Do influencers change our mind?

(1)

Do influencers change our mind?

How does the credibility of an influencer depend on the entertainment and educational value of a post and the type of the influencer?

by

Ezgi Genc

(2)

Do influencers change our mind?

How does the credibility of an influencer depend on the entertainment and educational value of a post and the type of the influencer?

by

Ezgi Genc

University of Groningen Faculty of Economics and Business

(3)

PREFACE

In front of you lies the thesis ‘Do influencers change our mind?’, a study about entertainment value, educational value and the type of influencer on the credibility of a post on Instagram. This thesis is my final project of the study Marketing Management at the University of Groningen.

Social media and online influencers have been my interest for the past couple of years, so I was very happy that I could combine this subject with my master thesis.

I would like to thank my supervisor Liane Voerman for her excellent guidance, support and feedback, the laughs and for challenging me to get the most out of this thesis. I am very grateful with a supervisor like her. Secondly, I would like to thank Susanne van der Velde for assisting me with SPSS. Without her, I still would not know how to perform a simple T-test. I want to thank my group members for the feedback and the pleasant working environments during our group meetings. I also would like to thank my friends and family for all the support I received the past six months.

I hope you enjoy reading this thesis as much as I had stress about it while writing it.

Ezgi Genc

(4)

EXECUTIVE SUMMARY

This study investigates the effect of different content on the credibility of the post of a macro- or micro-influencer. The posts of the influencer are uploaded on the social media platform Instagram. The aim of this research is to analyse whether there is a relationship between the educational value, entertainment value and the type of influencer on the credibility of the post. In addition, the susceptibility towards influencers and the social media usage that might lead towards a higher credibility are also carried out. The study uses 8 different conditions in a questionnaire to investigate the effect of the different variables. Insights about relations within the models are collected with ANOVA and a regression analysis. The entertainment value increases positively and significantly the credibility of the post. The educational value has the same effect. Furthermore, the micro-influencer has a higher credibility on a post than the macro-influencer. A high susceptibility towards influencers also leads to a higher credibility on a post, in contrast to a high social media usage. These results provide valuable insights for the online marketing industry and can be used for online advertising.

Keywords: social media • Instagram • influencers • macro-influencers • micro-influencers •

(5)

1 INTRODUCTION

Selena Gomez’s Coca-Cola ad on Instagram was the most liked picture of the year 2016: the picture has more than 6,9 million likes and almost 300,000 comments. The partnership seemed like a big win for Coca-Cola, who was able to use Selena’s 104 million Instagram followers at that time for their campaign. The popstar is Instagram’s most followed user (Statista, 2018), and in 2018 the 144 million followers easily surpass the likes and comments of the 114 million followers of Taylor Swift, Kim Kardashian with 124 million followers, and Ariana Grande, who has 142 million followers. People like Selena Gomez, Taylor Swift, Kim Kardashian and Ariana Grande are so-called celebrity influencers.

Lundahl (2014) state that in many cultures and countries, celebrities have become a special elite stratum, and that there is an overall trend of using celebrities in the media: next to Selena Gomez and Coca-Cola, there is also George Clooney with Nespresso, or Eva Mendez with Magnum. Schlecht (2003) defines a celebrity as ‘People who enjoy public recognition by a large share of

a certain group of people. Within a corresponding social group, celebrities generally differ from the social norm and enjoy a high degree of public awareness. Actors, models, sport athletes, entertainers, pop starts, business people, and politicians can all be celebrities.’

Celebrities that have a large share and public recognition, also have most times a large amount of following on social media.

Next to these celebrity influencers, there are people on Instagram like Negin Mirsalehi, a fashion enthusiast with her own clothing line and 5 million followers, Amrezy, a makeup artist and model with 5,6 million followers, and Jaclyn Hill, a YouTuber and makeup artist with 5,7 million followers. Their numbers of followers show that people on Instagram, next to celebrity influencers, can have a large following on social media. These people are called macro-influencers (Bijen, 2017). Celebrities are also a form of macro-macro-influencers.

(8)

Next to these macro-influencers on Instagram, there are also micro-influencers. Kalavrezos (2018) defines a micro-influencer as an everyday consumer who possesses a social media following between 1,000 and 100,000 followers. The definition of a micro-influencer is ‘any

YouTuber, Instagrammer, Snapchatter, or blogger with a relatively small (lower than 100,000) follower-base of highly-engaged, extremely attentive social media users’ (Mediakix, 2016).

Next to their follower amount, micro-influencers also differ in their approach and personal attachment to their followers: micro-influencers have a more authentic approach to their followers than a macro-influencer (Marwick, 2011). Micro-influencers are influencers like Kie, an artist with 1,500 followers, Nikki Fontaine, a makeup enthusiast with 7,000 followers, and Andrea Stinson, a model with 8,000 followers. Just like macro-influencers, a micro-influencer is also an everyday consumer who posts pictures of products, cities, restaurants and art they like, and interact with followers on social media.

These three types of influencers all have followers who are affected by what they post online (Bijen, 2017). The difference lies in if the influencer is a celebrity or not, and if the influencer is a macro-influencer or a micro-influencer.

According to De Veirman, Cauberghe, and Hudders (2017), brands have discovered the far-reaching impact and viral growth potential of influencers, people who built a large network of followers online, and who are regarded as trustworthy in one or several niches to promote their products. This corresponds well to the Coca-Cola ad on Selena Gomez’s Instagram, which led to a large exposure due to her following amount. De Veirman et al. (2017) mention a challenge for brands that aim to apply this type of WOM-marketing: identifying and selecting influencers whom may have a strong impact on their target audience and convince them to incorporate their products in their posts. Wong (2014) agrees with the statement that brands find it hard to find the right influencers that could address their target group.

1.1 Brands and influencers on paid media

The incorporation of a brand’s products in an influencer’s post can be used in several ways through different types of media. Lovett and Staelin (2016) distinguish between three types of media:

(9)

2. Earned media: influencers and followers can create buzz and e-WOM. Consumers can like the post, interact with the influencer and other consumers via Direct Messages or through comments on pictures. The influencer can mention, share, repost or review the bought or received products, or a brand itself, without getting paid for.

3. Owned media: through the brand’s website and other owned content. These owned contents are platforms like Instagram, Facebook and YouTube, where influencers have their own page, feed or webpage.

The most effective marketing tactics are the sponsored social media messages on Instagram (eMarketer, 2015). Sponsored posts are paid media. Out of the three mentioned types of media, this is the only commercial form that exists between an influencer and a company or brand (Wu, 2016). Those sponsored messages are mostly transmitted through an influencer (Zietek, 2016). People on Instagram are viewing a lot of pictures within a short period of time, which leads to an opportunity for brands to market their products and spread out effectively in a short period of time (Zietek, 2016). In this thesis, the focus is on paid media, because this is the form of media a company can have control over when reaching out to an influencer.

