• No results found

The Effect of Premium Promotions on Willingness to Pay

N/A
N/A
Protected

Academic year: 2021

Share "The Effect of Premium Promotions on Willingness to Pay"

Copied!
81
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

The Effect of Premium Promotions

on Willingness to Pay

Exploring the Value of Premium Customization

ROB JOKHORST June 27, 2013

(2)

The Effect of Premium Promotions on Willingness to Pay

Exploring the Value of Premium Customization

University of Groningen

Faculty of Economics and Business

Master thesis MSc Marketing

Marketing Intelligence

Author:

Rob Jokhorst

Raamstraat 8a

9711 CK Groningen

0610603774

Student number: 1534386

r.h.l.jokhorst@student.rug.nl

(3)

I Management summary

In an effort to measure the effects of premium promotions on consumers’ willingness to pay, this study uses choice based conjoint analysis to analyze the effects of specific types of premium promotions on the utility that consumers derive from these premium promotions. This study finds that the effect of premium promotions on willingness to pay is heterogeneous across consumers, by distinguishing a number of segments that are differently influenced by the various types of premium promotions tested in this study. The results of this study furthermore indicate that the effect of premium promotions on willingness to pay is moderated by the specific characteristics of the premium.

(4)

II Preface

After seven years of studying (I did a Bachelor in Marketing at a University of Applied Sciences next to my BSc. at the RUG), this thesis is the final ‘assignment’ of my Marketing Intelligence Master. At the same time as the thesis process started, I also started with my Master in Education at the RUG. The highly time-consuming educational internship (I taught Economics at a high school in Hoogezand) often made it a real challenge to make the intermediary deadlines, which are an inherent part of the new thesis program. While having successfully completed the first part of the Educational Master, I however decided not to continue this path; I am very enthusiastic about the contents of the Marketing Intelligence track and would like to experience this field of study in practice.

(5)

Table of Contents

I Management summary ... 2

II Preface ... 3

1. Introduction ... 7

1.1 Background of the study ... 7

1.1.1 Practical background ... 7

1.1.2 Theoretical background ... 2

1.2 Focus of the study ... 3

1.2.1 Problem identification ... 3

1.2.2 Problem statement ... 3

1.2.3 Research questions... 4

1.3 Relevance of the study ... 5

1.3.1 Theoretical relevance ... 5

1.3.2 Practical relevance ... 5

1.4 Structure of the study ... 5

2. Theoretical framework ... 6

2.1 Sales promotion ... 6

2.1.1 The negative effects of price discounts ... 7

2.1.2 Premium promotion ... 8

2.1.3 The main discrepancies between the effects of price discounts and free gifts ... 10

2.2 Determinants of premium promotion effectiveness ... 12

2.2.1 Promoted product-related determinants ... 12

2.2.2 Premium-related determinants... 13

2.2.3 Determinants related to consumer’s traits and perceptions ... 14

2.2.4 Interactions between determinants of premium promotion effectiveness ... 15

3 Hypothesis development ... 16

3.1 Research hypotheses ... 16

3.1.1 The effect of premium promotion on willingness to pay ... 16

3.1.2 The effect of premium customization on WTP ... 17

3.1.3 The effect of choice extensiveness... 18

3.2 Conceptual model ... 19

4. Research methodology ... 20

(6)

4.1.1 Research setting ... 20 4.1.2 Product category ... 20 4.1.3 Premiums ... 20 4.2 Research design ... 21 4.2.1 Conjoint design ... 22 4.2.2 Validity measures ... 29 4.2.3 Respondent sampling ... 30 4.2.4 Plan of analysis ... 30

5 Implications of the pretests for the CBC design ... 31

6. Main study results ... 32

6.1. Covariate analysis ... 32

6.1.1 Demographics ... 33

6.1.2 Brand and Premium Preference ... 33

6.1.3 Purchase behavior ... 33

6.2 Extreme response behavior... 34

6.3 Model selection ... 34

6.3.1 The aggregate model ... 35

6.3.2 The latent class model ... 36

6.4 Description of the segments ... 38

6.5 WTP estimates ... 40

6.6 Hypothesis testing ... 40

6.6.1. The effect of premium promotion on incremental utility for consumers ... 41

6.6.2 The effect of premium customization on incremental utility for consumers ... 42

6.6.3 The effect of customization extensiveness on incremental utility for consumers ... 43

6.7 Respondent representativeness and model validity ... 44

6.7.1 Representativeness of the respondent sample... 44

6.7.2. Model validity ... 44

6.7.3 Robustness of the results ... 45

6.8 Exploration of premium preference ... 46

6.8.1 Drivers of the preference for a specific premium ... 46

6.8.2 Drivers of the preference for a premium as opposed to no premium... 47

6.8.3 Influence of premium preference on utility derived from the premium promotion ... 48

7. Discussion ... 49

References ... 51

(7)

Appendix 1: Price Sensitivity Meter ... 58

Appendix 2: Holdout choice sets ... 59

Appendix 3: Pretest design ... 60

A.3.1 Pretest 1 ... 60

A 3.2 Pretest 2 ... 61

Appendix 4: Pretest results ... 64

A 4.1 Category familiarity and promoted product nature ... 64

A 4.2. Premium attractiveness, product nature, and fit ... 65

A 4.3 WTP estimates ... 65

Appendix 5: Premiums included in the study... 69

Appendix 6: Covariate summary analysis sample ... 71

Appendix 7: Explored latent class models ... 72

(8)

1. Introduction

1.1

Background of the study

1.1.1 Practical background

Most Dutch adults are familiar with the ‘Flippo’ promotion campaign, as executed in 1995 by Smiths1, a subsidiary of Frito-Lay. The sales promotion, in the form of a premium promotion, led to a sales increase of more than 25% and a market share increase of 11%2 for the company. It furthermore led to a 10% decline in sales of Smiths’ largest competitor, Croky3. More than 15 years later, premium promotions, defined here as “a product or a service offered free or at a relatively low price in return for the purchase of one or many products or services” (d’Astous & Jacob, 2002, p. 1746), are still on the rise according to Joop Holla, research director at GfK Panel Services Benelux. Holla argues that during a premium promotion campaign, supermarket sales are estimated to increase by 5%, which for market leader Albert Heijn implies a weekly sales increase of three million euro4.

In addition to premium promotions that are organized on the (supermarket) chain level, many FMCG5 brands have deployed premium promotions in an attempt to entice consumers; in the Dutch FMCG sector, brands such as Coca Cola, Heineken, and Unox have been involved in various forms of premium promotion. Presumably due to the relatively low prices of many FMCG products, premium promotions in this retail industry often take the form of a points or credits-based reward system. In the more expensive supermarket categories however, free gifts have been offered as well. A crate of beer of a premium brand such as ‘Heineken’6 has been promoted in the Netherlands with a free beer glass. Recently7, Heineken Netherlands offered a free plastic ice bucket with every crate of its freshly introduced Star Bottle beer8. The beer category is especially interesting due to the high promotion pressure in Dutch supermarkets; for various premium beer brands, more than two out of every three crates of beer are sold when on promotion, often even against a loss9.

