Who says what and where?
The influence of a message’s content, platform and source on electronic word-of-mouth behaviour and an organisation’s reputation
Master’s Thesis
Master’s programme Communication Science Graduate School of Communication
Name: Judith Bisig Student number: 10841644
Supervisor: Christine Liebrecht
Abstract
In the online environment, the corporate reputation as an important organisational asset gets harder to manage. Since the reputation is shaped by what people say about the organisation, the concept of electronic word-of-mouth (eWOM) gets inevitable. Using a corporate
approach, this study investigated the influence of a message’s content, source and platform on people’s intention to engage in eWOM, positive eWOM (pWOM) specifically, and on the corporate reputation. Therefore, 110 respondents filled in the online experiment using a 2 (platform: Facebook vs. corporate) x 2 (content: informative vs. entertaining) x 2 (source: MGC vs. UGC) mixed-subject experimental design. Results showed that people intend more to engage in pWOM in corporate communities than on Facebook. Also it was shown that the effects of a message’s content and source on pWOM as well as the emotional appeal towards the organisation depend on the platform. Providing these results, this study implements a clear approach on the concept of eWOM, which was situated on the blurring lines between the scholar perspectives and therefore offers valuable suggestions for future research.
Additionally, it offers some practical guidelines for reputation managers on how to manage what people say and feel about the organisation.
Key words. Electronic word-of-mouth (eWOM), pWOM management, corporate reputation, message characteristics
Introduction
The concept of electronic word-of-mouth (eWOM) has become increasingly important; especially in an era where people’s lives revolve more and more around the Internet, for organisations it is not only crucial to be present online and reach their audience, but also that people talk about the organisation, which builds and maintains a strong reputation (Bickart & Schindler, 2001). Reputation is an important asset to increase stakeholders’ loyalty as well as a protecting shield from a crisis’ harm (Coombs, 2007; Shamma, 2012). Thus, the
Research regarding persuasive communication has shown that word-of-mouth (WOM) off- and online is perceived as more credible and trustworthy than traditional advertising (Katz & Lazarsfeld, 1955; Engel, Blackwell, & Kegerreis, 1969) as well as marketing (Schiffman & Kanuk, 1995; Meuter, McCabe & Curran, 2013) and how electronic WOM influences attitudinal and behavioural consequences (Bickart & Schindler, 2001; Pavlou & Dimoka, 2006). Furthermore, current research has also investigated the sender’s motives that cause people to talk (positively) about an organisation online (Cheung & Lee, 2012; Men & Tsai, 2013; Muntinga, Moorman, & Smit, 2011; Enginkaya & Yilmaz, 2014; Hennig-Thurau, Gwinner, Walsh, & Gremier, 2004; De Matos & Vargos Rossi, 2008) and how that is perceived (Meuter et al., 2013; Weisfeld-Spolter, Sussan, & Gould, 2014). However, the Internet not only offers the possibility to influence its customers, but also to build a long-term relationship with them, resulting in blurring lines between the persuasive and corporate perspective on eWOM. Yet, the corporate point of view – how eWOM can be used to build long-term relationships with the consumers – has remained largely neglected, resulting in the question what marketers can do to make people talk positively with and about an
organisation. What marketers can manage is a message’s content, which was also called the “lynchpin” that connects the WOM givers and receivers (Nguyen & Romaniuk, 2014). Therefore, this study’s focus is on which content stimulates people the most to talk positively about the organisation. More specifically, acknowledging the multiple possibilities for
marketers to spread their messages – on social media or own corporate blogs (Kaplan & Haenlein, 2012) – and the increasing power of consumers online (Hennig-Thurau et al., 2010; Macnamara, 2010), two additional factors, platform and source, shall be included in this study. This leads to the following research question:
RQ: What kind of content influences positive word-of-mouth (pWOM) and an
organisation’s reputation the most positively and does that differ by the source of the message and the communication channel used?