According to Wu (2016), paid media has three underlying forms:

1. Explicit sponsorships, where the sponsoring company pays the influencer a flat fee, or a specified amount per number of views on a post specifically created to market a brand or product.

2. Affiliated links, where purchases are made though a link or coupon code provided by the influencer, where the influencer can earn a commission on each sale. Posts with affiliated links differ from the sponsored post in that the influencer doesn’t have to be partnered with the sponsor for the purpose of creating a specific post or video. Instead, the influencer is part of the company’s affiliate program, where purchases made through a specific URL, or a coupon code is attributed to the influencer and the influencer is compensated with a commission on the sale.

(10)

and review such products in a post or video. This form of paid media does not only exist in the form of free products, but also free trips and holidays.

Figure 1 - Three forms of paid media

Marketing in one of these three forms can be far more effective than traditional marketing, because consumers generally view the content of their preferred influencer content without guarding suspicions (Nielsen, 2013). Influencers on social media can make use of all three types of paid media.

1.2 The impact of an influencer on a sponsored post

Brands use celebrities and macro-influencers in the media to reduce the distance with the consumer: the main rationale here is let the follower be closer to the macro-influencer both physically and symbolically (Lundahl, 2014). If a consumer purchases an endorsed product, they become more like the celebrity, and thus more ‘celebrity-like’ (Goodman, 2010). A celebrity is ‘a person whose name has attention-getting, interested-rivetting, and

profit-generating value,’ which works in the interest of promoting development in some form or

(11)

Instagram influencers with high numbers of followers are considered more likeable and popular (Atkin & Block, 1983). Perceptions of popularity due to the influencer’s number of followers, leads to perceptions of opinion leadership. The number of followers, which reflects network size and an indication for popularity, is frequently used to identify these influential nodes (De Veirman et al. 2017). High numbers of followers may result in a larger reach of the commercial message and may strengthen the power of eWOM (Talavera, 2015). Thus, a post of a macro-influencer, whether it is sponsored or not, will have more exposure than a post of a micro-influencer. This will lead to a higher willingness to pay of that product, which is the eventual goal of a brand when reaching out to an influencer (Rebelo, 2017).

Macro- and micro-influencers are people that have the ability to influence the audience, because they know the best products and are willing to share information with other consumers. They connect the audience with the brand or company, while having partnerships and deals, and are motivated by financial rewards (White & Matthews, 2016). Zietek (2016) sees macro- and micro-influencers as opinion leaders within their specific field. Opinion leaders are ‘a diverse

group of individuals like columnists, commentators, artists, and scientists who shape opinions’

(Mourdoukoutas & Siomkos, 2009).

Zietek (2016) states that the main difference between a micro- and a macro-influencer is their credibility in their posts on Instagram. Schlecht (2003) mentions that the relationship between an influencer and a company can be influenced by the credibility and the attractiveness of a post. A lack of credibility leads to a decrease of influence, no matter what type of influence it is (Zietek, 2016).

1.3 The impact of an influencer on the credibility of a post

(12)

for online posts, to see what impact the posts have on the willingness to pay. To measure the credibility of the post, the entertainment and the educational value will be used in this thesis.

1.4 Educational value

Educational value can be defined as the extent to which the internet provides the audience with resourceful and helpful information (Chen, Clifford & Wells, 2002; Ducoffe, 1995). Educational value is described by Luo (2002) as an important motivator in a post of an influencer to interact with their audience. Previous literature underlines the importance of social media’s success to use educational content, since consumers use networking platforms to inform themselves (Lin & Lu, 2011). Educational posts contain information about the brand, products or services, the appearance of the company, and drive consumer’s motivation towards interaction (De Vries et al. 2012).

1.5 Entertainment value

Entertainment value is described as more engaging with the audience of the influencer than other types of content (Cvijikj & Michahelles, 2013). Entertainment value refers to the extent to which the online content is fun and entertaining to the audience (Eighmey & McCord, 1998). Entertainment value does not specifically relate to the product, service or brand (Rahman, Suberamanian, Zanuddin, Moghavvemi & Nasir, 2016). Entertainment value in a post is supposed to work as an emotional relief, increase the customer engagement whilst delivering consumers enjoyment (Muntinga, Moorman & Smit, 2011).

1.6 Problem statement

The research question is as followed: How does the credibility of an influencer depend on the

entertainment and educational value of a post and the type of influencer?

(13)

1.7 Contribution

The difference between micro- and macro-influencers has not been researched yet. Research exist on celebrity marketing and their influence on their audience, but at this day and age, everyone on social media could be an influencer. Next to the difference between micro- and macro-influencers, the effect of those two influencers has not been researched on the credibility of posts on social media, in this case on Instagram. The educational and entertainment value have also not been applied to posts on Instagram. Instagram was chosen as the medium of communication because of its visually engaging nature and its closeness to influencers. (Rebelo, 2017). Instagram is the best performing channel for social action, with an average rate of 3.2% engagement, which is above all the other social networks (Rhythm One, 2015). This all can lead to a guidance for brands to develop new marketing strategies, not only for macro-influencers, but also for micro-influencers. A firm’s use of social media ultimately seeks to successfully create sales (Hudson, Huang, Roth & Madden, 2016; Baecker, Barrot & Becker, 2016), the relationship between the educational and entertainment value on the credibility of the post of a micro- or macro-influencer is still inconclusive and a matter of research.

(14)

2 LITERATURE REVIEW

In the literature review the relationship between the dependent variable credibility, the independent variables entertainment and educational value and the type of influencer are explained.

2.1 The entertainment value on the credibility of a post of an influencer

According to Wu (2016) social media creates more value for customers than the traditional form of marketing like TV commercials, and that is because there is a better entertainment value from watching or seeing an Instagram personality. A sponsored post on social media differs from traditional commercials on television where the spokesperson in the advertisements appears more distant than the influencer on social media interacting with their audience (Cvijikj & Michahelles, 2013). Sponsored posts using entertainment value are perceived as fun, exciting and positively influence the attitude toward advertisements (Taylor et al. 2011). The consumer may find the micro- and macro-influencer more relatable and genuine, because the influencer is closer to their audience by showing their everyday life, giving the Instagram post a modest and down-to-earth sentiment, which is a valuable marketing asset (Wu, 2016). Followers of a micro- or macro-influencer have a relationship with this personality and are possibly interacting with the influencer via the comment section, which has a high entertainment value. The elements of trust and friendliness of the influencer brings the follower closer to them (Marwick, 2011). Consumers follow this influencer because of their personality and because they are interested in their lives. Entertaining content is a form of media gratification that is offered to escape problems or the daily routine and is supposed to work as an emotional relief and can service as a relaxation tool (Muntinga et al. 2011). Entertaining content is the most influential form of branded content, and this type of content is described as more engaging and leads to more credibility in a post than other types of content (Cvijikj & Michahelles, 2013). That is why the entertainment value will probably have a positive effect on the credibility of a post, because the consumer feels a personal connection with the influencer. In line with this reasoning, the following hypothesis is introduced:

(15)

2.2 The educational value on the credibility of a post of an influencer

Next to the entertainment value, there is an educational value attached to a specific post or to following an influencer on Instagram. Consumers can follow an influencer to learn how to do something. The consumer gets information from a tutorial or a review of products by the influencer (Wu, 2016). Lopes and Galletta (2006) state that the literature on educational value has emphasized the use of information for utilitarian purposes, for example, to save time or create inflows of cash. However, much of the content available on the internet has no direct application to decision making or other utilitarian purposes (Lopes & Galletta, 2006). Social network advertising that focuses on informational content tends to elicit more positive attitudes towards the advertising itself, than less-informational content (Taylor, Lewin & Strutton, 2011). Consequently, informational content might lead to a higher credibility of the post than non-informational content (Muntinga et al. 2011). The following hypothesis has been introduced:

Hypothesis 2: The higher the educational value, the higher the credibility of a post of a macro- or micro-influencer.