1

Now Lays

2

From 51 to 62%, ondernemingsdatabank.indicator.be/verkoopidee/de_flippos_achterna__/VLTAVKAR_EU010603, last accessed at 21.05.2013

3

www.volkskrant.nl/vk/nl/2844/Archief/archief/article/detail/395421/1995/07/06/Delen-in-het-flippo-succes-kan-alleen-voor-veel-geld.dhtml, last accessed at 22.05.2013

4

www.rnw.nl/nederlands/article/spaaracties-supermarkten-kinderen-bepalen-het-succes, last accessed at 22.05.2013

5

Fast Moving Consumer Goods

6

http://www.biernet.nl/nieuws/gratis-ellipse-heineken-glazen-bij-een-krat-heineken, last accessed at: 29.05.2013

7

During April and May 2013

8

http://www.heineken.com/nl/star-bottle/ice-bucket.aspx, last accessed at 24.05.2013

9

(9)

2 More popular than premium promotions are price discounts. At Appie.nl, the online retail department of Albert Heijn, a consumer can find a discounted premium beer brand almost every week, with the discount amounting to approximately 25%10. The price discounts on premium beer are often even larger than 30% in many Dutch supermarkets11. It is therefore questionable to what extent retailers are pursuing optimal price promotions (in terms of profits); the beer category is used by many retailers as a traffic builder12, which is likely to lead to the erosion of prices13.

1.1.2 Theoretical background

While appreciating the promising sales potential that price discounts offer (price discount elasticities often exceed -2 (Bijmolt et al., 2005)), negative consequences of price promotions such as increased price sensitivity (Ortmeyer et al., 1991) and decreased reference prices (Mela et al., 1997) have long been acknowledged by scholars. These drawbacks of price-discounts have spurred empirical work on EDLP14 and premium promotion strategies (Darke & Chung, 2005). Whereas EDLP strategies suffer from a number of similar drawbacks as price promotions (such as negative quality inferences), premium promotions offer a number of benefits, some of which are conservation of perceived product quality and reinforcement of brand equity (e.g. Darke & Chung (2005) and Palazo & Delgado

(2005)).

The focus of the premium promotion research has been on two specific forms of premium promotions: Cause-related marketing (e.g. Koschate-Fischer et al., 2012) and free gift offerings (e.g.

Buil et al., 2013). Cause-related marketing is defined here as “an offer from the firm to contribute a

specified amount to a designated cause when customers engage in revenue-providing exchanges that satisfy organizational and individual objectives” (Varadarajan & Menon 1988, p. 60). Empirical Cause-related marketing research shows that this form of premium promotion can potentially exert a stronger influence on sales than price discounts (i.e. a 20 cents charity donation can have a larger effect on sales than a 20 cents price discount (Strahilevitz (1999) & Arora & Henderson (2007)). Due to the high levels of heterogeneity between premium characteristics, such as the nature of the product and retail prices, premiums are however more difficult to juxtapose with price discounts (as compared to charity donations). Furthermore, due to the nonmonetary nature of a premium, it is often unclear what the monetary costs are for the manufacturer or retailer. As a consequence, the focus of the premium promotion literature has been on how consumers’ reactions to premium-based

10

ah.nl/appie/zoeken?rq=bier

11

bier.studentenkorting.nl & www.biernet.nl/bier/aanbiedingen/kratten

12

A product or category for which people prefer a certain supermarkt (also known as destination category)

13

symphonyiri.nl/portals/0/articlePdfs/Foodmagazine%20Categorie%20Dranken%20april%202010.pdf

14

(10)

3 promotional offersare influenced by attitudinal variables such as product interest, premium interest, and brand attitude (e.g. d’Astous & Jacob, 2002 & Buil et al., 2013).

1.2 Focus of the study

1.2.1 Problem identification

Scholars and marketers seem to agree that price promotions are likely to exert detrimental effects on sales (Darke & Chung, 2005) and brand-based equity (Keller, 2013). However, the high level of ambiguity regarding the effects of (highly heterogeneous) premium promotions and marketers’ familiarity with price discounts are likely to inhibit a shift from price to premium promotions. Especially in the era of marketing accountability (e.g. Rust et al., 2004), the ambiguity surrounding premium promotions is likely to instill hesitation regarding premium promotions in the minds of marketers. The effect of premium promotion is highly dependent on the specific premiums and their respective retail prices, providing a real challenge for both marketers and scholars to draw inferences from the outcomes of premium promotions.

The surge in online shopping over the last decade (Rigby, 2011) has however expanded the toolbox of marketers and scholars interested in the effects of a firm’s promotional activities; the increasing availability of data covering online shopping behavior has opened up the possibility to better determine the effects of promotions in general (Leeflang et al., 2002). Furthermore, online retailing allows for increased customization of a firm’s offerings (Blattberg et al., 2009).

Since most consumers prefer more over less choice, customization should lead to incremental utility for consumers. Incremental utility for consumers should consequently lead to an increased willingness to pay (WTP) (Gensler et al., 2012). WTP is defined here as the price at which a consumer is indifferent between purchasing and not purchasing (Moorthy et al. (1997) and Gensler et

al.(2012)). Empirical evidence from the field of Cause-related marketing suggests that offering

consumers a choice with regard to the selection of a charity15, increases the incremental utility of a premium promotion for consumers (Arora & Henderson, 2007). The value of a customization strategy has however not been explored in the premium promotion literature, leaving marketers in the dark with regard to the value of a customization strategy in a premium promotion context.

1.2.2 Problem statement

The central theme of this study is therefore the investigation of the incremental value derived from providing consumers with a choice regarding the selection of the premium that is associated with the

15

(11)

4 promoted product (a strategy referred to as ‘customization’ by Arora & Henderson (2007)). Whereas the incremental utility is a vague term and due to its latent nature by definition a highly abstract concept, WTP is a more intuitively appealing measure of the effect of a premium promotion. This study aims to kill two birds with one stone by indirectly measuring WTP through the estimation of the incremental utility of premium promotions for consumers. As Gensler (2012) shows, incremental utility for a consumer should in general logically result in a higher WTP. The problem statement can thus be formulated as follows:

To what extent does premium customization increase a consumer’s willingness to pay?

The following section will present the research questions accompanying the problem statement

1.2.3 Research questions

Elaborating on the lack of evidence regarding the effect of premium promotion on the incremental utility for consumers, this section presents a number of research questions.

