Brand authenticity – The new road to success for National Brands?

(1)

1

Brand authenticity – The new road to success for National Brands?

A study on the influence of the perceived brand authenticity gap between National Brands and Private Labels on the willingness to pay a price premium for National Brand food products and

the moderating role of the perceived hedonic and utilitarian product value.

(2)

2

Brand authenticity – The new road to success for National Brands?

A study on the influence of the perceived brand authenticity gap between National Brands and Private Labels on the willingness to pay a price premium for National Brand food products and

the moderating role of the perceived hedonic and utilitarian product value.

January, 2017

Master Thesis

MSc Marketing Intelligence University of Groningen Faculty of Economics and Business

Department of Marketing PO Box 800, 9700 AV Groningen

Melissa Bolks s2152592

Herman Colleniusstraat 144 9718 KZ Groningen, the Netherlands

0657253134

m.s.e.bolks@student.rug.nl

Supervisor: Dr. J.E.M. van Nierop

(3)

3

MANAGEMENT SUMMARY

In Dutch supermarkets, National Brands (NBs) face intense competition from Private Labels (PLs). PLs have developed themselves as being a worthy competitor for NBs with product quality levels that are approximately the same. Consequently, it is becoming more difficult for NB manufacturers to charge a price premium over PLs, which causes a drop in the sales of NBs. Therefore, this study aims at investigating a relative new concept in the marketing literature to increase NB performance again, namely brand authenticity. More precisely, the influence of the brand authenticity gap between NBs and PLs on consumers’ willingness to pay a price premium for NBs over PLs (WTPPP) is investigated. Furthermore, the moderating role of the perceived hedonic and utilitarian product value is examined. In addition, three different consumer segments were identified.

This study adopts a quantitative research approach and an online survey is distributed. In total, 214 people participated in this study. In the survey the participants were randomly distributed to three out of the four product categories used in this research (i.e. chocolate, cola, spaghetti and yogurt). The respondents were asked to rate their preferred NB and the PL of their most preferred supermarket in terms of brand authenticity. Furthermore, the respondents were asked to state their WTPPP and their perception of the hedonic and utilitarian product value of the product.

The research resulted in some interesting findings. First of all, a positive brand authenticity gap is found for NBs over PLs for all product categories. This implies that consumers perceive NBs as more authentic than PLs. However, it is also found that this brand authenticity gap is not very large. Furthermore, the results show that the brand authenticity gap between NBs and PLs has a positive effect on the WTPPP. This means that an increase in the brand authenticity gap leads to an increase in WTPPP. These results suggest that it is important for NB managers to investigate ways to maintain and increase the brand authenticity gap to increase the WTPPP.

(4)

4 Furthermore, three segments are identified, which are: 1) the wealthy youngsters and elderly, 2) the rational youngsters and 3) the loyal adults and elderly. Segmenting the respondents helps managers in identifying differences in consumer responses. Between segments, only the socio-demographics age and income were significantly different. However, all segments respond positively to the brand authenticity gap in terms of WTPPP and this is the highest for segment 3. This segment contains respondents older than 34 years with a relatively low income. Segment 1 scores the second highest in terms of the effect between brand authenticity and WTPPP. This segment contains young and elderly consumers with relatively the highest income. Segment 2 contains mostly youngsters who have an average income, but are not very willing to pay a price premium. In addition, the segments suggest that the level of income is not that important in the relationship between brand authenticity and WTPPP, because the segment with on average the lowest income responds the strongest to the brand authenticity gap in terms of WTPPP. This is an important finding, because it suggests that consumers either with a high or low income respond positively to the brand authenticity gap in terms of WTPPP. Since the effect between the brand authenticity gap and WTPPP is the highest for segment 3, NB managers could start with targeting consumers older than 34 years old with marketing strategies relating to brand authenticity.

All in all, brand authenticity could be a fruitful new way to increase NB performance again and face the intense competition with PLs.

Keywords: brand authenticity gap, willingness to pay a price premium, national brands, private

(5)

5

PREFACE

Before you lies the result of an interesting period full of hard work and commitment. After five and a half years of studying, where I had a very social and inspiring student life, finished a bachelor degree in Business administration at Groningen University and a master degree in Business administration at Lund University, I am about to receive my master degree in Marketing Intelligence. At the same time, this event will also mark a new phase in my life: it is time to say goodbye to my student life and enter a new world focusing on building my professional career. However, I could not have completed this research project without the help of some important persons. First of all, I would like to thank my supervisor Dr. Erjen van Nierop for his valuable feedback and help during the process of writing my thesis. My gratitude also goes to my fellow group members for their feedback and suggestions, and to the participants in my study without whom I could not have conducted this research. Furthermore, I am thankful for my best friends for their support and times of laughter and joy. When it was needed, you were always able to cheer me up and keep me motivated.

My special thanks go to my parents, to whom I am very grateful for their support during my entire period as a student. They have always believed in my capabilities, which really means a lot to me.

Melissa Bolks,

(6)

6

1. INTRODUCTION

National Brands (NBs) face intense competition from Private Labels (PLs) in the fast moving consumer goods sector (FMCG). The global average PL share in this sector is 16.5%, however this percentage differs much between countries (Nielsen, 2014). In western Europe, the competition between NBs and PLs is the highest, where the average PL share is 30%. The Netherlands, with a PL share of 27% scores just below this average, but it still means that PL sales in this country accounts for almost €1,- of every €3,- spent (Nielsen, 2014). These percentages already show the high popularity of PLs, but the actual PL share is even higher, because Nielsen (2014) excludes the PL performance in hard discount stores, while these kind of stores sell the most PLs. In addition to the high PL sales, it is expected that PL share will increase even more, only in the period of 2007-2012, global PL sales increased with 24% in the FMCG (Euromonitor International, 2013).

The high popularity of the often lower priced PLs has a negative influence on NB performance. Therefore, it is crucial that NBs address this problem effectively, but they still struggle to find an adequate response to the PLs threat. To this date, NB manufacturers often increase their product prices to compensate for the drop in sales (Steenkamp, van Heerde & Geyskens, 2010), however it is likely that this response only results in boosting PL sales, because consumers might perceive the large price premium of NBs over PLs as unjustified. Consequently, NB sales will decline even more, which illustrates that increasing product prices will only lead to reinforcing the downward spiral of NB performance. Therefore, new strategies need to be developed to guarantee sustained consumer preferences for NBs, however there is only limited research on whether and when consumers continue to be willing to pay a price premium for NBs over PLs (Steenkamp et al. 2010; Kadirov, 2015).

(9)

9 In the past, high product quality was an important reason why NBs were able to charge a price premium over PLs. However, the product quality advantage of NBs is eroding. Since the introduction of PLs, retailers developed their brands to be worthy competitors for NBs and improved their objective and perceived quality. Therefore, the objective and perceived quality differential between NBs and PLs is very small (Apelbaum, Gerstner & Naik, 2003; Steenkamp et al. 2010; Nielsen, 2014). As a consequence, quality –which is traditionally perceived as the dominant influencer- is losing its strength and only explains a small part of consumers’ willingness to pay a price premium for NBs over PLs (WTPPP) (Sethuraman, 2000, 2003).

