Graduate School of Communication Master’s Program Communication Science
Corporate Communication
Master’s Thesis
“Instagram is the new black: How celebrity endorsement affects perception of CR”
written by Olga Zhulidova (11175621) e-mail: olga.zhulidova@student.uva.nl
supervisor: dr. I.R. (Iina) Hellsten e-mail: I.R.Hellsten@uva.nl
Abstract
Nowadays, all companies from Fortune 500 list use social media in their internal and external corporate communication. Organizations use social media to engage with customers, to build and maintain reputation and image, or just simply to share the relevant information with stakeholders. Instagram is relevantly new medium for organizations and though 45% of Fortune 500 companies have a corporate Instagram page, this medium is fairly under researched in academic literature. In this study, the effect of celebrity endorsement on the perceptions of corporate reputation on Instagram has been examined, more precisely how message and celebrity type affect the perceived reputation. A 2x2 (celebrity type vs post type) between subjects experiment design was used. 199 participants were first exposed to the article related to the legalization issues that case company (Airbnb) faced in the
Netherlands. After the article the perceptions of Airbnb corporate reputation were measured; after the article participants were exposed to the stimuli material and after that the perception of Airbnb reputation has significantly improved, showing that celebrity endorsement in general had a significant effect on the perception of Airbnb corporate reputation. The results of the study suggest that both post type and celebrity type have a direct positive effect on the perceived reputation. In addition, results suggest a moderation effect of celebrity type on the relationship between post type and perception of Airbnb reputation. The study contributes to the limited body of research about celebrity endorsements on Instagram and its connection with the perceived corporate reputation. The implications of the study, limitations and the future research suggestions are discussed.
Key words: celebrity endorsement, corporate reputation, Instagram, celebrity type, message type.
Abbreviations
CR – Corporate Reputation
eWOM – Electronic Word of Mouth IC - Influential Celebrity
NIC – Non-Influential Celebrity PIC – Post by Influential Celebrity PNIC – Post by Non-Influential Celebrity
Introduction
Recently organizations have shown great interest in celebrity endorsement as a part of their marketing strategies (Shrimp, 2000). However, research into celebrity endorsement through the lens of corporate communication received a little attention in academia and more
interestingly, to the author’s knowledge, has never been researched on Instagram. This study aims to discover whether Instagram celebrity endorsements can affect stakeholders’
perceptions of corporate reputation (CR).
Digital research agency, We’re Social, highlight the shift in people’s digital behaviors, and report a 482 million rise active social media users (21% from 2016). Interestingly, in Europe, there are 412 million social media users (almost half of total population) with 54% social media penetration1 rate in Western Europe, which is second biggest global index after North America (We’re Social, 2017). Additionally, worldwide the number of Instagram active users reached 700 million in April 2017 (Statista, 2017). These numbers illustrate the reason for corporations’ interest in using social media to engage with stakeholders all over the world. With the help of social media, organizations can easily communicate with stakeholders and potentially activate a two-way dialogue with the public (Grunig, 2009).
Every Fortune 500 company use social media both in internal and external
communication practices, mainly focusing on LinkedIn (97%), Twitter (86%) and Facebook (84%). Instagram (45%) closes top five social media channels used by highly successful firms (Statista, 2016). Therefore, this shows the need for research in this area.
Social media, like any social environment, consists of popular users, opinion leaders and influencers, which brings us to the concept of electronic “word of mouth” (eWOM).
1
eWOM is defined as “any positive or negative statement made by potential, actual, or former customers about a product or company, which is made available to a multitude of people and institutions via the Internet” (Hennig-Thurau, Gwinner, Walsh, & Gremler, 2004, p. 39). One of the features of traditional WOM and eWOM is celebrity endorsement which is defined as “use of celebrities by advertisers as spokespeople for their brands” (Erdogan, 1999, p. 292). Celebrity endorsement have been proven to affect brand awareness, purchase intention, brand recognition, purchase behavior (Erdogan, 1999). Shrimp (2000) claims that 25% of American advertisers use celebrity endorsement in order to optimize advertising efforts and reach target audiences. Tripp, Jensen and Carlson (1994) documented the tendency of consumers to believe that the motives of celebrity endorsers are based on actual attachment to the brand or product, which makes this practice to be perceived as more natural brand endorsement, than advertising, for example. Choi & Rifon (2007) explain popularity of celebrity endorsement by celebrities’ entertainment function, due to the media overloading people with celebrity news. This leads to many people desiring to be associated with celebrities and their lifestyles.
To summarize, celebrity endorsement is potentially a very powerful promotional tool and can be of particular interest to marketers and advertisers. Celebrity endorsement mainly has been researched under marketing communication (Van Noort, Antheunis, & van
Reijmersdal, 2012) or business management paradigms (Holt, 2016). However, there is room to believe that this communication tool has a strong effect not only in marketing and business practices, but also on CR. Celebrity endorsement has received very little attention in
corporate communication literature. Even less attention has been paid to its influence in CR. The purpose of this study is to close this gap and shed light on the opportunities in the use celebrity endorsement while building and managing CR.
CR is the public perception of an organization, and positive reputation may determine organization’s success, lead to positive word of mouth and be a main criterion for loyalty
bonding (Fombrun, 1996; Walsh, Mitchell, Jackson, & Beatty, 2009). Therefore, a positive reputation can improve organization’s financial performance while negative reputation can lead stakeholders toward distrust in organizational products and services (Roberts & Dowling, 2002; Groenland, 2002). Hence, it is vital for organization to manage their reputation (Van Norel, Kommers, Van Hoof & Verhoeven, 2014).
This study claims that the credibility of celebrity endorsement is crucial when it comes to the CR. Roberts (2010) highlights importance of distinction between messenger credibility and message credibility. Messenger credibility is defined as the degree to which “recipient believes that the information source provider himself has considerable knowledge, skills and experiences, and that the information provided by him is objective without bias” (Ho & Chang Chien, 2010, p. 396). This research will focus on the different perception of celebrity endorsements between influential or non-influential celebrities (IC; NIC) (Amigó E. et al., 2014). Bakshy, Mason, Hofman, and Watts (2011) define influential social media users, or influencers (which includes but not limited to celebrities) as those who
“disproportionately impact the spread of information or some related behavior of interest”. They claim that influential strength on social media is related to the number of followers one has. Thus, it is argued that celebrities who have small amount of the followers could be considered as NIC. Nevertheless, they still have an impact on the people’s opinion among their followers. Message credibility is the degree to which a message is accurate, trustworthy and believable (Flanagin & Metzger, 2003; Roberts 2010). Flanagin and Metzger (2000) claim that online information credibility depends on the information type. They found that commercial information is perceived less credible than news, or entertainment information. Hence, it will be examined if the message type, in addition to celebrity type, affects
The main goal of this study is to research if celebrity endorsements affect stakeholders’ perception CR and whether it can be adapted to corporate communication paradigm.
RQ: To what extent does celebrity endorsement (particularly celebrity type and post type) affects stakeholders’ perception of organizational reputation?
Scientific relevance of this paper lies in the attempt to research whether celebrity endorsement could be adapted by corporate communication practitioners and used to build and maintain corporate communication; meantime social relevance of this paper lies in the attempt to shed light on peoples’ perceptions of celebrity endorsement with regard to IC and NIC, and which is perceived more positively when the celebrity endorsement contains personalized message (experience, emotion, etc.) or general (statement) message related to the company. In addition, this study’s social relevance is strongly related to Instagram as one of the fastest-growing visual social media networks. Additionally, visual content on social media has been proven to be more engaging than textual information (Djafarova &
Rushworth, 2017).