First, it should be tested whether indeed a premium promotion results in incremental utility for a consumer. The first research question is thus stated as follows:

1. To what extent does a premium promotion result in increased utility for consumers and consequently WTP?

Second, in line with the problem statement, it is of utmost interest for this study to explore the effect of a customization strategy in a premium promotion context. The second research question is therefore stated as:

2. Does a customization strategy lead to a higher WTP as compared to a preselected premium?

Finally, the question of choice extensiveness (i.e. how many premiums should a manufacturer or retailer offer?) has remained unanswered in both the Cause-related marketing and premium promotion literature. Whereas Arora & Henderson (2007) represent a customization strategy as a choice between five charity donations, one could also argue that the provision of only a single additional option can already be termed a customization strategy. More importantly, a marketer concerned with premium promotions would want to how many premiums to offer. The fourth research question is thus formulated as follows:

(12)

5

1.3 Relevance of the study

1.3.1 Theoretical relevance

As briefly discussed in section 1.1.2, the literature on premium promotion focuses on the attitudinal variables influencing the effectiveness of the promotion. Although a call for premium heterogenization (i.e. targeting different segments) has long been made (Simonson et al., 1994), I am not aware of any published empirical work that has addressed this issue within the field of premium promotions. Empirical evidence of a positive effect of premium customization on premium promotion effectiveness will contribute to the existing literature, by unveiling an additional determinant of premium promotion effectiveness. Finally, whereas the ‘depth’ dimension of premium promotion customization (i.e. offering consumers a choice between various promotions of the same type) has been researched (and found to exist) in a Cause-related Marketing setting, the ‘breadth’ dimension of premium promotion customization (i.e. offering consumers a choice between different types of premium promotions) has not been studied in the literature. This thesis attempts to make a first step in the empirical validation of a premium customization strategy.

1.3.2 Practical relevance

If convincing evidence is found that premium customization results in incremental utility for consumers, the message to managers involved in premium promotions will be straightforward: provide consumers with a choice regarding the premium. This implication will have special relevance for online retailers; the online retail environment allows for much more flexibility with regard to the implementation and execution of a customization strategy, as compared to a brick-and-mortar retail environment. One could for example think of the advantages of non-physical shopping environment (e.g. for Albert Heijn, premiums do not have be in stock in all of its local supermarkets, but can be centrally stored if sold through its webstore Appie.nl). Furthermore, for organizations (online retailers in specific) planning to reap the benefits of effective premium promotion, providing the consumer with increased choice regarding the premium is likely to lead to increased WTP.

1.4 Structure of the study

(13)

6

2. Theoretical framework

In this chapter, sales promotions will be briefly introduced. More specifically, this chapter will elaborate on price discounts (as the most popular form of monetary sales promotions) and free gifts (as the most popular form of nonmonetary sales promotions). Whereas section 2.1 elaborates on the disadvantages of price promotions and the benefits of premium promotion, section 2.2 provides an overview of the literature with regard to the determinants of premium promotion effectiveness.

2.1 Sales promotion

In line with Huff (1999), sales promotions are defined here as ‘promotional tools designed to achieve short-term objectives’, and can be subdivided into two main categories; price promotions and non-price promotions (Gedenk et al., 2006). Non-price promotions can be further subdivided into ‘supportive’ and ‘true’ price promotions (figure 1). Of the price promotions, price discounts are most often employed by marketers (Darke & Chung, 2005). The remainder of this section will elaborate on the drawbacks of price discounts and present premiums as an alternative form of promotion that is immune to some of the drawbacks inherent to price discounts.

Figure 1: A sales promotions taxonomy16

Source: adapted from Gedenk et al., 2006

16

(14)

7

2.1.1 The negative effects of price discounts

The main focus of the sales promotion literature concerns monetary promotions (Darke & Chung, 2005). Apart from the ease of implementation, price promotions have been found to provide some distinct benefits to marketers (e.g. Blattberg et al., 1995). However, especially during the last decade, monetary promotions have been criticized for their negative long-run implications (e.g. Chandon et

al. (2000), Srinivasan et al. (2004), and Darke & Chung, (2005)).

2.1.1.1 Long-term disadvantages of price discounts

The negative long-term effects of price discounts that are expected to have the strongest negative influence on consumers and are hence most prominent in the literature, are: increased price sensitivity, deterioration of brand equity, decreased reference price, and negative-price-quality inferences.

Mela et al. (1997) find that sales promotions increase price sensitivity in the long run; the nonloyal,

price-sensitive consumer segment has increased over time, partially as a consequence of price promotions. These authors furthermore find a negative influence of price promotions on brand loyalty, being partially a consequence of increased price sensitivity. Price promotions can even destroy market share, when used by hedonic high-equity brands which are competing against low priced brands (Chandon et al., 2000). Ortmeyer et al. (1991) find that price promotions decrease the reference price of consumers: instead of using the ‘real’ price as a mental cue for the purchase decision, consumers often use the discounted price when making the decision whether or not to purchase a product. Finally, price promotions often have a negative effect on price-quality inferences; consumers often use the lower (discounted) price to make inferences about the quality of the discounted product (Rao & Monroe (1989), Ortmeyer et al. (1991), Darke & Chung (2005)). A potential underlying mechanism of the negative price-quality inferences is the price-quality heuristic, which often (negatively) moderates the relationship between the price discount and respectively deal value and price perception, outweighing the positive effects of price discounts and thereby negatively influencing the acquisition and transaction utility (Darke & Chung, 2005). More specifically, the negative effect of price discounts on acquisition utility is generated by the consumer’s belief that for the lower price she is getting a lower-quality product, instead of an average-quality product. Negative price-quality inferences are most likely to occur when brand-equity is low (Rao & Monroe, 1989).

2.1.1.2 Moving away from price promotions?

Chandon et al. (2000) argue that both the academic and business field have long been aware of the

(15)

8 promotions, regardless of the empirical evidence stressing its harmful effects, might be a result of a myopic management approach of surviving in a price competition. Darke & Chung (2005) show in an experiment that lowering prices, as part of the EDLP formula or temporarily lowering prices as part of a price promotion, is very likely to lead to negative quality inferences by consumers with regard to the promoted product. Consequently, the authors advise marketers to rather use free gifts instead of price discounts, since free gifts are more likely to lead to positive quality inferences and therefore increased deal value, as perceived by consumers (Darke & Chung, 2005).

2.1.2 Premium promotion

As a result of the increased awareness of the negative long-term consequences of price promotions, scholars have been increasingly exploring the effects of non-price promotions, with a focus on premium promotions. As already mentioned in section 1.1.1, a premium is defined here as “a product or a service offered free or at a relatively low price in return for the purchase of one or many products or services” (d’Astous & Jacob, 2002 p.1746)17. The main focus of the premium promotion literature is however on premiums that are offered for free, as opposed to premiums that are only offered at a (relatively low) additional fee. This last type of premium is referred to as a ‘purchase-with-purchase’ (Raghubir, 2004). The effects of free gifts and purchase-with-purchase premiums are however not the same (Simonson et el., 1994). Therefore, the following section will discuss only free gift promotion. Section 3.1.1 will however briefly discuss the limited research on purchase-with-purchase premiums, as part of the hypothesis development.

