“Bloggers or celebrities?”
An analysis on consumers’ identification with bloggers and brands
Study: MSc. in Business Administration - Marketing track
Institution: University of Amsterdam, Amsterdam Business School Course: Master Thesis – final version
Thesis supervisor: M. Mossinkoff PhD
Student: Nila Nallathamby Student number: 10894357 Date: June 20th, 2016
2
STATEMENT OF ORIGINALITY
This document is written by Nila Nallathamby who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
ABSTRACT
The Internet has shaken the Marketing world. The rise of Social network sites, review sites and blog platforms has made consumers more sceptical and powerful. The consequence is a decrease in the effectiveness of mass media channels. As a response to this shift, traditional marketing strategies for mass media channels such as celebrity endorsements are gradually being combined or replaced with blogger endorsements as bloggers are seen as authentic and credible sources by consumers. Moreover, it is a very accessible format for companies due to its little cost compared to celebrities. It has already been found that bloggers affect brand attitudes and purchase intentions. However, little is known whether bloggers establish the same effect as celebrities do: facilitating brand identification. Even more so, little is known about how the effects of endorsers’ characteristics differ on consumer-brand identification. Therefore, this study has researched to what extent the blogger evokes consumer
identification with the said blogger, to what extent this leads to consumer-brand identification with the brand the blogger endorses and whether the characteristics of the blogger have an influence on this relationship. Using data from an experimental survey from 244 women in the age of 14 to 35, I found that consumers identify with fashion bloggers and that this process leads to consumer-brand identification, even when controlled for previous brand identification or memorable brand experiences. Furthermore, I identified that the
characteristics of the fashion blogger – source credibility, social influence and writing style – do not have a significant effect on this relationship. However, although it was not significant, unexpected results were found. To state, low credibility (vs. high credibility) of the blogger and a blogpost without a story (vs. with a story) both appeared to evoke a slightly higher value of consumer-brand identification. Implications of the findings are discussed. This thesis ends with suggestions for future research.
4
TABLE OF CONTENTS
1. INTRODUCTION ...7
2. LITERATURE REVIEW ... 10
2.1 Expressing identities by using brands ... 10
2.2 Consumer-brand identification ... 12
2.2.1 The CBI construct ... 14
2.3 Celebrities as facilitators of CBI ... 15
2.4 Bloggers as possible facilitators of CBI... 17
2.5 Operationalization of consumers’ identification with bloggers ... 19
2.5.1 Source credibility ... 22
2.5.2 Social influence ... 24
2.5.3 Writing style ... 28
2.6 Literature gaps and research question ... 30
2.7 Theoretical and managerial contribution ... 32
3. CONCEPTUAL FRAMEWORK... 33
3.1 Hypotheses ... 33
4. RESEARCH METHODOLOGY ... 36
4.1 Sample ... 36
4.2 Research design... 37
4.2.1 Experimental survey and procedure ... 38
4.2.2 Measures... 40
4.3 Procedure data collection... 44
4.3.1 Pre-test... 44
4.3.2 Pilot study ... 46
5
5. RESULTS AND ANALYSIS ... 47
5.1 Preliminary analysis ... 47
5.2 Reliability and factor analysis and scale descriptives ... 48
5.3 Analysis results ... 49
5.3.1 Testing hypotheses ... 49
5.3.2 Analysis with control variables ... 52
6. DISCUSSION AND CONCLSUION ... 56
7. MANAGERIAL IMPLICATIONS ... 61
8. LIMITATIONS AND FUTURE RESEARCH ... 63
9. REFERENCES ... 66
10. APPENDICES ... 73
I. Experimental survey main-study (English version) ... 73
II. Experimental survey main-study (Dutch version)... 79
III. Introduction page Nadina Gorano in experimental survey ... 85
IV. Blogger homepages in experimental survey ... 89
V. Blogposts in experimental survey ... 93
VI. Description of operationalization blogger characteristics ... 95
VII. Experimental survey pre-test ... 96
6
LIST OF FIGURES AND TABLES
Figure 1. Traits of the ideal social influencer (Langner, Hennigs & Wiedmann, 2013) ... 25
Figure 2. Marketing Influence Value model (Li, Lai & Chen, 2011) ... 26
Figure 3. Conceptual framework ... 33
Table 1. Gaps in the current literature ... 31
Table 2. 2 x 2 x 2 Factorial Design ... 38
Table 3. Results of pre-test: perceived source credibility and social influence by treatment. 45 Table 4. Descriptives of independent and dependent variables and correlation (n=244) ... 48
Table 5. ANCOVA: Test between Subjects effects - main effects and interaction effects ... 50
Table 6. Regression of consumer-blogger identification on CBI (n=244) ... 50
Table 7. ANCOVA: Test between Subjects effects controlled for previous brand identification - main effects and interaction effects (n=244) ... 53
Table 8. Regression of consumer-blogger identification on CBI controlled for previous brand identification (n=244). ... 53
Table 9. ANCOVA: Test between Subjects effects controlled for memorable brand experiences - main effects and interaction effects (n=244). ... 54
Table 10. Summary table of hypotheses. ... 56
Table 11. Factor loadings for the 7 items which belong to source credibility and the 9 items which belong to social influence (n=81) ... 100
Table 12. Factor loadings for the 9 items which belong to consumer-blogger identification and the 7 items which belong to CBI (n=244). ... 100
7
1. INTRODUCTION
Bloggers or celebrities? That is the question that inspired me to write this Master thesis. For decades, marketers have been using celebrities as endorsers for their brands. Think of Nicole Kidman for Chanel No. 5 or Michael Jordan for Nike. The effects of celebrity endorsement are proven to lead to greater purchase intentions (Dwivedi, Johnson & McDonald, 2015; Till, Stanley & Priluck, 2008). A reason for this is that consumers perceive themselves congruent to the celebrity, because they want to identify with their role model. Consumers choosing brands that are endorsed by a role model will get their identity updated, because the endorsed brands are associated with celebrities and hence express a desired and meaningful identity (Escalas & Bettman, 2003; Escalas & Bettman, 2005). It could be in the end that these consumers will identify themselves with the brand alone. This is called consumer-brand identification which is proven to lead to greater brand loyalty (Stokburger-Sauer, Ratneshwar & Sen, 2012).
A disadvantage of celebrity endorsement is the high cost paired with it. Since the invention of Web 2.0 diverse online social networks have arisen. Consumers generate online content and many of them are being followed by millions of other consumers. Consumers listen more to each other now than ever before. Many firms acknowledge this and there is a viewable shift of firms collaborating with these social influencers and role models: the bloggers. Are bloggers effective for brand identification? Do consumers choose brands to identify with bloggers? If that is true, bloggers would save companies a lot of money. Smaller companies with low marketing budgets could profit from this as well.