To answer this research question, an online experiment will be conducted with celebrity endorsement as independent variable and stakeholders’ perception of CR as dependent variable. In the experiment, Airbnb will be used as a case organization. Airbnb is an online sharing economy service platform which offers its users to rent or lease short-term lodging. Shared economy is defined as “a sustainable economic system built around the sharing of private assets” (Zekanović-Korona & Grzunov, 2014). Airbnb was founded in 2008 and expanded rapidly, resulting 100 million users worldwide and 2.3 million active listings by 2016 (Statista, 2016). The online property rental market is full of opportunities, with web-sites like tripping.com, FlipKey and VRBO adapting Airbnb’s economy sharing business model. In such a saturated market, a positive reputation can serve as a competitive
advantage (Fombrun et al., 2000). Moreover, Airbnb faced a wave of negative publicity recently related to its property rent legalization procedure and tax avoidance (Speier, 2017). It is crucial for Airbnb to maintain a positive reputation, as shared economy services become more popular (Zekanović-Korona & Grzunov, 2014). Another key thing to remember is that Airbnb already uses celebrity endorsements in their communication strategy, such as
providing top celebrities a free luxury property stay in exchange for positive eWOM (Beyoncé, Lady Gaga, Kylie Jenner, Gwyneth Paltrow, etc.). Lastly, one of the import Airbnb features are the photos of the apartments/rooms people can see. Customers write their review based on the match between photos and real apartment. Therefore, visual trait of the platforms play an important role in attracting new customers and user experience. Airbnb has an Instagram page with 2 million followers count and Airbnb hashtag (#airbnb) has been featured 2 million times among Instagram users. Thus, it is claimed that Airbnb case suits the present study’s research design, because of being a platform that needs to maintain their positive reputation, usage of celebrity endorsements and visual content preferences. Theoretical Framework
Corporate reputation
“It takes many good deeds to build a good reputation, and only one bad one to lose it” – these famous words by Benjamin Franklin, even decades after been said, are still a relevant
description of the modern business industry’s attempts to build and maintain CR. Fombrun (1996) claims that the reputation is a crucial aspect for the corporations to consider in their business strategy, as a good reputation may determine a company’s success.
The concept of CR has been in great interest of academics since the middle of the 20th century (Berens & Van Riel, 2004). Reputation is defined as a combination of “[…]
everything that is knowable about a firm. Empirically, it is a judgment of the firm made by a set of audiences on the basis of perceptions and assessments” (Schultz, Mouritsen, &
Gabrielsen, 2001, p. 24). Additionally, Barnett, Jermier and Lafferty (2005, p.34) define CR as “collective judgments of a corporation based on assessments of the financial, social, and environmental impacts attributed to the corporation over time”. This definition shows that building and maintaining CR is not only one single time point practice, but long term strategical commitment, that corporations should implement into their business plans. The Dutch CR expert Cees van Reil in his recent interview to Erasmus University journal “RSM Discovery: Management Knowledge” highlights the importance of organizations to create such a perception of corporation among its stakeholders, that it will have an “[…] added value for their personal lives […]” (van Riel, 2014, p. 5). Van Reil (2014) claims that this is one of the most influential future CR trends called social relevance. Positive communication of social relevance may have a great impact on CR, according to Van Reil (2014) if
companies focus on the positive impact they, their products or/and services have on a societal level, public will perceive the organization positively as well. For example, Philips and its LED technology, which decreases electricity costs and supports sustainable energy (Van Reil, 2014). Van Reil (2014) claims that social relevance could be also achieved without having socially relevant products or services, but by stressing the organizational relevance in their communication strategies, for example booking.com established Booking Cares initiative, which allows employees of the company to spend a certain amount of fully reimbursed working days on volunteering projects supported by the program, such as environmental and community related projects.
In addition to social relevance theory, the argumentation is based on the fact that the creation and maintaining of good reputation is also related to social expectations of CR. Berens and Van Riel (2004) define three mainstream ideas of people’s perception of corporations’ reputation and image: social expectations, corporate personality and trust. Social expectations mainstream stands for the public perception of firm’s likelihood to act in
a certain way; corporate personality mainstream reflects the personality attributes that people endue to organizations; and trust is the public’s reflection of corporation’s honesty, reliability and generosity. In this study, the focus is on the social expectations theory, as it is argued that social relevance of the company and social expectations by public could be better expressed and measured in the environment observers have an access to – social media networks. Verhezen (2015) in his work related to CR in a digitalized world, claims that acting responsibly and in accordance with social expectations is a crucial long term decision corporations should make, when building their business model. Additionally, Verhezen (2015) highlights importance to invest in so-called social ties corporation may have with its stakeholders. He claims, that these ties could be built in social media environment and argues that the digital age corporate giants, such as Apple, Amazon, Airbnb and others, should invest in social media communication (Verhezen, 2015).
In the present study, the interest also lays in the measurement of the reputation. According to Highhouse, Brooks and Gregarus (2009), CR is a multidimensional concept based on various associations, coming from various sources such as direct experience of a firm; or advices from the friends or family members. Stakeholders also can form their attitudes toward the organization based on media coverage (Highhouse et al., 2009). The one of the oldest ways of information transmission called “word of mouth”, in a digital age has transformed into the “electronic word of mouth”. WOM is a widely-used marketing tool and WOM is based on consumers’ opinion of company’s product or service. Word of mouth perceived as more credible and trustworthy source of information compared to corporate messages, for example (Flanagin & Metzger, 2000). WOM allows customers to decide if they are eager to say something about the corporation (Allsop, Bassett., & Hoskins, 2007). In this case, customers are seen as being independent from the corporations, in other words
customers perceived as an independent and credible source of information about company’s services or products (Silverman, 1997).
eWOM (Electronic Word of Mouth)
With the rise of Internet and digital technologies, more and more companies are using electronic practices to reach their audiences. In early practices, word-of-mouth (WOM) have been proven to affect customers purchase intention (Richins & Root-Shaffer, 1988).
However, with the evolving digital landscape and rise of social media, WOM have
transformed into electronic word-of-mouth (eWOM) (Hennig-Thurau, Gwinner, Walsh, & Gremler, 2004). Many scholars researched effects of eWOM on consumer behaviors, for example Park and Kim (2008) have found that customers tend to rely on electronic reviews when considering to buy some products, however they found that cognitive fit (e.g. alignment between message type and information processing strategy) plays a crucial role for the
purchase intention.
Traditional WOM occurred in conversations and discussions among friends and family, but with the fast development of digital technologies, WOM have evolved into the social media environment. This made eWOM easier to measure and research (Park & Kim, 2008). Additionally, eWOM have been found to be perceived as more natural and socially relevant type of communication (Park &Kim, 2008). Balasubramanian and Mahajan (2001) argued that customer can extract both economic and social value from eWOM, and have different expectations and motivation for engaging in eWOM. These traits of eWOM are in line with social expectations mainstream in CR research, with the help of eWOM companies can translate social expectations into the actual messages by credible sources.
eWOM hardly been researched in connection with corporate communication. Van Norel et al. (2014) one of the first attempted to research if eWOM may affect CR damage. They were interested if eWOM of highly influential Twitter users (e.g. celebrity
endorsement) can restore the CR damage and have found that indeed tweets by celebrities can restore stakeholders’ perceptions. Authors argued that when it comes to eWOM
measurement, the message type and user type should be taken into the account (Van Norel et al., 2014). Additionally, many researches of eWOM were mainly focused on electronic reviews or blogposts and their relation to various consumer behavior types (Godes & Mayzlin, 2009; Chu & Kamal, 2010; Forman, Ghose, & Wiesenfeld, 2008), however not so many researches dived into the social media environment. Social media environment for eWOM from public’s perspective is perceived as an easier way to express one’s opinion about products or services and engage with the eWOM, than engaging with customer service dialogues or catching media attention; from corporation perspective, in social media
environment it is easier to monitor and measure eWOM, however nearly impossible to control it (Van Norel et al., 2014).
