The effect of Assortment Allocation on the Reference
Price and Product Choice
Master thesis
University of Groningen
Faculty of Economics and Business
Msc Business Administration
Marketing Management
Annemieke Westerdijk
The effect of Assortment Allocation on the Reference
Price and Product Choice
University of Groningen
Faculty of Economics and Business
Msc Business Administration
Marketing Management
Author
Annemieke Westerdijk
1716069
Abel Tasmanplein 76
9726 EP Groningen
annemiekewesterdijk@hotmail.com
06-29341102
Supervisors
First supervisor: Prof. dr. L.M. Sloot
Second supervisor: dr. J.E.M. van Nierop
Management summary
One of the most basic functions of a retailer is determining the product and service assortment to offer to their consumers (Levy and Weitz, 2008). Assortment planning is an important decision for retailers and brand manufacturers because the product selection that a retailer offers to their consumers is one of the main determinants of consumers´ store choice and purchasing decisions (Kok and Xu, 2011).
In this study the effect of assortment allocation and the moderating effect of product type have been investigated in relation to product choice and reference price. The main goal of the study is to get more insight in these effects of assortment allocation. This lead to the following research question:
‘’ What is the effect of assortment allocation on the product choice and reference price?’’
The study discusses three key issues in relation to product choice and reference price. These three issues are shelf location, shelf space and the design of a product category. Also a moderating effect, product type, is investigated. In the literature review these aspects are further described and explained in relation to the study. The hypotheses are tested with use of online questionnaires which were filled in by 188 respondents.
During the analysing part of the study multicollinearity is discovered and one of the independent variables, shelf location, had to be omitted from the residual of the study.
Preface
This master thesis is part of the final phase of the study Marketing Management at the Rijks Universiteit Groningen. Writing the thesis was not always even easy and stress-free. Therefore, I want to thank my first supervisor Laurens Sloot for his support and his valuable provided knowledge and feedback. Also Kasim Sietsma played a great role with his advice and support during the writing parts.
Moreover, I want to thank my second supervisor Erjen van Nierop for his valuable and helpful comments during the final phase of the writing process.
Furthermore, I appreciate the helpful provided comments by the rest of the thesis group, Roosmarijn Luitjes, Frank Beke and Victor Jiang during the meetings of the thesis group. By writing our theses simultaneously, we kept each other motivated and inspired to finish the thesis.
Finally, I want to thank all the respondents who filled in the questionnaires. Without their help this thesis could not have been finished.
Table of Content
Management summary ... 3
Preface ... 4
Table of content ... 5
1.0 Introduction ... 7
1.1 Introduction of the topic ... 7
1.2 Research question ... 10
1.3 Contribution and Relevance ... 11
1.4 Methodology ... 11
1.5 Structure ... 11
2.0 Literature review ... 13
2.1 Assortment selection ... 13
2.2 Assortment allocation ... 15
2.2.1 Introduction in assortment allocation ... 15
2.2.2 Focus of the study ... 18
2.3 Product choice ... 20
2.4 Reference price ... 21
2.5 Moderator ... 25
2.7 Research of this study ... 26
3.0 Conceptual Framework ... 27 3.1 Conceptual Model ... 27 3.2 Dependent variables ... 27 3.2.1 Product choice ... 28 3.2.2 Reference price... 28 3.3 Independent variables ... 29
3.3.1 The effect of changes of shelf space on product choice and reference price ... 29
3.3.2 The effect of changes of location on the shelf on product choice and reference price .. 30
3.3.3 The effect of the atmosphere of the shelf on product choice and reference price ... 31
4.5 Plan of analysis ... 37
5.0 Results ... 39
5.1 Descriptive analysis ... 39
5.2 Cronbach’s Alpha ... 40
5.3 Hypotheses testing regarding product choice ... 41
5.3.1 Distribution of the product choices ... 42
5.3.2 Multicollinearity ... 43
5.3.2 Logistic regressions regarding the product choice ... 43
5.3.3.4 Less expensive brand ... 45
5.4 Hypotheses testing regarding reference price ... 47
5.4.1 Multiple regression ... 47 5.4.2 One-Sample T-Test ... 49 6.0 Conclusion ... 53 6.1 Product choice ... 53 6.2 Reference price ... 54 6.3 Overall Conclusion... 55 6.4 Managerial implications ... 55
7.0 Limitations and Future Research ... 57
References ... 58
Appendix A Questionnaire 1 ... 64
Appendix B Questionnaire 2 ... 72
Appendix C Questionnaire 3 ... 79
1.0 Introduction
In this first chapter, the master thesis will be presented. It will start with an introduction of the topic, which will leads to the formulation of the research question. Thereafter, the contribution and relevance of the study will be discussed. In the fourth section, the research method will be described. Finally, the last section will show the structure of the thesis.
1.1 Introduction of the topic
One of the most basic functions of a retailer is determining the product and service assortment to offer to their consumers (Levy and Weitz, 2008). Assortment planning is an important decision for retailers and brand manufacturers because the product selection that a retailer offers to their consumers is one of the main determinants of consumers´ store choice and purchasing decisions (Kok and Xu, 2011). Assortment is a key factor of the marketing mix because it is a strategic positioning tool for customer acquisition and retention (Grewal et al., 1999; Kahn, 1999; Koelemeijer and Oppewal, 1999; Stassen et al., 1999).
In a study of Bauer et al. (2012), the researchers state that quality, price levels and variety of the assortment determines the retailer’s market position and image. Mantrala et al. (2009) state that retailers attempt to offer a balance among variety (number of categories), depth (number of stock-keeping units within a category), and service level (the number of individual items of a particular SKU). According to Mantrala et al. (2009), retailers are constrained by the amount of money they can invest in inventory and by their physical space. For example, offering more variety may limit the depth within the categories. Retailers have to make the right decisions with respect to variety, depth and service levels. Otherwise, consumers will defect, which will result in losses in both current and future sales. In line with these strategic issues is the fact that assortment selection is one of the most challenging tasks in retailing due to the dynamic in consumer perceptions and preferences and the constraints and changing environmental factors (Mantrala et al., 2009).
displayed on the shelves that have the greatest value to customers to maximize the store’s total profit.