This thesis gives insights into the ability of bloggers in facilitating consumer-brand identification. Moreover, this study investigates to what extent the characteristics of bloggers – i.e. source credibility, social influence and writing style – have an influence on consumer-brand identification. The study is focused on fashion blogs as nowadays fashion blogs
8
represent the largest segment of the blogosphere (Halvorsen, Hoffman, Coste-Manière & Stankeviciute, 2013). Moreover, fashion brands naturally act as symbols of personal and social identity and reduce social risk to consumers (Tynan, McKechnie & Chhuon, 2010). The operational research question of this thesis is: “To what extent does the consumers’ perceived similarity with the fashion blogger lead to consumer-brand identification and to what extent do characteristics of the fashion blogger moderate this relationship?”. The answer of this question is highly relevant, because bloggers can make company advertisements less influential (Gensler, Völckner, Liu-Thompkins & Wiertz, 2013). Consumers rely on the honesty of bloggers because bloggers are allegedly helpful and independent from firms. Moreover, identification with bloggers increases the likelihood that the message about a brand is perceived as persuasive (Babic, Sotgiu, de Valck & Bijmolt, 2015). It is highly interesting for firms to know to what extent such collaboration can enhance the overall consumer evaluation of their brands and which characteristics of bloggers are of importance. Further, this study gives clarity in the extent of the bloggers’ ability to contribute to brand personality management, which includes the imagery of brand users, brand emotional values and brand identity (Demirbag, Kaplan, Yurt, Guneri & Kurtulus, 2010).
Apart from the managerial implications, this thesis also contributes to the literature in several ways. First, since the popularity of bloggers has increased in the last five years (Li, Lai & Chen, 2011), not much research is done in this field. This thesis brings more insights into the phenomenon of bloggers and especially, in their role as facilitators of consumer-brand identification. Second, current literature on consumer-brand identification including
endorsers, has only focused on the congruity between the endorser’s identity and the brand identity (Dwivedi, Johnson & McDonald, 2015). To deepen the current literature, this thesis has looked at characteristics of the blogger that may moderate the relationship between consumer-blogger and consumer-brand identification.
9 The research has been conducted via an experimental online survey. Each respondent has been enrolled in one treatment, where three characteristics of bloggers, i.e. source
credibility, social influence and writing style, were manipulated. In this way, the effect of bloggers’ characteristics on consumer-brand identification has been identified. Most importantly, the research has shown whether bloggers facilitate consumer-brand identification.
Thesis overview
This thesis is structured as follows: chapter 2, which is the next chapter, provides a comprehensive literature review of the key concepts in this study. Based on the literature review, the conceptual model and hypotheses are described in chapter 3.
In chapter 4 the research methodology of the study is described. The results of the study are presented in chapter 5.
The discussion and conclusion of the results are given in chapter 6. This is followed by a discussion of the managerial implications in chapter 7. This thesis ends with the limitations and future research suggestions in chapter 8.
10
2. LITERATURE REVIEW
In this chapter a comprehensive literature review is given. The review aims in providing what has already been studied about identification between persons and brands. First, general information about identification with a brand is given. This is followed up by a review of the consumer-brand identification literature. Second, an analysis on the process of identification with a celebrity is given. This describes the way how consumers may identify with bloggers. Last, a detailed analysis of bloggers is given and characteristics of bloggers that may
moderate consumer-brand identification are identified. Based on the literature review, a research gap is identified and the resulting operational research question is formulated. This chapter ends with the theoretical and managerial contribution of this study.
2.1 Expressing identities by using brands
What is a brand? Several definitions of a brand can be found in the literature. One of the most used definitions is from Kapferer (1992). The author states that a brand is not a product, but it is the product essence, its meaning and its direction and it defines its identity in time and space. According to Bhattacharya & Sen (2003) a brand identity should be “the distinctive and relatively enduring characteristics of a focal brand” (He, Li & Harris, 2012, p.649). Another definition of a brand which is often used is from Franzen (1990), who defines a brand as a ‘network of associations’ in the brain of the consumer. According to Demirbag, Kaplan, Yurt, Guneri & Kurtulus (2010), these associations are based on experiences, impressions and perceptions of the functional, emotional, and symbolic benefits of the brand. Therefore, brands generate emotional feelings towards consumers (Stokburger-Sauer, Ratneshwar & Sen, 2012) as brands convey psychological, symbolic and experiential values (Escalas & Bettman, 2003).
11 A brand identity can be appropriated by consumers (Stokburger-Sauer, Ratneshwar & Sen, 2012; Bhattacharya & Sen, 2003). The process of appropriation is called consumer-brand identification and will be discussed in section 2.2.
Especially luxury brands express an identity because of their strong symbolic value and therefore high brand meaning (Escalas & Bettman, 2003; Escalas & Bettman, 2005). According to Bhattacharya & Sen (2003) a brand has a strong and attractive identity when the identity is distinctive and prestigious. To state, consumers buy luxury brands especially to self-present or to get recognized by members of a society (Siverstein & Fiske, 2003 in Hwang & Kandampully, 2012). Thus, except for the ability of brands to construct or express the consumers’ identity, brands also express a social identity. A main explanation can be found in the theory of brand associations. To give an example, when members of a cultural group use a brand, certain brand associations arise which are related to the members of the cultural group (Escalas & Bettman, 2003). Liu, Li, Mizerski & Soh (2012) found that elements of brand associations - brand user and brand usage imagery - are strong predictors of brand loyalty when these elements are congruent with the consumer. Brand user imagery congruity refers to the degree of perceived similarity between the typical user of a brand and the consumer (Sigry et al, 1997 in Liu, Li, Mizerski & Soh, 2012). Brand usage imagery congruity refers to “the association between consumers’ perceptions of the typical use of a brand and how the brand is perceived appropriate regarding this situation of use” (Sotiropoulos, 2003 in Liu, Li, Mizerski & Soh, 2012, p.923). Obviously, a brand that is used by different types of people will not clearly communicate the consumer’s identity (Escalas & Bettman, 2005). Hence, a brand meaning is partly created by the society as identity always has a socio-cultural origin (Elliott, 1997). However, this study focuses on personal identity.
12
2.2 Consumer-brand identification
When the identity of a brand is being appropriated by a consumer, consumer-brand
identification occurs. Brand identification has a significant impact on consumer behaviour. Namely, brand preference, consumer loyalty and the psychological sense of brand community are found to be positively affected when consumers identify with a brand (Tuškej, Golob & Podnar, 2013).
As described in the previous section, consumer-brand identification, hereafter: CBI, is the concept of a consumer who identifies with a brand. These consumers use brands to construct and express their identities (Belk 1988; Holt 2002 in Gensler, Völckner, Liu-Thompkins & Wiertz, 2013). Although scholars have given the concept of CBI different definitions, CBI has always been viewed as a process of definition. The process of self-definition is a process in which consumers enhance their self-identity or verify their identity by using a brand. Stokburger-Sauer, Ratneswar & Sen (2012), who were the first that transferred the Social Identity Theory to the marketing field, defined CBI as: “a consumer’s perceived state of oneness with a brand” (Stokburger-Sauer, Ratneswar & Sen, 2012, p.407). According to the authors, identification occurs when there is congruency between the
perceived brand identity and the self-identity of the consumer (Stokburger-Sauer, Ratneshwar & Sen, 2012). Tuškej, Golob & Podnar (2013) defined CBI as: “the individual’s sense of sameness with a particular brand” (Tuškej, Golob & Podnar, 2013, p.53).