Cheung and Thadani (2012) conducted a meta-analysis of the recent literature related to eWOM and developed a conceptual framework of eWOM practices based on numerous factors such as type of stimuli (e.g. message type), responses (consumer behaviors),
communicators (e.g. user credibility), receivers (consumer prior involvement with eWOM) and contextual factors (platform). In the present study, the type of the message and user will be researched, and in addition customers’ responses will be studied, operationalized as a perception of CR. In addition, contextual factor also will be taken into the account, by
measuring eWOM practices on Instagram. Users previous involvement with eWOM practices will be also taken into the account. In other words, a comprehensive research of eWOM effects on CR will be conducted, taking into the account majority of the factors that have been previously presented in academic literature.
Celebrities have been found to have an impact on people’s opinion. Primarily celebrity endorsement has been researched in academic literature as having effects on brand awareness (Premeaux, 2005) both between males and females, having a slightly greater effect on males; effectiveness and usefulness of celebrity endorsements in advertising and its effect on brand image (Hakimi, Abedniy, & Zaeim, 2011); the effects of celebrity trustworthiness, expertise, and attractiveness on customers’ brand attitudes (Amos, Holmes, & Strutton, 2008); or celebrity source attractiveness and its relationship with purchase intention (Kahle & Homer, 1985). Celebrities are often considered as opinion leaders: ‘people who influence the
opinions, attitudes, beliefs, motivations, and behaviors of others’ (Valente & Pumpuang, 2007, p. 881). However, with the rise of social media platforms celebrity communication also migrated into the new environment. When Twitter was launched, it gained popularity very fast because of wide range of celebrities starting to use the platform. The same could be applied to Instagram, many celebrities are using Instagram nowadays and the “blue tick” feature (officially verified page) gives people opportunity to easily find and follow their favorite celebrities. Celebrities are very influential in an online world, the example of that is that Kim Kardashian gets more than 20 000 dollars per tweet and around 330 000 dollars for Instagram post (Forbes, 2016). The annual earnings of Kim Kardashian in 2016 was fifty-one million dollars and according to Forbes (2016) around 20% of Kardashian earnings come from endorsements on social media. This shows a great interest of brands to engage with highly influential users on social media. In addition, the difference in Kardashian
endorsement costs on Twitter and Instagram shows that advertisers are in particular interest of Instagram platform as they are willing to pay more for communication there. Lastly, according to Djafarova and Rushworth (2016) the purpose of celebrity endorsement is to add value to a brand, product or service, which is also in line with CR social relevance theory.
This study emphasizes the importance of celebrity endorsement credibility when it is used in corporate communication practices. In his early research, Erdogan (1999)
distinguished four core models of celebrity endorsement: the source credibility model, the source attractiveness model, product match model and the meaning transfer model. This study focuses on the source credibility model and product match model, because these models are linked to the social relevance and social expectations theories of CR. According to the social relevance theory of CR, organizations and their products/services have an extra value in public’s life, the product match model of celebrity endorsement emphasizes
importance of the perceived “fit” between the company and celebrity. For example, if celebrity who has a long history of emphasizing his/her fast food passion, suddenly starts endorsing healthy lifestyle brands, it is emphasizes the extra value of the brand, as besides the nutrition advantages, the brand also affects a lifestyle of the celebrity and company’s extra value is highlighted. The source credibility model refers to the fact that effectiveness of celebrity endorsement is correlated with endorsers’ to be perceived as trustworthy and expert (Erdogan, 1999). People are interested in following celebrity advices, as this makes them feel closer to celebrity lifestyles, and gives an additional value of the brand (Choi & Rifon, 2007), which is according to social relevance theory is an important factor for the maintaining and positive reputation. The product match model refers to the fact that for effective celebrity endorsement practice, the brand and the celebrity image should fit on numerous factors, such as match between the values of the company and celebrity image. This argumentation is also in line with social expectation theory of CR. Both product match and source credibility, according to Erdogan (1999) have potential to be perceived more natural and positive
perception of the endorsement. In this study, it is augured that the differentiation between IC and NIC is important. However, the influence type of celebrities hasn’t been researched academically and therefore, the argumentation is based on the fact that number of followers
even for celebrities will play a role when determine their influence (Bakshy et al., 2011). Bakshy et al. (2011) argued that the level of influence on Twitter is correlated with the number of followers. To imply this argument in this study a pre-experiment survey will be conducted, in this survey 20 female celebrities will be chosen randomly (Appendix A) with different number of followers on Instagram and participants will be asked to rate the celebrities on 10-point scale by answering to the question: “If you have to make a decision about house/apartment rent, please indicate to what extent would you rely on the opinion of below listed celebrities?”. The most and least IC will be chosen for the experiment design. It is argued that the more IC the more people will rely his/her opinion (Bakshy et al., 2011). In addition, it is argued that if celebrity is perceived credible and has a great number of
followers, this will indicate her/his level of influence:
H1: Endorsement by IC will lead to more positive perception of Airbnb reputation than endorsement by NIC.
Moderation and mediation effects. Roberts (2010) argue that credibility of celebrity endorsement depends on message credibility to the same degree as a source credibility. The message credibility refers to the extent of the message believability (Roberts, 2010) which is reflected in message accuracy, trustworthiness and completeness (Flanagin & Metzger, 2003). Message credibility may depend on the structure of a message and the language attractiveness (Roberts, 2010). Flanagin and Metzger (2000) claim that Internet information credibility dependent on the type of information transmitted. For example, commercial information is perceived less credible than news or entertainment information (Flanagin and Metzger, 2000). Therefore, similarly to Van Norel et al. (2014) research the message type will be tested by distinguishing between personalized and general messages. Van Norel et al. (2014) claimed that personalized message had a greater effect on the CR perception than general message. In addition, it is argued that the celebrity type (IC vs NIC) will moderate
the effect of the message type, due to the Instagram environment, where the communication is based on the visual aspects. Furthermore, it is claimed that the source type will mediate the relationship between the message type and perception of Airbnb reputation, as it has been found that personal messages are more attractive for the audience and play a great role when it comes to brand endorsement (Marwick & Boyd, 2010). Hence, it is claimed:
H2: a) Personalized Instagram posts will lead to more positive perception of Airbnb reputation than general Instagram posts, b) and this will be more pronounced for posts by influential celebrity (PIC) than posts by non-influential celebrity (PNIC).
H3: Post type (personalized vs general) will mediate the relationship between celebrity type (IC vs NIC) and perception of Airbnb corporate reputation.
The interest of this study lays in testing hypothesis in social media environment, and particularly on Instagram. Instagram has been found a new “manna from heaven” for celebrity endorsement, as it emphasizes the visual context of the message and it has been found that visual posts are creating more engagement than just text posts (Djafarova & Rushworth, 2017). In the next chapter the Instagram social network and the important of measuring celebrity endorsement in this particular environment will be reviewed.