The declining role of product quality in WTPPP urges NB manufacturers to focus on non-quality related strategies to increase NB performance. This is in line with the finding that non-quality related aspects is the most important influencer of WTPPP (Sethuraman, 2000; 2003). Since NBs are commonly perceived to be the ‘image’ brands and PLs as the ‘no-frills’ brands (Sethuraman, 2003) this could especially be a fruitful new direction for NBs. One such an alternative strategy that could be regarded as a non-quality related aspect and as a part of a brand’s image is brand authenticity (Bruhn, Schoenmüller, Schäfer, & Heinrich 2012). Therefore, this study aims to get a better understanding of the relative underexplored role of brand authenticity in WTPPP.

(10)

10 Beverland & Farrelly, 2010). Secondly, it is a consequence of the loss of traditional sources of meaning and self-identity in the postmodern markets (global markets) (Arnould & Price, 2000). Lastly, the standardization and homogenization in the marketplace leads to the search for authenticity in consumption (Thompson, Rindfleisch, and Arsel 2006).

The ascribed importance of brand authenticity raises the question to what extent it could be a valuable strategy to enhance NB performance. Therefore the purpose of this research is to investigate whether brand authenticity also translates into an increase of WTPPP in supermarkets. More specifically, the study is interested in whether the difference in brand authenticity perceptions between NBs and PLs leads to a higher WTPPP. Furthermore, research found that non-quality related aspects are more important for products that are perceived as more hedonic than for products that are perceived as more utilitarian (Sethuraman, 2003). This finding could suggest that brand authenticity, identified as a non-quality related aspect, is valued more in product categories which are perceived as more hedonic. Therefore, it is interesting to study to what extent hedonic and utilitarian product perceptions influence the relationship between brand authenticity and WTPPP. All in all, the following research question could be identified:

To what extent does the perceived brand authenticity gap between NBs and PLs in supermarkets influence consumers’ willingness to pay a price premium for NBs over PLs?

To get an adequate answer to this research question, the following sub-questions will be addressed: 1) To what extent do consumers perceive a difference in brand authenticity between NBs and PLs? 2) What influence do perceptions about hedonic and utilitarian product value have on the relationship between the perceived brand authenticity gap and willingness to pay a price premium for NBs over PLs?

Theoretical and managerial relevance

(11)

11 carried out in the context of luxury brands, however there is little research on the concept involving mass-marketed products (Alexander, 2009), such as NBs and PLs in the FMCG. Lastly, this study addresses the lack of quantitative research involving the concept of brand authenticity (Schallehn, Burmann & Riley, 2014; Kadirov, 2015) by adopting a quantitative research method to investigate the concept and its influence on WTPPP.

Apart from the theoretical contributions, the study also provides managerial relevance. First of all, the study provides a better understanding of the concept of brand authenticity that could help marketers to identify and create new opportunities for brand positioning and value creation for consumers. Secondly, it provides managers with knowledge on whether they should focus on products that score high on either hedonic or utilitarian product perceptions when exploiting marketing strategies relating to brand authenticity. Furthermore, the study helps to identify consumer segments that respond similar to the brand authenticity gap in terms of WTPPP. These insights could be a starting point for marketers to create new advertising and targeting strategies. Report structure

(12)

12

2. THEORETICAL FRAMEWORK

2.1 Brand authenticity

Brand authenticity is not an easy to define concept and people have different views on what brand authenticity means (Beverland, 2005a). In general, brand authenticity could be divided into three perspectives (Morhart, Malär, Guevremont, Girardin & Grohmann, 2015).

(13)

13 ‘worthy of acceptance, authorative, trustworthy, not imaginary, false or imitation, conforming to an original’, while Napoli et al. (2014) define it as ‘a subjective evaluation of genuineness ascribed to a brand by its consumers. These conceptualizations reflect the objectivist perspective in de study of Beverland (2009) versus the constructivist perspective in the study of Napoli et al. (2014). Despite these different ways of viewing brand authenticity, there is some general consensus and consistency in all conceptualizations about the concept, namely that brand authenticity involves what is genuine, real and true (Beverland & Farrelly, 2010; Kennick, 1985). However, this is still a quite ambiguous and broad definition of brand authenticity and therefore, there is recently an increase in interest to quantify the concept and to develop brand authenticity measurement scales. In table 1, the studies that made a serious attempt to develop brand authenticity measurement scales are listed. In these studies, drivers of the concept are identified and quantified. However, to this date, there are far more qualitative studies than quantitative studies on the brand authenticity concept (Kadirov, 2015).

Authors Drivers of brand authenticity

Morhart et al. (2015) Credibility, integrity, symbolism, continuity Napoli et al. (2014) Quality commitment, sincerity and heritage Schallehn et al. (2014) Consistency, continuity and individuality Eggers et al. (2013) Brand congruence, brand consistency and brand

orientation

Table 1: Drivers of brand authenticity

(14)

14

Continuity

This dimension suggests that authentic brands should reflect timelessness, historicity, and an ability to transcend trends (Morhart et al. 2015). Therefore this dimension relates to past-related aspects, brand heritage, brand history, stability over time and the probability that the brand will sustain in the future. An example of a brand that reflects continuity is Coca-Cola. This brand developed many new versions of the original product, but at the same time they created a sense of stability and timelessness through consistent design features such as the iconic brand name design.

Credibility

‘Brand credibility is defined as a brand’s willingness and ability to deliver on their promises’ (Morhart et al. 2015:202). This means that it is important to consumers that the brand matches their expectations. To give an example, the Dutch supermarket chain Jumbo has the following unique Jumbo formula: The best service, the largest assortment and the lowest price (www.jumbo.com, 25-10-2016). Therefore, consumers expect that the supermarket chain meets these expectations on all three aspects and do not set expectations they cannot meet. This is important because honesty, transparency, trustworthiness and sincerity are crucial factors when judging on brand credibility (Morhart et al. 2015).

Integrity

(15)

15

Symbolism

This dimension reflects that brands should help to construct consumers with who they are (Morhart et al. 2015). Therefore, symbolism could be seen as the potential of a brand to serve as a source for identity construction, by presenting self-referential cues that relates to values, roles and relationships (Morhart et al. 2015). Key to this dimension is the preference for proximity to place, people and culture, because these aspects represent a higher ideal- i.e. valuing the fellowship within a community (Beverland & Farrelly, 2010). For example, Starbucks created an online community, which makes consumers involved and connected with the company and fellow Starbucks visitors. A quote by the CEO chairman of Starbucks Schultz emphasizes the importance of this community: ‘I was taken by the power that savoring a simple cup of coffee can have to connect people and create community’ (Kaplin, 2011).

2.2. Brand authenticity gap between NBs and PLs

To date, little research is conducted on the brand authenticity gap between NBs and PLs. Only the study of Kadirov (2015) did research into this question and found a positive mean brand authenticity gap for NBs in supermarkets. However, validating the brand authenticity gap in this research is also important because NBs should be perceived as more authentic than PLs to yield favorable results in terms of WTPPP. The fact that Kadirov (2015) found a positive brand authenticity gap could be related to the fact that NBs are considered as ‘image brands’ while PLs are considered as ‘no-frills brands’ (Sethuraman, 2003). Consumers are likely to attach higher perceptions of authenticity to NBs because it is part of a products image. In addition, the fact that PLs are for a long time seen as inferior in terms of quality to NBs and only offering a good price could further suggest that NBs could be perceived as more authentic.