Retailers expanded their assortments dramatically in the past decades and have reached exceptional levels of variety. According to Brat et al., (2009) a typical American grocery store carries more than 47,000 items, which is more than 50 percent higher than a decade ago. An important aspect of a retailer’s decision about how many items to carry in each product category involves understanding the impact of assortment size on consumers’ choice of a retailer (Chernev & Hamilton, 2009). Thus, a retailer might want to know whether reducing the number of items in its assortment will lead to a decline in store attractiveness and lower the likelihood of consumers choosing the store. Or, on the other hand a retailer might want to know whether increasing the assortment will result in greater store preference.
An important transition in retail management is the transition from traditional brand management to category management (CM) as a recent retail management initiative introduced new challenges to assortment planning in grocery retailing (Gruen and Shah, 2000; Bauer et al., 2012). Rather than managing brands across products categories, most grocery retailers today have adopted CM. CM shifts the focus of assortment planning from the brand level to the category level (Gruen and Shah, 2000). Under CM, a category manager will be in charge of the product mix, pricing, and merchandising of all brands within a specific category (Basuroy et al, 2001).
Regardless of all discussed issues, consumers expect retailers to offer the right mix of products, at the right price, with the right promotions, at the right time, at the right place (Gruen and Shah, 2000). In this way retailers serve the expectations of the customers. However, what contains the right mix of products or a good assortment? Bauer et al. (2012) investigated how consumers form judgements of what constitutes ‘’a good grocery assortment’’. Consumers use only a limited number of informational cues to form perceptions about four higher-level assortment dimensions; the assortment’s pricing, its quality, its variety, and its presentation.
(1) Consumers form perceptions about the pricing of an assortment through three distinctive informational cues: - grocery products of good value for the money, - various price ranges, - private label products.
(2) Consumers use three informational cues to form perceptions about the quality of an assortment offered in a grocery category. These three are number of organic products, -freshness of the products offered in the category, and -number of well-known national brands.
grocery assortment are -variety of different brands, -flavours, -package sizes and -quality ranges. These aspects not only enhance consumers’ likelihood to find the right product, but also provide consumers with the desired choice flexibility (Bauer et al., 2012).
(4) There is considerable empirical support for the assumption that many consumer decisions are influenced by contextual stimuli in the shopping environment (e.g. Inman et al., 1990). Promotional activities like scale signs, the shelf layout and the space devoted to a product category have been shown to significantly influence consumers’ purchasing decisions, as well as retailers’ sales and profits (Dreze et al., 1994; van Nierop et al., 2008). The way that products are presented inside the store (at the point of purchase) determines consumers’ overall evaluation of the grocery category’s attractiveness. Also here, consumers use three informational cues specifically to form assortment presentation perceptions, which are -appeal of the products’ display, -logic of the products’ arrangement and -product organization that supports consumers in finding what they are looking for.
This knowledge can be used by category managers regarding decision making in choices of assortments. Category managers are not only required to provide the right mix of products but also to manage its pricing, promotional activities and in-store presentation (Gruen and Shah, 2000). They should aim to periodically re-evaluate key categories, which represent the retailer’s primary drivers of store traffic and sales.
Pricing
An important issue regarding assortment is pricing. Even though assortments are more important than retail prices in store choice decisions (Briesch et al., 2009), prices are still very important in consumer shopping. According to Bauer et al. (2012), consumers form perceptions about the pricing of an assortment at the category level. This result is further supported by findings from reference price research, which suggest that consumers have internal, memory-based reference prices for each relevant product category (Monroe and Lee, 1999; Mazumdar et al., 2005). However, rather than knowing the precise prices of all available products, consumers form general price perceptions of the assortment and evaluate the expensiveness of shopping in the grocery category in relative terms. A consensus in marketing is that consumers arrive at such price evaluations by comparing the offered price of a product to a reference price (Blattberg et al., 1995; Kalyanaram and Winer, 1995). If the price is lower than the reference price, consumers are more likely to consider the current price as attractive and thus are more likely to purchase the product (Monroe, 1990).
judgements and choices based on the comparison of observed phenomena to an internal reference point (Kalyanaram and Winer, 1995).
According to different operationalizations of reference price, it can be stated that reference prices have a consistent and significant impact on consumer demand (Kalyanaram and Winer, 1995), which results in influencing the choices of consumers. Consumers react differently to price decreases relative to the reference price. Consumers react more strongly to price increases than to price decreases.
Two broad types of reference prices have been identified in literature: internal and external reference prices (Mayhew and Winer, 1992).
Internal reference price (IRP) generally represents inherent characteristics related to a
product and is the price in a buyer’s mind that serves as a fair and normal basis for judging or comparing actual prices (Grewal et al., 1998). Consumers may have some expectation based on their knowledge and memories of the historic prices of the product and therefore, enter the purchase environment with an idea of how much they should pay for it (Kalwani et al., 1990; Kalyanaram and Little, 1994).
External reference prices (ERP) are provided by observed stimuli in the purchase
environment (Mayhew and Winer, 1992), which is based on the prices of all brands in the category at the time of the consumer’s purchase. Consumers are often exposed to price information of competing products at the point of purchase. The readily available price information for similar products may be used as an external standard of comparison to assess the value of the target price.
Both internal and external reference prices are compared to the actual price of a brand in determining whether to choose that brand (Rajendran and Gerald, 1994). In other words, consumers consider the internal and external reference price discrepancy in their brand selection. For this reason, both reference prices are used for a more and comprehensive understanding of reference pricing.
1.2 Research question
The purpose of this study is to investigate the relation between assortment allocation and reference price & product choice. The following central question will be researched:
‘’ What is the effect of assortment allocation on the product choice and the reference price?’’
- How can an assortment be characterized? - What contains assortment allocation?
- What is a reference price and how does it work?
- What is the influence of assortment allocation on the product choice and reference price? - What is the influence of the moderator on the reference price and on product choice?
(further explained below)
The relationship between assortment allocation and reference price is influenced by a moderator. This study contains the moderator product type of the category. A product category can enclose more utilitarian properties or more hedonic properties. The literature review in the next chapter will describe this moderator more accurately.
1.3 Contribution and Relevance
This study contributes to literature in two ways. First, these two concepts (assortment and reference prices) have not been studied in a similar manner in the existing reference price literature. Although the effect of characteristics of assortment on consumer behaviour has been researched, its effect on the reference price has not been investigated. Secondly, this study will provide retailers with new knowledge regarding assortment and reference prices which will help them making the right decisions regarding assortment allocation.