2.1.2.1 Benefits of free gifts

The effects of price promotions that are expected to have the largest negative influence and are hence most discussed in the literature are: conservation of product quality and reinforcement of brand equity.

Darke & Chung (2005) find that consumers are more likely to have equal or increased perceptions of

the quality of a product when it is combined with a free gift offer, as opposed to a price discount. More specifically, the authors refer to the concept of attribute framing as the underlying (psychological) mechanism of the effects of free gifts on quality inferences. Darke & Chung (2005) define attribute framing as “superficial variations in the description of a single aspect of the offer [which] leads consumers to make either additional positive or negative associations about the product, which then influence the overall valence of further product evaluations” (p. 37).

17

(16)

9 The authors find that the aforementioned difference in promotion framing (as a premium, instead of a price discount of the same value) results in consumers not correcting for the value of the promotion premium. Consequently, negative inferences with regard to the quality of the base product are inhibited. In the experiments undertaken by Darke & Chung (2005), consumers used a product’s full price to base their quality judgments on. The monetary value of the gift was thus not taken into account, which implies that consumers maintained perceptions of quality which consequently increased perceptions of deal value. An illustrative example of this is provided by the others: “if a product priced at $59.99 was offered with a free gift worth $20, quality inferences should be based on the full price of $59.99 rather than the corrected price of $39.99. According to the attribute framing perspective, this is because the full price of the item is more salient to consumers than the price that corrects for the value of the gift” (Darke & Chung (2005), p.37). Whereas the free gift increased deal value, the price discount did not. It is however worth noting that the deal value is partly a function of the attractiveness of the gift; Simonson et al. (1994) find that free gifts that are perceived as unattractive have a negative effect on purchase likelihood. d’Astous &

Landreville (2003) partially support this claim; they find that an unattractive premium can have a

negative effect on consumers’ evaluation of the promotion. However, if there exists a good (perceived) fit between the base product and the free gift, the unattractive free gift nevertheless leads to a positive evaluation of the promotion.

In a study on the differential effects of monetary and non-monetary promotions on consumer-based brand equity, Palazo & Delgado (2005) show that non-monetary promotions are likely to positively influence brand equity (whereas monetary promotions are not). In line with the conclusions of Palazo

& Delgado (2005), Montaner & Pina (2011) find in more recent research that, whereas the frequent

use of price promotions has a negative effect on brand equity, non-monetary promotions such as free gifts result in increased brand equity.

2.1.2.2 Risks of premium promotion

Although many authors (e.g. Chandon et al. (2000), Darke & Chung (2005), Arora & Henderson

(2007), and Montaner et al. (2011)) argue that in most situations marketers should favor premium

promotions over price discounts, scholars have also pointed out potential drawbacks of premium promotions. This section elaborates on the two risks related to free gift promotion most referred to in the literature: value-discounting and low gift attractiveness.

(17)

10 free gift in a premium promotion, consumers will decrease their evaluation of this product, leading to both lower purchase intentions and willingness to pay.

A second risk of a free gift promotion is that such an offer has been found to lead to a decreased (brand) choice probability by consumers when the premium is perceived to be unattractive or unneeded (Simonson et al., 1994). More specifically, these authors find that consumers often need a justification for buying the product-premium combination, since they believe that their product choice is evaluated by others. Linking the promoted product to an unattractive premium thus leads to justification difficulties and hence a negative evaluation of the promoted product, as compared to the same product without the premium.

2.1.3 The main discrepancies between the effects of price discounts and free gifts

As discussed in section 2.1.1.1, price promotions are likely to lead to undesired results in the long run. Many authors argue that most of these long-run implications are likely to be avoided by the use of premium promotions (as opposed to price promotions), with as additional benefit that the inclusion of a premium has the potential to increase consumer’s evaluation of the product. Promotion effectiveness especially differs between both forms of sales promotion, when one considers:

(1) Nonmonetary factors (2) The nature of the product Nonmonetary factors

Whereas price discounts are found to decrease the promoted product’s reference price and deal value, free gifts are found to maintain the reference price for the promoted product, while at the same time increasing deal value (Darke & Chung, 2005). The main mechanism underlying this effect is the consumers’ quality inferences about the promoted product; Darke & Chung (2005) demonstrate that in the case of free gifts, consumers based their quality inferences on the original price of the promoted product, rather than the original price minus the estimated value of the free gift. Premium promotions are furthermore more likely than monetary promotions to lead to positive brand associations (Chandon et al. (2000)), which form the core of customer based brand equity

(Keller, 2013).

The nature of the product

(18)

11 products. Chandon et al. (2000) furthermore explored the interaction between promotion type and product nature and find that monetary promotions prove to be more effective when linked to a utilitarian product than when combined with a hedonic product. More specifically, as figure 2 illustrates, price discounts can potentially decrease market share when associated with a hedonic product. Whereas the nature of the promoted product (utilitarian or hedonic) is of great importance with regard to the effectiveness of price discounts, the nature of the promoted product has been found to play an even larger role in the context of premium promotions; in the case of a premium promotion, consumers’ perceptions of the promotion are not limited to the promoted product, but to the (preferably hedonic) nature of the free gift as well (Chandon et al., 2000). The authors find that free gifts are generally perceived as providing relatively high hedonic, but low utilitarian benefits. Price reductions, on the other hand, are found to provide relatively low hedonic benefits, but high utilitarian benefits. A more thorough discussion on the role of product nature is provided in section 2.2.1

Figure 2: The effects of benefit congruency for high- and low-equity brands

(19)

12 discussed in Chapter 4. The factors that are found to influence the premium promotion effectiveness will have to be held constant or incorporated in the experimental design in order to increase the validity of the study.

2.2 Determinants of premium promotion effectiveness

This section discusses the determinants of promotion effectiveness related to the promoted product (section 2.2.1), the premium (section 2.2.2), Consumer traits and perceptions (section 2.2.3), and the interactions between promoted product and premium (section 2.2.4). Table 1 provides an overview of the main determinants of premium promotion effectiveness (where italics indicates that the authors find no effects of the specific determinant on premium promotion effectiveness). The determinants ‘brand attitude’ and ‘overall evaluation’ will only be very briefly discussed, since these are not incorporated in the study design.

Table 1: Determinants of premium promotion effectiveness

italics= authors find no significant effect

2.2.1 Promoted product-related determinants

The two most discussed determinants of premium promotion effectiveness that are related to the promoted product are the ‘nature of the promoted product’ and the ‘brand attitude’.