2.3 The interaction effect of the entertainment and educational value on the credibility of a post of an influencer

(16)

Galletta (2006) also state the strength of combination of the educational and entertainment value. The authors state that both factors had a significant indirect effect on the willingness to pay. The prediction is that the entertainment aspect in combination with educational aspects will have a positive impact on the credibility of posts.

Hypotheses 3: When entertainment value is low, credibility is higher when the educational value is higher. When entertainment value is high, an enhancement of this effect is expected, so the increase in credibility is even higher.

2.4 The influence of the type of influencer on the credibility of a post

The number of followers indicates if the influencer functions as an influencer and attracts the right audience (Zietek, 2016). It implies that the influencer is trustworthy and attractive to its audience and therefore is able to reach more engagement. The quality of an influencer is not only measured by reviewing their number of followers, but also their reach and engagement with the audience (Brodie, Hollebeek, Juric & Ilic, 2011). Brodie et al. (2011) state that engagement is an important KPI for brands to measure the success of their marketing campaigns used by the influencer. The more followers an influencer has, the more popular the influencer is.

Credibility and trustworthiness also define the difference between a micro- and a macro-influencer. When it comes to a decrease of credibility the potential to influence also gets reduced (Atkin & Block, 2016). Authenticity is the key for successful influencer marketing, because it involves personality and exclusivity. According to Zietek (2016) it is a hazard for Instagram influencers to work with too many brands and post too many sponsored posts. By doing so, they create distance with their audience and become less credible (Scheibehenne, 2010). Generally speaking, a macro-influencer will get more possibilities to work with a brand than a micro-influencer does, because of their reach.

(17)

The prediction is that the micro-influencer will have more credibility than the macro-influencer, because the distance with the audience is lower than with the macro-influencer. Authenticity involves personality and exclusivity, and the prediction is that micro-influencers are seen as more involved and exclusive when comparing this to macro-influencers. Viewers can get the impression that macro-influencers are branded too much, which creates more distance between the influencer and the audience. The following hypothesis has been introduced:

Hypothesis 4: The micro-influencer has a higher credibility on a post than the macro-influencer.

2.5 The type of influencer on the credibility of an educational post

The type of influencer that uploads an educational post has also an effect on the credibility of the post. Zietek (2016) mentions that if a macro-influencer posts a strictly educational post, the distance between the influencer and the audience will be more than when a micro-influencer uploads an educational post. In other words, the credibility will be high for a macro-influencer when the educational value is higher, but the increase in credibility is even higher for a micro-influencer when the educational value is high. The micro-micro-influencer is more authentic to the audience because of the lower personal distance. As Leung (2009) stated: the social aspects are the added value for the audience for a micro-influencer.

The prediction is that the micro-influencer is more credible than the macro-influencer when it comes to strictly educational posts. The following hypothesis has been made:

Hypothesis 5: When the type of influencer is a macro-influencer, the credibility is higher when the educational value is higher. When the type of influencer is a micro-influencer, an enhancement of this effect is expected, so the increase in credibility is even higher.

2.6 The type of influencer on the credibility of an entertainment post

(18)

The prediction is that the micro-influencer is more credible than the macro-influencer when it comes to strictly entertainment posts. This, because the followers of the micro-influencer will experience more resemblance with a micro-influencer.

Hypothesis 6: When the type of influencer is a macro-influencer, the credibility is higher when the entertainment value is higher. When the type of influencer is a micro-influencer, an enhancement of this effect is expected, so the increase in credibility is even higher

2.7 The susceptibility towards influencers on the credibility of the post of an influencer The moderator ‘susceptibility towards influencers’ has been added to the conceptual model, because the more susceptible followers are towards influencers, the more they rely on the influencers, the more credible the post will be in the eyes of the follower (Phau & Teah, 2009). If followers are not susceptible towards influencers, the influencer will not make a difference in the credibility of the post. The more susceptible the follower is towards influencers, the more the follower can be influenced and the more credible the post seems (Langner, Hennigs, Wiedmann, 2013).

Hypothesis 7: The higher the susceptibility towards the influencer, the higher the credibility of the post.

2.8 The social media usage of the follower on the type of influencer

The moderator ‘social media usage’ has been added to the conceptual model, because it could have an effect on the type of influencer. The higher the social media usage of the consumer, the higher the knowledge about influencers will be (Leonardi, 2014). When the knowledge is high, the follower can easily differentiate between the content of a micro- and a macro-influencer. When the social media usage of the follower is high, the more critical the follower will be of the content of the influencer, because the follower will be more aware of certain posts (Phau & Teah, 2009) The macro-influencer will seem less sincere, and the follower will be more suspicious, than in comparison with a micro-influencer, who happen to be more sincere.

(19)

2.9 The social media usage of the follower on the credibility of the post of an influencer The prediction here is, the higher the social media usage of the influencer will be, the more aware the influencer will be of certain posts. The higher the social media usage is, the better the follower is in recognizing if the influencer is more genuine about the post or not, regardless of whether the influencer is a micro-influencer or a macro-influencer (Leonardi, 2014). When the follower knows how companies work with influencers, sponsored posts or free products, the less the follower will trust the opinion of the influencer.

Hypothesis 9: The higher the social media usage of the follower, the lower the credibility of the post.

2.10 Conceptual model

The dependent variable is the credibility; how believable a post on Instagram is. The two independent variables are the entertainment and the educational value. These values depend on the number of followers an influencer has, so if the influencer is classified as a micro- or macro-influencer. The research question is as followed: how does the credibility of an influencer

depend on the entertainment and educational value of a post and the type of influencer?

The following conceptual model has been made:

Type of influencer (macro/micro)

Educational value of a Social media usage

post of a macro- or micro influencer

Credibility of the influencer

Entertainment value of a

post of a macro- or micro Susceptibility towards

(20)

3 METHODOLOGY

The methodology part consists of the preparation and the implementation of the research.