Assortments and pricing are two of the most important issues for a successful business strategy. The prices of an offered good or service have a great deal of influence on the general choice- and buying behaviour of the consumer. Therefore it is very important to investigate the way consumers react to price information and how this perception consequentially influences consumers’ behaviour.
1.4 Methodology
The methodology of this research consists of a literature review and empirical research. The research questions will be answered by a literature review and application of this literature in order to be able to set up testable hypotheses for the empirical part of the research. These hypotheses will be tested with a questionnaire. This questionnaire will be allocated under regular grocery shoppers of supermarkets in the Netherlands in order to collect the data. The data will be analysed using the statistical program SPSS. The literature review and the analyses together, will answer the sub-questions and thereby the main research question.
1.5 Structure
2.0 Literature review
This study discusses three key issues in related category management areas; assortment allocation, reference pricing and product choice. Since assortment allocation is strongly related with assortment selection, some key characteristics of assortment selection are discussed first. Thereafter, the areas assortment allocation, product choice, reference prices and the moderator product type are described.
Retail management is to develop a retail mix to satisfy consumers’ demand and to affect consumers’ purchasing decisions. The factors in retail mix include store location, product assortment, pricing, advertising and promotion, store design and display, services and personnel selling (Levy and Weitz, 2008).
2.1 Assortment selection
Assortment selection is a key issue faced by retail managers and category managers. Due to the limited shelf space, the question remains which items should be added to or deleted from the assortment. The retailer is concerned with the optimum use of shelves to provide a complete assortment of products, which meets the preferences of the consumers. The variety of products being displayed on the shelves should have the greatest value to consumers so as to maximize the store’s total profit (Hariga et al., 2007).
Assortment can be characterized according to the (1) number of brands, (2) number of stock keeping units (SKUs) per brand, (3) number of sizes per brand, (4) proportion of SKUs that are unique to the retailer (a proxy for private label) (5) availability of household’s favourite brands (Briesch, Chintagunta and Fox, 2009).
The key findings of the study of Briesch, Chintagunta and Fox (2009) contain the number of brands in an assortment and the presence of a household’s favourite brands increase the probability of choosing a store. However, the number of SKUs per brand, the number of sizes per brand, and the number of unique SKUs do not. In general, they argue that store choice decisions are more responsive to changes in assortment than to changes in price.
Broniarczyk and Hoyer (2006) state in their research that shoppers may prefer smaller grocery store assortments. Assortment is fundamentally different from price and convenience in that lower prices and more convenience are uniformly preferred, but larger assortments are not.
space devoted to the category, and the availability of the consumer’s favourite item. Furthermore, the research of Broniarczyk, Hoyer and McAlister (1998) also shows that shoppers’ perceptions of a retail assortment depend on the availability of their favourite items. It may be clear that favourite items vary by individual.
An important decision that retailers make involves selecting the number of items constituting their assortments in each product category (Chernev and Hamilton, 2009). They argue that this decision involves optimizing the benefits and costs of the assortment size for both buyers and sellers. Thus, from a retailer’s standpoint, smaller assortments are often considered more desirable for cost related reasons, for example inventory, shelf space and financing costs.
Moreover, not only retailers benefit from smaller assortment. Consumer preference for retailers offering larger assortments tends to decrease as the attractiveness of the options in their assortments increases and can even lead to a reversal of preferences in favour of retailers offering smaller assortments (Chernev and Hamilton, 2009).
The study of Diehl and Poynor (2010) demonstrates that even when consumers make a purchase, the same item may generate lower satisfaction when chosen from a larger rather than a smaller assortment. This effect is explained in terms of an expectation-disconfirmation mechanism. When assortments are small, consumers have low expectations about their ability to match their preferences. The opposite is true for large assortments. Consumers may experience greater negative expectation disconfirmation or less positive expectation disconfirmation when a chosen item comes from a larger rather than a smaller set of items.
Contribution to this research
In this study, the focus is placed on assortment allocation. Since assortment allocation is strongly related with assortment selection, a short description of this area is provided. For this study a given assortment will be used, which will be adapted to the purposes of the research.
2.2 Assortment allocation
2.2.1 Introduction in assortment allocation
Shelf management has been considered as an important decision to retailers (Hwang et al, 2005). Shelf space is a limited resource that must be optimally divided among the different categories and their various brands (Amrouche and Zaccour, 2007). Retailers must make frequent decisions on which products to include in the assortment and how much shelf space to allocate to them, because of fixed and scarce shelf space and the number of potentially available products increases steadily (Irion, Lu and Tsao, 2011). When retailers develop a well-designed shelf space management system, they can attract customers, prevent stock-outs and, more importantly, increase the financial performance of the store while reducing operating costs (Yang and Chen, 1999).
Assortment allocation models
Different models have been proposed in literature for optimizing shelf space allocation across product categories. Relevant literature starts in the 1960s and 1970s, when a number of experiments were conducted to measure the effect of shelf space on sales. In the literature, the space elasticity has been widely used to estimate the relationship between sales and allocated space.
Space elasticity is a ratio of relative change of sales to relative change of display space. The measurement of this elasticity is divided into two types:
- Direct elasticity is designed to measure the effect on demand by changing the display space for an individual product. The increase of display space for a product may stimulate the demand of products, but in turn, it may decrease the demand of substitute and/or complementary products.
The first researchers only considered direct elasticity. Anderson and Amato (1974) modelled the impact of display on sales as a logistic function and derived optimally conditions for two products.
Other authors who also only considered the direct space elasticity were Hansen and Heinsbroek (1979). They constructed optimization models to select and allocate products, where profit, inventory cost, and cost for allocating a product on a shelf were taken into consideration.
Corstjens and Doyle (1981, 1983) were the first to optimize store profitability with respect to space allocation. They considered both the direct and the cross-space elasticities in their multiplicative demand function, and specified a cost function that moderates the profitability of the allocation. They explicitly captured the substitution and complementarity effects among products with a power, multiplicative function of the display areas, and proposed a geometric programming algorithm to solve their model. An extensive comparison with alternative procedures suggests that their general model leads to significantly different allocation rules and better profit performance (Corstjens and Doyle, 1981).