H1: There is a positive brand authenticity gap between NBs and PLs in favor of NBs.

2.3 Consumers’ willingness to pay a price premium for authentic brands

(16)

16 higher purchase intention (Napoli et al. 2014) and a stronger emotional brand attachment and more positive word-of-mouth (Morhart et al. 2015). These findings show that perceived brand authenticity is a successful branding strategy, however less is known about its capability of influencing WTPPP. It is interesting to investigate WTPPP because it is the best measure to interpret consumer choices for brands (Agarwal & Rao, 1996). Since this study tries to contribute to increasing NB performance, WTPPP could be regarded a valuable measure.

2.3.1 Consumers’ willingness to pay a price premium

WTPPP is the ability of a brand to charge a higher price than its relevant competitors and it is often defined synonymously with customer-based brand equity (Sethuraman, 2003; Blackston, 1995; Ailawadi, Lehmann & Neslin, 2003; Agarwal & Rao, 1996; de Chernatony & McDonald, 2003; Aaker, 1996). While some researches did study the relationship between perceived brand authenticity and WTPPP, these researches were mainly based on creating authenticity perceptions by communicating isolated appeals on different drivers of the concept of brand authenticity. For example, Skuras & Vakrou (2002) found a link between WTPPP and authentic wine perceptions. They created brand authenticity perceptions by focusing on putting the product’s origin on the package, which indeed yielded a higher WTPPP. Furthermore, Beverland (2005b) created perceptions of brand authenticity by creating a sincere story of the wine brand. He also found that appearing authentic was critical in order to command price premiums. While these studies suggest that certain ways of creating brand authenticity messages could indeed increase WTPPP, this research takes a specific interest in investigating the influence of the total brand authenticity construct on WTPPP by taking into account all its relevant aspects found by the study of Morhart et al. (2015) (i.e. credibility, integrity, symbolism, continuity).

(17)

17

Figure 1: Drivers of WTPPP for packaged food (Anselmsson et al. 2007)

From the identified drivers by Anselmsson et al. (2007), they claim that only 20 percent of consumers’ willingness to pay for different packaged food products can be explained by perceived quality. Therefore, non-quality attributes are very important in understanding why consumers are willing to pay a price premium. Consumers may perceive no quality differential between products, but they still are willing to pay a price premium for a particular product, due to reasons such as familiarity, imagery or other positive associations (Sethuraman, 2000). Of particular interest to this study are the dimensions associations and uniqueness (figure 1).

2.3.2 Brand associations and uniqueness

(18)

18 good or bad, however it is impossible to judge whether the wine is authentic or not. Furthermore, brand associations are primarily created by communications (Ophuis and Van Trijp, 1995; Grunert, Bech-Larsen & Bredahl 2000; Brunso et al., 2002). In the case of creating brand authenticity perceptions, marketers utilize appeals to heritage, origin, production methods, credibility, self-verification and moral values to create authentic brand images (Morhart et al. 2015).

For brand associations to have an impact on brand equity, they should be unique, favorable and strong (Keller, 1993). Uniqueness plays a central role because nowadays, business is highly characterized by price competition, copy-cat activities and successful me-too brands getting higher and higher market shares. Especially in the retail sector, uniqueness is considered to be important since this in this sector, consumers often have to choose between many different products and brands. It is suggested that uniqueness is an essential aspect of brand authenticity and is mainly attributed to brands to give the product a set of values that differentiates itself from its commercialized competitors (Beverland 2005a; Rose & Wood 2005; Lewis & Bridger, 2001). This implies that authenticity endows uniqueness to a brand, which is the basis of differentiation (Porter, 1980). This is exemplified by the study of Beverland (2005b) who describes the use of a specific claim for wine brands with the aim of differentiating themselves from competitor wine brands and providing a difficult to replicate competitor advantage.

All in all, perceived brand authenticity could be seen as a relevant brand association and part of endowing uniqueness to a brand, which are two important reasons why consumers are willing to pay a price premium for consumer packaged goods (Anselmsson et al. 2007). Furthermore, research attributed important other positive consequences of having an authentic brand perceptions such as higher brand credibility and trust, a better brand status and corporate reputation, a higher purchase intention and stronger emotional brand attachment and more positive word-of-mouth. These favorable outcomes could also imply a positive effect on WTPPP. Lastly, some researchers established the link between isolated appeals on brand authenticity and WTPPP.

(19)

19 H2: The perceived brand authenticity gap between NBs and PLs has a positive influence on consumers’ willingness to pay a price premium for NBs over PLs.

2.4 Perceived brand authenticity and product category

Previous research has suggested that there are differences between food product categories in terms of WTPPP (e.g. Sethuraman & Cole, 1999; Sethuraman, 2003). Therefore, it is interesting to take a closer look at this when studying brand authenticity and its effect on WTPPP to get insights into the specific category for which it is desirable to have an authentic brand positioning (Napoli et al. 2014).

Perceived hedonic product value describes the extent to how much consumption pleasure a product gives the consumer, i.e. it is about providing a more experiential consumption, fun and excitement. In contrast, perceived utilitarian product value describes the extent to which products are consumed for their usefulness and functionality (Richins, 1994; Dhar & Wertenbroch, 2000). It is important to note that the hedonic and utilitarian product value are two different dimensions of attitude towards the products (Voss, Spangenberg & Grohmann, 2003). This means that consumers rate a product on both an utilitarian and a hedonic value scale. Inherent to these different dimensions is the characteristic that no product scores maximum on hedonic and minimum on utilitarian and vice versa. This indicates that no product could be labeled completely utilitarian or hedonic, because it always involves aspects of both dimensions (Batra & Ahtola, 1990).

(20)

20 perceptions about brand authenticity. However, this effect will be smaller than for products with high perceptions of hedonic value. This leads to the following hypotheses:

H3a: The level of hedonic product value positively moderates the relationship between the perceived brand authenticity gap and consumers’ willingness to pay a price premium for NBs over PLs.

H3b: The level of utilitarian product value positively moderates the relationship between the perceived brand authenticity gap and consumers’ willingness to pay a price premium for NBs over PLs.

To indicate that the moderating effect of H3a is expected to be stronger than of H3b, the following hypothesis is formulated:

H3c: The moderating effect of the level of hedonic product value on the relationship between the perceived brand authenticity gap and consumers’ willingness to pay a price premium for NBs over PLs, is higher than for the level of utilitarian product value.

2.5 Conceptual model

From the theoretical part the following conceptual model is created (figure 2). The perceived brand authenticity gap is the independent variable, where consumers’ willingness to pay a price premium for NBs over PLs is the dependent variable. To investigate the effects of perceived level of utilitarian and hedonic product value, those variables are put as moderators in the model.