3.1 Type of research

This study is a causal research, because this study is about the relationship between the independent variables entertainment value, educational value and the type of influencer, and the dependent variable credibility. The study also considers conclusive research. Causal research explains the cause and effect relationship between the given variables (Malhotra, 2009). This research type has two objectives: to understand which variables the cause and which variables the effect is, and to determine the nature of the relationship between the causal variables and the effect to be predicted (Malhotra, 2009).

In this research the effect of the educational and entertainment value of a post on Instagram of respectively a macro- and micro-influencer has been investigated.

3.2 Method

The method of data collection that is used for this analysis type is a questionnaire in Qualtrics. This fits well with the causal research, because this type of research is quantitative in nature, pre-planned and structured in design (Malhotra, 2009). The questionnaire was posted on LinkedIn, a networking website, with a link to the questionnaire and the request that only women could fill in the questionnaire. Next to LinkedIn, respondents were collected at the University Library of the Rijksuniversiteit in Groningen and the Public Library in Groningen. Female students of the Rijksuniversiteit and Hanzehogeschool were asked to fill in the questionnaire on a mobile phone, a tablet or a laptop.

3.3 Design of the experiment

(21)

Entertainment value

High Low

Educational value

High 1. Micro 2. Macro 5. Micro 6. Macro

Low 3. Micro 4. Macro 7. Micro 8. Macro

Table 1 – 2x2x2 factorial design.

3.4 Population and sample

To make sure every condition is equally distributed, the questionnaire is randomly assigned to the respondents in the eight conditions. This gives every condition an even number of respondents. 305 respondents have filled in the questionnaire. 5 respondents did not finish the questionnaire, so these were excluded from the questionnaire. Out of the 300 respondents, 10 male respondents were excluded as well, since the focus is only on women, which leaves the research with 290 female respondents. Out of the 290 female respondents, 145 respondents saw a post of a micro-influencer, and 145 respondents saw a post of a macro-influencer. The distribution over the different conditions is shown in table 3.

Entertainment value High Low Educational value High 1: Micro n=38 2: Macro n=36 5: Micro n=34 6: Macro n=35 Low 3: Micro n=38 4: Macro n=37 7: Micro n=35 8: Macro n=36

Table 3 – Number of respondents per condition.

3.5 Design of the stimulus

Influencer: micro vs. macro

(22)

Figure 2 - Left: micro-influencer. Right: macro-influencer.

In this case, the micro-influencer has 2,500 followers and 1,013 likes and 20 comments on the picture. The macro-influencer has 3,850,000 followers and 197,991 likes and 1,226 comments on the picture. The picture of the influencer will be the same for the micro- and for the macro-influencer, just like the amount of posts on the Instagram of the macro-influencer, which is 1,000 posts. The only difference will be the number of followers, likes and comments that is shown on the picture. The assumption here is that the respondent will know if the influencer is a macro- or micro-influencer by looking at the number of followers, likes and comments, and will assume the personal distance or the resemblance with the influencer.

(23)

Educational value: high vs. low

The respondent will see a picture of the micro- or macro-influencer where the influencer has a high education value, or a low education value. The description will be as followed:

 High educational description: ‘ZIMMI Dress: white embroidered laser cut dress. Off the shoulder. Lace up bust detail. Bell sleeves. 100% cotton. $99,95. Available at Tigermist.com. Instagram: @Tigermist.’

 Low educational description: ‘Available at Tigermist.com’

The high educational description gives the name and the type of the dress, the characteristics of the item, the material, the price, what the website of the brand is and what the Instagram of the brand is. The dress is chosen because this item fits with the conceptual variable of this objective. When consumers want to find the dress, they have to know the name of the clothing piece and on what website they can buy it from. The Instagram tag of the brand is also added, so that followers can be linked to Tigermist and can possibly find more clothing pieces to buy. The extra information like the material and the characteristics of the dress is also available on the website of the brand but is added in the description of the post on Instagram to give make a difference between the high and low educational value. The low educational description only states the website of the brand.

Entertainment value: high vs. low

The respondent will also see a picture of the micro- or macro-influencer where the influencer has a high entertainment value, or a low entertainment value. The description will be as followed:

 High entertainment description: ‘Because I got so many questions on where I got my

dress, it’s from Tigermist! It’s the ZIMMI dress! I love the off shoulder and wide sleeves, it’s perfect for summer when you are having a drink with your friends at the beach, or when you are dancing the night away in a fancy club. Be careful with the white though…’

 Low entertainment description: ‘Love this dress’

(24)

Combining the influencer with the entertainment and educational values

Combining the educational and entertainment descriptions in the post of a micro- and a macro-influencer will give eight conditions, because there are eight different combinations. These eight conditions are added in Appendix 1. To give an example, two conditions with different characteristics are shown in figure 3 and in figure 4.

Figure 3 shows an example of a micro-influencer with a high entertainment and low educational description.

Because I got so many questions on where I got my dress, it’s from Tigermist! It’s the ZIMMI dress! I love the off shoulder and wide sleeves,

it’s perfect for summer when you are having a drink with your friends at the beach, or when you are dancing the night away in a fancy club.

Be careful with the white though…

Available at Tigermist.com’

Figure 3: Micro-influencer with high entertainment, low educational value

(25)

Love this dress.

ZIMMI Dress: white embroidered laser cut dress. Off the shoulder. Lace up bust detail. Bell

sleeves. 100% cotton. $99,95. Available at Tigermist.com. Instagram: @Tigermist.

Figure 4: Macro-influencer with low entertainment, high educational value

Pre-test

Pre-testing refers to testing the questionnaire on a small sample of respondents to identify and eliminate potential problems (Malhotra, 2009). The pre-test group should be similar to the respondents in terms of their background characteristics, familiarity with the topic, and attitudes and behaviors of interest (Malhotra, 2009). The pre-test of the questionnaire for this thesis is done to a test-group of 20 respondents, which were all women. This, because women on social media, especially on Instagram, are more interested in clothes and fashion than men do. The population of this research is therefore solely women. That is why men are excluded from this interview. The respondents in the pre-test were shown pictures of a macro-influencer or a micro-influencer, a description of a high or a low informational post in combination with a description of a high or a low entertaining post, like the example in figure 4. The pre-test showed that the respondents saw a clear difference between the two influencers and the high and low values of the educational and entertainment posts.

3.6 Operation of the experiment

(26)

influencers and the usage of social media are asked to the respondent. The questions are asked with a 7-point Likert scale, where the score 1 stands for ‘Strongly agree’ and the score 7 stands for ‘Strongly disagree’.

Subject Source Items Scale FA/RA

Credibility of the post (DV) Venkatesh et al. (2003) Dimoka et al. (2010) Pavlou et al. (2006) Messner et al. (2004).

1. I believe the information that was given in the post is genuine.

2. It felt like the influencer is genuine about the product.

3. The brand the influencer was sponsored by, has to be good, because they choose this influencer.

4. It felt like the influencer only wanted to sell me the clothing item.

5. I felt a personal connection with the influencer. 7-point Likert-scale EV=79.30 α=.934 Manipulation check educational value (IV)

This post was very informative 7-point Likert-scale

Manipulation check entertainment value (IV)

This post was very entertaining 7-point Likert-scale

Attention check 1. The person who uploaded the post had many followers.