Building on the work Corstjens and Doyle (1981, 1983), the authors Bultez and Neart (1988) elaborated a general, theoretical shelf space allocation model, which focuses on the demand interdependencies prevailing across and within product-groups. Their model is called Shelf Allocation for Retailer’s Profit (SH.A.R.P). They applied the model to experimental data with six brands and found that the proposed model improves on current profit levels and that it is better than the rules of thumb. However, the optimization focuses only on the shelf space devoted to an item, and does not include other shelf layout descriptors such as shelf height and the horizontal position of an item on the shelf, nor does it include marketing instruments such as feature and price.
Furthermore, Borin et al. (1994) developed a category management model to aid retailers in the space constrained decisions of which products to stock (assortment) and how much shelf space to allocate to those products. Their model is formulated as a constrained optimization problem, with assortment and allocation of space as the decision variables. The non-linearities in the objective function and the zero-one decision variables disallow a closed form solution.
Other streams of research have investigated optimization routines for shelf allocation. Although several routines have been proposed to optimize shelf layouts, they often only consider the shelf number and the number of facings to allocate.
Van Nierop et al. (2008) presented a new approach to measure the impact of shelf layout on sales and to optimize shelf arrangements. After estimating the model parameters on experimental data, van Nierop et al. (2008) found that the shelf layout has significant effects on baseline sales and marketing effectiveness. This not only holds for the number of facings allocated to an item, but also for other shelf descriptors such as shelf height and distance to the end of the aisle. The results of van Nierop et al. (2008) are important for the study of this paper and are discussed below.
Firstly, a higher number of facings cause a higher intercept, which in turn, results in more sales. However, this effect levels off as the number of facings increases. The effect of the number of facings on sales exhibits diminishing returns.
Secondly, for a low number of facings, the price elasticity appears to be higher than -0,5. This could result from the fact that these items are not very visible on the shelf, and therefore consumers do not easily observe prices. For these items, price discounts may have a limited effect. However, as the number of facings goes up, the price sensitivity also increases. Having more facings increases awareness of the price level, and thus creates more price sensitivity among customers. Also here, this effect appears to level off as the number of facings increases further.
Thirdly, not only the number of facings correlates positively with sales, also shelf height does. Just as with facings, a higher value (i.e., a higher shelf location) makes consumers more price-sensitive. This results from the fact that consumers see the prices for higher located products more easily than the prices of those products at the bottom shelf. However, this effect on baseline sales levels off as you move toward the top shelf.
the expected sales. A small positive effect is found by the horizontal distance to the middle of the shelf. For the optimal result, being close to the racetrack is best.
In conclusion, besides the fact that an increased number of facings can gain profit, this can also result from the fact that the item is put on a different location on the shelf, closer to the racetrack, or a combination of these things.
Other possibilities
Retailers currently use a wide range of methods to choose assortments and allocate space. Besides academic literature, commercial software programs such as Apollo (Information Resources) and Spaceman (ACNielsen) have contributed to the area of allocation of shelf space through non optimal methods. The widespread availability of scanner data and commercial shelf management models allow retailers to quickly detect and eliminate unprofitable items (Borin et al., 1994). An increasing number of these models attempt to maximize retailer profits by allocating space according to item profitability and shelf-space elasticity.
In conclusion, many studies have been conducted in the field of assortment allocation. However, there are still many gaps in the literature to develop the excellent model to allocate the assortment. In this study, the relevant knowledge is used to provide assortment allocations, which are in line with the study.
2.2.2 Focus of the study
After discussing the relevant literature of assortment selection and assortment allocation the focus of the study will be addressed. Assortment allocation can be divided into multiple aspects. In this study three aspects of assortment allocation are chosen to focus on. These three aspects are shelf space, shelf location, and shelf atmosphere. These three factors may or may not influence the relation between assortment allocation and reference price and product choice.
Shelf space
The amount of shelf space is the number of facings which are allocated to a product. The choice of which items to stock and how many space allocated to the items is difficult because of the limited shelf space available.
On the other hand, an SKU with more facings has higher sales. However, also here the additional benefits of one extra facing does decrease. Furthermore, price effects appear to be weaker for items with few facings (Martinez de Albeniz and Roels, 2007).
There is a difference between shelf space in larger assortments and in smaller assortments. Larger assortments place more demands on shelf space, or, in case of constant shelf space, necessitate a rearrangement of space across items, with fewer facings per item on average (Chiang and Wilcox, 1997). As this reduces, the inventory can be stocked per item (both in the store’s storage room and on the shelf). Larger assortments may not only reduce the items’ visibility on the shelf, but may also lead to an increased risk of stock-out occurrence (Bultez & Naert, 1988; Bultez, Naeart, Gijsbrechts, & van den Abeele, 1989).
As mentioned before, for a low number of facings, the price elasticity appears to be higher than -0,5 (van Nierop et al., 2008). This could result from the fact that these items are not very visible on the shelf, therefore the price is not easily observed by consumers. For these items, price discounts may have a limited effect. However, as the number of facings goes up, the price sensitivity also increases. Having more facings increases awareness of the price level, and thus creates more price sensitivity among customers. Also here, this effect appears to level off as the number of facings increases further.
Location on the shelf
Literature is limited regarding the location of the shelf area. One of the few researchers is Dreze et al. (1994). In their research they conducted a series of field experiments and found that location had a large impact on sales, whereas changes in the number of facings allocated to a brand had much less impact as long as a minimum threshold (to avoid out-of-stocks) was maintained. In particular, location appears to have a large impact on sales. For example, in most categories, products perform best when placed at eye level.
An improper location or an under-allocation of space might kill a product before it achieves full sales potential. Retailers work hard to maximize return on their investment: allocating too many facings is a waste, while allocating too few will result in lost sales due to out-of-stocks (Dreze et al, 1994).
Finally, Hwang et al. (2005) state that the demand rate is assumed to be a function of the displayed inventory level and the display location within the shelves of each item. They concluded that the level of the shelves, on which items are displayed, has an effect on sales and that the position of eye-level is the best.
Design of the shelf
This area is under-researched in the existing literature. This study will contribute to the knowledge of the impact of the atmosphere of the shelf. In the case of this study, design of the shelf presents the layout of the shelf of a particular category.
Figure 1 shows a normal shelf layout of the wine category in a Dutch supermarket (Albert Heijn). In figure 2 a more special wine category layout is displayed in the same Dutch supermarket. The question in this study is what effect the shelf atmosphere has in relation with assortment allocation, reference price and product choice.