By answering this research question, this study provides relevant guidelines for reputation managers on how to communicate with its customers in order to build and strengthen the relationship with them and how to increase the reputation. Furthermore, it adds to the scholar understanding of eWOM by demonstrating the blurring lines between research fields and clarifying the other (so far neglected) corporate perspective on eWOM. Moreover, due to its explorative character, this study proposes new factors to be considered in future research.
Theoretical framework
Reputation, as one of an organisation’s main assets to attract customers and a source of competitive advantage (Fombrun, Gardberg, & Sever, 2000; Hall, 1992), has been the focus of practitioners as well as academics in the last decades (Fombrun & van Riel, 1997;
Fombrun & van Riel, 2004; Abbratt & Klein, 2012). It has been found to enable organisations to realize opportunities, neutralize threats (Argenti & Druckenmiller, 2004) and even be a protective shield in certain crisis contexts (Sohn & Lariscy, 2012). In this study, it shall be referred to the definition by Fombrun et al. (2000) that defines corporate reputation as a “collective representation of a firm’s past behaviour and outcomes that depicts the firm’s ability to render valued results to multiple stakeholders” (p. 243). Also, this representation is one in stakeholders’ minds that is evaluated over time (Gotsi & Wilson, 2001) and consists of a rational as well as an emotional component (Dijkmans et al., 2015). Despite this long-term character, participants’ perception of the corporate reputation based on a single stimulus will be investigated (Dijkmans, Kerkhof, & Beukeboom, 2015). However, because of the brief (and therefore heuristic) exposure to the stimuli that information might affect more the emotional component (Petty & Cacioppo, 1986). Additionally, when those mental
representations are uttered, corporate reputation is being shaped by what people say and think about an organisation (Brønn, 2010) and is therefore closely linked to the concept of word-of-mouth.
Traditional word-of-mouth (WOM) has been demonstrated to have a significant effect on consumers’ behaviour along with a higher impact than traditional advertising types (Katz
& Lazarsfeld, 1955; Engel et al., 1969). With the expansion of the Internet this consumer-to-consumer behaviour also shifted online (Gruen, Osmonbekov, & Czaplewski, 2006). This offered organisations the opportunity to join those conversations and try to impact consumers’ attitudes and behaviour. However, joining electronic WOM (eWOM) conversations cannot only be regarded as a tool to “manipulate” consumers, but is also a valuable way to build a relationship with them. Like this, eWOM allows organisations to adapt a symmetrical two-way model of communicating with its public and is therefore one of several options (e.g. issue management, framing, agenda setting) to manage the corporate reputation (Grunig & Grunig, 1992).
EWOM is referred to 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 et al., 2004, p.39). Especially in the last decade – with the rise of social media - eWOM has got increasingly important for consumers and conjointly in academic research (see King, Racherla, & Bush, 2014). Particularly social networking sites (SNSs), such as Facebook, are popular platforms to engage with brands, i.e. liking, sharing and commenting on brands’ posts (King et al., 2014). Since people articulate their thoughts and feelings publicly by engaging with brands, those utterances (likes, comments & shares) are equated with eWOM (Men & Tsai, 2013), which is connected to the corporate reputation. Moreover, following the typology of Muntinga et al. (2011) who distinguished brand-related behaviour online in consuming, contributing and creating, eWOM mainly refers to the latter two; contributing and creating (i.e. commenting on and sharing a brand’s content).
Because WOM off- as well as online is perceived to be more reliable, credible and trustworthy than traditional marketing (King et al., 2014; Meuter et al., 2013; Weisfeld-Spolter et al., 2014; Brown, Broderick, & Lee, 2007; Buttle, 1998; Godes & Mayzlin, 2004; Schiffman & Kanuk, 1995), it is an important factor to be considered in building and maintaining an organisations’ reputation. Especially since eWOM has been found to be an important driver of consumers’ engagement (Algesheimer et al., 2010; Nambisan & Baron,
2007; Schau & Muniz, 2002) next to the effects on trust and loyalty (Awad & Ragowsky, 2008; Ba & Pavlu, 2002; Gauri, Bhatnagar, & Rao, 2008), and customers’ purchase behaviour (Chevalier & Mayzlin, 2006; East, Hammond, & Wright, 2007; Bickart & Schindler, 2001; Pavlou & Dimoka, 2006). With all those effects of eWOM the importance of it for reputation managers gets striking.