In the literature, another version of CBI exists, namely: self-brand connections. Similar to CBI, self-brand connections result in a process of self-definition (Escalas & Bettman, 2003). Escalas & Bettman (2003) define self-brand connection as: “the extent to which an individual has incorporated a brand into his or her self-concept” (Escalas &
Bettman, 2003, p.339). Self-brand connection is conceptualized slightly different from CBI, because self-brand connections include “the potential motivations guiding self-brand
13 connections, such as communicating one's identity to others” (Stokburger-Sauer, Ratneshwar & Sen, 2012, p.407). Moreover, interesting enough, CBI and self-brand connections have a different theoretical origin. Namely, self-brand connection is built upon the theory of brand user associations developed by Keller (1993) and upon the theory of transferring meaning to brands developed by McCracken (1986) (Escalas & Bettman, 2003). The theory of Keller (1993), applied to self-brand connections, states that when attractive people use a brand, positive associations will arise about the brand because attractive people use it. And to follow up, the theory of McCracken (1986) states that the meaning of the brand will change in the direction of the personality and lifestyle meanings of the attractive people. As a result, other consumers will be able to enhance their self-concept by using the brand, or the opposite: people will not use the brand because they do not want to be associated with the typical user of the brand.
CBI originates from the Social Identity Theory (SIT) and is developed by Stokburger-Sauer, Ratneswar & Sen (2012). SIT is developed by Tajfel (1979) and Tajfel and Turner (1986). SIT addresses identity on individual as well as on social level. When identity on individual level is applied on the CBI concept, it means that consumers identify with a specific brand and develop feelings of affinity towards the brand (Tuškej, Golob & Podnar, 2013). He, Li & Harris (2012) found that a strong brand identity leads to higher degrees of CBI. When identity on the social level is applied on the CBI concept, it means that “brands act as a communication instrument to manifest the desire to associate with or to dissociate from other individuals” (Del Rio et al., 2001, p.412 in Tuškej, Golob & Podnar, 2013).
Therefore, the SIT is composed of three elements: (1) categorisation, (2) identification and (3) comparison. This means that a consumer makes imagined categories to put other people and himself into, identifies with a certain group (the in-group) and compares this group with other groups (the out-group) (Stokburger-Sauer, Ratneshwar & Sen, 2012). Several studies have
14
shown that brands which are used by an aspiration group enhance the brand usage of the consumer with the result of a higher loyalty towards the brands (Tuškej, Golob & Podnar, 2013; Stokburger-Sauer, Ratneshwar & Sen, 2012; Escalas & Bettman, 2003).
In conclusion, CBI and self-brand connections are both formed when an attractive group or person with whom someone (wants to) identifies uses a brand. The brand will gain a personal meaning to the consumer because of the positive associations with the brand. In the following, the antecedents and consequences of CBI are explained.
2.2.1 The CBI construct
In this section, CBI as construct is explained to get a rich understanding about the concept. The antecedents, moderator and consequences of CBI which are identified by Stokburger-Sauer, Ratneshwar & Sen (2012) are discussed in the following paragraphs.
According to Stokburger-Sauer, Ratneshwar & Sen (2012), the antecedents of CBI are (1) brand–self similarity (i.e. the brand’s personality is similar to the person), (2) brand distinctiveness (i.e. unique or distinctive brand), (3) brand social benefits (i.e. the brand leads to social connections), (4) brand warmth (in emotional terms) and (5) memorable brand experiences (Stokburger-Sauer, Ratneshwar & Sen, 2012). Product category involvement is found as a moderator on CBI. According to Stokburger-Sauer, Ratneshwar & Sen (2012), product category involvement is the consumer’s perceived relevance of a product category, which is based on personal values.
Consequences of CBI are (1) brand loyalty (i.e. “a deeply held commitment to rebuy or re-patronize a preferred brand in the future” (Oliver, 1999, p.34)) and (2) brand advocacy on social and physical level (Stokburger-Sauer, Ratneshwar & Sen, 2012). On the social level, brand advocacy can result in recommendations about the brand or defences when the brand is being attacked. On the physical level, brand advocacy can result in different forms, from
15 buying and using products with the brand logo to putting tattoos with the brand logo
(Stokburger-Sauer, Ratneshwar & Sen, 2012).
Providing the fact that CBI leads to higher brand loyalty and advocacy, it is interesting to search for ways that facilitate CBI. In the current literature, it is proven that aspirational reference groups, in particular communities and celebrities, facilitate CBI (Dwivedi, Johnson & McDonald, 2015; Escalas & Bettman, 2003). In terms of operationalization, the scale self-brand connections has mostly been used in studies as this concept consists of additional items that measures consumers’ motivations that guides self-brand connections (e.g. “I use this brand to communicate who I am to other people”). Therefore, it is decided to operationalize CBI in this study as a combination of CBI and self-brand connections. So, how do celebrities facilitate CBI? This is discussed in the following section.
2.3 Celebrities as facilitators of CBI
Celebrities are symbolic aspirational reference groups (Escalas, 2004). Originally, an
aspirational reference group is an individual or group where the consumer does not belong to, although the consumer wants to because this group possess a value-expressive appeal (Grier and Deshpande, 2001). Aspirational reference groups influence opinions, beliefs, attitudes and behaviours and assist consumers in gaining a desired identity (Dwivedi, Johnson, &
McDonald, 2015). This theoretical perspective derives from SIT as described before and suggests that a brand used by an aspirational reference group will be evaluated more positively (Dwivedi, Johnson & McDonald, 2015). This is, because aspirational reference groups evoke positive associations about themselves. Therefore, they generate positive associations towards the brands they use. Hence, consumers who (want to) identify with a celebrity, will choose the same brands used by the celebrity. As a result, consumers will meet their self-related needs (Escalas & Bettman, 2003), because the celebrity’s personality- and lifestyle-related meanings are transferred into the endorsed brands (McCracken, 1989).
16
Elaborating on this, it should be likely that celebrities facilitate CBI, because celebrities establish a valued identity for the brand. Moreover, Dwivedi, Johnson &
McDonald (2015) have shown that when the celebrity is perceived as credible, consumers will perceive the brand that is endorsed by the celebrity also as credible. To have effective
celebrity endorsement, studies have found that the celebrity’s identity has to be congruent with the brand identity (Dwivedi, Johnson & McDonald, 2015; Till, Stanley & Priluck, 2008).
In addition to SIT, the theory of perceived similarity can also explain the power of aspirational reference groups. Perceived similarity promotes the desire to be like the other that is perceived as favourable or rewarding, because certain similarities signal that it is possible for the consumer to become like the other (Hoffner and Cantor, 1991 in Hoffner & Buchanan, 2005). The similarity between an information source, in this context a celebrity, and an information seeker is called homophily. This concept is discussed in section 2.5. It has been found that the greater the homophily between an information seeker and an information source, the more persuasive the message of the information source is (Li & Du, 2011). This clarifies the fact that endorsing celebrities can persuade consumers to use the endorsed brands to express their personal identity. So, to summarize, celebrities represent aspirational
reference groups to consumers. Consumers use endorsed brands if they (want to) identify with celebrities. As a result, consumers may identify with the brand alone – the celebrity has facilitated CBI. As stated earlier, it is interesting to look at other ways that can facilitate CBI. Bloggers are the latest phenomenon of social influencers (Booth & Matic, 2011) and act as role models to consumers. That means that bloggers may also serve as aspirational reference groups. Companies, especially luxury brands already collaborate with bloggers to promote their brands (Halvorsen, Hoffman, Coste-Manière & Stankeviciute, 2013).However, it is not yet known whether bloggers serve as aspirational reference groups and if they are able to
17 facilitate CBI. The following section emphasizes the current literature on bloggers to deal with this question.