Figure 1. Conceptual Model
One of Instagram slogans says: “We imagine a world more connected through photos”. This company’s motto describes the main business idea behind the platform, to share moments through the visual attributes. The platforms that use visual feature as their main component, such as Instagram, Pinterest, Tumblr, Snapchat and others are becoming more and more popular among users (Thomas & Thomas, 2012). Implementation of new features such as Instagram stories, photo and video stickers, location filters and disappearing photos/videos that one can send via private messages on Instagram, have tremendously raised the number of Instagram active users. In the first quarter of 2017, Instagram reported 300 million active daily users, 80% of Instagram users are from outside of USA, Instagram user base forms 20% of all Internet users and up to January 2017, 4 billion of photos have been shared through the Instagram (omnicoreagency.com, 2017). It is not surprise that Instagram is used actively by corporations for both marketing and corporate communication purposes. However, Instagram as a corporate communication tool received very little attention in academic literature, mostly academics were interested in social media in general, missing the opportunity to look into the depth of Instagram - the second biggest content sharing social media network in the world (smartinsights.com, 2017).
According to Bakhshi, Shamma and Gilbert (2014) photos and videos are becoming significantly more important way of sharing information than text posts. eWOM is now a separate practice in marketing communication, called influencer marketing (Ferguson, 2008). Majority of the influencer content is produced for Instagram and YouTube (Tapinfluence, 2017). According to Tapinfluence report (Tapinfluence, 2017) Instagram plays an import role in customer-company communication and is one of the most desirable platforms for
influencer marketing. Additionally, influencer marketing has been shown to outperform traditional practices, for instance, 11% higher ROI (return on investment) levels compared to other marketing practices, more likely to be positively perceived by new generation,
(TapInfluence, 2017) etc. This also shows organizations’ interest in using Instagram for engaging with its customers. Lastly, according to Forbes (2017) influencer marketing will gain even more popularity among practitioners in 2017.
Dijkmans, Kerkhof & Beukeboom (2014) found a positive relationship between consumer social media engagement and their attitudes towards CR. However, their research was generic, and based on various social media activities, rather than only one specific network. Additionally, Djafarova and Rushworth (2016) claim that celebrities' pages are at the top of the list of the most-followed profiles on Instagram and it is not surprising that organizations use these users as their voice among their followers. People tend to think that users with a high number of followers are more trustworthy (Jin & Phua, 2014), which is in line with source credibility theory.
eWOM on Instagram is constantly emerging and mostly based on the popularity of Instagram social network itself and accessibility of Instagram via smart phones and other devices (Thoumrungroje, 2014). Celebrity endorsement is perceived, by organizations, as credible source in generating a positive eWOM regarding company, its products or services (Spry, Pappu, & Bettina Cornwell, 2011). The impact of celebrity endorsement as marketing communication tool has been extensively researched previously, however it has not been researched under the corporate communication paradigm. For example, numerous of researches have been conducted to connect celebrity endorsements and various marketing outcome variables, such as brand awareness and purchase intention; however, celebrity endorsement did not receive an attention in corporate recommunication literature in
connection with CR or brand image, CSR practices and etc. Additionally, Instagram as one of the fast-growing social media networks also did not receive enough attention in the literature. To close these gaps our study attempts to research impact of celebrity endorsement on the perception of CR among its stakeholders on Instagram.
Method
Design. To investigate the effect celebrity endorsement could have on CR, an online experimental design has been adopted. The 2x2 between subjects true experiment was conducted with the celebrity endorsement, operationalized as celebrity type (influential vs non-influential) and type of the Instagram post (personal vs general) as independent variable; and stakeholders’ perception of Airbnb reputation as dependent variable (Appendix B).
After the set of demographic questions participants were exposed to the article about Airbnb legalization problems (Appendix C). After the article participants were asked to evaluate Airbnb reputation and were randomly assigned to one of the four experimental conditions (see Table 1).
Table 1. Experimental Conditions Distribution.
Condition N %
PIC with personalized message 41 20.6
PIC with general message 39 19.6
PNIC with personalized message 39 19.6
PNIC with general message 40 20.1
Control condition 40 20.1
Total 199 100
Sample. The participants for the experiment were recruited by the means of convenience snow ball sampling. An invitation to participate in the experiment was sent to the potential participants by e-mail, posted on the Facebook and LinkedIn. The invitation to participate in the study, besides the general information about the experiment, contained the information about the sampling procedure, asking participants to pass the experiment to the relevant potential participants. The recruitment process occurred between 8th and 20th May 2017. In
total 199 people participated in the experiment and on average it took 19 minutes to
participate in experiment, 51 participants have been excluded from the final dataset, because not completing the survey, not fitting the requirements of the experiment (past and current Airbnb stakeholders), resulting in response rate = 79%. The majority of participants were females 51.8%, and 47.2% were males. The average age interval was between 18 and 24 years old (48.2%). 83.4% of participants had a higher education (Bachelor degree and higher, including doctorate), 13.6% were college graduates and only 3% were high school graduates. Among all the participant 46.7% were students and 36.7% were employed (20.1% full time and 16.6% part time), 11.6% were unemployed and looking for a job, 4.5% of participants stated their occupation level as freelancer. 42.7% of participants were single, 31.7% were in relationship and 10.1% were in civil partnership, 13.1%were married and 2.5% were divorced. Majority of participants were from the Netherlands (74.9%), Azerbaijan (12.6%) and France (2%).
Stimuli. Before the main experimental conditions, a pre-test have been conducted to measure the celebrity influence levels and test perceptions of the general and personalized Instagram posts. The survey has been distributed via Facebook between 21st and 29th April 2017 by the means of snowball sampling. In total 69 people participated in survey, average time to fill survey was 7 minutes, 1 person have been excluded from the final dataset, because not completing the survey, resulting in response rate = 98.6%. Sample demographics presented and detailed pre-test procedure presented in Appendix D.
The pre-test survey was divide in two blocks. First was related to the celebrity
influence and participants were asked to express reliable celebrity source by answering to the question: “If you have to make a decision about house/apartment rent, please indicate to what extent would you rely on the opinion of below listed celebrities?”, participants were proposed to rate their choice on 5-point Likert scale ranging from “Extremely likely” to
“Extremely unlikely”, including addition answer option “I don’t know this celebrity”.
Penelope Cruz and Lindsay Lohan have been selected for the main experiment design, due to the low levels of “I don’t know this celebrity” answers.
The second block was investigation whether the Instagram posts caption have been reflecting personalized or general information. The participants have been exposed to four Instagram posts containing personalized and general information. Participants were asked to indicate degree to which they agreed that post was containing general of personalized information (see Appendix E) on a 5-point Likert scale ranging from “Strongly agree” to “Strongly disagree”. Based on the analysis it was decided to include the captions of Post #3 and Post #4 as personalized captions; and Post #5 and Post #6 as general captions; for the main experiment, resulting in stimuli material presented in Appendix F.
Measures. eWOM and Celebrity Endorsement. The independent variable of this research is celebrity endorsement on Instagram. This variable consists of two dimensions: post type and celebrity type. To create this variable a pre-test has been conducted to choose the celebrity type (influential vs non-influential) and post type (personalized vs general). The celebrity endorsement variable is nominal level variable and consists of following categories: PIC containing personalized caption (1), PIC containing general caption (2), PNIC containing personalized caption (3), PNIC containing general (4). These four categories were
operationalized into the Instagram posts.