(21)

21

3. RESEARCH DESIGN

3.1 Data collection

As mentioned before, this research adopts a quantitative research approach to measure the hypotheses of the study. To accommodate this, an online survey is created in the software program Qualtrics. The research explicitly focusses on Dutch consumers, because consumers from other countries might evaluate the questions in a different way. For example, respondents from the Netherlands could have a different attitude towards WTPPP and attach different meanings to brand authenticity than respondents from other countries. Therefore, to eliminate the bias of differences in perceptions between countries, which may lead to less robust and reliable results, the survey will only be distributed among Dutch consumers.

Furthermore, specific attention is paid to getting a representative sample of the Dutch shopper population. Therefore, the survey is not only put on the social media platform Facebook, but also on online forums that target a relatively older group of people. Furthermore, in each of the supermarkets AH, Jumbo and PLUS, the researcher distributed 25 website links to the survey to try to increase the response rate and to reach a more representative sample of the Dutch shopper population.

3.2 Translating of the constructs

(22)

22 Back-translation is the most commonly approach used in marketing to test the translation accuracy of theoretical constructs (Douglas & Craig, 2007; Brislin, 1970). When using back-translation, the researcher first translates the English constructs to the Dutch language and after that several peers were asked to translate this back to the Dutch language. On basis of those responses, the original and back-translated versions were compared. After careful evaluations, a decision is made on how to translate the constructs.

3.3 Survey design

In the beginning of the survey, the respondents are asked to choose the supermarket they visit most often or in cases where they do not visit one of these stores, they have to choose the one they are most familiar with. This choice is reflected throughout the survey, where respondents are asked to evaluate products of both a NB and a PL, where the PL brand is adjusted to the supermarket choice. Respondents could choose between the supermarkets AH, Jumbo and PLUS. These supermarkets were chosen because they have the highest market share when excluding hard discount stores. Hard discount stores are excluded from the study because they do not sell much NB products, and often do not have their own PL, which makes comparisons between NBs and PLs not feasible. Furthermore, the NBs and PLs are in the survey are shown next to each other, which accommodates the respondents to easily compare the NBs and PLs and to create a more realistic shopper situation. Besides the flexibility of PLs in the survey, NBs could also vary. Respondents are randomly asked to rate three out of the four product categories used in the survey (i.e. chocolate, cola, spaghetti and yogurt) and they are asked to state their preferred NB of each of the products. This question was also aimed to establish a more realistic evaluation of the NBs versus PLs. The choice of these particular four NBs are based on the total number of products of that brand in the particular product category according to the website AH.nl (table 2). For more insight on how the survey is designed, see Appendix A.

Chocolate Cola Spaghetti Yogurt

Milka Coca-Cola Honig Almhof

Verkade Pepsi Grand’Italia Campina

Côte d’Or Dr. Pepper Soubry Arla

Table 2: NBs per product category

(23)

23 and education were asked. Lastly, to increase the survey rate, the incentive to win a €20,- VVV- giftcard is used.

3.4 Design and measurements of the construct

3.4.1 Perceived brand authenticity

The construct of perceived brand authenticity is investigated with the measurement scale developed by Morhart et al. (2015). As already mentioned in the theoretical framework, the construct involves four dimensions: continuity, credibility, integrity and symbolism. Each of these four constructs have several sub-questions, with a total of fifteen items (Appendix B). Since the brand authenticity gap needs to be measured, it is important that brand authenticity perceptions are measured for both NBs and PLs. To measure the brand authenticity gap, an adjustment to the measurement scale of Morhart et al. (2015) has been made. While the original measurement scale consists of fifteen different questions divided over the four dimensions, this study will measure the construct with only four questions reflecting the different dimensions (Appendix B). This decision is made based on two important reasons: First of all, a 15-item measurement scale is very large when taking into account that respondents need to rank these questions for both NBs and PLs and for three different product categories. However, reducing the amount of product categories was not desirable since the moderators perceived hedonic and utilitarian product value needed to be measured as well. Consequently, the many questions could lead to survey fatigue, which leads to lower response rates and respondents drop out of the survey (Porter, Whitcomb & Weitzer, 2004). Secondly, the questions in the original measurement scale are very similar to each other and when translated to the Dutch language, it may become difficult for respondents to see differences between the questions, which could lead to less accurate results. Furthermore, this scale adjustment could to some extent be justified by a study conducted by Bergkvist & Rossiter (2007) who showed that single-item measures are as equally valid as multiple-item measures in terms of predictive validity. However, they studied the variables attitude towards a brand attitude and attitude towards an ad, so security of generalizations of the results is not guaranteed.

3.4.2 Hedonic and utilitarian product characteristics

(24)

24 likely reflect either hedonic or utilitarian perceptions. This is needed, because it is interesting to have differences in perceptions of hedonic and utilitarian product value in order to explain differences in WTPPP. However, as mentioned in the theoretical framework, no product could be either regarded as completely hedonic or utilitarian, however there are some products that are mainly bought because of their functional (i.e. utilitarian) purpose or their characteristic of providing pleasure (i.e. hedonic). Taking this into account, four product categories are used, where two products are more likely to be perceived as more hedonic and two products which are likely perceived to be more utilitarian. As briefly mentioned before, these products are chocolate, cola, yogurt and spaghetti. This decision is also supported by previous studies: chocolate and cola is suggested by Sethuraman & Cole (1999) as a relatively hedonic product, while yogurt and spaghetti is regarded as relatively utilitarian by Coelho do Vale and Verga Matos (2015). While the product categories are chosen to accommodate for differences in hedonic and utilitarian product value, it is still important to ask consumers their perception of the product’s utilitarian and hedonic product value. For doing this, the measurement scale of Okada (2005) is used (Appendix B). Respondents had to rate their hedonic product value perception on a scale from 0 (‘not at all hedonic’) to 6 (‘extremely hedonic’) as well as their utilitarian product value perception on a scale from 0 (‘not at all utilitarian’) to 6 (‘extremely utilitarian’). However, for respondents the terms of hedonic and utilitarian are difficult and it is likely that they do not understand these terms. Therefore, these terms are described to make sure that respondents understand the question. This description is based on the studies of Richins (1994) and Dhar & Wertenbroch (2000) who state that: ‘Some products are primarily consumed for their usefulness and functionality (i.e. utilitarian goods) and other products are primarily consumed because it provides the consumer with pleasure, a more experiential consumption, fun and excitement (i.e. hedonistic goods)’.

3.4.3 Price premium

(25)

25 category ‘more than twice as much’ will be recoded to 125% in the analyses, the other categories will be coded as the actual percentages that consumers state to be willing to pay more.

3.5 Plan of analysis

The main software program that is used to analyze the data is Latent Gold. The advantage of this program over other statistical programs is that it can carry out different types of analyses and at the same time specify different numbers of clusters in the data (Vermunt & Magidson, 2005). Therefore, the first step is to carry out a regression analysis with the predictor variables, the moderator variable and the covariates. This ‘naïve model’ is compared with the model of numbers of clusters that fits the regression model best. It is likely that a model with n numbers of clusters will be favored against the ‘naïve model’, which means that the hypotheses are tested for each of the classes found.