2.4 Bloggers as possible facilitators of CBI
According to Segev, Villar & Fiske (2012) a blogger is a mix of a role model, an opinion leader and social influencer. The latter two concepts are nearly the same. Opinion leaders are defined as “the individuals who were likely to influence other persons in their immediate environment” (Katz & Lazarsfeld, 1955, p.3 in Watts & Dodds, 2007). Influentials are “a minority of individuals who influence an exceptional numbers of their peers and are important to the formation of public opinion” (Watts & Dodds, 2007, p.441). Bloggers are highly
informed individuals who are connected with several ties. They have a direct influence on consumers as they write in an informal, conversational style (Segev, Villar & Fiske, 2012).
There are a variety of bloggers as they write about a variety of topics, for example food, fashion or technology. Bloggers use their blog to diffuse information about their particular topic of interest. This can be in the form of photos and informative text or
emotionally loaded stories. To give a short explanation of a blog, a blog is an interactive form of online media, where bloggers provide their readers with up-to-date information and the opportunity to interact with other readers and the blogger (Segev, Villar & Fiske, 2012). Features of blogs that provide interaction and facilitates conversations are hyperlinks, comments and blog rolls (Willi, Melewar & Broderick, 2013). Readers do not have to join discussions on the blog, which makes it easy for readers to follow a blogger. Blogs can have a far reach, as it is online and easy accessible (Segev, Villar & Fiske, 2012).
By diffusing information via blogs, bloggers create electronic word of mouth (eWOM). Word of mouth is described as: “informal communications directed at other consumers about the ownership, usage, or characteristics of particular goods and services or their sellers” (Westbrook, 1987, p.261 in Berger, 2014). This can appear in the form of mere
18
mentions to active recommendations (Berger, 2014). Because (e)WOM is disseminated by peers rather than companies, (e)WOM is perceived as more credible (Godes, Mayzlin, Chen, Das, Dellarocas, Pfeiffer, Libai, Sen, Shi and Verlegh, 2005). Hence, bloggers communicate in a natural and personal way which provides a certain degree of perceived credibility (Segev, Villar & Fiske, 2012). However, it is not sure what influential impact blogger have on
consumers, since the study by Watts & Dodds (2007) revealed that mostly large cascades of influence are not driven by influentials, but by a critical mass of easily influenced individuals. Anyhow, some bloggers are being perceived as experts on a particular topic and get invited on television shows and as writers for magazines. Hence, bloggers can influence opinions,
attitudes, beliefs, motivations and behaviours of others.
There are different type of blogs, such as individual blogs, community blogs and blogs set up by companies. Also, there are a variety of bloggers as their particular topic of interest can differ. In this thesis, the focus is on individual fashion bloggers using inside blogs. Inside fashion blogs are blogs written by a fashion blogger with knowledge of the fashion industry (Willi, Melewar & Broderick, 2013). Fashion blogs focuses on fashion brands, fashion products, street style and personal style (Halvorsen, Hoffman, Coste-Manière &
Stankeviciute, 2013). It is decided to focus on fashion bloggers, because fashion bloggers tend to have an important role as they can reduce social risk for consumers. Moreover, fashion brands and products act as symbols of personal and social identity (Tynan, McKechnie & Chhuon, 2010) which makes it more interesting to study CBI facilitated by fashion bloggers.
An individual blogger may be able to facilitate CBI if the blogger represent an
aspirational reference group to consumers. Several studies have demonstrated that homophily increases WOM with as consequence adoption of behaviours (Berger, 2014; Gensler,
Völckner, Liu-Thompkins & Wiertz, 2013). This would mean that consumers identifying with fashion bloggers would also identify with the brand the blogger wears. So in conclusion, the
19 study focuses on identification with the individual fashion blogger who is possibly facilitat ing CBI.
Next, the operationalization of identification with bloggers is discussed by reviewing literature on consumer’s identification with celebrities as this is already studied.
2.5 Operationalization of consumers’ identification with bloggers
How should the consumer’s identification with bloggers be measured? Current literature on brand identification through celebrities exists, which could be applied on identification with bloggers. However, the literature is limited. To date, studies on celebrity endorsement have mostly focused on two models of effective celebrity endorsement rather than identification with celebrities. The models are: (1) the credibility model (e.g. honesty, reliability of the endorser) and (2) the attractiveness model (e.g. beautiful, classy endorser) (Till, Stanley & Priluck, 2008). These models suggest that the celebrity should be perceived as credible and attractive by the consumer in order to get an effective brand endorsement. However, Basil (1996) did focus on consumers’ identification and conducted a study on consumers’
identification with celebrities to study message effects. The author found that a celebrity with whom consumers identified, insured the greatest likelihood of achieving lasting attitude or behaviour change. Basil (1996) measured identification through the concept of parasocial relationship. Parasocial relationship is widely used in literature concerning media
personalities and is described as an individual’s identification with a media personality. The identification occurs because individuals think or feel as if they know the media personalities to which they are exposed (Rubin & McHugh, 1987 in Basil, 1996). Basil (1996) measured parasocial relationship through a scale with 8 items (e.g. “How much do you like celebrity X”, “Do you have feelings for celebrity X”, “Do you feel that celebrity X can be a friend of you”) (α=.84). This scale is reliable, because the Cronbach’s alpha (α) is above 0.8 (Field, 2013). The study of Theran, Newberg & Gleason (2010) also measured identification through
20
parasocial relationship. This study looked at teenage girls’ identification with fictional television characters. The authors developed two parasocial interaction (PSI) subscales, ‘involvement’ and ‘emotional interaction’, to measure the parasocial relationship. The involvement scale consists of 20 items to measure the extent of adolescents’ involvement in PSI. Items describe behaviours and feelings toward a favourite media figure (e.g. “I think this person is like an old friend”) (α=0.93). The PSI scale emotional intensity consists of 6 items (e.g. “How much would you like to be like her?”, “How connected do you feel to her?”, “How much would you like to meet her?” (α=0.82).