Figure 1 Wine category Albert Heijn Figure 2 Wine category Albert Heijn
2.3 Product choice
If consumers are completely brand loyal and the product is available at the store, the space allocated to an item has no effect on its sales (Anderson, 1979). If the favourite product is present, an individual always buys the same product, and either delays the purchase or proceeds to another location if it is not available.
Hwang et al. (2005) assumed that the level of shelf on which the product is displayed significantly influences the sales of products.
Chiang and Wilcox (1997) note useful facts about the relationship between key category-level decisions that retailers make. They state that retailers offer a greater product variety in categories where consumers indicate that they are loyal to a particular brand. Retailers also allocate less shelf space to categories that have higher average margins in order to try to influence consumers’ store choice decision.
2.4 Reference price
Reference prices are standards against which the purchase price of a product is judged (Monroe, 1973). Reference prices are used by consumers to compare actual selling prices to a price standard during a purchase situation. Reference price applies only to those consumers who respond to differences in price (Rajendran and Tellis, 1994).
The concept of reference price is that it is an internal standard against which observed prices are compared. There is a significant body of theory to support the notion that individuals make judgements and choices based on the comparison of observed phenomena to an internal reference point (Kalyanaram and Winer, 1995).
Reference price is an important construct from a managerial perspective. An example of this importance is the timing of sales promotions. This can be greatly affected by whether or not it is assumed consumers form reference prices (Kalyanaram and Winer, 1995). Price promotions can influence the reference prices when products are regularly sold by promotion. They can lower the reference price and have two subsequent impacts: (1) the later promotions are not perceived to be as such a good deal as the earlier ones, and (2) a return to the ‘’normal’’ price may look to the consumer like a price increase (Kalyanaram and Winer, 1995).
Behavioural foundations
The behavioural foundations for the reference price concept come from several different areas of psychology. The most commonly used rational for reference prices is Helson’s Adaptation-Level theory (1964).
would be as follow: if a person repeatedly lifts a weight of 200 grams, he or she then becomes accustomed to this weight, i.e., adapts to it, and the adaption level becomes 200 grams.
Besides the Adaptation-Level theory, other psychological theories strongly support the reference price concept. One of these other theories is the Assimilation-Contrast theory which is most closely identified with Sherif and Hovland (1958).
Sherif and Hovland (1961) developed the Assimilation-Contrast theory based on results from a series of experiments they undertook with weights and numerical scales. They noted that their participants used two processes when making psychophysical judgments, which they termed contrast and assimilation. Contrast denotes difference, repulsion and movement away from an object's attributes or position, whereas assimilation denotes likeness, attraction and movement toward objects and quality (Helson, 1964). They generalized their original results in these terms: When an anchor is introduced at the end or slightly removed from the end of the series, there will be a displacement of the scale of judgment toward the anchor and assimilation of the new reference point in the series. When, however, the reference point is too isolated, there will be displacement in the opposite direction (i.e., away from the anchor) (Sherif, Taub, & Hovland, 1958). Their theory is similar to adaptation-level theory since it posits that an individual compares new stimuli to a background of previous experience. Sherif and Hovland (1961) proposed there was latitude of acceptance for new stimuli that were tolerable, latitude of rejection for those considered being objectionable, and latitude of non-commitment for those not evaluated as either acceptable or objectionable. Assimilation and contrast were viewed as complementary, not independent processes. Thus, while Adaptation-Level theory introduces the notion of a reference point and offers an explanation on how it is derived, Assimilation-Contrast theory complements it by introducing the concept of latitude of acceptance around that adaptation level.
The last psychological theory which supports the reference price concept is the Prospect theory of Kahneman and Tversky (1979). This theory offeres further refinement to explaining reference price formation in that it suggested why adaptation resulted in different anchor points being established when the potential outcomes of the adaptation were framed differently. Like the other two theories, it recognized that perceptions and judgments are relative, and evaluations of the acceptability of a price increase are made by comparing it to a reference point.
pleasure associated with gaining the same amount’’. This was their seminal contribution and it extended the other two theories’ explanations of reference price formation.
In the context of price, a gain occurs when a given price is lower than the reference price. In this situation, there is a high probability that a purchase will be made. When a loss occurs because a price is higher than the reference price, not only is a purchase less likely, but also the size of the effect is substantially larger. The proclivity for loss aversion means that ‘’ People feel the pain associated with a price increase more sharply than they feel the joy associated with a price decrease. They react more strongly to a negative change than they do to a positive change’’ (Raghubir, 2006).
This recognition explains why the latitude of acceptance is asymmetric with a smaller zone of acceptance for price increases than for price decreases. Prospect theory undergirds reframing efforts designed to change a participant’s evaluation of an outcome through identifying who gains and who loses from changes in price. It also explains the effectiveness of several of the reference price related strategies.
Empirical generalizations
Considerable theoretical justification for consumers’ use of psychological reference points exist from the research literature. Kalyanaram and Winer (1995) note three empirical generalizations that are well-supported in the marketing literature.
First, there is ample evidence that consumers use reference prices in making brand choices. Reference prices have a consistent and significant impact on consumer demand. This is a strong finding in the existing reference price literature.
Second, the empirical results on reference pricing also support the generalization that consumers rely on past prices as part of the reference price formation process. ‘Internal’ reference prices utilize past prices as part of the customer’s information set. All empirical studies on references prices have assumed that past prices are important components of the reference price formation process. Depending upon how the choice models are estimated, these studies have found either explicitly that past prices are significant predictors of reference price or, based on the overwhelming evidence produced in generalization one, reference prices have been found to significantly influence choice, have inferred that past prices form the basis of reference price.
react differently to price increases and price decreases relatively to the reference price. Consumers react more strongly to price increases than to price decreases.
Internal and external reference price
Two broad types of reference prices have been identified in literature: internal and external reference prices (Mayhew and Winer, 1992).