2. The person who uploaded the post had: - 1 – 10,000 followers - 10,001 – 100,000 followers - 100,001 – 1,000,000 followers - 1,000,001 – 2,500,000 followers - 2,500,001 – 5,000,000 followers 7-point Likert-scale Susceptibility towards influencers Pronin et al. (2007) Dimoka et al. (2010) Ryan et al. (2000)

1. I follow more influencers on social media than my own personal friends and family. 2. I follow influencers on social media because I value their opinion.

3. I follow influencers on social media because I want to be inspired.

7-point Likert-scale EV=66.24 α=.737 Social media usage Lenhart et al. (2010) Hughes et al. (2012)

1. Social media is very important to me. 2. Instagram is a very important type of social media for me.

3. Clothes and fashion interest me.

7-point Likert-scale EV=65.27 α=.731 Control variables

What is your age? What is your gender?

(27)

Credibility of the post

The scale for the credibility of the posts of the influencer is based on the articles of Venkatesh, Morris, Davis and Davis (2003), Dimoka (2010), Pavlou and Dimoka (2006), and Messner, Baumer and Rosenfeld (2004). The authors talk about the influence of celebrities and other social influencers. Pavlou et al. (2006) discuss if the influencer is likely to be reliable and credible, if the influencer is trusted by the audience, and if the product matches the posted description. Dimoka (2010) researched if the influencer is credible during the transaction and if the influencer has the expertise to understand the needs and preferences of the audience. Based on the items these authors asked to the respondents, the items in table 2 are formed for the dependent variable.

Manipulation check educational value

Respondents will see a post of a micro- or a macro-influencer. The educational value will be low or high. Based on the type of influencer in the questionnaire, and the amount of information that is given in the post, the respondents can indicate with a 7-point Likert scale how much information is given in the post. The manipulation check has worked if the respondents can indicate the right amount of value (high/low) that has been given in the post.

Manipulation check entertainment value

The same goes for the entertainment value. Respondents will see a post of a micro- or a macro-influencer and based on the type of macro-influencer, and the amount of entertaining information that is given in the post, the respondents can indicate how much entertainment value the post has. The manipulation check has worked if the respondents can indicate the right amount of value (high/low) that has been given in the post.

Attention check

The attention check is done to indicate the difference between a micro- and a macro-influencer. If the respondent is able to answer how many followers the influencer has, they indicate if the influencer is a micro-influencer, or a macro-influencer. The attention check has worked if the respondents can indicate how many followers the influencer has.

Susceptibility towards influencers

(28)

about the effect of celebrities on social and moral values. Ryan et al. (2000) asked their respondents about the different social influencers that could play a role in the decisions a respondent has to make.

Social media usage

The questions about the usage of social media are based on the items in the research of Lenhart et al. (2010) and Hughes et al. (2012). The questions are based on the importance of social media, on Instagram as an own platform, and on the susceptibility towards clothes and fashion.

Control variables

The only control variable that has been used is the age of the respondent. Because the population for this research is solely women, men are excluded from this interview. Because the questionnaire is partly conducted online, this question is added to exclude possible men in who filled in the questionnaire.

3.7 Plan of analysis

In the plan of analysis, first the manipulation checks of the educational value, the entertainment value and the type of influencer have been tested, then the credibility of the post, the susceptibility and the social media usage. The complete SPSS output has been added in Appendix 2.

To see if the manipulation check of the educational value, the entertainment value and the type of influencer has worked, a T-test has been conducted. The T-test is a procedure that can be used to test whether the levels or groups of the mentioned independent variables differ on the dependent variable, in this case credibility. The Eigenvalue has to be larger than 1 (Malhotra, 2009).

Manipulation checks: results Educational value

(29)

There was a significant difference in the scores for the post where educational value was absent (M=5.59, SD=1.07) and where the educational value was present (M=2.04, SD=1.01),

t(287.76)=29.19, p<.001. This difference is 99% significant. This means the manipulation

check was successful, the picture that was educational was also seen by participants as more educational than the one that was not.

Entertainment value

The question ‘This post was very entertaining’ was asked with a 7-point Likert scale, where the score 1 stands for ‘Strongly agree’ and the score 7 stands for ‘Strongly disagree’.

There was a significant difference in the scores for the post where entertainment value was absent (M=5.76, SD=1.34) and where the entertainment value was present (M=1.95, SD=1.00),

t(258.73)=27.26, p<.001. This difference is 99% significant. This means the manipulation

check was successful, the picture that was educational was also seen by participants as more educational than the one that was not.

Type of influencer

The question ‘The person who uploaded the post had many followers’ was asked with a 7-point Likert scale, where the score 1 stands for ‘Strongly agree’ and the score 7 stands for ‘Strongly disagree’.

There was a significant difference in the scores for the post of a micro-influencer (M=4.59,

SD=1.89) and for the post of a macro-influencer (M=1.80, SD=.78), t(191.58)=16.47, p<.001.

This difference is 99% significant. This means the manipulation check was successful, the respondents who saw the micro-influencer thought the influencer had not many followers, and the respondents who saw the macro-influencer thought the influencer had many followers.

Overall, the manipulations worked and the factor and reliability analysis came out significant.

Credibility of the post

(30)

analyses. All questions are asked with a 7-point Likert scale, where the score 1 stands for ‘Strongly agree’ and the score 7 stands for ‘Strongly disagree’.

A factor analysis has been conducted. KMO has a value of .888. Item 1 explains 79.30% of the variance. The further items explain 8.79% or less. The Component Matrix shows that item 4 is negatively formulated, -.818. Reversing this item leads to a positive score of .826. Now, the minimum score is .818 and the maximum score is .946. Uni-dimensionality is proven.

After this, a reliability analysis has been conducted (Malhotra, 2009). The Cronbach’s Alpha is .934. The Cronbach’s Alpha if the item is deleted lies between .903 and .937, so no item has been deleted. Therefore, a new variable has been made for ‘credibility of the post’, computed by summing the 5 items and dividing this score by 5. So, the internal consistency of the 5 items is high enough to combine them into 1 new variable.

Susceptibility towards influencers

A factor analysis has been conducted for the susceptibility towards influencers. All questions are asked with a 7-point Likert scale, where the score 1 stands for ‘Strongly agree’ and the score 7 stands for ‘Strongly disagree’.

KMO has a value of .615. Item 1 explains 66.24% of the variance. The further items explain 22.72% or less. The Component Matrix shows that all items are positively formulated with a minimum score of .717 and a maximum score of .889. Reversing of an item is not necessary. Uni-dimensionality has proven.

After this, a reliability analysis has been conducted. The Cronbach’s Alpha is ,737. The Cronbach’s Alpha if the item is deleted lies between ,505 and ,703, so no item has been deleted. Therefore, a new variable has been made for ‘susceptibility towards influencers’, computed by summing the 3 items and dividing this score by 3. So, the internal consistency of the 3 items is high enough to combine them into 1 new variable.