Determinants of premium promotion effectiveness

Section Related to Source

Nature of the promoted product

2.2.1 Promoted product Chandon et al. (2000), Zheng & Kivetz (2009) vs

Kwok & Uncles (2005); Palazon and Delgado (2005), Buil et al. (2013)

Brand attitude Promoted product d’Astous and Jacob (2002), Buil et al. (2013)

Nature of the gift 2.2.2 Premium Palazon (2009), Buil et al. (2013)

Attractiveness of the gift Premium d’Astous and Jacob (2002), d’Astous and Landreville (2003), Buil et al. (2013)

Deal proneness 2.2.3 Consumer’s traits and perceptions Lichtenstein et al. (1995), Chandon et al. (2000), Delvecchio (2005), Buil et al. (2013)

Overall evaluation Consumer’s traits and perceptions Vaidyanathan et al. (2000), Laroche et al.(2003), Buil et al., 2013

Perceived fit / benefit congruency

(20)

13 Product nature of the promoted product

Chandon et al. (2000) classify sales promotions as utilitarian when it helps consumers maximize

“utility, efficiency, and economy of their shopping and buying” and as hedonic when it provides “intrinsic stimulation, fun, and self-esteem” (p.67). Using principal component analysis, they further classify these sales promotion benefits by relating them either to their utilitarian or hedonic dimension and subsequently find that (monetary) savings, quality18, and convenience19 are perceived by consumers’ as utilitarian benefits, whereas the value expression20, entertainment21, and exploration22 benefits are perceived by consumers as hedonic benefits. The authors find that the effectiveness of a sales promotion is influenced by the nature (utilitarian or hedonic) of the promoted product. More specifically, they find that a hedonic product nature positively influences the effectiveness of the premium promotion. These findings are supported by more recent work of

Zheng & Kivetz (2009), who find that “consumers find it more difficult to justify spending money on

hedonic luxuries for themselves and therefore have a stronger need for and are more sensitive to external justifications such as promotions” (p. 565). Other authors however fail to find a different effect of hedonic products on premium promotion effectiveness, as compared to premium products

(Kwok & Uncles (2005); Palazon and Delgado (2005), Buil et al. (2013)). Hence, the effect of product

nature on premium promotion effectiveness is highly ambiguous. Brand attitude

In line with Wilkie (1990), brand attitude is defined here as ‘a consumer’s overall evaluation of a brand’. d’Astous and Jacob (2002) find that brand attitude positively correlates with the evaluation of the premium promotion. Building on these findings, Buil et al. (2013) find that premium promotion is more effective for favorably evaluated brands.

2.2.2 Premium-related determinants

The two most discussed determinants of premium promotion effectiveness related to the premium are the ‘nature of the gift’ and the ‘attractiveness of the gift’.

Nature of the gift

Whereas the evidence regarding the promoted product nature is ambiguous, research on the nature of the gift learns that hedonic gifts are more effective in a premium promotion setting, as compared

18

Defined by Chandon et al. (2000) as the ‘increase in the quality of the product bought’

19

Defined by Chandon et al. (2000) as the ‘reduction in search and decision costs’

20

Defined by Chandon et al. (2000) as expression and enhancement of self-concept and personal values

21

Defined by Chandon et al. (2000) as ‘amusement and esthetic value’

22

(21)

14 to utilitarian gifts; Palazon (2009) and Buil et al. (2013)find that consumers evaluate the gift more positively when the gift is perceived as more hedonic (as compared to more utilitarian).

Attractiveness of the gift

Simonson et al. (1994) and d’Astous & Jacob (2002) find that consumers tend to question the

underlying motivations of the free gift promotion, but that they do less so when they find the gift more attractive. d’Astous and Landreville (2003) find that a premium promotion including an attractive premium is better evaluated by consumers than one that includes a less attractive premium. Finally, Buil et al. (2013) also find that consumers’ evaluations of the gifts are more favorable when the gift is perceived as more attractive.

2.2.3 Determinants related to consumer’s traits and perceptions

The two most discussed determinants of premium promotion effectiveness that are related to consumers’ traits and perceptions are ‘deal proneness’ and ‘overall evaluation of the promotion’. Deal proneness

Deal proneness is defined here as “a general proneness to respond to promotions predominantly because they are in deal form” (Lichtenstein et al., 1990, p. 55). Chandon et al. (2000) find that a consumer’s deal proneness positively influences her perceptions of the promotion. Consequently, deal prone consumers are more likely to respond to premium promotions. Delvecchio (2005) finds that these deal-prone consumers derive more benefits from the promotion, compared to the ‘deal-neutral’ consumer. As a consequence, deal proneness influences purchase intentions (Buil et al.

2013) and are found to have a higher likelihood of purchasing the promoted product than the ‘deal

neutral’ consumer.

Overall evaluation of the promotion

The overall evaluation23 or attitude that consumers have toward an offering is strongly correlated to behavior or intentions (Buil et al., 2013). More specifically, Vaidyanathan et al. (2000) find that a consumer’s attitude toward promotions strongly influences her willingness to buy. In a similar vein,

Laroche et al. (2003) find that purchase intentions towards promotions increase when consumers

feel more positive about them. In more recent research, Buil et al. (2013) alsoshow that a positive overall evaluation of a gift promotion results in higher purchase intentions.

23

(22)

15

2.2.4 Interactions between determinants of premium promotion effectiveness

This section elaborates on the interactions between the promoted product and the premium. Perceived fit/benefit congruency

Due to the presence of two products included in a premium promotion (the promoted product and the premium), the nature of both the promoted product and the premium play an important role with regard to consumers’ evaluation of the premium promotion. Benefit-congruency, sometimes simply referred to as fit24, boils down to whether a utilitarian product is combined with a utilitarian gift (good fit) or with a hedonic gift (bad fit) and vice versa. The underlying reasoning here is that the promoted product and the free gift should provide benefits of a similar nature. Chandon et al. (2000) find that premium promotions are more effective when the free gift provides benefits that are congruent with the benefits provided by the promoted products. It must however be noted that these benefit-congruency effects are only found for high-quality brands. d’Astous & Landreville

(2003) find that a premium promotion including a premium that has a strong link with the promoted

product is better evaluated by consumers than a premium promotion that is perceived to have a bad fit between the promoted product and the premium. Montaner et al. (2011) find that only for high-equity brands, the nature of the promoted product (disordinally) interacts with the nature of the gift. For medium-equity brands, promoted product nature and premium nature do not interact. Hence, for high-equity brands, consumer evaluation of the promotion can be increased when the promoted product is combined with a hedonic gift. For medium equity brands, gifts of a hedonic nature lead to higher consumer evaluation than gifts of a utilitarian nature, regardless of promoted product-gift congruency.

24

(23)

16

3 Hypothesis development

3.1 Research hypotheses

This section consists of three sub-sections, where each section presents a hypothesis directly related to a specific research question (formulated in section 1.2.3 (and repeated hereunder):

1. To what extent does a premium promotion result in increased utility for consumers and consequently WTP?

2. Does a customization strategy lead to a higher WTP as compared to a preselected premium? 3. Does increasing the extent of customization result in a higher WTP?