While Basil (1996) and Theran, Newberg & Gleason (2010) measured identification through the concept parasocial relationship, there are also studies that have measured
identification through perceived similarity (i.e. homophily). Hoffner & Buchanan (2005) also conducted a study on consumers’ identification with television characters, but measured consumers’ identification through perceived similarity. These authors define identification as: “the desire to be like or act like the character” (Hoffner & Buchanan, 2005, p.325). The authors argue that individuals who identify with for instance a television character, will adopt traits, attitudes or behaviours of the television character or will incorporate the characteristics of the character into his sense of self. As described before, perceived similarity or homophily promotes the desire to be like a person as certain similarities signal that it is possible to become like the other person (Hoffner and Cantor, 1991 in Hoffner & Buchanan, 2005). Hoffner & Buchanan (2005) measured perceived similarity on the attitude similarity subscale (e.g. “thinks like me”) of the Perceived Homophily Measure by McCroskey, Richmond & Daly (1975) (α=0.81 for female characters). The construct perceived similarity is also used in literature concerning organizational identification, which derives from SIT. Namely,
Bhattacharya & Sen (2003) identified perceived identity attractiveness and perceived identity similarity as drivers of consumer-company identification. The authors argue that consumers
21 should identify more with a company if they find the company attractive and consumers should find self-relevant information that are similar to their own easier to focus on.
Moreover, Hahn & Lee (2014) conducted a study to test whether the psychological closeness of the consumer with a fashion blogger led to a positive attitude toward the fashion blog and purchase intention toward the fashion products endorsed by the blogger. The concept “psychological closeness” is defined as perceived similarity (Tesser & Campbell, 1980 in Hahn & Lee, 2014). In this study, Hahn & Lee (2014) measured psychological closeness by combining items from two other studies. These are the subscale ‘appearance’ of the Perceived Homophily Measure by McCroskey, Richmond & Daly (1975) and the PSI scale ‘emotional intensity’ by Theran, Newberg & Gleason (2010).
In addition, it is found that consumers’ identification with communities also can result in CBI. Surprisingly, studies concerning the consumer’s identification with communities also used items from the Homophily Measure by McCroskey, Richmond & Daly (1975) and the PSI subscale ‘emotional intensity’ by Theran, Newberg & Gleason (2010). Specifically, the concept was operationalized by consumers’ self-assessed fit with each type member group (e.g. “I like the people in type X”, “I fit in the type X category”) and through aspiration group fit (e.g. “I would like to be a part of the type X group”) (Escalas & Bettman, 2003).
To conclude, it seems that consumer-blogger identification can best be measured on the basis of consumer’s perceived similarity with the fashion blogger. This could be measured through items from the Perceived Homophily Measure by McCroskey, Richmond & Daly (1975) and through the PSI subscale ‘emotional intensity’ by Theran, Newberg & Gleason (2010). So, the consumer’s perceived similarity with the fashion blogger may lead to CBI.
It is interesting to look at characteristics of bloggers that may affect the strength of this relationship. Therefore, in the following three sections, three characteristics of bloggers that could moderate the relationship between consumers-blogger identification and CBI are
22
discussed. The characteristics discussed are source credibility, social influence and writing style.
2.5.1 Source credibility
Source credibility is an important characteristic, because it is associated with reducing risk. Individuals with high perceived credibility are more able to persuade consumers (Kang, 2010). Credibility has been discussed in the literature from three perspectives: medium credibility, content credibility and source credibility (Metzger et al., 2003 in Kang, 2010). Medium credibility is the perceived level of credibility of a specific medium, in the context of this study: the blog. Content credibility is the perceived credibility of the communicated message. Indicators for content credibility could be accuracy and informational quality. The last, source credibility is the perceived credibility of the sender of the message, the blogger. Source credibility consist of several dimensions, however, research on source credibility has mainly focused on the expertise dimension and the trustworthiness dimension (Kang, 2010). Source expertise is based on skills, competences, or knowledge through experience (Ohanian, 1990). An individual with high expertise can be described as knowledgeable, reputable and competent. Trustworthiness refers to goodness or morality of the source. An individual who is perceived as trustworthy can be described as well-intentioned, truthful or unbiased (Tseng & Fogg, 1999 in Rubin & Liddy, 2006). It could be argued that bloggers are generally seen as trustworthy individuals, because they spread information online which has a far reach. If their generated eWOM goes viral, it is because they are perceived as credible. It is shown, namely, that (e)WOM has a high power due to source reliability and interpersonal communication (Chung & Darke, 2006; Day, 1971 in Buttle, 1998). Interpersonal communication is argued to be an important driver of credibility, since interpersonal channels have been shown to be more effective in changing brand attitudes and purchase intentions than mass media channels
23 Source credibility is a component of reliability. However, for this study it is decided to study source credibility rather than reliability. This is because reliability also encompasses the dimension: “attractiveness of the source” (Ohanian, 1990). The attractiveness dimension is mainly applied in celebrity endorsements studies. Furthermore, manipulation of attractiveness is not possible since it is decided to work with one blogger to insure consistency. Moreover it is not ethical because it involves a real person. This means that the attractiveness of the blogger will not be taken into account. It should also be noted that the expertise dimension will not be taken into account in the source credibility construct, since this thesis also studies the social influence of the blogger. Social influence is hypothesized as another moderator on the relationship between consumer-blogger identification and CBI. The expertise of the source is regarded to be of more importance in the social influence construct than in the source credibility construct. Separation of expertise from source credibility is possible, because Willemsen, Neijens & Bronner (2012) concluded that online product reviews can induce differential effects on both perceived expertise and perceived trustworthiness.
However, these dimensions have to be examined separately because experts were perceived as having more expert knowledge and at the same time as having less trustworthiness than peers. Therefore, in this thesis the social influence construct will encompasses the expertise
dimension, and the source credibility construct will encompass interpersonal communication and trustworthiness. The dimensions will be used to design and to measure the concepts.
Why should source credibility drive consumers’ identification with the blogger? This has to do with associations with the source. According to Ohanian (1990), celebrities who are perceived by consumers as credible carry favourable associations of attractiveness,
trustworthiness and expertise. According to SIT, positive associations about other individuals or groups can lead to identification with them (Stokburger-Sauer, Ratneshwar & Sen, 2012). Thus, when the blogger is perceived as more credible, consumers may identify more with the
24
blogger. Moreover, prior research has shown that signals of homophily and trustworthiness are of high importance in the effectiveness of eWOM (Babic, Sotgiu, de Valck & Bijmolt, 2015). This means that messages coming from senders who are perceived as similar and as trustworthy to the receiver are more persuasive. Also Dwivedi, Johnson & McDonald (2015) demonstrated that when a celebrity is perceived as credible, the celebrity carried multiple favourable associations that consumers would like to identify themselves with.
Source credibility is considered as a moderator on the relationship between consumer-blogger identification and CBI. There are several studies that have looked at source credibility and brand attitudes. For instance, scholars have found that when the communicator was perceived as highly trustworthy, the message was more effective in changing attitudes (Miller & Baseheart, 1969 in Pornpitakpan, 2003). Therefore, a broad range of research has
examined credibility as a determinant of message effectiveness by celebrity endorsers (Basil, 1996). Studies have demonstrated that credible sources produce more attitude change than less credible sources (Willemsen, Neijens & Bronner, 2012, Chu & Kamal, 2008). Studies have also shown that credibility affects argument quality (Chu & Kamal, 2008). Furthermore, Hsu and Tsou (2011) found that source credibility provides positive associations to the blog reader and that this in turn affects the purchase intention. Therefore, it is argued that source credibility moderates the relationship between consumer-blogger identification and CBI. Next, the blogger’s social influence as a moderating effect is discussed.