Social media usage

(31)

KMO has a value of .666. Item 1 explains 65.27% of the variance. The further items explain 20.46% or less. The Component Matrix shows that all items are positively formulated with a minimum score of .565 and a maximum score of .710. Reversing of an item is not necessary. Uni-dimensionality has proven.

(32)

4 RESULTS

The results chapter consists the results of the ANOVA analysis and the reliability analysis.

4.1 ANOVA

To test whether the entertainment value of a post, the educational value of a post and the person who uploaded the post influence the credibility of a post, an ANOVA was conducted with entertainment value (high/low), educational value (high/low) and the type of influencer (micro/macro) as independent variables and credibility as the dependent variable. The total SPSS output of ANOVA can be found in Appendix 3.

Entertainment value High Low Educational value High 1: Micro μ=2.03 2: Macro μ=2.47 5: Micro μ=4.89 6: Macro μ=5.04 Low 3: Micro μ=2.18 4: Macro μ=2.79 7: Micro μ=5.42 8: Macro μ=5.56

Table 4 – μ of the credibility per condition.

High/Macro Low/Micro F

Educational value M=3.61 (SD=.09) M=3.99 (SD=.09) F(1,282)=8.79, p=.003

Entertainment value M=2.37 (SD=.09) M=5.23 (SD=.09) F(1,282)=500.05, p<.001

Type of influencer M=3.97 (SD=.09) M=3.63 (SD=.09) F(1,282)=6.74, p=.010

Table 5 – μ of the credibility for the main effects.

All main effects were significant and in the expected direction.

Interaction effects F

Entertainment value * Educational value F(1,282)=1.29, p=.256 Educational value * Type of influencer F(1,282)=.10, p=.752 Entertainment value * Type of influencer F(1,282)=2.22, p=.138

Table 6 – μ of the credibility for the interaction effects.

(33)

4.2 Regression analysis

To test the hypotheses, a regression analysis was conducted, as shown in table 7 and table 8.

Model 1 Y = ß0 + ß1*Ent + ß2*Edu + ß3*Type + e

Model 2 Y = ß0 + ß1*Ent + ß2*Edu + ß3*Type + ß4*(Ent*Edu) + ß5*(Ent*Type) + ß6*(Edu*Type) + e

Model 3 Y = ß0 + ß1*Ent + ß2*Edu + ß3*Type + ß4*(Ent*Edu) + ß5*(Ent*Type) + ß6*(Edu*Type) + ß7*Smu + e

Model 4 Y = ß0 + ß1*Ent + ß2*Edu + ß3*Type + ß4*(Ent*Edu) +

ß5*(Ent*Type) + ß6*(Edu*Type) + ß7*Smu + ß8*(Smu*Type) + e Model 5 Y = ß0 + ß1*Ent + ß2*Edu + ß3*Type + ß4*(Ent*Edu) +

ß5*(Ent*Type) + ß6*(Edu*Type) + ß7*Susc + e

Full model Y = ß0 + ß1*Ent + ß2*Edu + ß3*Type + ß4*(Ent*Edu) +

ß5*(Ent*Type) + ß6*(Edu*Type) + ß7*Smu + ß8*(Smu*Type) + ß9*Susc + e

Full model adjusted Y = ß0 + ß1*Ent + ß2*Edu + ß3*Type + ß4*(Ent*Edu) + ß5*(Ent*Type) + ß6*(Edu*Type) + ß7*Smu + ß8*Susc + e

Where:

Ent Entertainment value Edu Educational value Type Type of influencer Smu Social media usage

Susc Susceptibility towards influencers

Table 7 – Formulas regression model

Multicollinearity

(34)

Findings regression model

Model 1 shows the main effects, the educational value, the entertainment value and the type of influencer. Model 1 is for all variables significant, which was also proven in the ANOVA analysis.

Model 2 shows the interactions between the first three variables. The interactions are not significant, which was also proven in the ANOVA analysis. What is remarkable, is that the type of influencer as a main effect is not significant anymore.

In model 3 the covariate social media usage has been introduced. Still, only the entertainment value and educational value are significant. The R² adjusted has been decreased.

Model 4 shows the interaction between social media usage and the type of influencer, which is also not significant, apart from the entertainment value and the educational value.

In model 5 the covariate susceptibility towards influencers has been introduced, covariate social media usage and the interaction between social media usage and the type of influencer has been removed. What is remarkable, is that the susceptibility towards influencers is significant. The R² adjusted has been increased, just like the F value.

The full model shows every variable and every covariate with the desired interactions. Again, only the entertainment value, educational value and the susceptibility towards influencers is significant. The R² adjusted stays the same, but the F value has decreased.

(35)

Variable M1 M2 M3 M4 M5 Full model

Full model adjusted

Beta Sig. Beta Sig. Beta Sig. Beta Sig. Beta Sig. Beta Sig. Beta Sig.

Ent -2.86 .000 * -3.20 .000 * -3.20 .000 * -3.20 .000 * -3.23 .000 * -3.24 .000 * -3.23 .000 * Edu -.38 .004 * -.48 .032 * -.48 .033 * -.49 .029 * -.53 .019 * -.55 .015 * -.539 .017 * Type .34 .009 * .18 .412 .19 .404 -.17 .678 .16 .479 -.22 .592 .145 .514 Ent*Edu .29 .256 .29 .261 .31 .231 .32 .208 .35 .169 .332 .194 Ent*Type .38 .138 .39 .134 .38 .135 .40 .116 .40 .119 .399 .118 Edu*Type -.08 .752 -.09 .718 -.11 .675 -.05 .844 -.04 .899 -.025 .924 Smu .04 .556 -.05 .677 -.15 .200 -.063 .443 Smu*Type .15 .310 .15 .297 Susc .13 .025 * .16 .022 * .159 .022 * F (ANOVA) 173.33 87.45 74.83 65.62 76.76 59.82 67.14 Sig. (ANOVA) .000 .000 .000 .000 .000 .000 .000 R² adjusted .642 .643 .642 .642 .648 .648 .648

Table 8 – Regression analysis

(36)

4.3 Discussion of the results

According to previous results, H1 and H2 do show significant results, and are therefore supported. The more entertaining post (M=2.37, SD=0.09) was seen as more credible than the less entertaining post (M=5.23, SD=.09), F(1,282)=500.05, p<.001. The more educational post (M=3.61, SD=.09) was seen as more credible than the less educational post (M=3.99, SD=.09),

F(1,282)=8.79, p=.003. H3 has been rejected, because the results were not significant. There is

no interaction or enhancement of this effect. H4 is accepted, because there was a significant main effect of the type of influencer on the credibility of the post. The micro-influencer was seen as more credible (M=3.63, SD=.09) than the macro-influencer (M=3.97, SD=.09),

F(1,282)=6.74, p=.010. This is also proven in the regression model. H5 and H6 are rejected,

because the results were not significant. There is no interaction of enhancement of this effect. H7 has been accepted, because this result was significant and has a positive value, as shown in the regression model. H8 and H9 are rejected. Social media usage was not significant. Social media usage with the type of influencer as an interaction had also no significant effect. Table 9 shows the hypotheses and if the hypothesis has been rejected or accepted.