CR (corporate reputation). The perception of the CR was measured with the scale developed by Ponzi, Fombrun and Gardberg (2011) - RepTrak™ Pulse. Participants were asked to assess four following statements on a reversed 5-point Likert scale (from “strongly agree” to “strongly disagree”) two times: first, after reading the article about Airbnb and seconds, after the Instagram manipulation:
Airbnb is a company that I trust.
Airbnb is a company that I admire and respect. Airbnb has a good overall reputation
For both scales the Principal Component Analysis with corresponding varimax rotation have been conducted. For the before-manipulation scale factor analysis indicated that there was one component (eigenvalue = 3.61) which explained 90.2% of variance in 4-items and factor loadings ranged from .94 to .96. As a next step, scale reliability analysis was performed and showed the high reliability level, reflected in Cronbach’s a = .96 (M=3.30; SD=1.02). For the after-manipulation scale factor analysis also indicated only one component (eigenvalue = 3.64) which explained of 91.0% variance and factor loadings ranged from .94 to .97. Scale reliability analysis indicate Cronbach’s a = .97 (M=2.57; SD =.95) (see Appendix G, Table 1 and 2). Both scales were computed and proceeded for the further analysis.
Control variables. Several control variables were used in the experiment. Control variables were related to the participants’ perceptions of shared economy services, social media use and previous experience with celebrity endorsements (Table 3).
Demographic variables. In the beginning of the experiment participants were asked to indicate several demographics traits such as their gender, age, education level, marital status and occupation status. Those questions were asked to check the normal distribution of the variables in the sample and to test the randomization.
Randomization check. Randomization of the experimental conditions has been checked. First, a one-way ANOVA was conducted with condition, as independent variable and age, as dependent variable. ANOVA indicated no main effect of independent variable on age (F (4,194)=3.82, p=0.100, η2= 0.07), which confirms no differences among experimental conditions on the basis of age. Thus, randomization of age has been successful. Second, one-way ANOVA with education level as dependent variable was conducted. ANOVA analysis
indicated there were no differences of education level in experimental conditions (F (4,194)=1.959, p=0.102, η2=0.04). Hence, randomization was successful on the basis of education level. Additionally, marital status (F (4,194)=5.416, p=0.070 η2=0.03) and
occupation level (F (4,194)=2.058, p=0.088, η2=0.05) also successfully met the requirements of randomization based on one-way ANOVA results. Lastly, the Chi square test has been conducted to check if there were any differences in experimental conditions for males or females. Chi square test indicated no differences between conditions because of gender (Χ2 (8) = 4.83, p = 0.775). hence, randomization of gender was successful. To conclude all above, randomization of gender, age, education level, marital status and occupation level was
successful and there was no need to include mentioned variables in the analysis as control variables.
Manipulation check. First, to check if participants perceived the Airbnb article topic
correctly the question “In the beginning of the experiment the news article about Airbnb have been presented. Do you recall what article was about?” was asked in the end of survey. The answers were: 1. Airbnb new travel destination; 2. Airbnb subletting crisis; 3. Airbnb
investment crisis; 4. Airbnb web site new feature. 93% of respondents answered correctly to this question choosing a “Airbnb subletting crisis” option (M=2.00; SD=0.36).
Secondly, participants were asked a question about celebrities use in the experiment: “Do you believe that Instagram users used in this study were celebrities?” with respective “yes” and “no” answers.
One-way ANOVA analysis have been conducted to test if there is a difference between experimental conditions and control condition with regards to this question.
Levene’s test indicated that the requirement for the equal variances between the groups have not been met (F (4,194)=6.97; p=.000). Thus, Welch test have been conducted and showed that there is a significant difference between the groups (F (4, 96)=44.61; p=.000). The
post-hoc test showed that the control group significantly differed from the groups with
personalized post by IC (Mdifference=.80; p=.000); with general P IC (Mdifference=.0.80; p=.000), with personalized post by NIC (Mdifference=.77; p=.000) and with general PNIC (Mdifference=.63; p=.000). To put it in other words, participants in experimental conditions indicated that Instagram users used in the experiment were actual celebrities, and participants in control condition indicated that the users used in experiment were not celebrities. Thus, it is assumed that manipulation worked.
Thirdly, participants were asked if they would rely on the opinion of the celebrity they have seen in the experiment, they were able to rate their answer on 5-point Liker scale from 1 - “extremely likely” to 5 - “extremely unlikely”. The reverse scale types were used in this research on constant basis.
One-way ANOVA analysis have been conducted to check whether celebrity type manipulation worked. Levene’s test indicated the equal variances between the groups were not assumed (F (4,194)=3.61; p=.007). Hence, Welch test have been performed and
indicated that there is a significant difference between the groups (F (4, 96)=30.6; p=.000). The results post-hoc test, presented in Appendix G, Table 4, show that PIC significantly differed from PNIC and control group. However, there was not a significant difference between control group and participants in the group where general post by NIC have been exposed. This result will be addressed in limitations of this study and it is concluded that manipulation based on celebrity type partially worked.
Lastly, to check if the manipulation based on post type worked, a one-way ANOVA have been conducted. Levene’s test showed that requirement of equal variances between the variables the was not met (F (4,194)=3.78; p=.006). Accordingly to the statistical standards, Welch test have been conducted and showed that there is a significant difference between the
groups (F (4, 96)=49.9; p=.000). The post hoc test results, showed in Appendix G, Table 5, indicated that there was a significant difference between groups based on the post type criteria: groups with general posts significantly differed from the groups with personalized posts and control group. Hence, it is concluded that manipulation based on the post type also worked.
Procedure. The online experiment has been distributed by the means of convince snowball sampling to participants by sending anonymous link to participants’ e-mails and private messages on Facebook, additionally experiment survey has been posted on Facebook and LinkedIn. After the clicking on the link, participants were exposed to the general
information about the experiment, terms of anonymity, contact details of researcher, as well as informed about the possibility win 25 Euro gift-card from bol.com in the end of the experiment. Next step was an agreement statement. As the main interest of this study lays in Airbnb stakeholders (past and current) next question asked participants if they ever used Airbnb services, those who answered no were transmitted to the end of the survey and excluded from the data analysis. After this question, a series of demographic questions were asked related to the participants’ gender, age, education level, occupation and marital statuses.
In the next section participants were exposed to the series of questions related to the sharing economy services. In the next section participants were exposed to the article about Airbnb legalization crisis and after reading the article they were asked to rate the Airbnb reputation. Next participants were exposed to the series of questions related to the social media use (if they were not social media users, they were skipped to the manipulation), such as what social media platforms they use and how often. As a next step participants were randomly assigned to one of the four experimental conditions and one control condition. After the manipulation participants were asked to rate Airbnb reputation again. Lastly,
participants were exposed to the celebrity endorsement section (if they follow any celebrities on social media, on what platforms do they follow celebrities, if theses celebrities promote any brands, etc.) In addition, manipulation check questions were present.
In the end of the experiment participants were able to leave their e-mail addresses to win a gift-card from bol.com. After filling their e-mails participants were debriefed about the experiment. The debriefing statement contained information about study goal and that all materials used in the experiment were fictional. After the defining participants were ended the survey.