2.5.2 Social influence
As described in the previous section, social influence is a blogger’s characteristic and is hypothesized to affect the relationship between consumer-blogger identification and CBI. Similar to source credibility, individuals or groups with social influence evoke positive associations (Langner, Hennings & Wiedmann, 2013). Moreover, as discussed previously, individuals tend to identify with aspirational reference groups. Aspirational reference groups
25 are seen as sources with high social influence (Escalas & Bettman, 2003). It could be that the higher the social influence of the blogger, the higher the degree of consumer-blogger
identification.
There are studies that have demonstrated that social influence has a positive effect on purchase intentions. For example, Hahn & Lee (2014) and Kulmala, Mesiranta & Tuominen (2013) identified that bloggers with high fashion leadership positively moderated the purchase intention of readers of blogs. What does a blogger with high social influence look like?
Bloggers differ in social influence due to personal traits (Li, Lai & Chen, 2011). The role of the bloggers is mainly to create awareness and to signal benefits to others; they can be
influential in encouraging the trial and adoption of products (Li, Lai & Chen, 2011). Langner, Hennigs & Wiedmann (2013) identified traits of the ideal social influencer which are
presented in figure 1.
Figure 1. Traits of the ideal social influencer (Langner, Hennigs & Wiedmann, 2013)
As can be seen in figure 1, two traits are circled: (1) expertise & knowledge and (2) involvement. These two traits will be used in this study to operationalize and measure the blogger’s social influence. The trait “expertise and knowledge” is defined as: “the ability to perform product related tasks successfully” (Alba & Hutchinson, 1987, p.411). The trait
26
“involvement” determines the extent to which a person will be perceived as an expert
(Langner, Hennigs & Wiedmann, 2013). The remaining traits cannot be applied in this study as the framework is originally meant to conduct network analysis.
Wiedmann, Hennigs & Langner (2010) also identified traits of the fashion influencer. The personal trait that is useable from this model is “personality strength”. The trait
personality strength is related to the trait “involvement”, as this trait is described as coming into contact with many people, participating in discussions and social events. The authors also identified “empathy” as a trait of the fashion influencer, which refers to personal interaction. However, personal interaction in this study is considered as a dimension of source credibility. Personal interaction signals helpfulness, as it is defined as the ability to sense the reactions of other people and noticing clues (Greenberg and Greenberg, 1983 in Wiedmann, Hennigs & Langner, 2010). Therefore, it has less to do with social influence.
In addition to personal traits, blog characteristics also signal social influence. Li, Lai & Chen (2011) conducted a network analysis and developed the Marketing Influence Value model. The Marketing Influence Value model identifies the extent of social influence of bloggers based on three dimensions of blog characteristics, which are network-based, content-based and activeness-content-based values (see figure 2).
27
Network-based value. This dimension consists of social connection factors and social
interaction factors. Social connection factors are (1) in-links and out-links and (2) network externalities. In-links are links in other online sources that refer to the page of the blogger. The more the in-links, the better the blogpost is recognized and the more influential the blogger is. Out-links are the amount of links in the blogpost referring to other web pages. The authors argue that the more the out-links, the less original the blogpost and therefore the less influential the blogger is. Network externalities, the second social connection factor, are the number of visitors. The more visitors a blogger has, the more influencing the blogger. Social interaction factors are factors that signal interaction: feedback, citations, comments and blogroll relationships in terms of recommendations. According to the Li, Lai & Chen (2011), these factors reflect the interest of readers to take the advice to purchase a particular product. The second dimension is content-based value. This dimension is based on the content of the blogpost, and consists of (1) subjectiveness of the post, (2) the length of the post and (3) the living time on the blog. Subjectiveness is considered as an indicator of social influence, because blog visitors are more impressed by subjective comments (Li, Lai & Chen, 2011). According to the authors, the second characteristic length of the blogpost, is positively correlated with the number of comments and the last characteristic, the living time on the blogpost is positively related to more social interactions. Finally, the last dimension is
activeness-based value. This consists of (1) the amount of content posts, which helps to
attract users to revisit or to subscribe and (2) comment replies, which facilitate conversations. This latter can be interpreted as the level of the blogger’s willingness to answer readers’ questions (Li, Lai & Chen, 2011).
For this study, living time in network and content posts cannot be applied as this study does not focus on real-life bloggers and existing blogposts. In addition to this, the elements subjectiveness and length of a blogpost are both found in the concept writing style, which will
28
be discussed in the following section. The element comment replies which refers to trustworthiness of the source will be applied in the design of the source credibility,
specifically interpersonal communication. Therefore, based on the blog characteristics, only the network-based value dimension will be applied to the social influence concept.
To summarize, in this study the social influence of the blogger will be based on the network-based value dimensions and on the following personal traits: (1) expertise and knowledge, (2) involvement and (3) personality strength. Next, writing style as the last hypothesized moderator is discussed.
2.5.3 Writing style
The last characteristic of the blogger which will be studied is the writing style, specifically storytelling and the lack of storytelling. As described earlier, bloggers create content through words and photos and the way they do this differs. Some bloggers write stories about their product experiences, others post photos to show interesting products. Despite the fact that the concept storytelling has yet not been studied in the context of bloggers, the concept is broadly studied in the marketing field. Scholars have recognized that stories describing the brand generate more positive brand attitudes than brands without brand stories (Lundqvist, Liljander, Gummerus & Van Riel, 2013) as “[brand] stories can help build awareness, comprehension, empathy, recognition, recall, and provide meaning to the brand” (Singh and Sonnenburg 2012, p. 189 in Gensler, Völckner, Liu-Thompkins & Wiertz, 2013).
An effective brand story is structured with a beginning, middle and ending in order to engage the listener. The beginning of the story should provide a conflict, whereby questions arise to the listener who wants to restore the harmony. Then, the middle presents the central message, which is the solution to the conflict. Finally, the ending should evoke emotional feelings to the listener as this part is best remembered (Lundqvist, Liljander, Gummerus & Van Riel, 2013). Stories that include conflicts and consumer-brand relationships are often
29 successful brand stories as they are best remembered (Schank 1999; Woodside 2010 in
Gensler, Völckner, Liu-Thompkins & Wiertz, 2013).
