of .735. The Bartlett's test of Sphericity was significant: (χ2 (6) = 1138.12, p < .000) and the communalities were all higher than .7 which is above the recommended threshold of 0.4. The first factor had an Eigenvalue of 2.976 and explained 74.40% of variance. Cronbach’s Alpha was considered as excellent with a result of .885. Thus, analysis was continued with the 1-factor solution and a new variable was created: ‘Willingness for Exclusivity’.
Factor Analyses of Independent Variables
Table 3 - Overview Factor Analyses
Multiple Regression: Analyzing the Effects of the Independent Variables on the Willingness to Engage in Exclusivity
This section will investigate how much of the variance in the Dependent Variable (Willingness of the Manufacturers to Engage in Exclusivity) can be explained by the Independent Variables. Before a multiple regression analysis was conducted, several assumptions were checked that had to be met in order to ensure that the data suited this analysis. The extensive review of those assumptions can be found in Appendix 4. Since the assumptions have been met, a multiple regression analysis was conducted. However, the two variables ‘category size: market revenue in product group’ and ‘product group: large choice set’ were dropped from the model due to multicollinearity. Also, ‘past experience with exclusivity’ was not included anymore in the model due to a suspiciously high t-value and a too strong logical connection
Concept # of
Items Factors Names of New Variables
Cronbach’s Alpha Retailer Relationship 11 2 Positivity of Relationship Importance of Relationship α = .944 α = .861 Expected Support
from Retailer 8 1 Level of Expected Support α = .959
Retailer Image 7 2 Reputation & Capacity of Retailer Uniqueness of Retailer α = .906 α = .853 Competitive
Supplier Reaction 8 1 Level of Expected Reaction of Competitive Suppliers α = .935 Competitive Retailer
Reaction 9 2 Neg. Reaction: Anticipated Retribution Positive Reaction: Bandwagon Effect α = .916 α = .738 Brand Image
(Manufacturer POV) 6 1 Brand Image Positivity α = .849
Correlation Past Experience
a. Dependent Variable: Willingness Exclusivity Table 4: Refined Multiple Regression Modela
Variables Unstandardized Coefficients Standardized Coefficients Beta Collinearity Statistics B Std.
Error t Sig. Tolerance VIF
(Constant)
Manufacturer Characteristics
-.611 .463 -1.319 .188
Yearly revenue .097 .056 .093 1.742 .082 .482 2.076
Type of manufacturer (brand focus) -.016 .062 -.014 -.264 .792 .472 2.117 Market position of company -.065 .055 -.063 -1.167 .244 .470 2.126 Market share of company in product group -.002 .002 -.045 -.819 .413 .462 2.163 Market share of focal brand in product group -.037 .039 -.051 -.957 .339 .490 2.041 Distribution rate (in general) .045 .053 .049 .855 .393 .421 2.374 Availability at focal retailer .094 .117 .036 .801 .424 .689 1.451 Brand image: positivity -.175 .061 -.179 -2.853 .005 .350 2.860 Brand image: relatively high price -.041 .029 -.075 -1.421 .156 .494 2.025 Retailer Characteristics
Market share of retailers in % .419 .046 .455 9.178 .000 .562 1.781 Fair share performance of retailer .079 .030 .117 2.677 .008 .729 1.372 Price strategy of retailer .014 .026 .024 .539 .590 .708 1.413 Role of product group (differentiation) .113 .099 .050 1.142 .254 .715 1.398 Retailer image: reputation and capacities .051 .052 .055 .989 .323 .453 2.206 Retailer image: uniqueness .107 .044 .111 2.424 .016 .661 1.513 Level of expected support from retailer
Product Group Characteristics
-.143 .045 -.154 -3.196 .002 .594 1.685 Private label share in product group .006 .002 .164 3.086 .002 .492 2.033 Category growth: expected revenue dev. .074 .043 .083 1.720 .086 .599 1.670 Product group: intense competition .004 .033 .007 .112 .911 .391 2.559 Product group: similar quality -.048 .034 -.074 -1.386 .167 .484 2.067 Gross margin retailers in product group .044 .049 .045 .884 .377 .524 1.908 Manuf. & Retailer Relationship
Positivity of retailer relation .087 .041 .093 2.100 .036 .709 1.410 Importance of retailer relation .110 .045 .118 2.429 .016 .591 1.692 Anticipated Reactions of Manuf. & Ret.
The regression model statistically significantly predicted the “Willingness of Manufacturers to Engage in Exclusivity”, F(26, 327) = 15.234, p < .000, R2 = .548, adj. R2 = .512.