As shown in the definition above, eWOM can consist of positive as well as negative statements, whereby negative WOM (nWOM) is a type of complaining behaviour online that much research has focused on (Dekay, 2012; van Noort, Willemsen, Kerkhof, & Verhoeven, 2014) - with webcare and reputation management it is tried to forcome the corporate
reputation’s harm. However, unlike that reactive approach as in webcare, this study focuses on the proactive way of fostering a positive conversation about the brand and thereby building a strong corporate reputation. Moreover, it has been shown that positive WOM (pWOM) is more memorable than nWOM (Kimmel & Kitchen, 2014) and has the potential to attract new customers (Rosen, 2000; Stokes & Lomax, 2002). Moreover, pWOM is expected to build and maintain a strong corporate reputation and therefore prevents nWOM beforehand. Ergo, it will be investigated what practitioners can do in order to trigger that positive conversation with and about the organisation.
In order to know how to trigger pWOM, it is relevant to know what makes people turn to SNSs and engaging with organisations online, which has been investigated by a number of studies. Several motivations for SNS usage have been identified, such as entertainment, integration and social interaction, personal identity, information (Muntinga et al., 2011), which have later been extended by renumeration and empowerment (Men & Tsai, 2013). It is reasoned that those findings can be used as an orientation on what people might find
interesting, which might foster pWOM. However, reviewing the existing literature (and keeping the above-mentioned findings in mind) two points are striking: Firstly, the literature so far has mainly focused on self- and social-related motivations of eWOM (Cheung & Lee, 2012; Enginkaya & Yilmaz, 2014; Hennig-Thurau et al., 2004; Men & Tsai, 2013; Muntinga et al., 2013), but lacks a corporate point of view; it doesn’t indicate concretely what
practitioners can do in order to increase eWOM and thereby build and strengthen the corporate reputation. Besides other factors it’s mainly a message’s content that reputation managers can influence. Secondly, it strikes that a major part of the studies only focus on Facebook and neglect the Internet’s potential for an organisation’s own corporate community. Moreover, this study acknowledges that the Internet hands the power over to the users
(Hennig-Thurau et al., 2010; Macnamara, 2010), which offers consumers the possibility to either react to marketer-generated content (MGC) or place self-produced content about the brands.
Following, it shall be taken a closer look into those three (so far neglected) factors that might influence people’s eWOM behaviour; the content of the post, the platform it is posted on and the source it is posted by.
The content
A message’s content can be understood as the link between the eWOM givers and receivers (Nguyen & Romaniuk, 2014), resulting in the question which type of content consumers are more likely to forward. Furthermore, findings suggest that eWOM can be increased by providing interesting content about the brand that people will want to share with others (Nguyen & Romaniuk, 2014). It is therefore argued that a post’s content has to address consumers’ needs and motivations for it to spark off as much pWOM as possible.
Research to date has found a multiplicity of different motivations for people to engage in eWOM. Especially, because organisations may be perceived as intruders in the online environment, they should try to meet consumers’ needs and offer gratifications to those motivations that lead consumers to turn to media online (Katz, Gurevith, & Haas, 1973). Conspicuously, there are two motivations that recur in a number of studies; information and entertainment. Not only, have they been demonstrated to be important drivers for people to turn to social networks in the first place (Tsai & Men, 2012; Lin & Lu, 2011), but they have also been shown as important drivers for people’s eWOM behaviour (Men & Tsai, 2013; Muntinga et al., 2011; de Vries, Gensler, & Leeflang, 2012; Cvijlkj & Michahelles, 2013).