(26)

26

4. RESULTS

4.1 Descriptive statistics

4.1.1 Demographics

In total 310 respondents participated in this survey. After an initial analysis for outliers and missing data, 94 respondents were excluded from the analysis, which resulted in 214 responses that could be used for analysis. Most excluded responses were due to the fact that they started the survey, but did not finish it. And two participants were excluded because they showed to be significant outliers in terms of their responses. Further analysis showed no significant outliers or missing values for certain questions, which could be explained by the set-up of the survey since the questions required a forced response to continue.

This study puts an emphasis on the importance of having a representative sample of the total shopper population. Therefore, the descriptive statistics of the sample of this study is compared to the actual shopper population in the Netherlands in 2014 (EFMI Business School/CBL 2014). In this study the group of women were slightly underrepresented and the age distribution did also stroke not well with the age distribution of EFMI. Therefore, adjusted weighted of cases is applied for the gender and age categories to create a more representative sample of the shopper population in the Netherlands (table 3).

Concerning the demographic statistics of household size, it is acknowledged that the one person household size is underrepresented, while the household size of three persons of more is

Unweighted responses (n=214) Weighted responses (n=214) EFMI 2014 Gender Men 32,9% 29,0% 29.0% Women 67,1% 71,0% 71.0% Age 18-34 years 47,2% 24,0% 24.0% 35-54 years 31,5% 40,9% 41.0%

55 years and older 21,3% 35,1% 35.0%

(27)

27 overrepresented. Furthermore, regarding the income variable, CBS calculated that the modal gross monthly income of an average household consisting of 2.2 persons is €4.875,- While many respondents in the data set indeed indicate an income level that respond to the modal gross income found by CBS (2014) it is likely that household composition also has much influence on the gross household income. Therefore, this variable is more difficult to compare to average income levels because it depends on the number of people that are employed in the household. However, this research does not have information about this. Furthermore, regarding the educational level, in the sample the educational levels of HBO and WO are overrepresented and the lower levels of education are underrepresented (CBS, 2016). However, while it is notified that not all socio-demographic variables stroke well with the actual Dutch (shopper) population, by adjusting weighting cases for gender and age the overall representativeness increased (table 4).

Unweighted responses Weighted responses (n=214) EFMI, 2014 Household size 1 person 18.1% 21.3% 37.0% 2 persons 36.6% 35.7% 33.0% 3 persons or more 45.4% 51.9% 30.0%

Gross monthly income

< €1.000 11.1% 8.0% N/A €1.000 - €2.000 12.0% 10.1% N/A €2.001 - €3.000 25.0% 25.4% N/A €3.001 - €4.000 19.9% 21.2% N/A €4.001 - €5.000 16.2% 18.4% N/A > €5.000 15.7% 17.0% N/A Education

Primary school 0.5% 0.5% N/A

LBO/VMBO 2.8% 5.2% N/A HAVO 4.6% 5.2% N/A VWO 1.4% 0.7% N/A MBO 18.5% 21.3% N/A HBO 47.7% 48.0% N/A WO 24.5% 19.1% N/A

(28)

28

4.1.2 Product categories and brands

As mentioned in the research design, respondents were randomly presented with three out of the four products used in the survey, to avoid fatigue effects. The distribution of the respondents across these product categories are shown in table 5. Due to the fact that respondents were randomly assigned to three of the four categories, with a pre-specified command that this should be equally distributed, the distribution between the product categories is almost the same (ranging from 73% to 76,5%).

Furthermore, consumers could indicate their preferred supermarket and choose for each product category their preferred NB and supermarket. The preferences for supermarkets and NBs are presented in table 6 and table 7. The outcomes of preferred PLs are in line with the market share of these supermarkets, with AH as having the highest market share and PLUS with the lowest market share of the three supermarkets. Besides, the preferences for NBs show much differences, with specific low preferences for the brands Soubry (spaghetti) and Dr. Pepper (cola).

Product category Chocolate Cola Spaghetti Yogurt

Distribution percentage 73% 76% 75% 76%

Table 5: Respondent distribution across categories

Supermarket preference AH Jumbo PLUS

Percentage 39% 35% 26%

Table 6: Respondent preference for supermarket

Product Category Most often preferred Second often preferred

Last often preferred

Chocolate Verkade (43.9%) Milka (34.8%) Côte d’Or (21.3%) Spaghetti Grand’Italia (49.5%) Honig (47.0%) Soubry (3.4%) Yogurt Campina (74.2%) Arla (17.3%) Almhof (8.4%) Cola Coca-Cola (87.4%) Pepsi (11.1%) Dr. Pepper (1.4%)

Table 7: Respondent preference for NBs

4.1.3 Relative hedonic and utilitarian product value positioning

(29)

29 with the most hedonic value and yogurt together with spaghetti is perceived to have the highest utilitarian product value. All in all, the expectation that consumers indeed value chocolate and cola high on hedonic and spaghetti and yogurt high on utilitarian product value is met. This results in a variety of hedonic and utilitarian product perceptions which is useful for investigating the moderator effects of these variables.

Figure 3: Positioning of the product categories 4.2 Reliability and validity

4.2.1 Brand authenticity construct

(30)

30 indicated that the construct is reliable in terms of internal consistency, since generally it is perceived that it should be higher than 0,7 (Schilling, 2002). While this approach of examining the construct of brand authenticity violates the measurement of the construct proposed by Morhart et al. (2015), taking the questions together into one variable supports the hypotheses of the research better. Previous research suggest that hypotheses that are formulated at the concept level should also be analyzed on the concept level instead of the dimension level (Wong, Law & Huang, 2008). It is important to match the level of arguments made in the theory with the level of analyses, because otherwise it is very difficult to answer the formulated hypotheses. However, it is acknowledged that measuring brand authenticity as one variable leads to loss of richness of the data and possibly to violation of the validity scale proposed by Morhart et al. (2015), however for this study continuing with one variable seems the best option.

4.2.2 Pooling of the different product categories

Since this study investigated four different product categories, it is important to check whether the estimates of the parameters of these different product categories are significantly homogenous or not. This has implications for whether it is allowed to pool the responses of the different product categories. One way to test this is by using a F-test, in this case the Chow-test (Leeflang, Wieringa, Bijmolt & Pauwels, 2015). The mathematical notation of the test is shown below.

Fv1, v2 ~ (RSS1−RSS2)/v1_RSS2/v2

Where,

RSS1 = The residual sum of squares of the pooled regression RSS2 = The residual sum of squares of the unit-by-unit regressions

V1 = The difference in the degrees of freedom between the pooled regression and the unit-by-unit-regressions

V2= The total degrees of freedom which are unused of the unit-by-unit regressions.

(31)

31

V1 V2 RSS1 RSS2 F-value

Value 66 554 187563.582 168645.047 0.9416

Table 8: Chow test statistics

When looking at the critical values in the F-table distribution, the p-value associated with this test is 0.6085. Therefore, pooling is allowed.