3.1.1 The effect of premium promotion on willingness to pay

Thus far, no published research has investigated the effect of premium promotions on WTP. However, in the related field of cause-related marketing, Koschate-Fischer et al. (2012) find that a charity donation positively influences WTP. Arora & Henderson (2007) find that charity donations25 increase the effectiveness of the promotion, measured as the incremental utility for consumers26. As mentioned in section 1.2.1 and mathematically explicated in section 4.1.2.3, incremental utility should logically result in a higher WTP (Gensler et al. 2012). Since WTP is thus not directly measured, the hypotheses apply to the increase in utility derived from a premium promotion. In the same vein as charity donations lead to incremental utility and WTP, premium promotions are expected to increase a consumer’s willingness to pay. Thus:

Hypothesis 1: The presence of a premium promotion results in incremental utility for consumers

A distinction however has to be made between the effect of free gifts and a purchase-with-purchase premium (i.e. the option to purchase a premium, often against a reduced price); whereas free gifts have been found to be likely to have a positive effect on consumers (e.g. d’Astous & Jacob (2002) and

Darke & Chung (2005), purchase-with-purchase premiums have been much less researched, but

found to be likely to decrease perceived deal value (Simonson et al., 1994). Thus, whereas it is expected that a free gift will increase utility for consumers, the effect of a purchase-with-purchase premium is expected to be negative, leading to the following sub-hypotheses:

Hypothesis 1a: Combining the promoted product with a free gift results in incremental utility for consumers

Hypothesis 1b: Combining the promoted product with a ‘purchase-with-purchase’ option results in a loss of utility for consumers

25

Arora & Henderson (2007) refer to these charity donations as ‘embedded premiums’

26

(24)

17

3.1.2 The effect of premium customization on WTP

Arora & Henderson (2007) find that customized charity donations are a more effective form of sales

promotions than a preselected charity donation. The authors find the effect of the charity donations to be highly heterogeneous; consumers differ in their sensitivity towards charity donations based on their demographics, personal motivations and their affinity towards the charity that is linked to the promoted product. Heterogeneous effects of charity donations on consumers are also found by

Bendapudi et al. (1996) and Winterich & Barone (2011). Simonsen et al. (1994) find that the market

for free gifts can be divided into two segments: the first is made up by consumers who find the gift attractive and hence have a greater likelihood of buying the product, whereas the second segment represents those who do not find the gift attractive (and consequently are less likely to buy).

Simonsen et al. (1994) further argue that a large, third segment might exist: consumers who are

indifferent with regard to the promotion or simply do not consider it in their purchase decision. The effectiveness of a promotion consequently depends on both the size of the segments and their respective positive (negative) effects on buying behavior. Apart from the need for marketers to map the effects of the gift promotion on different (segments) of consumers, Simonson addresses the importance of targeting the gift promotion only to those (segments of) consumers who value the gift, while avoiding exposure of the other segments to these products.

Thus, based on the findings of Arora & Henderson (2007) and in line with the expectations of

Simonson et al. (1994), the effect of premium promotions on consumers is expected to be highly

heterogeneous. Similar to the strength of consumer affinity for the charity donation (being linked to the promoted product) playing an important role in the determination of effectiveness of the charity donation (Arora & Henderson, 2007), the strength of consumer preference for the premium promotion should be an important determinant of premium promotion effectiveness and consequently a consumer’s WTP. Providing a consumer a choice between different premiums should thus lead to a better fit between the premium promotion and individual consumer preferences, leading to the following hypothesis:

Hypothesis 2: Compared to a fixed premium, premium customization results in higher incremental utility for consumers

As Arora & Henderson (2007) find in a Cause-related marketing context, offering the consumer a

(25)

18

Hypothesis 2a: Offering a depth premium customization option results in higher incremental utility for

consumers than a preselected free gift

Whereas ‘depth’ customization has been researched in a cause-related marketing context, a ‘breadth customization’ (i.e. providing consumers with the option to choose between different types of premiums) has not been reported in the literature. The expected outcome is therefore highly ambiguous. However, in line with the expectations for hypothesis 1b (a purchase-with-purchase premium is hypothesizes to decrease utility for consumers), the effect of adding a purchase-with-purchase option to a preselected free gift is expected to decrease the utility for consumers. Thus:

Hypothesis 2b: Offering a premium customization option of a different type (i.e. offering both a free gift and a purchase-with-purchase option simultaneously), results in lower incremental utility for consumers than a

preselected free gift

3.1.3 The effect of choice extensiveness

The rational choice assumption underlying classic economic theory is that more choice is always better than less choice. However, when choice making becomes more complex, consumers are more likely to use heuristics in order to simplify the decision making process (Iyengar & Lepper (2000)). These authors reason that although increased choice extensiveness might initially be perceived as desirable by consumers, it may also lead to choice demotivation. This is also referred to as the choice overload hypothesis, which states that “an increase in the number of options to choose from may lead to adverse consequences such as a decrease in the motivation to choose or the satisfaction with the finally chosen option” (Scheibehenne et al. (2010), p. 409).Iyengar & Lepper (2000)tested this choice overload hypothesis by comparing preferences for a selection of 6 vs 24 different flavors of jam and found that the former has a stronger effect on consumers’ purchase intentions. Other authors presented similar findings, using a wide array of FMCG products (e.g. coffee (Mogilner et al., 2008) and chocolates (Chernev, 2003)). Other authors however do not find a positive effect of a reduction in choice extensiveness on product sales (e.g. Borle et al. (2005), using data from an online grocery/delivery service) and Sloot et al. (2006), collaborating with a Dutch retailer and focusing on the detergent category27). In a recent meta-analysis, Scheibehenne et al. (2010) find a mean effect size of choice extensiveness of approximately 028. The authors however find high variance between the results of the different studies, indicating that the validity of the choice overload hypothesis might depend on particular conditions. These specific conditions could however not be tested meta-analytically due to their distinct nature. In their study the authors therefore narrow the

27

The authors did however find a positive effect of a category reduction on the attraction of new category buyers.

28

(26)

19 preconditions down to three potential moderators of the effect of choice overload on purchase intentions, being the ‘environment of the assortment or choice’, the individual decision makers’ goals and strategies, and the interactions between these two. In conclusion, the authors state that “more choice is [found to be] better with regard to consumption quantity if decision makers had well-defined preferences prior to choice” (Scheibehenne et al. (2010), p. 421)).

Taking into account the ambiguous effects of the choice overload hypothesis, and the limited number choices that will be tested in the main study29, the choice overload hypothesis is expected not to be applicable here, implying that consumers are expected to derive the highest utility from a choice between all the available premiums. Thus,

Hypothesis 3: Increasing the customization extensiveness has a positive effect on premium promotion effectiveness

3.2 Conceptual model

The hypotheses are schematically represented in a conceptual model (figure 3). The covariates30 will

be addressed in the next chapter, which elaborates on the methodology of the research.

Figure 3: Conceptual model

29

in line with Arora & Henderson (2007), the maximum number of choices provided is 5

30

(27)

20

4. Research methodology

This chapter presents the research methodology and the plan of analysis underlying the experiments. Section 4.1 will introduce the research setting and the product category. Section 4.2 will elaborate on issues related to the research methodology, being among others the choice-based conjoint analysis (CBC), the pretests, and validity measures.