The importance of those two motivations results in the suggestion for organisations to
disseminate a mixture of informative and entertaining content (Men & Tsai, 2013; Katz et al., 1973). However, the literature hasn’t been very uniform about this distinction nor doesn’t it offer any examples for its operationalization. On the one hand, content shall be distinguished as follows: brand- or product-related posts (containing information about the company and/or its products) will be considered as informative posts, whereas content that is brand- or
product-unrelated shall be regarded as entertaining content (Cvijikj & Michahelles, 2013). On the other hand, some studies were found that operationalize informative content as providing useful and relevant information, whereas entertaining content is understood to deliver
enjoyable and entertaining content (Cho, Huh, & Faber, 2014; de Vries et al., 2012). It shall be noted that most of those ways of operationalization weren’t conducted in experimental studies and are only vaguely explained. This study, however, will combine the
afore-mentioned approaches and take the following; informative content is understood as useful & relevant information, which is explicitly linked to the products (i.e. naming the product name). Further, entertaining content is operationalized as enjoyable and entertaining with a link to the product category, but no explicit mention of the product.
However, even though people might turn to SNSs for both information and entertainment reasons, the exposure to informative vs. entertaining content may result in different types of activity. Regarding eWOM, the distinction between consuming,
contributing to and creating brand-related content can be made (Muntinga et al., 2011). Those authors also showed that entertainment is one of the drivers for brand-related consumption, contribution as well as creation, whereas information only drives people to consume that content. Consequently, it is expected that entertaining content leads to more pWOM and a more positive reputation than informative content.
The platform
Secondly, it shall be taken a closer look into the platform a message is posted on. SNSs, which allow individuals as well as organisations to construct profiles and interact with each
other (Boyd & Ellison, 2007), are important platforms for people to talk with and about brands, thus to engage in eWOM (e.g. brand fan pages on Facebook). Because of their high popularity and the lower costs of getting in contact with its customers more efficiently, organisations quickly turned to SNSs (Kaplan & Haenlein, 2012). Moreover, SNSs provide a suitable platform to adopt a symmetrical communication model to build up long-term
relationships with consumers (Grunig & Grunig, 1992; Grunig, 2009). But next to SNSs, brands also have the possibility to be present and interact with their customers on their own corporate platform, which could have the advantages of building up an own community and not being dependent on a platform that could be gone in a few years time (Kaplan &
Haenlein, 2012). Furthermore, it was drawn up that for users there are differences between the motivations to join traditional online brand communities (e.g. corporate communities) and brand fan pages on SNSs: While joining a corporate community is more driven by an interest of the brand that is the centre of the community, joining a brand fan page on Facebook is also a means to show one’s self-identity to one’s own “friends”, who don’t necessarily have to be fans of that brand as well (Jahn & Kunz, 2012). This even leads to the presumption that the motivation for the usage and engagement might differ between the two platform options (Jahn & Kunz, 2012).
That difference, however, hasn’t been researched so far. However, from the differing motivations to join either a corporate community or a brand fan page on Facebook, some expectations can be derived: Considering that people joining a corporate community are probably more committed to the brand, they are expected to engage more in pWOM and to have a more positive reputation of that brand than people who solely want to shape their image by joining a brand fan page on Facebook.
The source
Originally, the Internet was created to link people together so they could share information more quickly and more easily and engage in consumer-to-consumer (C2C) communication. So, it has been argued that organisations – discovering the potential of web 2.0 technologies
and trying to leverage that potential by playing a part in C2C communication – are the uninvited guest in the online environment (Fournier & Avery, 2011). Hence, whether it is a marketer generating content on behalf of the firms or consumers, might be an important factor to be considered when managing the corporate reputation. So far, it hasn’t been investigated if and how a content’s source impacts people’s intentions to engage in pWOM and the
oganisations effort to build a strong reputation. Followingly, this study examines whether there is a difference between marketer-generated content (MGC) or content that is generated by users (UGC), who thereby start a conversation with the organisation in question (Goh, Heng, & Lin, 2013).