The model had a R2 valueof .548 which implies that the model explains 54.8% of the
variability in the DV “Willingness to Engage in Exclusivity”. This can be considered as relatively high. However, since R2 can only increase when variables are added to a model, it is key to also consider the adjusted R2 value which accounts for the amount of variables and if they add explanatory power to the model (Malhotra, 2006). The adj. R2 is lower with a value of .512 which implies that some variables did not add much explaining power. Nevertheless, in order to have a good overview of the influencing factors for the manufacturers’ decision to provide exclusive assortment, all variables will be kept in the model.
However, a regression model should not only be evaluated by examining the R2 values, but also by investigating the residuals. The residuals (error term: difference between observed and predicted values). There should be a random distribution of the residuals as well as a linear relationship (Malhotra, 2006). Figure 3 shows that those
requirements are met. Figure 3: Scatterplot of Residuals Interaction Effect
effect that ‘yearly revenue’ (of the manufacturer) has on the ‘willingness to provide exclusive assortment’. This was an explorative analysis since no hypothesis was created for this relation beforehand.
Figure 3: Possible Interaction Effect of Market Share on Yearly Revenue
In order to determine if a moderating effect exists of the size of the retailer (‘market share retailer’) onto the effect of the size of the manufacturer (‘yearly revenue’), an interaction term was added to the multiple regression model. First, both variables (‘market share retailer’ and ‘yearly revenue’) were standardized into z-scores. Second, the two standardized variables were multiplied and the result was saved into a new variable named: ‘Interaction effect: yearly revenue x market share retailer’. The regression model was estimated once again including the interaction term. R2 remained to be the same with a value of .548. The adjusted R2 value dropped marginally from .512 to .511 for the model that included the interaction term. The interaction was not significant, b = .019, t(326) = .507, p < .613. There was no multicollinearity (VIF statistics of 1.043). This suggests that the effect of the yearly revenue of the manufacturer (on the willingness to provide exclusive assortment) did not depend on the size of the retailer (in terms of its market share). Thus, seemingly the saying did not apply for this data. Both small and large manufacturers prefer large retailer when it comes to providing exclusive assortment (‘market share retailer’: b = .455, t(327) = 9.178, p < .000).
The results of the multiple regression model and the implications for the hypotheses will be described in the next section.
DV: Willingness to provide Exclusive
Assortment IV: Yearly Revenue
of Manufacturer
Hypotheses Results
H1: The yearly revenue of the manufacturer is negatively related to the willingness to
provide exclusive assortment. ‘Yearly revenue’ did not significantly predict the ‘Willingness to Engage in Exclusivity’, b = .093, t(327) = 1.742, p < .082. There was no significant effect, thus the hypothesis H1 was rejected.
H2: Brand manufacturers (type of manufacturer: main focus branded products) are
less willing to provide exclusive assortment. ‘Type of manufacturer’ did not significantly predict the DV, b = -.014, t(327) = -.264, p < .729. Thus, H2 was
rejected.
H3: Market leaders (market position: first position) are less willing to provide
exclusive assortment. The ‘market position’ of the manufacturer did not significantly predict the DV, b = -.063, t(327) = -1.167, p < .244. H3 was rejected.
H4: The higher the market share of the manufacturer in the focal product group the
less likely the manufacturer will be to provide exclusive assortment. The ‘market share in the focal product group’ was not a significant predictor of the willingness to engage in exclusive assortment agreements, b = -.045, t(327) = -.819, p < .413. Thus, H4 was rejected.
H5: The higher the market share of the focal brand the less willing the manufacturer
will be to provide exclusive assortment. The ‘market share of the focal brand’ did not significantly predict the DV, b = -.051, t(327) = -.957, p < .339. H5 was rejected.
H6: The distribution rate of the manufacturers’ products is negatively related to the
willingness to provide exclusive assortment. There was no significant relation between the ‘distribution rate’ and the DV, b = .049, t(327) = .855, p < .393. Thus, H6
was rejected.
H7: A positive brand image of the manufacturers’ brand(s) has a negative influence
‘Brand image: positivity’ significantly predicted ‘Willingness to Engage in Exclusivity’, b = -.179, t(327) = -2.853, p < .005. Thus, the more positive the manufacturers’ perception of the own brand image, the less likely it is that exclusive assortment will be provided. H7 was accepted.
H8: The market share of the retailer is positively related to the willingness of a
manufacturer to provide exclusive assortment. ‘Market share of the retailer’ significantly predicted the willingness of the manufacturer to provide exclusive assortment, b = .455, t(327) = 9.178 p < .000. Thus, manufacturers seem to prefer large retailers for agreements about exclusive assortment. H8 was accepted.