4.1 Introduction of the research

This section will introduce and elaborate on the selection of the research setting, the product category, and the premiums selected for this study.

4.1.1 Research setting

The research conducted in this study will try to mimic an online retail setting, since mimicking an online setting is a necessary prerequisite in order to make premium customization (and hence the CBC) more realistic; for a brick-and-mortar supermarket the customization option will be highly resource-consuming and unpractical, making a premium customization strategy highly unrealistic. Appie.nl is therefore used as a benchmark for the CBC layout (concerning pictures, fonts and sizes) and the setting is presented as a general ‘online retail environment’. Consumers are told to imagine that they want to buy a crate of beer of their favorite brand online. Appie.nl is only mentioned here as an example, in order to stimulate the credibility of the choice tasks (i.e. respondents might realize that premium customization is a realistic situation). The reason for not explicitly mentioning Appie.nl as the retail setting is to exclude retailer preference as much as possible; i.e. respondents’ attitude towards Albert Heijn or Appie.nl in particular should not influence their choice-making.

4.1.2 Product category

In line with most of the literature on premium promotion, the experiment will focus on the FMCG sector. More specifically, the beer category is chosen as the preferred FMCG category for this study, due to both the relatively high retail prices (providing an opportunity to include somewhat more expensive premiums) and the high price competition intensity in this FMCG category31.

4.1.3 Premiums

The premiums included in the experiment (whether included as a free gift or a purchase-with-purchase) are expected to be of high importance with regard to the effectiveness of the premium promotion. Since the premiums will have to be both attractive and of a hedonic nature in order to

31

(28)

21 positively influence the effectiveness of the premium promotion (section 2.2.2), the largest (self-declared) Dutch ‘gadget’ website32 is used as the source for the premiums included in the study.

4.2 Research design

In order to measure the differences in a consumer’s utility and consequently derive individual WTP for the different premium promotions, conjoint-based analysis is applied. The input for the CBC will be determined by two pretests; the goal of the first pretest is the selection of ten (relatively) attractive premiums, which form the core input for the second pretest. The second pretest will then validate the hedonic nature of both the beer category and the premiums and estimate a realistic price range for both the beer category and the individual premiums. The main purpose of the second pretest is however to select the five premiums that are found most attractive and congruent with the nature of the beer category. The overall research design thus consists of two pretests and a main study (including the CBC), as summarized in figure 4.

Figure 4: Overview research design

Section 4.2.1 will discuss the conjoint design, including the derivation strategy regarding WTP estimates from the premium-price combinations. Section 4.2.2 will discuss the validity measures. The data collection process and the plan of analysis are described in respectively section 4.2.3 and 4.2.4. Figure 5 provides an overview of how various sections are directly related to the research design.

32

(29)

22

Figure 5: relationship between chapter sections and research design

4.2.1 Conjoint design

Conjoint analysis is a popular method to measure WTP, used mainly in marketing, transportation, and environmental economics (Jedidi & Jagpal, 2009). This section will discuss several issues related to conjoint analysis, concerning the type of conjoint methodology (4.2.1.1), the estimation of hypothetical versus real WTP (4.2.1.2), how a (customized) premium promotion is expected to lead to increased utility (and consequently to increased WTP) (4.2.1.3), extreme response behavior (4.2.1.4), the attributes and levels used in the CBC study (4.2.1.5), the presentation method, profiles and choice sets (4.2.1.6), and the covariates (4.2.1.7).

4.2.1.1 Choice-based conjoint analysis (CBC)

The main reasons for selecting CBC as the preferred form of conjoint methodology are the low number of attributes included in the conjoint design33(Hair et al., 2010). In addition, CBC is found to be more realistic; the provision of a choice between multiple options (including a no-purchase option), known as the choice set, increases the face validity of the method, as compared to traditional conjoint analysis (Hair et al., 2010). The inclusion of the no-purchase option in a choice-set has furthermore been found to increase the efficiency of the choice task and better allow for choice

33

(30)

23 simulation (Brazell et al., 2006). With regard to the WTP estimates, the no-purchase option furthermore facilitates the obtainment of “unambiguous monetary values for the WTP” (Jedidi &

Jagpal, 2009, p. 45), being of sheer importance in the context of this thesis.

4.2.1.2 Real vs hypothetical WTP

The estimation of ‘real’ WTP has some clear advantages over the estimation of ‘hypothetical’ WTP. Due to resource constraints however, incentive-aligned CBC (Ding et al. (2005) & Ding (2007)) is not feasible for this study and hence only a respondent’s hypothetical WTP will be measured. In a recent study however, Miller et al. (2011) find that the bias in hypothetical WTP, as compared to real WTP, is limited. The hypothetical WTP estimates derived from the CBC are thus expected to lead to valid outcomes.

4.2.1.3 Deriving incremental utility from premium promotion

The consumer choice process is often represented as a process of utility maximization. In order to model the choice process of an individual consumer, the following utility function is often used in marketing applications (Arora & Henderson, 2007):

where is the utility for a certain consumer ( ), which is a function of the attributes of the promoted product ( ) and the preference structure of the individual consumer ( ). Attaching a premium promotion to the product transforms the utility function of a consumer to where represents the premium promotion and the preference structure of the individual consumer for this premium promotion. Naturally, will depend on ; i.e. different premiums promotions will result in different preference structures for these premiums. Ideally, a retailer or manufacturer maximizes this utility function by tailoring the premium promotion offer to the individual consumer’s needs.

Deriving incremental utility from a customized premium promotion

Customization essentially entails providing the consumer with a choice between alternatives, thereby addressing preference heterogeneity among consumers. Preference for the premium promotion is logically expected to differ as well for different premiums. Customization, in a premium promotion context thus would imply that

(31)

24 Deriving WTP from the CBC parameters

In order to describe the choice-behavior of the respondents (and hence derive the part-worths for the different levels), a Multinomial Logit model (MNL) is applied. More specifically, latent-class methodology is applied, since it allows for the derivation of potential segments of respondents

(Karniouchina et al., 2009). By maximizing the likelihood function, latent-class analysis allows for the

use of the posterior segment membership probability to derive parameters for individual respondents, based on segment-specific parameters of the estimations (Wedel & Kamakura, 2000). In a MNL model, the probability that respondent chooses choice premium promotion in choice set equals (Gensler et al., 2012):

And the probability that consumer h chooses the no-purchase option is

(32)

25 After individual parameters are derived for the product attributes, the price, and the no-purchase option, WTP can be calculated as follows (Gensler et al., 2012):

Which is equivalent to:

( )

Since WTP is directly derived from the utility of the no-purchase option, choice-behavior related to the no-purchase option should be closely monitored; extreme response behavior should be mapped and avoided as much as possible (Gensler et al., 2012). The following sub-section will briefly discuss the implications of extreme response behavior for this study.