- Internal reference price
Internal reference prices (IRP) are stored in memory on the basis of actual, fair, or other price concepts (Mayhew and Winer, 1995). The existence of internal reference price is supported by Adaptation-Level theory, which suggests that people adapt to the level of past stimuli and judge new stimuli in comparison with the adaptation level. In other words, internal reference price is an internal standard against which observed prices are compared. The internal reference price generally represents inherent characteristics related to a product and is the price in a buyer’s mind that serves as a fair and normal basis for judging or comparing actual prices (Grewal, Monroe & Krishnan, 1998). Consumers may have some expectation based on their knowledge and memories of the historic prices of the product and therefore, enter the purchase environment with an idea of how much they should pay for it (Kalwani, Yim, Rinne & Sujita, 1990; Kalyanaram & Little, 1994).
- External reference price
External reference prices (ERP) are provided by observed stimuli in the purchase environment (Mayhew and Winer, 1992), and is based on the prices of all brands in the category at the time of the consumer’s purchase. Consumers are often exposed to price information of competing products at the point of purchase. The readily available price information for similar products may be used as an external standard of comparison to assess the value of the target price. External reference price often involves a comparison between the purchase price of a brand and the current prices of other competing brands observed. For example, it can be a function of the prices of all competing brands observed at the point of purchase (Rajendran and Gerald, 1994), or the extreme prices present (Krishna, Wagner, Yoon and Adaval, 2006).
Antecedents of reference price
One stream of research has identified various antecedents of reference price and these are described in this last subsection of reference pricing (Mazumdar et al., 2005).
- Prior purchase experience
The strongest determinant of a consumer’s IRP is the prior prices he or she observes. Prices encountered on recent occasions have greater effect on IRP than distant ones. Finally, the greater the share of prior promotional purchases, the lower is the consumer’s IRP.
- Purchase context moderators
Thaler (1985) demonstrates that reference points for an identical product differ simply because of differences in purchase contexts. Bell and Lattin (1998) demonstrate that large-basket shoppers are less price elastic in their individual category purchase incidence decisions but are more price-elastic in their store choice decisions. This suggests that the shopping occasions should moderate the influence of prior price and promotional history on IRP. The effect of purchase context moderators may also vary by shopping type.
- Store environment moderators
A brand’s IRP may vary by store because of the level of service provided, assortment offered, or store types (e.g., factory outlet, specialty store, mass merchandiser) (Berkowitz and Walton 1980; Biswas and Blair 1991).
- Product category moderators
IRPs for durable products are influenced by aggregate factors such as anticipated economic conditions (e.g., inflation) and household demographics. In addition, in the formation of IRPs for durable products, competitive prices and differences in attribute configurations and features across alternatives are more salient than historical prices; historical prices of durable products are used only to discern a price trend, if it exists. Finally, consumers’ price expectations are influenced by the technology used in a specific brand compared with other brands in the same durable product category.
2.5 Moderator
In this study the moderator ‘product type’ is included. Product type may or may not influence the relation between assortment allocation and product choice and reference price. In this section the moderator is discussed in more detail.
and utilitarian motives. Product categories differ in the extent to which their overall attitudes are derived from hedonic or utilitarian components. Hedonic products are desired for pleasure, fantasy, and fun; whereas utilitarian items tend to fulfil basic needs or help accomplish functional or practical tasks (Strahilevitz and Myers, 1998). Some categories consist only of hedonic components, whereas other categories consist of only utilitarian and there are also categories which have a mix of the two components.
Consumer evaluation of a consumption object is cognitively placed on both a utilitarian dimension of instrumentality (e.g., how useful or beneficial the object is), and on a hedonic dimension measuring the experiential affect associated with the object (e.g., how pleasant and agreeable those associated feelings are) (Khan and Dhar, 2010).
In this study we examine consumer choice of two product categories, one of which is seen as superior on a hedonic dimension and the other is seen as superior on a utilization dimension. An example of a hedonic product is chocolate, and an example of a utilitarian product is milk.
According to the research of Dhar and Wertenbroch (2000) products and brands that have a high score on the hedonic dimension relative to the score on the utilitarian dimension are able to charge a price premium. Thus, when consumers want to purchase a hedonic product, they are less price sensitive than when they want to purchase a utilitarian product. This can be explained by the fact that when consumers purchase a utilitarian product they are looking at the quality/price relation of the product, whereas when consumers want to purchase a hedonic product they only consider the pleasure, quality of the product.
In this study, we expect that the type of product moderates the effect of shelf space and shelf location on the product choice and the reference price of a category.
Consumers buy utilitarian products because these products have a special function to solve a need, and thus are more goal-oriented in their product choice. When buying utilitarian products, consumers are less influenced by external factors, like shelf allocation and location of the products on the shelf.
2.7 Research of this study
As can be seen from the introduction and the literature review, many research is performed in the areas of assortment selection, allocation, product choice and reference price. However, there is not been many research executed on linking these aspects together.
3.0 Conceptual Framework
In this chapter, the conceptual model of the study and the associated hypotheses is presented. First, the conceptual model is presented to provide a clear overview of the model presented in the subsequent subsections of this chapter. Thereafter, the dependent variables are discussed. Then, the independent variables are presented. These antecedents will be defined and the hypotheses will be drawn regarding the effect on the dependent variables, product choice and reference price.
3.1 Conceptual Model
By combining the literature and the hypotheses, the following framework is developed. The graphical representation shows the relations between the independent variables, moderator and the dependent variables.
Figure 1 Conceptual Model Assortment Allocation H1H2 H3 H4 H5 H6 H 7 H8 H9
Independent variables Dependent variables
H 10 H11 H12
3.2 Dependent variables
In this study there are two dependent variables: (1) product choice and (2) reference price of a category. In this subsection these two variables will be brief defined.
3.2.1 Product choice
Research has shown that the majority of the consumers are not brand loyal. They are willing to change their initial choice and switch to other products (Motes et al., 1985; Walter et al., 1975). Two reasons are (1) that the consumer’s brand was not available, or (2) that the shelf display changed consumers’ choice. For these consumers the final choice may be influenced by one or more in-store marketing factors.
This study will focus on the second reason; that the shelf display changed consumers’ choice of a product. The product choice of a consumer can be described as the actual choice of a product from a defined category made by a consumer. An example of a defined category is the milk category in a supermarket.
3.2.2 Reference price
The second dependent variable is reference price of a category. Reference prices are standards against which the purchase price of a product is judged (Monroe, 1973). As already is mentioned in the literature review, reference prices are used by consumers to compare actual selling prices to a price standard during a purchase situation.