Although evidence exists that visuals are more effective than texts for the
persuasiveness of messages (Parkinson, 2012), it is hypothesized that blogposts containing stories should lead to higher values of CBI than blogposts without stories. At least three explanations in the literature can be given to support this hypothesis. First, according to Polkinghome (1991) consumers tend to create their identity via self-stories (Escalas, 2004). Therefore, the meaning of a brand for a consumer can be the result of its being part of a story (Escalas, 2004). Advertisements that tell stories engage and entertain consumers. The
advertisements communicate how products can be used to create meaning (Escalas, 2004). Therefore, bloggers may establish a stronger brand meaning through stories. Furthermore, Escalas (2004) demonstrated that advertisements in which stories are told enhance the consumer’s self-concept and generate positive brand attitudes. A second theory that also supports the hypothesis is the elaboration likelihood model (ELM) developed by Petty & Cacioppo (1983). Petty & Cacioppo (1983) state that when people have high motivation and high ability to process a persuasive message – this is in the case of consumers identifying with a blogger – these people will follow the central route instead of the peripheral route. The central route leads people to high message elaboration. People who follow the central route will form their attitude towards the message, because these people combine and integrate issue-relevant information. The peripheral route is followed by people who lack motivation or ability to process a message. In the peripheral route, people rely on relatively less thoughtful processes with the consequence of a low message elaboration (Petty & Cacioppo, 1983). Therefore, it could be argued that when people identify with a blogger, they have high motivation and high ability to process the message of the blogger. A blogpost with a story may therefore have a stronger effect on CBI than in the case of a consumer who identifies
30
with the blogger, but does not get to read a story. Moreover, desirable consequences of a brand, which are derived from elaboration in the central route result in more positive
associations with the consumer’s life (Kugman, 1965 in Petty & Cacioppo, 1983). And last, it is important with storytelling that listeners are able to identify with the sender and that the message evokes positive brand associations (Mossberg & Johansen, 2006 in Lundqvist, Liljander, Gummerus & Van Riel, 2013). This means that consumers should identify with the blogger who tells the story, which consequently should lead to the process of brand
connections. Identification with the sender of the brand story should not be complex as Berger (2014) argued that sharing an emotional story increases feelings of homophily.
So in conclusion, the writing style, specifically the use of a story, is hypothesized as a moderator on the relationship between consumer-blogger identification and CBI, so that this relationship should be stronger when there is storytelling. To sum up, the hypothesized moderators on the relationship between consumer-blogger identification and CBI are source credibility, social influence and writing style. These three characteristics are hypothesized to increase the strength of the relationship between consumer-blogger identification and CBI.
2.6 Literature gaps and research question
Although several studies have examined the relationship between social reference groups and brand evaluations, little is known about the specific relationship between identification with reference groups and CBI. Member groups and aspiration groups influence CBI (Escalas & Bettman, 2003). In addition, it is known that celebrities can serve as aspirational reference groups. However, facilitating CBI through celebrities is only recently studied by Dwivedi, Johnson & McDonald (2015). The authors showed that credible endorsers have a positive influence on CBI. The authors looked at the perceived endorser-brand congruence as a possible moderator on the relationship between the credible endorser and CBI. Perceived endorser-brand congruence is the similarity in identity between the endorser and the brand
31 (Dwivedi, Johnson & McDonald, 2015). Anyhow, it was already known that perceived
endorser-brand congruence makes celebrity endorsements effective (Till, Stanley & Priluck, 2008). Thus, little is known about other factors that could have an influence on CBI.
Fashion bloggers are upcoming and are perceived as social influencers as they are reaching increasingly larger audiences (Booth & Matic, 2011). Because fashion bloggers communicate about fashion brands, it is relevant to understand the influence of fashion bloggers on brands. Studies on the influence of fashion bloggers on brands are limited, but existing studies have confirmed that fashion bloggers affect consumers’ attitudes towards brands (Halvorsen, Hoffman, Coste-Manière & Stankeviciute, 2013) and purchase intentions (Chu & Kamal, 2008). However, little is known about the underlying factors of fashion bloggers that influence consumers’ attitudes. In other words, little is known to what extent bloggers’ characteristics influence consumers’ attitudes. Furthermore, as fashion bloggers have a certain amount of social influence and act as role models, research is lacking on their role as aspiration groups who are able to facilitate CBI.
Table 1. Gaps in the current literature
Ways to facilitate CBI
Factors influencing CBI
Via reference groups: member groups and aspiration groups. Specifically, focused on cultural groups and celebrities, not on bloggers yet.
Only focused on source credibility and perceived endorser-brand congruence.
In conclusion, this study is aimed at closing the gaps in the current literature on reference groups facilitating CBI (see table 1). This study will focus on the influence of the fashion blogger who serves as an aspirational reference group. The study will investigate to what extent the fashion blogger facilitates CBI and to what extent the characteristics of the fashion blogger influence this relationship. The selected bloggers’ characteristics that will be studied
32
are source credibility, social influence and writing style. The formulated research question is: “To what extent does the consumers’ perceived similarity with the fashion blogger lead to consumer-brand identification and to what extent docharacteristics of the fashion blogger moderate this relationship?”.
2.7 Theoretical and managerial contribution
This study will contribute to the literature in the following ways. First, the main focus of this study is the influence of bloggers on CBI. Because the current literature has not focused on bloggers as reference groups, knowledge on the facilitation of CBI through reference groups will be broadened. Second, the current literature has only covered congruity of identity between reference groups and brands as moderator of CBI. By focussing in this study on characteristics of reference groups, a deeper knowledge will be added to the literature. As described earlier, the effects of source credibility, social influence and writing style will be covered. Lastly, by studying bloggers as aspirational reference groups and as facilitators of CBI, literature on the influence of bloggers will be supplemented in both ways.
This study will also contribute to firms and managers. Since bloggers are demanding more attention of companies (Booth & Matic, 2011), it is relevant to find out to what extent bloggers have an influence on CBI. Bloggers are, namely, gradually integrating into
consumers’ life (Uzunoğlu & Kip, 2014). Knowledge about the influence of fashion bloggers on CBI could give companies extra motivations to collaborate with a blogger in order to enhance brand loyalty and brand advocacy. Second, this research will examine the
effectiveness of blogger endorsements as substitute for celebrity endorsements. Findings of this study may give marketers opportunities to operate more cost-efficiently by replacing or adding blogger endorsements to their marketing tactics. Third, this study will investigate to what extent the different characteristics of bloggers affect CBI. Insights into bloggers’ characteristics may lead managers to achieve more effective brand endorsements.
33
3. CONCEPTUAL FRAMEWORK
Figure 3. Conceptual framework
Based on the literature review, the conceptual framework of this study is presented in figure 3. The conceptual framework presents the developed hypotheses, which are described in the following section.
3.1 Hypotheses
As mentioned in the literature review, deriving from SIT, bloggers represent aspirational reference groups to consumers. Consumers who want to look like, or want to identify with a reference group can choose to use brands endorsed by the reference group. As result, the identity of the consumers will be updated (Escalas & Bettman, 2003). It is hypothesized that this process leads to CBI. The reasoning behind this is that the endorser’s personality- and lifestyle-related meanings are now embedded in the endorsed brands (McCracken, 1989). To look like or to express the same identity as the endorser, the consumer will start to identify with the brand too.
H1. Consumer-blogger identification has a positive influence on consumer-brand identification.
SIT states that favourable associations about other individuals or groups can lead individuals to identify with them (Stokburger-Sauer, Ratneshwar & Sen, 2012). As result, individuals will
34
get their identity verified or enhanced. Several studies that are based on SIT have shown that individuals who are perceived as credible evoke desirable associations among other
individuals (Dwivedi, Johnson & McDonald, 2015; Ohanian, 1990). This means that bloggers with high credibility should lead individuals to identify with them.