Results
In the present experiment, the effect of celebrity endorsement on Airbnb CR perceptions have been studied. The perception of Airbnb CR has been measured twice in the experiment, after participants read the article about Airbnb (O1) and after they have been exposed to the Instagram intervention (O2). First, to check if there was a difference between O1 and O2, a paired sample T-test have been conducted. The results of T-test show that there was a significant difference (t (198)=9.72; Mdifference=.72; p=.000; 95% CI [1.05;.07]) between reputation before the manipulation (M=3.30; SD=1.02) and reputation after manipulation (M=2.57; SD=.95). These results confirm the assumption that celebrity endorsement affects perceptions of CR.
To test the first hypothesis of whether endorsement by IC led to more positive perception of Airbnb reputation than endorsement by NIC an independent sample T-test have been conducted with celebrity type (0-influential, 1-non-influential) as independent variable and Airbnb reputation after the manipulation as dependent variable. Levene’s Test showed that equal variances requirement was not met (F=83.3, p=.000). The results of T-test (t (116)=4.55, Mdifference=.58; p=.000, CI 95% [.33;.84]) indicated that there was a significant difference between means of PIC (M=2.08, SD=.52) and PNIC (M=2.67, SD=1.02) on stakeholders’
perceptions of Airbnb reputation. The results of independent sample T-test suggest that people who were exposed to PIC perceived Airbnb reputation more positive than those who have been exposed to the PNIC and this effect was of medium strength (d=.72). Thus, first hypothesis was accepted.
To test of whether personalized Instagram posts led to more positive perception of Airbnb reputation than general Instagram post (H2a) and the moderation effect of celebrity type on this relationship (H2b), a moderation analysis using PROCESS macro by Hayes (2014) have been conducted, using Model 1 (simple moderation analysis) with post type as independent variable, perceptions of CR after the manipulation as dependent variable and celebrity type as a moderator variable. The overall regression model was significant, F (3, 155) = 33.4, p = .000, R2 = .39. There was a significant direct effect (H2a) of post type on perceptions of Airbnb reputation, b = .65, SE = .11 t (155) = 6.08, p = .000, 95% CI [.44;.86], meaning that people who were exposed to the personalized Instagram posts tend to score higher on the perception of Airbnb reputation scale. Hence, hypothesis 2a have been accepted. As a next step, moderation effect has been examined. The moderation analysis found a significant interaction effect between post type and celebrity type (b = 1.23, SE = .21 t (155) = 5.78, p = .000, 95% CI [.81;1.65]), meaning that celebrity type has a significant moderation effect on the relationship between post type and perception of Airbnb reputation. Hence, people who were exposed to the personalized post tend to score higher on Airbnb reputation scale, and were scoring higher on the same scale when exposed to the PIC than by NIC. Thus, hypothesis 2b has been accepted.
Table 4. PROCESS macro Model Summary. Model Summary
R R2 F df1 df2 p
Model
CI
b SE t p LL UL
Constant 2.37 .053 44.4 .000 2.26 2.47
Post Type (personalized vs general) .65 .11 6.08 .000 .44 .86 Celebrity type (influential vs
non-influential)
.57 .11 5.35 .000 .36 .78
Interaction 1.23 .21 5.78 .000 .81 1.65
To test whether post type (general vs personalized) mediates the relationship between the celebrity type (IC vs NIC) and stakeholders’ perception of Airbnb reputation a linear regression analysis has been used. As a direct effect between celebrity type and perception of Airbnb CR has been established in hypothesis 1, according to MacKinnon, Lockwood, Hoffman, West and Sheets (2002), the regression analysis with celebrity type and Airbnb reputation perception as independent variable and post type as dependent variable should have been conducted. MacKinnon et al. (2002) argued that the requirement for the
establishing the full mediation effect with dichotomous mediator is the insignificant effect of celebrity type with the coefficient (b*) decrease and significant effect of the perception of Airbnb CR. However, the results of the regression analysis suggest that the overall model was not significant (F(2, 156)=15.32 p > .05). Thus, there were not a mediation effect of the post type on the relationship between the celebrity type and perception of Airbnb reputation and hypothesis 3 has been rejected.
Conclusion and discussion
In this study, the effect of celebrity endorsement on CR was examined. More precisely, the study focused on the effect that Instagram posts by celebrities have on the perceptions of CR and, if there an effect, which celebrity type and which message type plays a greater role. The findings of the research suggest that, in general, Instagram posts by celebrities have a
significant effect on stakeholders’ perceptions of Airbnb CR. These Instagram posts could be potentially used by corporation to maintain their reputation or to mediate reputational damage if such happen (Van Norel et al., 2014). This study sheds the light into the perceptions of CR in an environment, where corporations have a little power to control the communication flow – social media. Existing researches has already shown that Instagram is a platform that may affect brand perceptions (Djafarova & Rushworth, 2017). Celebrity and influencer
endorsements are widely used as marketing promotional tool for various business objectives such as brand awareness (Premeaux, 2005), brand image (Hakimi et al., 2011), consumer behavior (Godes & Mayzlin, 2009), etc. For brands, positive eWOM is in a great interest, as it has been found that people show a tendency to trust social media networks as sources of information (Jansen, Zhang, Sobel, & Chowdury, 2009). Another key thing to remember is that this study attempted to understand the effect of celebrities spreading a positive word about corporation via Instagram and their influence on perceptions of CR.
In this study, the celebrity type and the type of the message have been researched. The main effect has been found between the celebrity type and perceptions of CR. Celebrity endorsement by IC led to higher perceptions of Airbnb CR. This may be explained by the fact that the IC in this research, according to pre-test results has been perceived as a trustworthy in terms of property rent, which demonstrates so-called fit between the celebrity and the brands, which according to Fink, Parker, Cunningham, and Cuneen (2012) is a crucial aspect when it comes to celebrity use in corporate communication practices.
In the present study, in addition to celebrity type, effects of message type were
investigated. The findings suggest a significant main effect of message type on perceptions of CR, meaning that personalized messages affect perceptions of CR more than general
messages. In addition, moderation effect of celebrity endorsement has been found, meaning that PIC moderated the relationship between the message type and stakeholders’ perceptions
of corporate information. Finally, there were no mediation effect of the message type on the relationship between the celebrity type and perceptions of CR.
Theoretical Implications. The present study contributes to the academic literature on a numerous topic. First of all, it is a first study to research effect of celebrity endorsement on perceptions of CR on Instagram. Several researches have been conducted to research effects of celebrity endorsements, however, most of them were related to marketing communication paradigm (Godes & Mayzlin, 2009) or CR during the crisis (Van Norel et al., 2014). This is a first research attempting to establish a general effect between these two concepts. Secondly, similar to Van Norel et al. (2014) this study researched message type and celebrity type as a dimensions of celebrity endorsement and similar findings have been examined. The type of celebrity significantly improved CR perceptions. However controversially to Van Norel et al. (2014), who did not find a significant effect of the message type on the perceptions of CR, in this research effect of message type on the CR perception has been found. Additionally, in the present study different type of celebrities has been used, compared to Van Norel et al. (2014) attractive and intelligent celebrities, in the present study IC and NIC has been used, with relevant pre-test measures.