However, pooling the data from all the cross sections has implications for the quality of the analysis, because it assumes homogeneity of all the parameters and intercepts. Therefore, it could lead to very generalized results in terms of to what extent the different variables in the analysis explain WTPPP. To test whether it is more favorable to adopt a pooling option that could take into account some more heterogeneity, it is tested whether the dependent variable of WTPPP is significantly different across the product categories. To test this, an one-way ANOVA is conducted (table 9). The ANOVA test is used to test whether the mean of the dependent variable significantly differs between categories. The output showed that the Levene statistic was significant at a significance of p=0,000. This test is used to check whether there is homogeneneity in the variances across the groups, however the significant result of the test shows that there is heterogeneity in variances. Therefore, the different product categories show significant different results in the mean WTPPP. Furthermore, from table 9 could be seen that cola yields the highest WTPPP, while yogurt yields the lowest results. To dive deeper into these results, the Tukey post-hoc test is carried out to check whether all categories do significantly differ in terms of WTPPP. However, the product category combinations of chocolate and cola (p=0.139 ) and spaghetti and yoghurt (p=0.939) did not differ significantly from each other. The other product combinations did differ significantly from each other (p<0.05).

Product category Mean Standard deviation

Chocolate 18.44% 19.32

Spaghetti 12.64% 15.04

Yogurt 11.37% 16.35

Cola 23.25% 27.03

Total 16.38% 20.47

Table 9: Mean willingness to pay a price premium for NBs

(32)

32 adopting an OLSDV model, which is a specific partially pooled model. Partially pooling is a way to accomplish a balance between on the one side the benefits of pooling, and on the other side accommodating heterogeneity between the cross sections, which is also referred to as fixed effect models (Leeflang et al. 2015). This is also a relevant technique in marketing applications where the dependent variable (in this case WTPPP) is significantly different between the cross sections, but where it is reasonable to assume that the response of the other variables in the study is about the same across the different cross sections (Leeflang et al. 2015). The mathematical notation of an OLSDV model is:

Yit = αit + Xitβ + ԑit

To conclude, an OLSDV model is used in this study to conduct the different analyses.

4.3 Perceived brand authenticity gap

It is interesting to dive deeper into the question whether consumers’ perceive NBs as more authentic than PLs (table 10). In this table the percentage of consumers who either perceive an authenticity gap advantage for NB, an authenticity gap advantage for PLs, or no difference between authenticity perceptions for NBs and PLs are displayed. To put these results into context, the results are compared to the findings in the study of Kadirov (2015).

Variable statistics Perceived authenticity gap (This study)

Perceived authenticity gap (Kadirov, 2015) X>0, NBs advantage over PLs 66.60% 68.90% X=0, No difference 23.50% 20.20% X<0 PLs advantage over NBs 9.90% 10.90% Mean 1.67* 2.12 Coefficient of variation 2.07* 1.32

* Adjusted to the measurement scale of Kadirov (2015). Table 10: Brand authenticity gap evaluations

(33)

33 variation. The mean authenticity gap in this study is much lower as measured by Kadirov (2015), while the coefficient of variation is way larger in this study. The coefficient of variation of 2.07 means that the standard deviation is equal to 2.07 of the average. Working with the coefficient of variation makes it possible to compare results measured on different scales. To interpret the differences regarding the average authenticity gap and the coefficient of variation, it is important to note that the study of Kadirov (2015) is carried out in New-Zealand which has a PL market share of only 13% compared to the Netherlands with a market share of 27%. Therefore, the differences could be a consequence of country differences. For example, it is possible that due to the high popularity of PLs in the Netherlands and the fact that PLs are further developed than in New-Zealand, consumers in the Netherlands might still perceive an authenticity gap but this gap becomes smaller when PL popularity increases.

Besides the analysis of the brand authenticity gap in terms of a positive, a negative or a neutral evaluation, it is also interesting to investigate whether the brand authenticity evaluations are significantly different for NBs and PLs in general, and for the different product categories. To test for this, a paired samples t-test was conducted (table 11).

Mean Standard deviation t-value Significance Authenticity NBs 4.88 1.39 t(640)= 19.75 p=0,00 Authenticity PLs 3.82 1.33 -Chocolate- Authenticity NBs 5.00 1.31 t(159)= 11.09 p=0,00 -Chocolate- Authenticity PLs 3.82 1.23 -Spaghetti- Authenticity NBs 4.75 1.55 t(155)=8.17 p=0.00 -Spaghetti- Authenticity PLs 3.86 1.38 -Yogurt- Authenticity NBs 4.78 1.36 t(166)=6.97 p=0.00 -Yogurt- Authenticity PLs 4.19 1.25 -Cola- Authenticity NBs 4.99 1.34 t(158)=14.14 p=0.00 -Cola- Authenticity PLs 3.38 1.36

(34)

34 From table 11, it can be concluded that there is a significant difference in the scores between the authenticity perceptions for NBs and PLs. Furthermore, it is also confirmed that for all the product categories in this research there is a significant positive difference in favor of NBs between authenticity perceptions for NBs and PLs. Therefore, the first hypothesis is confirmed and can be accepted.

4.4 Willingness to pay a price premium

To get a first feeling of the dependent variable WTPPP, the average WTPPP is measured and compared with to study of Kadirov (2015) (table 12).

This study Kadirov (2015)

Average willingness to pay a price premium

16.38% 26.35%

Percentage of consumers willing to pay ≥10% for NBs

70.50% 87.4%

Percentage of consumers willing to pay ≥50% for NBs

3.00% 21.2%

Table 12: Willingness to pay a price premium evaluations

(35)

35 confirmed by the study of Steenkamp et al. (2010), who found that in the Netherlands, the WTPPP is the lowest compared to 22 other countries in the research.

In the following section, the effect of the brand authenticity gap on WTPPP is investigated, together with the moderators perceived hedonic product value and perceived utilitarian product value.

4.5 Regression analysis

(36)

36

4.5.1 Regression analysis including interaction terms

As mentioned above, the first step is to determine the optimal class solution (table 13).

LL BIC CAIC Class error R2

1 class -2749.8726 5553.3918 5563.3918 0.0000 0.2568 2 class -2458.2095 5114.9117 5151.9117 0.0289 0.5757 3 class -2369.4465 5082.2316 5146.2316 0.0607 0.6363 4 class -2308.0215 5104.2277 5195.2277 0.0516 0.6441 5 class -2260.7340 5154.4987 5272.4987 0.0441 0.6368 6 class -2216.1957 5210.2681 5355.2681 0.0409 0.6459 7 class -2172.4073 5267.5373 5439.5373 0.0451 0.7201

Table 13: Class specific statistics

When looking at the BIC and the CAIC scores, the BIC and the CAIC scores are the lowest for the 3-class model. Furthermore, the classification error is small with 0.0607 and the R2 is relatively high with 0.6363. Despite that the classification error of the 3-class segment is the highest of the models, it is still acceptable and the R2 does not improve much from changing to a model with another number of classes. Therefore, the 3-class solution seems the best model for the data.

By choosing the 3-class model, the related regression results for this model are interpreted (Appendix C). However, the regression results showed that both the interaction effects for perceived hedonic and utilitarian product value are not significant with respectively p-values of 0.91 and 0.76. This means that the level of hedonic or utilitarian product value does not significantly influence the relationship strength between perceived brand authenticity and WTPPP. Therefore, hypotheses H3a, H3b and H3c cannot be accepted.