4.2.1.4 Extreme response behavior

(33)

26

4.2.1.5 Attributes and levels

In order to derive the respondents’ individual WTP for the various types of premium promotions, two attributes are included: ‘price’ and ‘premium’.

Premium

Whereas the price attribute is included in most conjoint studies, the ‘premium’ attribute is of specific interest for this thesis; regardless of price, the preference for the different types of premium

promotions will provide insights regarding the ‘effectiveness’ of these different types of premium promotions. More specifically, the premium promotions are included in the CBC design as follows: The baseline option represents the absence of a premium promotion (‘NoPremium’). The second level represents the inclusion of a pre-pecified (free) gift (‘FixedGift’). The specific free gift will be determined by the pretests: in line with Arora & Henderson (2007), the most attractive premium will be selected for the ‘fixed gift option’. The third option provides a limited choice regarding the fixed gifts; this level represents a choice between two fixed gifts (‘FixedChoice’). The fourth option then provides for an extended free gift choice, where a choice can be made between 534 free gifts (‘FreeChoice’). The fifth level provides a purchase option (‘PwP’). The purchase-with-purchase option is framed as a ‘free choice’ option, since Simonsen et al. (1994) found that a purchase-with-purchase option is unlikely to lead to increased perceived deal value. This finding has not been explicitly questioned in the literature. By framing the purchase-with-purchase option as a free choice option (again representing 5 premiums), the chances of exercising a positive effect on perceived deal value are presumed to be larger. The main reason for not including a fixed purchase-with-purchase option and/or a fixed choice purchase-purchase-with-purchase option is the efficiency of the CBC design; the number of potential combinations would rise from 30 to respectively 35 and 40, which would considerably increase the number of choice tasks to be performed by the respondents. Thus, if the customized purchase-with-purchase option is found to result in incremental utility, additional research should be conducted with the objective of more thoroughly exploring the effects of different purchase-with-purchase premiums on perceived deal value. The last level, level six, represents the ‘breadth customization’ option and provides a fixed choice option, including both a free gift and a PwP option (‘BreadthCustomization’)35.

Table 2 shows the relation between the levels of the promotion attribute and the hypotheses as formulated in section 3.1

34

Again, in line with Arora & Henderson (2007)

35

(34)

27

Table 2: Relation between the premium attribute levels and the hypotheses

Premium (level) Hypothesis

NoPremium(1) FixedGift(2) 1a ( ) NoPremium(1) PwP (5) 1b ( ) FixedGift(2) FreeChoice(4) 2a ( ) FixedGift (2) Breadth Customization(6) 2b ( ) FixedGift (2) FixedChoice (3) FreeChoice (4) 3 Price

Since the estimation of WTP is strongly affected by the price parameter (eq. 2), relatively many price levels are included, so as to increase the estimation accuracy of consumer’s real WTP. However, even though the ‘premium’ attribute consists of six levels, only five are used for the price attribute. This is in line with the principles advocated by Orme (2006), who suggests not to include more than five levels to describe the price attribute; more data per price point is preferred over sparser data over more price points. The price levels will be determined by a combination of the WTP ranges derived in the second pretest (using van Westendorp’s (1976) Price Sensitivity Meter (PSM)) and ‘realistic prices’ (as used by Appie.nl), in order to ensure sufficient overlap between estimated WTP for a crate of branded beer and marketplace prices and consequently avoid extreme response behavior as much as possible. The PSM is explained in more detail in appendix 1.

(35)

28

4.2.1.6 Presentation method, profiles, choice sets, and prohibitions

Due to the small number of attributes (2), the full profile method is chosen as the presentation method36. The minimum number of profiles that have to be evaluated by each respondent is 1037. The minimum number of estimated parameters will thus amount to 1038 as well. However, calibrating the design parameters using an efficiency level of 95%, leads to the conclusion that a minimum of 14 choice sets should be evaluated by an individual respondent in order to derive reliable parameter estimates. This is however only the case when a single version of the CBC design is used. Randomly distributing an additional version of the CBC allows the number of choice tasks to drop to only 9 per version (thus amounting to a total of 18 different choice sets). Important to note here is that the aforementioned minimum number of choice tasks are only applicable when four profiles (not counting the ‘no-purchase’ option) are presented at the same time. The presentation of three profiles (again excluding the ‘no-purchase’ option) results in a minimum of respectively 98(!) (single design version) and 30 (two design versions) choice tasks per respondent, if the desired efficiency level of 95% is to be maintained. Due to the limited number of attributes, the simultaneous presentation of four profiles is not expected to lead to serious difficulties. Actually, Orme (2006)finds that a typical study includes about three to five product concepts per task (The simultaneous presentation of a larger number of choice tasks is even likely to be beneficial to the respondent, since it takes respondents considerably less than twice as long to answer choice tasks with four concepts as with two concepts (Orme, 2006).

Finally, it is important to note that no prohibitions are implemented; all price-premium combinations are thus possible. The trade-off made here is between design effectiveness and credibility of the choice tasks. In general however, Orme (2006) states that ‘prohibitions, if at all possible, should be avoided’; the specification of unnecessary or excessive prohibitions is mentioned as one of the most common mistakes. Even though some attribute combinations are less likely to appear in practice (e.g. the lowest price with the customized free gift), “it is highly recommended to exercise serious restraints when considering prohibiting pairs”39. Prohibiting for example the lowest price

(PriceLevel1) with the free choice gift option (FreeChoice), would increase the minimum number of choice tasks to be evaluated by respondents to 90(!) when simultaneously presenting four profiles and using two versions with a minimum efficiency level of 95%. For this reason, no prohibitions are implemented in the CBC experiment. In order to ensure as much credibility as possible however, the lowest price level should still be above a price that is perceived as a clear real-world discount or ‘too

36

The full profile method is recommended when the number of factors < 6 (Hair et al., 2010)

37

Minimum number of profiles = Total number of levels across all factors (11) – number of factors (2) + 1 (Hair et al., 2010) 38

Number of estimated parameters = Total number of levels (11) – number of attributes(2) +1 (Hair et al., 2010) 39

Referenties

GERELATEERDE DOCUMENTEN

For Case F there is a founder team of four people. The interviewee is responsible for the internationalization and to bring the product on the market. He mentioned that

examined the effect of message framing (gain vs. loss) and imagery (pleasant vs. unpleasant) on emotions and donation intention of an environmental charity cause.. The

discount depth on return probability becomes stronger for hedonic categories, compared to utilitarian

As described in section 4.3.3, model 3 is the fullest version of the models as its aim is to discover whether interaction effects are present between hedonic versus

Specifically, the present thesis focuses on two main symbolic dimensions, namely, environmental self-identity and environmental social identity, that could influence

The goal of study I was to investigate consumers’ purchase intention in a social media environment that is often used for FGC, namely Facebook. The aim was to

To what extend will these combinations between base and premium products, considering the match of both hedonic and utilitarian product associations, have different influences

It is showed that the WTP for the single elements (design and functional customization) is higher than the WTP for both elements combined. Women would pay 15.6 EUR