Managerial Implications. This study demonstrates findings that could be advantageous for the modern corporations. Firstly, the study shows that celebrity endorsements could be adapted from the marketing communication and used by practitioners to maintain the CR, however, as it was discussed in theoretical framework section, corporations need to make sure to establish fit between the corporation and celebrity (Hakimi et al., 2011). Secondly, the study shows that if corporation decide to engage with the celebrity endorsement for the sake of the improving their CR both celebrity type and message type should be taken into the account. For example, if corporation decides to engage with NIC, in other words celebrity who does not perfectly fit to organizational image, it is better that this celebrity to use
personalized messages to endorse the corporation. Finally, this study sheds a light of how corporations can use Instagram for the sake of building and maintaining CR. It shows, that the usage of celebrity endorsement on Instagram not only possible as a corporate
communication tool, but also this tool can be advantageous when it comes to the CRs maintenance (Van Norel et al., 2014). For instance, this tool could be used similar to Van Norel et al (2014) for maintaining a reputation while reputational damage or to establish social relevance of the company, when company does not have a product/service portfolio, which allows them to establish social relevance without celebrity endorsement (Van Reil, 2014)
Limitations and suggestions for future research. First of all, the study implied an experimental design which is designed to replicate the real-life conditions, however the reality for Instagram users will be the scrolling of Instagram feed, which was not possible in this study’s design. In addition, in this experiment people were exposed to the Instagram posts at one particular moment, however when it comes to reality if celebrities collaborate with the brands they produce sponsored content continuously. Future research should focus more on the implementing of those limitations, for example, by creating an interactive Instagram feed and imply longitudinal design, using the same sample for several times. Secondly, manipulations check related to the celebrity type worked only partially, which also should be taken into the account when research’s results will be replicated. Thirdly, in this study only female celebrities have been used. To strengthen the results of the research it would be plausible to use male celebrities as well. Finally, the sample type does not allow the results of the study to be generalized to broader population, as in this study a convince
snowball sampling method has been used.
The presents study was a first step into the research of celebrity endorsement effect on the perceptions of CR on Instagram. According to the research results both celebrity type and
message type play a great role in this relationship. Additionally, celebrity type significantly moderated relationship of post type and perception of CR. Future research is crucial to strengthen the results of this study.
References
Active Instagram users from January 2013 to April 2017 (2017). Statista. Retrieved from: https://www.statista.com/statistics/253577/number-of-monthly-active-instagram-users/ Airbnb Statistics & Facts. (2016). Retrieved from:
https://www.statista.com/topics/2273/airbnb/
Albu, O. B., & Etter, M. (2016). Hypertextuality and Social Media A Study of the
Constitutive and Paradoxical Implications of Organizational Twitter Use. Management Communication Quarterly, 30(1), 5-31. doi: 10.1177/0893318915601161
Allsop, D. T., Bassett, B. R., & Hoskins, J. A. (2007). Word-of-mouth research: Principles and applications. Journal of Advertising Research, 47, 398–411. doi:
10.2501/S0021849907070419
Amigó E. et al. (2014). Overview of RepLab 2014: Author Profiling and Reputation
Dimensions for Online Reputation Management. In: Kanoulas E. et al. (eds)
Information Access Evaluation. Multilinguality, Multimodality, and Interaction. CLEF 2014. Lecture Notes in Computer Science, 8685, 3017-322, doi: 10.1007/978-3-319-11382-1_24.
Amos, C., Holmes, G., & Strutton, D. (2008). Exploring the relationship between celebrity endorser effects and advertising effectiveness: A quantitative synthesis of effect size. International Journal of Advertising, 27, 209–234. doi:
10.1080/02650487.2008.11073052
Bakhshi, S., Shamma, D. and Gilbert, E. (2014). Faces engage us. Proceedings of the 32nd annual ACM conference on Human factors in computing systems - CHI '14. doi: 10.1145/2556288.2557403
Bakshy, E., Mason, W. A., Hofman, J. M., & Watts, D. J. (2011). Everyone’s an influencer: Quantifying influence on Twitter. In Proceedings of the 4th ACM international
conference on web search and data mining, Hong Kong, CN, 9–12 February, 2011 (pp. 65–74). Hong Kong: ACM.
Balasubramanian S., Mahajan V. (2001). The economic leverage of the virtual community. International Journal of Electronic Commerce, 5(3), 103 – 138.
Barnett, M., Jermier, J., & Lafferty, B.A. (2005). Corporate Reputation: The Definitional Landscape. Corporate Reputation Review, 9(1), 26-38, doi:
10.1057/palgrave.crr.1550012.
Berens, G & van Riel, C. B. M. (2004). Corporate Associations in the Academic Literature: Three Main Streams of Thought in the Reputation Measurement Literature. Corporate Reputation Review, 7 (2), 161–178, doi: 10.1057/palgrave.crr.1540218
Berens, G., & Van Riel, C. B. M. (2004). Corporate associations in the academic literature: Three mains streams of thought in the reputation measurement literature. Corporate Reputation Review, 7, 161–178. doi: 10.1057/palgrave.crr.1540218
Cheung, C. and Thadani, D. (2012). The impact of electronic word-of-mouth communication: A literature analysis and integrative model. Decision Support Systems, 54(1), 461-470. doi: 10.1016/j.dss.2012.06.008
Choi, S.M. & Rifon, N.J. (2012). It is a Match: The Impact of congruence between celebrity image and consumer ideal self on endorsement effectiveness. Psychology and
Marketing, 29(9), 639–650. DOI: 10.1002/mar.2055
Chu S. & Kim Y., (2011) Determinants of consumer engagement in electronic word-of-mouth (eWOM) in social networking sites, International Journal of Advertising, 30, 47-75. doi: dx.doi.org/10.2501/IJA-30-1-047-075
Chu, S., & Kamal, S. (2010). The effect of perceived blogger credibility and argument quality on message elaboration and brand attitudes: An exploratory study. Journal of Interactive Advertising, 8, 26–37.
David H. Silvera Benedikte Austad, (2004). Factors predicting the effectiveness of celebrity endorsement advertisements. European Journal of Marketing, 38 (11/12),1509 - 1526. doi: dx.doi.org/10.1108/03090560410560218
Digital in 2017: Global Overview (2017). We’re Social. Retrieved from: http://wearesocial.com/uk/blog/2017/01/digital-in-2017-global-overview
Dijkmans, C., Kerkhof, P. and Beukeboom, C. (2015). A stage to engage: Social media use and corporate reputation. Tourism Management, 47, 58-67. doi:
10.1016/j.tourman.2014.09.005
Djafarova E., Rushworth C. (2017) Exploring the credibility of online celebrities' Instagram profiles in influencing the purchase decisions of young female users. Computers in Human Behavior, 68, 1-7. DOI: 10.1016/j.chb.2016.11.009
Doh S.J. & Hwang J.S., (2009). How consumers evaluate eWOM (electronic word-of-mouth) messages, CyberPsychology & Behavior. 12(2), 193-197. doi:10.1089/cpb.2008.0109 Erdogan B. Z. (1999). Celebrity Endorsement: A Literature Review, Journal of Marketing
Management, 15(4), 291-314. doi: 10.1362/026725799784870379
Ferguson, R. (2008). Word of mouth and viral marketing: taking the temperature of the hottest trends in marketing. Journal of Consumer Marketing, 25(3), 179-182.doi: 10.1108/07363760810870671
Fink, J., Parker, H., Cunningham, G. and Cuneen, J. (2012). Female athlete endorsers: Determinants of effectiveness. Sport Management Review, 15(1), 13-22. doi: 10.1016/j.smr.2011.01.003
Flanagin, A. J., & Metzger, M. J. (2000). Perceptions of internet information credibility.