Consumers use reference prices in making brand choices; they have a consistent and significant impact on consumer demand (Kalyanaram and Winer, 1995). Consumers rely on past prices as part of the reference price formation process. All empirical studies on reference prices have assumed that past prices are important components of the reference price formation process. Past prices are significant predictors of reference price.
Both the internal and external reference prices (see literature review) are compared to the actual price of a brand when consumers determine whether to choose a brand (Rajendran and Gerald, 1994).
3.3 Independent variables
In this study components of assortment allocation are the independent variables, namely shelf space, location on the shelf and the atmosphere of the shelf. In this section the variables will be briefly described and the related hypotheses are proposed.
3.3.1 The effect of changes of shelf space on product choice and reference price
According to the literature review, the amount of shelf space is the number of facings which are allocated to a product. Due to the limited shelf space the choice of which items to stock and how many space allocated to the product is difficult.
A central premise to shelf space allocation models is that as shelf space increase, unit sales increase. Also in practise you can see that items which sell well tend to get more facings. Martinez de Albeniz and Roels (2007) have confirmed this premise by stating that items with more facings have higher sales. However, the additional benefits of one extra facing decreases. This effect is defined as laws of diminishing returns which states that in all productive processes, adding more of one factor of production, while holding all others constant, will at some point yield lower per-unit returns (Samuelson and Nordhaus, 2001).
In this study direct elasticity will be measured. Direct elasticity is designed to measure the effect on demand by changing the display space for a SKU. Kok and Fisher (2007) state that substitution takes place within a category but not across categories. This entails that a change in the amount of facings assigned to an SKU, within a category, influences demand of the other products within the same category. So, the increase of shelf space may stimulate the demand of this particular SKU, but may decrease the demand of a substitute product. Therefore:
H1: The number of facings of an SKU is positively related to the product choice of that particular SKU.
H2: The effect of the number of facings of an SKU on the product choice of that particular SKU follows a diminishing returns scale.
In this study, we expect that the price effects as stated above will also exist in relation to the reference price of a category. More numbers of facings of a SKU results in more awareness in the category and more awareness of the price of this SKU. In relation to reference price, we expect that the more facings are allocated to a particular SKU, the more consumers are aware of that particular product price which will result in a higher effect of this SKU on the reference price of a category. Therefore:
H3: The number of facings of an SKU is positively related with the effect on the consumer’s reference price of a category.
H4: An increase of a SKU’s number of facings, which is higher priced than the consumer’s reference price of the category, is expected to increase the consumer’s reference price. H5: An increase of a SKU’s number of facings, which is lower priced than the consumer’s
reference price of the category, is expected to decrease the consumer’s reference price. 3.3.2 The effect of changes of location on the shelf on product choice and reference price
Another interesting field of research is the location on the shelf of an SKU. Dreze et al. (1994) have stated that location had a large impact on sales. An improper or an under-allocation of space might kill a product before it achieves full sales potential. Moreover, Hwang et al. (2005) conclude their research with the findings that the level of the shelves, on which items are displayed, has an effect on sales and that the position of eye-level is the best.
For this study, the finding of Hwang et al. (2005), that eye-level is the best position for an SKU to be allocated, will be used for measuring the effect of location on product choice and reference price. When a product is more central on the shelf, consumers are more aware of this product and the probability of being chosen will be higher. Therefore, the following hypothesis will be tested:
H6: The more central (eye-level) an SKU is placed, the higher the possibility that this SKU will be chosen from the category.
Regarding reference prices, we expect that products which are more central located on the shelf will have a greater influence on the reference price than products which are less central located in a category. This is due to the fact that consumers are more aware of the prices of the central located products than the less central located products in a category. This leads to the following hypothesis:
H7: The more central (eye-level) an SKU is located, the higher the influence of this SKU on the consumer’s reference price of the category.
3.3.3 The effect of the atmosphere of the shelf on product choice and reference price
Retailers today face increasing competition in their market, stimulating them to focus on in-store merchandising (e.g., aisle and display placement decisions) and promotion (e.g., price and deal decisions) strategies to improve their shares of consumer purchases and wallets (Bolton, Shankar, and Montoya 2007; Kumar, Shah, and Venkatesan 2006).
One way of executing this is creating distinctive categories. Retailers must differentiate their categories from those of competing retailers through distinctive product assortments (Bolton et al., 2007). Retailers can provide large assortments with high levels of price inconsistency and less promotion, or by providing smaller assortments with less choice in price and products. Another way is developing a distinctive category presentation which distinguishes the category from that of other categories in the store.
In this study, a distinctive category presentation will be developed to measure whether or not there is an effect of this presentation on the product choice of a category or on the consumer’s reference price. The expectation is that when consumers are exposed to the presentation of this special category, which is very distinctive in comparison to other categories (within the store or across stores), and experience it, that this will influence their price perception of this particular category. For example, store A has a distinctive beer category in their store which is different than other categories and categories of other stores. When consumers are exposed to this special category they experience a unique situation. It is likely that consumers will take more time to see the products and choose a special beer product than the usual beer bottle product.
The following hypothesis is proposed:
Assuming that hypothesis H8 will hold, we expect that an increased reference price will lead to a higher probability of accepting higher priced products which leads to a higher purchase probability of the higher priced products in a category. Thus, the following hypothesis will be measured:
H9: A distinctive presentation of a category is expected to create a product shift in the product choice of consumers.
3.4 Moderator
When consumers choose a product from an assortment, they are driven by hedonic and utilitarian motives. Product categories differ in the extent to which their overall attitudes are derived as hedonic or utilitarian product categories.
Utilitarian products tend to fulfil the basic needs of consumers or help them accomplish functional or practical tasks (Strahilevitz and Myers, 1998). On the other hand, hedonic products are desired for pleasure, fantasy and fun.
According to the research of Dhar and Wertenbroch (2000) products and brands that have a high score on the hedonic dimension relative to the score on the utilitarian dimension are able to charge a price premium. Thus, when consumers want to purchase a hedonic product, they are less price sensitive than when they want to purchase a utilitarian product. This can be explained by the fact that when consumers purchase a utilitarian product they are looking at the quality/price relation of the product, whereas when consumers want to purchase a hedonic product they only consider the pleasure and quality of the product.