(37)

37

4.5.2 Regression analysis without the interaction terms

When looking at the BIC and the CAIC scores, the BIC and the CAIC scores are the lowest for the 3-class model (table 14). Furthermore, the classification error is small with 0.0613 and the R2 is relatively high with 0.6337. Despite that the classification error of the 3-class segment is the highest of the other models estimated, it is still acceptable and the R2 does not improve much from changing to a model with another number of classes. Therefore, the 3-class solution seems the best model for the data.

LL BIC CAIC Class error R2

1 class -2750.7306 5544.3785 5552.3785 0.0000 0.2548 2 class -2459.2934 5095.6207 5128.6207 0.0286 0.5731 3 class -2370.3977 5051.9460 5109.9460 0.0613 0.6337 4 class -2325.0154 5095.2981 5178.2981 0.0470 0.6353 5 class -2276.2195 5137.1876 5246.1876 0.0510 0.6346 6 class -2243.8131 5211.8562 5346.8562 0.0421 0.6441 7 class -2158.4846 5180.6806 5341.6806 0.0284 0.7136

Table 14: Class specific statistics

By choosing the 3-class model, the related regression results for this model are interpreted (table 15). In this table the different coefficients, wald-statistics and the significance of the variables are displayed. To avoid confusion, the first Wald-statistic explains the significance of the specific variable on the dependent variable willingness to pay a price premium for NBs over PLs. The second Wald (=) statistic explains whether the variable is significantly different between the three segments.

Class 1 Class 2 Class 3 Wald Wald (=) Mean

(38)

38 authenticity gap Hedonic product value 2.0957 0.9437 2.4819 58.1038*** 8.4055** 1.7047 Utilitarian product value 0.6477 -0.0994 4.6479 11.6302*** 10.6825*** 0.9717 Covariates Gender Man -0.0293 -0.0759 0.1052 0.4155 Vrouw 0.0293 0.0759 -0.1052 Age 25-34 years 0.3500 -0.8967 0.5467 10.6103** 34-55 years -0.0652 0.5512 -0.4860 > 55 years -0.2848 0.3455 -0.0607 Income <€ 1.000 -0.0472 -1.4294 1.4766 17.1468* €1.000- €2.000 -0.4957 -0.1016 0.5973 €2.001-€3.000 0.1791 0.6027 -0.7818 €3.001-€4000 -0.1840 0.6056 -0.4216 €4.001-€5000 0.3134 0.3316 -0.6459 > €5.000 0.2334 -0.0088 -0.2245 Education Primary school -1.7645 2.1879 -0.4234 11.4989 LBO/ VMBO 1.0828 -1.1131 0.0304 HAVO 0.5688 -0.2594 -0.3094 VWO -2.9455 3.0439 -0.0983 MBO 1.2331 -1.7400 0.5069 HBO 0.8126 -0.9007 0.0881 WO 1.0128 -1.2186 0.2059 Household size 1 person 0.0631 0.3288 -0.3920 2.5151 2 persons -0.1111 -0.0021 0.1132 3 or more persons 0.0479 -0.3267 0.2788

*p< 0.1, **p<0.05, ***p <0 .001 Table 15: 3-class regression model

Interpretation of the regression analysis results

(39)

39 WTPPP. For segment 2, an increase of 1 point in the brand authenticity gap leads to an increase of 2.46 percent point in WTPPP, while for segment 3 this effect is even higher where an increase of 1 point in the brand authenticity gap leads to an increase of 6.63 percent point in WTPPP. Therefore, hypothesis 2 is accepted.

(40)

40 When looking at the wald (=) statistic, the variables brand authenticity gap, hedonic product value and utilitarian product value do significantly differ between the three segments. Furthermore, the covariates in the model predict class membership and in this model the covariates are effect-coded. However, only the covariates age and income do significantly predict class membership. Therefore, when describing the segments, the insignificant covariates of gender, education and household size will not be discussed or only briefly.

Description of the different segments

To get more insights in how the three classes differ from each other, the profile of the segments is investigated. First of all it is important to be aware of the class sizes of each segment (figure 4). The first segment is the largest, while the third segment is considerably the smallest segment.

Figure 4: Class size

(41)

41

Willingness to pay a price premium

Furthermore, the differences in the WTPPP between the three segments are displayed in table 5.

Figure 5: average WTPPP per segment

Figure 5 shows that there are differences in terms of the average WTPPP. For segment 3, the average WTPPP is by far the highest. While for segment 2, the lowest average WTPPP is found. This means that for segment 3, consumers are on average willing to pay approximately 40% more for NBs than for PLs. However, for segment 1 and 2, this is way smaller. Respondents in segment 1 are on average willing to pay approximately 15% more for NBs than for PLs, while respondents in segment 2 are willing to pay approximately 6% more for NBs over PLs.

Age

In figure 6, the differences in age distribution between the segments are shown.

Figure 6: Age distribution

0,00% 10,00% 20,00% 30,00% 40,00% 50,00%

Segment 1 Segment 2 Segment 3

% WTPPP

% WTPP 0,00% 10,00% 20,00% 30,00% 40,00% 50,00% 60,00%

(42)

42 At first sight, the differences in age distribution is not that clear. The probability that a respondents between 25-34 years belongs to segment 2 is the highest, while for the other segments this probability is also quite high (i.e. approximately 30% for segment 1 and 28% for segment 3). For respondents between 35-54 years old, the probability of belonging to segment 2 is also the highest, while for segment 3 this probability is also quite high with approximately 42%. While for 55 years and older, the probability of belonging to segment 1 is the highest, followed by segment 3 with 31%.

Income

For the variable income, the distribution between the segments are shown in figure 7.

Figure 7: Income distribution

Figure 7 shows that for income, a respondent that has a monthly gross household income less than €1.000 or between €1.000-€2.000 the probability of belonging to segment 3 is the highest. For a respondent that has a monthly gross household income between €4.001-€5.000 or higher than > €5.000 the probability of belonging to segment 1 is the highest. For a respondent that has a monthly gross household income between €2.001-€3.000 and between €3.001-€4.000 the probability of belonging to segment 2 is the highest.

0,00% 5,00% 10,00% 15,00% 20,00% 25,00% 30,00% 35,00%

(43)

43

Overview of segment information

To make the differences between the segments more clear, a table with all the segment information is displayed below (table 16).

Segment 1 2 3 R2 0.5003 0.4423 0.3562 Class size 45.49% 39.11% 15.40% Mean % WTPPP 14.67% 6.11% 40.27% Predictors Authenticity gap ++ + ++

Hedonic product value + +/- +

Utilitarian product value +/- +/- ++

Covariates

Age 25-34 years old/ 55

years and older

25-34 years old 35-54 years old/55 years and older

Income Higher than €4.001 Between €2.001 and €4.000

€2.000 or less

+/- = neutral + = important ++ = really important Table 16: Segment specific description

Detailed segment description

Segment 1: The wealthy youngsters and elderly

This largest segment contains the most wealthy consumers in the age category 25-34 years old and 55 years and older. However, with 14.67%, they do not display the highest WTPPP for NBs over PLs. Therefore, it seems that they are willing to pay a price premium for NBs over PLs, but this premium should not be too high. Further, the brand authenticity gap and the hedonic value is important to this segment in terms of a higher WTPPP. On the other hand, the utilitarian product value does not play a large role in a higher WTPPP.