Journalism and Mass Communication Quarterly, 77, 515–540. doi:
10.1177/107769900007700304
Flanagin, A. J., & Metzger, M. J. (2003). The perceived credibility of personal Web page information as influenced by the sex of the source. Computers in Human Behavior, 19, 683–701, doi: 10.1016/S0747-5632(03)00021-9.
Fombrun, C. J. (1996). Reputation: Realizing value from the corporate image. Boston: Harvard Business School Press
Fombrun, C. J., Gardberg, N. A., & Sever, J. M. (2000). The reputation quotient: A multi-stakeholder measure of corporate reputation. Journal of Brand Management, 7, 241– 255. doi: 10.1057/bm.2000.10
Forman, C., Ghose, A., & Wiesenfeld, B. (2008). Examining the relationship between reviews and sales: The role of reviewer identity disclosure in electronic markets. Information Systems Research, 19, 291–313. doi: 10.1287/isre.1080.0193 Gibbs, J.L., Rozaidi, N.A., & Eisenberg, J., (2013). Overcoming the ideology of
openness: Probing the affordances of social media for organizational knowledge sharing. Journal of computer-mediated communication, 19, 102-120. doi: 10.1111/jcc4.12034
Giese, J. L., Spangenberg, E. R., & Crowley, A. N. (1996). Effects of product-specific word-of-mouth communication on product category involvement. Marketing Letters, 7, 187– 199. doi:10.1007/BF00434909
Godes, D. and Mayzlin, D. (2009). Firm-Created Word-of-Mouth Communication: Evidence from a Field Test. Marketing Science, 28(4),721-739. doi: 10.1287/mksc.1080.0444 Godes, D., & Mayzlin, D. (2009). Firm-created word-of-mouth communication: Evidence
Groenland, E. A. G. (2002). Qualitative research to validate the RQ dimensions. Corporate Reputation Review, 4, 308–315. doi: 10.1057/palgrave.crr.1540152
Grunig, J. E. (2009). Paradigms of global public relations in an age of digitalisation. PRism, 6(2), 1-19.
Hakimi, B. Y., Abedniy, A., & Zaeim, M. N. (2011). Investigate the impact of celebrity endorsement on brand image. European Journal of Scientific Research, 58, 116–132. Hennig-Thurau, T., Gwinner, K. P., Walsh, G., & Gremler, D. D. (2004). Electronic
word-of-mouth via consumer-opinion platforms: What motivates consumers to articulate themselves on the Internet? Journal of Interactive Marketing, 18, 38–52. doi: dx.doi.org/10.1002/dir.10073
Hennig-Thurau, T., Malthouse, E. C., Friege, C., Gensler, S., Lobschat, L., Rangaswamy, A., et al. (2010). The impact of new media on customer relationships. Journal of Service
Research, 13, 311–330. doi: 10.1177/1094670510375460
Highhouse, S., Thornbury, E.E., & Little, I.S. (2007). Social-identity functions of attraction to organizations. Organizational Behavior and Human Decision Processes, 103, 134-146, doi: 10.1016/j.obhdp.2006.01.001
Ho, H., & Chang Chien, P. (2010). Influence of message trust in online word-of mouth on consumer behavior. In International conference on electronics and information (pp. 395–399). Los Alamitos: IEEE.
Holt, D. (2016). Branding in the age of social media. Harvard Business Review, 94(3). Instagram statistics (2017). Omnicore Agency. Retrieved from:
https://www.omnicoreagency.com/ blog/instagram-statistics
Jansen, B., Zhang, M., Sobel, K. and Chowdury, A. (2009). Twitter power: Tweets as
electronic word of mouth. Journal of the American Society for Information Science and Technology, 60(11), 2169-2188. doi: 10.1002/asi.21149
Jin, S. and Phua, J. (2014). Following Celebrities’ Tweets About Brands: The Impact of Twitter-Based Electronic Word-of-Mouth on Consumers’ Source Credibility Perception, Buying Intention, and Social Identification With Celebrities. Journal of Advertising, 43(2), 181-195. doi: 10.1080/00913367.2013.827606
Lee M. & Youn S., (2009) Electronic word of mouth (eWOM), International Journal of
Advertising, 28(3), 473-499. doi: http://dx.doi.org/10.2501/S0265048709200709
MacKinnon, D., Lockwood, C., Hoffman, J., West, S. and Sheets, V. (2002). A comparison of methods to test mediation and other intervening variable effects. Psychological Methods, 7(1), 83-104. doi: 10.1037/1082-989X.7.1.83
Marwick, A. and boyd, d. (2011). To See and Be Seen: Celebrity Practice on Twitter. Convergence: The International Journal of Research into New Media Technologies, 17(2), 139-158. doi: 10.1177/1354856510394539
Muda M., Musa R. & Putit L. (2012) Breaking through the Clutter in Media Environment: How Do Celebrities Help? Procedia - Social and Behavioral Sciences, 42, 374-382. doi: 10.1016/j.sbspro.2012.04.201
Muntinga, D.G. (2016). Managing brands in the age of DIY-branding: The COBRA
approach. In T. Langner et al. (Eds.), Handbuch Sozialtechniken der Kommunikation,1-20. Wiesbaden: Springer Fachmedien.
Page, G., & Fearn, H. (2005). Corporate Reputation: what do consumers really care about?
Journal of Advertising Research, 45, 305–313. doi:
doi.org/10.1017/S0021849905050361
Park, D. and Kim, S. (2008). The effects of consumer knowledge on message processing of electronic word-of-mouth via online consumer reviews. Electronic Commerce
Ponzi, L. J., Fombrun, C. J., & Gardberg, N. A. (2011). RepTrak™ Pulse: Conceptualizing and validating a short-form measure of corporate reputation. Corporate Reputation Review, 14(1), 15–35, doi: 10.1057/crr.2011.5.
Premeaux, S. R. (2005). The attitudes of middle class male and female consumers regarding the effectiveness of celebrity endorsers. Journal of Promotion Management, 11, 33–48. doi: 10.1300/J057v11n04_04
Richins M.L., &i Root-Shaffer, T. (1988). The Role of Evolvement and Opinion Leadership in Consumer Word-Of-Mouth: An Implicit Model Made Explicit, Advances in
Consumer, 15, 32-36.
Robehmed N. (2016, November 16). Kim Kardashian West's Earnings: $51 Million In 2016. Forbes. Retrieved from:
https://www.forbes.com/sites/natalierobehmed/2016/11/16/kim-kardashian-wests-earnings-51-million-in-2016/#7dcab9326a5b
Roberts, C. (2010). Correlations among variables in message and messenger credibility scales. American Behavioral Scientist, 54 (1), 43–56, doi: 10.1177/0002764210376310. Roberts, P. W., & Dowling, G. R. (2002). Corporate reputation and sustained superior
financial performance. Strategic Management Journal, 23, 1077–1093. doi: 10.1002/smj.274
Sanders K. (2017, June 26). What Is The Future Of Influencer Marketing? Forbes.
Retrieved from: https://www.forbes.com/sites/forbesagencycouncil/2017/06/26/what-is-the-future-of-influencer-marketing/#2c23283e28ca
Schultz, M., Mouritsen, J., & Gabrielsen, G. (2001). Sticky reputation: Analyzing a ranking system. Corporate Reputation Review, 4, 24–4, doi: 10.1057/palgrave.crr.1540130. Shrimp, T.A. (2000), Advertising Promotion: Supplemental Aspects of Integrated Marketing