H1 The higher the entertainment value, the higher the credibility of a post of a macro-

or micro-influencer.



H2 The higher the educational value, the higher the credibility of a post of a macro- or

micro-influencer.



H3 When entertainment value is low, credibility is higher when the educational value

is higher. When entertainment value is high, an enhancement of this effect is expected, so the increase in credibility is even higher.

x

H4 The micro-influencer has a higher credibility on a post than the macro-influencer.  H5 When the type of influencer is a macro-influencer, the credibility is higher when

the educational value is higher. When the type of influencer is a micro-influencer, an enhancement of this effect is expected, so the increase in credibility is even higher.

x

H6 When the type of influencer is a macro-influencer, the credibility is higher when

the entertainment value is higher. When the type of influencer is a micro-influencer, an enhancement of this effect is expected, so the increase in credibility is even higher.

x

H7 The higher the susceptibility towards the influencer, the higher the credibility of the

post.



H8 The higher the social media usage, the lower the credibility of the post. x H9 When the type of influencer is a macro-influencer, the credibility of the post will

be lower with higher social media usage. When the influencer is a micro-influencer, the credibility of the post will be higher with higher social media usage.

(37)

5 CONCLUSION AND DISCUSSION

The conclusion of this thesis and an answer to the research question is given. Afterwards, the managerial recommendations and limitations are given.

5.1 Conclusions

This study provides insights into the credibility of a micro- or macro-influencer depending on the entertainment and educational value of the post. The role of the type of influencer, the educational value and the entertainment value define a new view for the relation between the content of a post and the credibility of a post of an influencer. The research question ‘How does

the credibility of an influencer depend on the entertainment and educational value of a post and the type of influencer?’ is partially answered.

The effectiveness of educational and entertainment value in a post has been confirmed multiple times and in multiple online metrics. According to Wu (2016), Taylor et al. (2011), Muntinga et al. (2011) and Cvijikj and Michahelles (2013) it is true that a higher entertainment value in the post of an influencer, will lead to higher credibility. Posts using a high entertainment value are perceived as fun, exciting, more genuine and trustworthy and this leads to a positive influence towards the credibility of the post. This statement is also true for the educational value, which has been confirmed multiple times in this research as well. When the educational value of a post is high, followers like to believe the information in the post is trustworthy. According to Taylor et al. (2011) posts that focus on giving information to the follower, tend to achieve more positive attitudes towards the post than less-informational content.

Furthermore, an enhancement of the combination of entertainment and educational value on the credibility of a post was expected, because of the research conducted by Dolan et al. (2016) and Leung et al. (2009). However, in this research there was no interaction effect between the entertainment and educational value. The educational value and the entertainment value were significant on its own, but there was unfortunately no enhancement, regardless of what Dolan et al. (2016) and Leung et al. (2009) said about the social aspects that would come along.

(38)

audience. According to the research in this thesis, there is a significant difference in the type of influencers: the micro-influencer was seen as more credible than the macro-influencer. As a brand or company, it is recommended to use a micro-influencer, because this might give the audience of that influencer more trust in the brand, which can lead to a higher willingness to pay.

The susceptibility towards influencers on the credibility of the post also has a significant effect. The more susceptible followers are towards influencers, the more they rely on the influencers, the more they trust what the influencer has to say, which will lead to more credibility of the posts in the eyes of the follower. The more susceptible the follower is towards the influencer, the more the follower can be influenced (Langner et al. 2013). The interaction of susceptibility towards influencers and micro-influencers has not been researched in this thesis but might lead to interesting results.

A surprising result was the change in significance level of the main effect type of influencer when adding the interactions with the other variables. The type of influencer became suddenly not significant, due to the interactions. The significance level of the interaction with the entertainment value is however higher than the interaction with the educational level. It is however in this case not significance, but further research into this combination is advised.

The type of influencer on the credibility of an educational post shows a contradiction with the findings of the study of Zietek (2016) and Leung (2009). This also corresponds to the credibility of the influencer on an entertaining post. According to the researchers, the micro-influencer would be more credible for the educational and entertainment post. This contradicts the findings in this thesis, because there was no significant difference between the two types of influencers. The prediction was also that the macro-influencer would have more distance with the audience than the micro-influencer while posting an educational or entertaining post, but there was unfortunately no difference. What this means is that the follower only pays attention to the type of influencer, and not so much to the difference in the content of the post of the two influencers. The same recommendation is applicable to brands: a micro-influencer would have more impact on the credibility, regardless of the content of the post.

(39)

post will be lower when the follower has a high social media usage, which was not the case. If the influencer is a micro-influencer, the expectation would be that the credibility of the post would be higher with higher social media usage, but this statement was also not significant. The results for this statement were surprising. For future research it is recommended that the interaction between social media usage and the susceptibility towards brands is measured, because a high social media usage can lead to more knowledge about certain influencers, which can affect the susceptibility towards influencers.

The social media usage of the follower on the credibility of the post was also not significant. The prediction was that a high social media usage of the follower, could lead to more awareness of certain posts, regardless of whether the influencer is a micro-influencer or a macro-influencer. This statement was however not significant.

5.2 Management recommendations

To relate the findings in this research to managerial implications, the use of micro-influencers is highly recommended. Educational value and entertainment value in a post both lead to good results, but the best result came from posts with entertaining content.

Using a micro-influencer has its own benefits. When a brand uses a smaller influencer for their products, it enters another niche group with opportunities. Followers of the micro-influencer are more likely to identify that brand with something positive. A micro-influencer also has a strong connection with their audience. If a brand decides to work along with this type of influencer, it will gain the trust of the followers.

Online influencers are nowadays more credible than old-fashioned celebrities on TV (Khamis, Ang & Welling, 2017). Feeling a connection with the influencer is an important driver in this day and age, so the recommendation for companies is to work with online influencers instead of celebrities.

5.3 Limitations and future research

(40)

media platform and a broader sampling frame might increase the generalizability of the findings.

The topic that was chosen was in the fashion field. The brand Tigermist is a real brand and is popular on a social media platform like Instagram. However, when using a well-known online brand, it is hard to control the other variables and attitudes towards the brand, which could have influenced the results. Using an unknown or made up brand name might be the solution for this problem. When measuring trust or purchase intention, it is however recommended to use a brand that is known to the public.

(41)

REFERENCES

Atkin, C., & Block, M. (1983). Effectiveness of celebrity endorsers. Journal of

Advertising Research, Vol 23 (1), 57-61.