In this study the expectation is that the kind of product moderates the effect of shelf location and shelf space on the product choice of a category. Consumers buy utilitarian products because these products have a special function to solve a need, and thus are more goal-oriented in their product choice. When buying utilitarian products, consumers are less influenced by external factors like shelf allocation and location of the products on the shelf. Therefore, the following hypothesis will be tested:
H10: Utilitarian products negatively moderate the effect of the number of facings of an SKU on that of the product choice of that particular SKU.
And finally, in this study the expectation is that a distinctive category is positively related with the reference price what is stated in hypothesis 8. The influence of the moderator aspect is applied to this hypothesis and result in the final one of this study:
4.0 Research Design
The goal of this research is to test whether the independent variables (shelf location, shelf space and shelf design) have an influence on the product choice and the reference price of a product category.
In the previous chapters the literature and the hypotheses are developed and in the upcoming chapter the research design is described. First, the research method is presented which will be used in order to test the hypotheses. Second, the design of the questionnaire will be described. Then, the data collection method will be discussed. Finally, the plan of analysis is presented.
4.1 Research method
In order to answer the research question, an experimental study is performed to test the hypotheses. An experimental study is an appropriate method since experimental studies explore the potential causal relationships between claims and consumer psychological or behavioral responses (Xie and Boush, 2011). The experimental study is performed with use of a questionnaire.
The study consists of four questionnaires on which will be elaborated later on. In appendix A-D, you can find these questionnaires.
Product categories
In order to test the hypotheses four product categories are used in the study. Product categories can exist of utilitarian or hedonic products. In this study two utilitarian (butter and milk) and two hedonic (chocolate and chips) product categories will be used.
Each category consists of four brands in order to adapt the categories to the needs of the research. This entails for example changing the shelf location or shelf space of the products in the product category. In the following table the product categories and the associated products are depicted.
Table 1
Product categories of the research
Product
category Volume Most expensive brand
Premium
brand Store brand Less expensive brand Butter 500 grams Becel Blue Band AH butter Euroshopper
Milk 1 liter Campina Friesche Vlag AH milk Zaanse Hoeve
Chocolate 100 grams Verkade Milka AH chocolate Euroshopper
The chosen products have the same amount of volume, which result in a fair comparison of the different products in the product category. The chosen brands have the range of expensive brands to less expensive brands.
Manipulation check
In order to check whether the moderator conveys the correct product type (utilitarian or hedonic) the questionnaire contains questions about the feeling the product expresses to the respondent. Respondents were asked to rate the product categories on three characteristics of utilitarian and three characteristics of hedonic feelings. Generally utilitarian products score high on usefulness, necessary and rational, where hedonic products score high on pleasant, pleasurable and enjoyable. Using a seven-point Likert-scale, ranging from totally agree to totally disagree, tests this.
For example: To what extent conveys the product category butter the following feelings? The product is:
Useful
Totally agree 1 2 3 4 5 6 7 Totally disagree Pleasant
Totally agree 1 2 3 4 5 6 7 Totally disagree
4.2 Questionnaire design
In order to answer the problem statement the following experimental design is conducted. The study consists of four questionnaires (appendix A-D). The questionnaires consist of the same questions, but with different pictures of the manipulated shelves. Every respondent was confronted with one shelf of every product category, so in total every respondent filled in four cases for the research.
Each questionnaire starts with seven demographical questions. Examples are: ‘’what is your gender’’ and ‘’ of how many people exists your household? ‘’ These questions serve to receive more insights into the background of the respondents.
After the demographical questions the respondents are confronted with a particular situation in which they need to do their grocery shopping at the nearest supermarket.
After investigating the product choice and the reference price respondents are showed a picture of a shelf of a particular product category. In order to investigate whether this shelf has an impact on the product choice and reference price the following questions are asked: ‘’ Which product(s) will you considering? More answers are possible) ‘’ and ‘’ Which product will you choose? ‘’
The sequence of the questions regarding a particular product category varies by questionnaire but the questions are the same.
4.3 Data collection
The data for this study is collected using an online survey via Thesistools. Thesistools is an online program where you can develop and with which you can distribute questionnaires. In order to distribute the questionnaires, the links are spread primarily through the personal network of the researcher. By sending emails and using social networks, respondents were asked to fill in the questionnaire and send them to their personal network. In this way, a snowball effect is created. A sampling procedure may be defined as snowball sampling when the researcher accesses respondents through contact information that is provided by other respondents (Noy, 2008).
There is one condition to participate in the survey. To be an adequate respondent, the respondents have to do grocery shopping regularly. This in controlled by the question: ‘’ Do you regularly do grocery shopping’s?’’ In this way, the respondents who aren’t adequate can be easily deleted from the list of respondents.
4.4 Measurement
In order to get more insight in the way the independent and dependent variables will be tested, the following table provides a clear overview.
Table 2
Definition of the constructs of the study
Independent variables
Shelf space The shelf space can be 1 facing or 2 facings.
Shelf location The shelf location consists of height 1 (at the top), height 2 (in
the middle) and height 3 (at the bottom). Design The questionnaires 1 and 3 contain a shelf with a distinctive
design.
Product type The product types of this study are utilitarian (butter and milk) and hedonic (chocolate and chips)
Dependent variables
Product choice The product choice is defined as the product which is chosen by the respondent
Reference price The reference price will be measured as a combination of the products under consideration (x0,4) and the product choice (x0,6)
The reference price will be measured by combining two items; the first one is the average price of the products under consideration and the second item is the actual product choice of the respondents of a product category. Since the actual product choice is a clear, conscious choice which the respondent makes, this item is weighted higher in the calculation than the other item. This result in the following equation:
Reference price= ((average of products under consideration) x 0,4) + (product choice x 0,6)
4.5 Plan of analysis
The analyses of this study consist of the Cronbach’s alpha, logistic regression, multiple regression and One-Sample T-test. First, the Cronbach’s alpha will be used in order to check the reliability of the multi-items scales of the questions regarding the product type of a product category.
After conducting the Cronbach’s alpha the next analysis is to check whether there is multicollinearity between the variables which are used in the logistic regression (Field, 2009).
In order to test the hypotheses regarding the product choice of a SKU a logistic regression method is used. Logistic regression is multiple regression but with an outcome variable that is a categorical variable and predictor analysis variables that are continuous or categorical (Field, 2009). In this study it is known as binary logistic regression, because this study tries to predict membership of only two categorical outcomes (the product is chosen or is not chosen). The logistic regression model predicts the probability of an event occurring for a given person.