H9: A high fair-share of the retailer is positively related to the manufacturer being
willing to provide exclusive assortment for the focal retailer. ‘Fair-share’ significantly predicted the DV, b = .117, t(327) = 2.677, p < .008. H9 was therefore accepted; the
willingness of manufacturers to provide exclusive assortment increases with the fair-share performance of the focal retailer.
H10: The higher the retailers’ prices in the product group (price strategy of retailer),
the more likely manufacturers are to provide exclusive assortment. The ‘price strategy of the retailer’ did not predict the DV significantly, b = .024, t(327) = .539, p < .590. Thus, H8 was rejected.
H11: The more positive the retailer image, the more likely it is that a manufacturer
will be willing to provide exclusive assortment. The Factor Analysis showed that the items measuring the retailer image could be split into two underlying dimensions. ‘Retailer image: reputation and capacities’ did not significantly predict the willingness to provide exclusive assortment, b = .055, t(327) = .989, p < .323. Thus, H11A was rejected. However, ‘Retailer image: uniqueness’ was a significant predictor,
H12: A higher level of (expected) support is positively related to the willingness of a
manufacturer to provide exclusive assortment. ‘Level of expected support’ significantly predicted the willingness to provide exclusive assortment, b = -.154, t(327) = -3.196, p < .002. However, the relationship was contrary to the expected relation. The higher the expectations for support by the retailer, the less likely the manufacturer was to be willing provide exclusive assortment. Thus, H11 was rejected.
H13: If the product category of the manufacturer is used for differentiation (role of
product category: differentiation) at the focal retailer, it increases the willingness of
the manufacturer to provide exclusive assortment. The ‘role of the product category for the retailer’ did not significantly predict the DV, b = .050, t(327) = .989, p < .323. Thus, H12 was rejected.
H14: ‘Competition intensity in product group’, ‘similar quality in product group’,
and ‘large choice set in product group’ are positively related to the willingness of a manufacturer to provide exclusive assortment. ‘Large choice set in product group’ was dropped from the regression model due to multicollinearity. ‘Competition intensity in product group’ did not significantly predict the DV, b = .007, t(327) = .112, p < .911 and also ‘similar quality in product group’ was not significant, b = -.074, t(327) = -1.386, p < .167. Thus, H14A and H14B were rejected.
H15: The category size is positively related to the willingness to provide exclusivity.
‘Category size’ was dropped from the regression model due to multicollinearity. H16: The expected category growth is positively related to the willingness to provide
exclusive assortment. The expected ‘category growth’ was no significant predictor for the DV, b = .083, t(327) = 1.720, p < .086. Thus, H16 was rejected.
H17: The higher the private label share in the product group, the higher the
willingness of a manufacturer to provide exclusive assortment. ‘Private label share in the product group’ significantly predicted the willingness to engage in exclusivity, b = .164, t(327) = 3.086, p < .002. The willingness of manufacturers to provide exclusive assortment increases with the private label share in the product group. Thus, H17 was
H18: Gross margin (for the retailer) is positively related to the willingness to provide
exclusive assortment. The ‘gross margin’ for the retailer was not significantly related to the DV, b = .045, t(327) = .884, p < .377. Thus, H18 was rejected.
H19: The relationship (relationship quality) with the retailer is positively related to
the willingness of the manufacturer to provide exclusivity. The Factor Analysis showed that the relationship items were split into ‘relationship positivity’ and ‘relationship importance’. ‘Positivity of retailer relationship’ was a significant predictor, b = .093, t(327) = 2.100, p < .036. The more positive the manufacturer perceives the relation to the retailer, the more the manufacture is willing to provide exclusive assortment. Thus, H19 was accepted.
H20: Relationship dependence is positively related to the manufacturers’ willingness
to provide exclusive assortment. The ‘importance of retailer relationship’ significantly predicted the willingness of manufacturers to engage in exclusive assortment agreements, b = .118, t(327) = 2.429, p < .016. The more important the manufacturer perceives the relation to the focal retailer to be, the more likely it is that exclusive assortment will be provided. H20 was accepted.
H21: A strong anticipated competitive reaction of excluded retailers is negatively
related to the willingness of the manufacturer to provide exclusive assortment. The Factor Analysis showed that the items measuring the anticipated competitive reaction of excluded retailers could be split into two underlying dimensions. The ‘competitive reaction of other retailers: retribution” significantly predicted the DV, b = -.230, t(327) = -4.505, p < .000. This means that the stronger the anticipated negative reaction of other excluded retailers, the lower the willingness of manufacturers to provide exclusive assortment to the focal retailer. H21A was accepted. However, there