Bijen, Y. (2017). #AD: the effects of an influencer, comment and product combination on brand image. Master’s thesis, University of Twente. Conducted through: https://essay.utwente.nl/72265/1/Bijen_MA_BMS1.pdf

Brodie, R., Hollebeek, L., Juric, B., & Ilic, A. (2011). Customer engagement: conceptual domain, fundamental propositions, and implications for research. Journal of Service Research, Vol. 14 (3), 252-271.

Chen, Q., Clifford, S.J., & Wells, W.D. Attitude toward the site II: new information. J Advert Res, 2002: 33-45.

Cvijik, I.P., & Michahelles, F. (2013). Online engagement factors on Facebook brand pages. Social Network Analysis and Mining, Vol. 3 (4), 843-861.

De Veirman, M., Cauberghe, V., & Hudders, L. (2017). Marketing through Instagram influencers: impact of number of followers and product divergence on brand attitude.

International Journal of Advertising, Vol. 36 (5), 798-828.

De Vries, L., Gensler, S., & Leeflang, P. (2012). Popularity of brand posts on brand fan pages: an investigation of the effects of social media marketing. Journal of Interactive

Marketing, Vol. 26 (2), 83-91.

Dimoka, A. (2010). What does the brain tell us about trust and distrust? Evidence from a functional neuroimaging study. MIS Quarterly, Vol. 34, 373-396.

Dolan, R., Conduit, J., Fahy, J., & Go, S. (2016). Social media engagement behaviour: a uses and gratifications perspective. Journal of Strategic Marketing, Vol. 24 (3-4), 261-227.

Edelman, D.C. (2010). Branding in the digital age: you’re spending your money in all the wrong places. Harvard Business Review, Vol. 88 (12), 62-69.

Eighmey, J., McCord, L. (1998). Adding value in the information age: uses and gratifications of sites on the world wide web. Journal of Business Research, Vol 41 (3), 187-194.

eMarketer Inc. C.em and Reserved, A.R. (2015). Are sponsored social posts the most

effective marketing channel? Conducted through:

http://www.emarketer.com/Article/Sponsored-Social-Posts-Most- Effective-Marketing-Channel/1013242

(42)

Goodman, M. (2010). The mirror of consumption: celebritisation, developmental consumption and the shifting cultural politics of fair trade. Geoforum, Vol. 41: 104–116.

Hudson, S., Huang, L., Roth, M., & Madden, T. (2016). The influence of social media interactions on consumer-brand relationships: a three-country study of brand perceptions and marketing behaviors. International Journal of Research in Marketing, Vol. 33 (1), 27-41.

Hughes, D.J., Rowe, M., Batey, M., & Lee, A. (2012). A tale of two sides: Twitter vs. Facebook and the personality predictors of social media usage. Computers in Human Behavior, Vol. 28 (2), 561-569.

Kalavrezos, A. (2018). How to find micro-influencers on Instagram to boost sales? Conducted through: https://keyhole.co/stories/how-to-find-micro-influencers-on-instagram-to-boost-sales/

Khamis, S., Ang, L., & Welling, R. (2017). Self-branding, micro-celebrity and the rise of social media influencers. Celebrity Studies, Vol. 8, 191–208. Conducted through: https://doi.org/10.1080/19392397. 2016.1218292.

Langner, S., Hennigs, N., & Wiedmann, K.P. (2013). Social persuasion: targeting social identities through social influencers. Journal of Consumer Marketing, Vol. 30 (1), 31-49.

Lenhart, A., Purcell, K., Smith, A., & Zickuhr, K. (2010). Social media and mobile internet use among teens and young adults. Washington, DC: Pew Internet and American Life Project. Conducted through: http://pewinternet.org/Reports/2010/Social-Media-and-Young-Adults.aspx.

Leonardi, P.M. (2013). Social media, knowledge sharing, and innovation: toward a theory of communication visibility. Information Systems Research, Vol. 25 (4), 796-816.

Lin, K.Y., & Lu, H.P. (2011). Why people use social networking sites: an empirical study integrating network externalities and motivation theory. Computers in Human Behaviour, Vol. 27 (3), 1152-1161.

Littler, J. (2008). I feel your pain: Cosmopolitan charity and the public fashioning of the celebrity soul. Social Semiotics, Vol. 18 (2), 237-251.

Lopes, A.B., & Galletta, D.F. (2006). Consumer perceptions and willingness to pay for intrinsically motivated online content. Journal of Management Information Systems, Vol. 23 (2), 203-231.

Lovett, M.J., & Staelin, R. (2016). The role of paid, earned, and owned media in building entertainment brands: reminding, informing, and enhancing enjoyment. Marketing

(43)

Lundahl, O. (2014). Fashionalising sustainable consumption in lifestyle media. In NA –

Advances in Consumer Research, Duluth, MN: Association for Consumer Research, Vol. 42,

343–347.

Luo, X. (2002). Uses and gratifications theory and e-consumer behaviors: a construal equation modelling study. Journal of Interactive Advertising, Vol. 2, 34-41.

Maecker, O., Barrot, C., & Becker, J. (2016). The effect of social media interactions on consumer relationship management. Business Research, Vol. 9 (1), 133-155.

Malhotra, K.N. (2009). Marketing Research, an applied orientation. Pearson Education (Us).

Marwick, A. (2011). To see and be seen: celebrity practice on Twitter. International Journal of Research into New Media Technologies, Vol. 17 (2), 139-158.

Mediakix Press Release. (2016). What are micro-influencers: definitions, trends and advantages. Conducted through: http://mediakix.com/2016/06/micro-influencers-definition-marketing/

Messner, S.F., Baumer, E.P., & Rosenfeld, R. (2004). Dimensions of social capital and rates of criminal homicide. American Sociological Review, Vol. 69, 882-903.

Mourdoukoutas, P., & Siomkos, G.J. (2009). The Seven principles of WOM and buzz marketing: crossing the tipping point. Germany: Springer-Verlag Berlin and Heidelberg GmbH & Co. K.

Muntinga, D.G., Moorman, M., & Smit, E.G. (2011). Introducing COBRA’s: exploring motivations for brand-related social media use. International Journal of Advertising, Vol. 30 (1), 13-46.

Nielsen, A. (2013). Global trust in advertising and brand messages. Conducted through: http://www.nielsen.com/us/en/insights/reports/2013/global-trust-in-advertising-and-brand-messages.html

Phau, I., Teah, M. (2009). The devil weirs (counterfeit) Prada: a study of antecedents and outcomes of attitudes towards counterfeits of luxury brands. Journal of Consumer

Marketing.

Pavlou, P.A., & Dimoka, A. (2006). The nature and role of feedback text comment in online marketplaces: implications for trust building, price premiums, and seller differentiations.

Information Systems Research, Vol. 17, 392-414.

Do influencers change our mind?

Do influencers change our mind?

Do influencers change our mind?

PREFACE

EXECUTIVE SUMMARY

TABLE OF CONTENTS

1 INTRODUCTION

2 LITERATURE REVIEW

3 METHODOLOGY

4 RESULTS

5 CONCLUSION AND DISCUSSION

REFERENCES