Chu and Kamal (2008) found that when a blogger is perceived as more credible, the blogger will influence the reader more. Thus, the degree of perceived credibility of the
blogger affects the extent of message elaboration. This points towards the possibility that CBI should enhance when the blogger is perceived as credible. The blogger’s message about the brand should be more persuasive. Moreover, Hsu and Tsou (2011) have found that blogger credibility provides positive associations to readers of blogs, which in turn positively affects purchase intention. Therefore, it is argued that a blogger with high credibility will increase feelings of homophily, and consequently generate positive brand associations for the consumer, so that a more valuable brand meaning is established for the consumer. As result the relationship between consumer-blogger identification and CBI will be stronger, than in the case of a blogger with low credibility.
H2. The relationship between consumer-blogger identification and consumer-brand
identification is moderated by the source credibility of the blogger, so that this relationship is stronger for a high value of source credibility of the blogger.
Similar to source credibility, social influence generates positive associations towards individuals (Langner, Hennings & Wiedmann, 2013). Social influencers are role models and therefore represent aspiration groups; individuals want to look like them and identify with them. As consequence, bloggers can influence the purchase intention of consumers (Hahn & Lee, 2014; Kulmala, Mesiranta & Tuominen, 2013). Moreover, Li, Lai & Chen (2011) claim in their study that several authors have recognized that influential blogs affects sales and advertisements more positively. Therefore, it is hypothesized that bloggers with high social
35 influence have a greater effect on the consumers’ identification with the blogger and the brand, than bloggers with low social influence.
H3. The relationship between consumer-blogger identification and consumer-brand
identification is moderated by the social influence of the blogger, so that this relationship is stronger for a high value of social influence of the blogger.
And lastly, as discussed in the literature review, it is shown that advertisements including stories can enhance the self-concept of a consumer because a brand meaning is established (Escalas, 2004). Also, people with high motivation and high ability to process a message, engage in message elaboration (Petty & Cacioppo, 1983). This means that
consumers who identify with the blogger, should have high motivation to elaborate the blogger’s message about the brand. As result, the consumer will evaluate the brand as having desirable consequences on his own life (Kugman, 1965 in Petty & Cacioppo, 1983).
Therefore it is hypothesized that bloggers will increase CBI if their blogpost consist of a story, because then the brand meaning should be more salient. Thus, consumers should identify themselves more with the endorsed brand in the case of a storytelling blogpost, than in the case of a blogpost without a story.
H4. The relationship between consumer-blogger identification and consumer-brand identification is moderated by the writing style of the blogger, so that this relationship is stronger for storytelling by the blogger.
36
4. RESEARCH METHODOLOGY
This chapter describes the research methodology, including sample, design, measures and the procedure of data collection. Because the research question points towards a confirmatory study, the research question has been answered through a quantitative, cross-sectional study.
4.1 Sample
In 2014, blogs were read or self maintained by 23.0% of Dutch Internet users (Akkermans et al., 2015). This means that in 2014, around 3.7 million Dutch people used blogs. Looking at characteristics of blog users in the Netherlands, men were using blogs 4.0% more often than women. The age group between 25 and 44 years old had the highest penetration rate (31.0% in their group), followed by the age group between 12 and 24 years old (26.0% in their group). Lastly, regarding education level, it appeared that the higher the education level, the more users (Akkermans et al., 2015).
Due to the enormous research population of blog users, a sampling frame was not available (Saunders & Lewis, 2012). Therefore, the sample was drawn from a non-probability sample consisting of 244 cases so that each treatment counted at least 30 participants.
According to Saunders & Lewis (2012), 30 participants in a treatment is sufficient. The sample consisted of women between the age of 12 and 35 years. This sample was selected for three reasons: (1) this age group has a high penetration rate for blog users, (2) the study used a female blogger, which means that female respondents (vs. male respondents) should have a higher likelihood to identify with the blogger, and (3) the age group should have a high likelihood to identify with the blogger, because the blogger is 22 years old.
An online experimental survey was distributed via social media and lectures. The invitees decided themselves to take part in the research (self-selection sampling). According to Saunders, Lewis & Thornhill (2009), the main risk of a self-selection sampling is that people may participate because of their feelings or opinions about the research topic, with the
37 consequence of limiting the generalization. To attack this risk, surveys were distributed in lectures at universities. By motivating students to fill in the survey, a little coercion was used to persons who do not have a strong opinion about the research topic.
The response rate of the main study was about 50%, which means that about 480 people had tried to fill in the survey. The main reason for the low response rate is that respondents did not pay attention to the information regarding the timer. A timer from one minute was set on the page on which the homepage of the blogger was presented, in order to prevent respondents from not paying careful attention to this page. Respondents who wanted to proceed to the next page within one minute and did not read the information about the one-minute timer, stopped the survey as they had the feeling that continuing to the next page was not possible.
The respondents which were used for the analyses were women (n=244). The average age was 21.2 years (SD = 2.86) and the most respondents were 19 years old (15.2%). The youngest respondent was 12 years old (n=1) and the oldest respondent was 34 years old (n=1). Regarding the education level, most respondents had achieved a Bachelor’s degree (36.5%), followed by high school graduates (34.0%).
4.2 Research design
To identify the relationship between consumer-blogger identification and CBI, and to analyze the effects of the blogger’s characteristics (source credibility, social influence and writing style) on the relationship between consumer-blogger identification and CBI, a quantitative study seemed most appropriate. An online experimental survey was designed, with scenario’s in which the level of the blogger’s characteristic differed. Respondents had to elicit their judgments about these scenarios, which made the experiment a vignette study. The study consisted of a 2 x 2 x 2 factorial design (see table 2) to test the effects of the blogger’s characteristics on the relationship between consumer-blogger identification and CBI.
38
Table 2. 2 x 2 x 2 Factorial Design
High source credibility Low source credibility
Storytelling No storytelling Storytelling No storytelling
High influence Treatment 1 Treatment 2 Treatment 5 Treatment 6
Low influence Treatment 3 Treatment 4 Treatment 7 Treatment 8
As could be seen from table 2, the manipulation of the three blogger’s characteristics (i.e. source credibility, social influence and writing style) resulted in a 2 x 2 x 2 factorial design, meaning that the research contained eight treatments. Since each participant was assigned to only one treatment, the experiment was a between subjects design. In the next section, the description and the design of the treatments are explained.
4.2.1 Experimental survey and procedure
A structured questionnaire was made regarding the identification with a fictive female fashion blogger, named Nadina Gorano, and the brand Chanel. It was decided to focus on fashion, because of two reasons: (1) fashion bloggers represent the largest segment of the blogosphere (Halvorsen, Hoffman, Coste-Manière & Stankeviciute, 2013), and (2) fashion brands have the power to signal personal and social identity (Tynan, McKechnie & Chhuon, 2010).
The experimental survey was divided into three parts (please consult appendix I for the English version of the questionnaire and appendix II for the Dutch version of the
questionnaire). The first part began with three statements regarding the fashion involvement of the respondent (e.g. “I like to shop for clothes”). These statements functioned merely as controlling variables and were measured on a 5-point Likert scale, ranging from: strongly disagree to strongly agree. Hereafter, the respondents were introduced to Nadina Gorano, the fashion blogger. Respondents were introduced via an introduction page, on where information regarding the credibility and social influence of Nadina Gorano was provided. Therefore, the