Segment 2: Rational youngsters

(44)

44 over PLs, with only 6.11%. Therefore, this group seems to be the most price sensitive and not per se wanting to buy NBs. This groups also shows the lowest positive response for the brand authenticity gap in terms of WTPPP. Furthermore, the hedonic and utilitarian product value is not really that important for determining WTPPP. This segment therefore raises the assumption that they make very rational buying decisions and are price conscious.

Segment 3: Loyal adults and elderly

In this segment the most consumers are older than 35 years old and have relatively the lowest household income. However, interestingly, this segment displays the highest average WTPPP for NBs over PLs with 40.27%. This indicates that this group is very loyal to buying NBs, despite that they have a lower income and the availability of often lower priced PLs. They also respond the most positive compared to the other segments to the brand authenticity gap, hedonic product value and utilitarian product value in terms of WTPPP.

4.6 Evaluating the hypotheses

Apart from the first hypothesis, which was confirmed before, the other hypothesis are evaluated per segment (table 17).

Hypotheses Segment 1 Segment 2 Segment 3

H2: The perceived brand authenticity gap between NBs and PLs has a positive influence on consumers’ willingness to pay a price premium for NBs over PLs.

Confirmed Confirmed Confirmed

H3a: The level of hedonic product value positively moderates the relationship between the perceived brand authenticity gap and consumers’ willingness to pay a price premium for NBs over PLs.

Not confirmed Not confirmed Not confirmed

H3b: The level of utilitarian product value positively moderates the relationship between the perceived brand authenticity gap and consumers’ willingness to pay a price premium for NBs over PLs.

(45)

45 H3c: The moderating effect of the level of

hedonic product value on the relationship between the perceived brand authenticity gap and consumers’ willingness to pay a price premium for NBs over PLs, is higher than for the level of utilitarian product value.

Not confirmed Not confirmed Not confirmed

(46)

46

5. CONCLUSIONS AND RECOMMENDATIONS

5.1 Conclusions

NBs face intense competition from PLs in the FMCG (Nielsen, 2014). The high popularity of PLs leads to a drop in NB sales, and therefore NB manufacturers often choose to increase prices to compensate for the decrease in sales (Steenkamp et al. 2010). However, the increase is NB prices only leads to a perceived unjustified larger price difference between NBs and PLs and consumers are not willing to pay this premium. In the past, NBs could justify their price premium because the quality of the products were higher, nowadays this advantage is eroded (Apelbaum, et al. 2003; Steenkamp et al. 2010; Nielsen, 2014). Since quality is not a point of differentiation for NBs and PLs anymore, this study aims to investigate an alternative strategy that might yield a higher WTPPP for NBs over PLs, namely brand authenticity.

To get more insights in the role of brand authenticity and WTPPP a quantitative research was executed with 214 participants. In an online survey the participants compared brand authenticity aspects of NBs and PLs of four different product categories (i.e. chocolate, cola, spaghetti and yogurt) to each other and relating this to the WTPPP. The product categories in the research were chosen based on their inherent higher expected hedonic (chocolate and cola) or utilitarian product value (spaghetti and yogurt). Carefully choosing the products was important, because it was expected that the level of perceived hedonic or utilitarian product value had a moderating effect on the relationship between the brand authenticity gap and the WTPPP.

(47)

47 12.64% and 11.37%. This could be explained by the assumption that consumers are willing to pay more for hedonic than for utilitarian products (Sethuraman & Cole, 1999). Despite these differences, WTPPP is still low in all categories, and this low percentage increases the need for NBs to have insights into the factors that drive WTPPP to increase NB performance again. Therefore, it is important to investigate the possibilities of alternative strategies such as brand authenticity instead of only focusing on price and quality.

Furthermore, an expected positive significant brand authenticity gap between NBs and PLs is found. The average authenticity for NBs is approximately 1 point. While this number do not have a concrete meaning, it could be concluded that the difference between the scores of NBs and PLs is not very high. Between product categories, the differences in the brand authenticity gap does not deviate too much from the approximately 1 point difference between NBs and PLs, however cola has the highest brand authenticity and yogurt the lowest. This could mean that there are in basis some differences in authenticity ratings which could be product related. When comparing these results to the study of Kadirov (2015), the mean gap in this study is smaller and the coefficient of variation is larger, while the percentage of consumers that perceive no quality gap, a negative quality gap or a positive quality gap are approximately the same. This could indicate that a higher popularity of PLs results in a decrease of the perceived brand authenticity gap. While this finding may be biased by the products and the related NBs chosen, it gives an indication that NBs are perceived as more authentic, which is an important finding and conclusion because it gives NBs the opportunity to act on this advantage. However, NB managers should also focus on increasing the brand authenticity gap since the increase of PL popularity possible leads to a decrease in the authenticity gap. For increasing the authenticity gap, NB managers could dive into literature that investigated the effect of isolated authenticity appeals on the increase in WTPPP, e.g. Skuras & Vakrou (2002) and Beverland (2005b).

(48)

48 therefore, when implementing strategies to increase perceptions of brand authenticity this finding contributes to the confidence that at least a lot of consumers do respond positively to brands that are perceived more authentic.

(49)

49 As mentioned briefly in the previous paragraph, the study found 3 different consumer segments, which significantly differ in terms of their responses towards the brand authenticity gap, hedonic product value, utilitarian product value. Furthermore they could be separated based on the demographic variables age and income. Segment 3 showed the highest response to the brand authenticity gap in terms of WTPPP and segment 2 the lowest. When looking at segment 3, most consumers are older than 34 years old and have a relatively low income. However they do also display the highest average WTPPP. Therefore, this segment is likely to be loyal to NBs already and WTPPP could be increased by increasing the brand authenticity gap. Therefore, this segment is quite attractive where brand managers could focus on with advertisements about authenticity. To conclude: Brand authenticity- the new road to success for NB manufacturers? Yes, brand authenticity could definitely be a new fruitful strategy for NBs in their battle against the popularity of PLs. NBs are perceived as more authentic and consumers are willing to pay a higher price premium when the brand authenticity gap between NBs and PLs is larger. Therefore, NB manufacturers should focus on how to maintain and increase the brand authenticity gap between the NBs and PLs to create a new strategy to enhance NB performance again.

5.2 Limitations and recommendations for further research

While in this study interesting insights are provided in terms of the brand authenticity gap and WTPPP, this research does have some limitations.

(50)

Brand authenticity – The new road to success for National Brands?

Brand authenticity – The new road to success for National Brands?

Brand authenticity – The new road to success for National Brands?

MANAGEMENT SUMMARY

PREFACE

TABLE OF CONTENTS

1. INTRODUCTION

2. THEORETICAL FRAMEWORK

3. RESEARCH DESIGN

4. RESULTS

% WTPPP

5. CONCLUSIONS AND RECOMMENDATIONS