Earlier research showed that a source’s intentions are an important factor for the audience’s evaluation of the trustworthiness (Eagly, Wood, & Chaiken, 1978). Hence, a marketer’s intention to improve the corporate reputation (and thereby increase also the monetary value of the organisation [Fombrun et al., 2000]) might have a different impact on the credibility than a user’s intention to get involved with a brand. Furthermore, it has been argued that people’s perception that a source is similar to the receiver can also lead to an increased persuasive influence (Hass, 1981; Price, Feick, & Higie, 1989). Thus, because UGC is originated by a source that’s independent of the organisation or brand (Bickart & Schindler, 2001) and has the potential to generate greater empathy, trustworthiness and relevance than MGC (Eagly et al., 1978), it is expected to result in a higher credibility. Due to those outlined advantages of UGC and the more positive emotions and heuristics towards it, it is also
expected to result in a higher amount of pWOM and a higher reputation than MGC.
In conclusion, there are several expectations that shall be investigated in this study: pWOM and a positive reputation are expected to be higher for entertaining compared to informative content, on a corporate community compared to Facebook and when originated by a user instead of a marketer. However, derived from the above outlined research so far, some interaction effects are expected as well: Because of the higher commitment in corporate
communities compared to Facebook groups, the effects of a message’s content and its source are expected to be stronger for the corporate community posts than the Facebook posts.
Methodology Design
To answer this study’s research question, a 2 (platform: Facebook vs. corporate) x 2 (content: informative vs. entertaining) x 2 (source: MGC vs. UGC) mixed-subject experimental design was used (see table 1). In order to reduce the needed total of participants, platform was chosen as the only between factor. Hence, participants were assigned to one of two questionnaires (Facebook vs. corporate) where they saw the 4 constructed posts alternated with the fillers, thus, keeping possible confounds over the 4 conditions constant and thereby increasing internal validity (Bryman, 2012). Moreover, the order of the 4 conditions was randomized in order to rule out order-effects (Bryman, 2012). So, once assigned to either the Facebook or corporate community condition, each participant randomly saw informative and entertaining posts, generated by marketers and users.
Table 1. Experimental design.
Between-subject factor: Within-subject factors
Platform Content: informative Content: entertaining
Source Source
MGC UGC MGC UGC
Questionnaire I Corporate platform
Condition 1 Condition 2 Condition 3 Condition 4 informative / MGC informative / UGC entertaining/ MGC entertaining/ UGC Questionnaire II Facebook
Condition 5 Condition 6 Condition 7 Condition 8 informative / MGC informative / UGC entertaining/ MGC entertaining/ UGC
Pre-test
In order to guarantee internal validity regarding the manipulation of the stimuli and to select 4 brands that wouldn’t significantly differ regarding the attitudes towards them, a pre-test was conducted. To minimize effects of product-type and/or the brands’ branch, only brands from the FMCG market were chosen, because it can also be expected that participants know those brands and show a comparatively low engagement (Cvijikj & Michahelles, 2013). Therefore, the 20 most popular FMCG Facebook brand pages were selected (Socialbakers, 2015, April 15) whereof the ones not relevant for the European market were deleted, resulting in a total of 16 brands, towards which the participants had to indicate their attitude on three 7-point
semantic differential items (unfavourable – favourable, negative – positive and bad – good) (MacKenzie & Lutz, 1989). In order to test the content’s manipulation, for two random – but representative – brands (i.e. Snickers and Milka), two posts manipulated in their informative vs. entertaining content were shown to participants who rated the content on how informative and entertaining it is, using a single semantic differential item (1 = informative, 7 =
entertaining).
The stimuli were constructed as follows: To manipulate the posts it was orientated on existing posts of the brand fan pages of the two chosen brands. Looking at the brand fan pages, it appeared that several posts contained either a photo or video, which is why it was chosen to make posts with a picture (relating to the content and showing the brand) and aligning content (text). The informative content was based on relevant information found on the brands’ websites and explicitly named the brand, whereas the entertaining content
consisted of more enjoyable content, not explicitly naming the brand name, but related to the product type (i.e. chocolate) (see figure 1 and 2). All the stimuli can be found in Appendix A.
Figure 2. Pre-tested informative post of Snickers. Figure 1. Pre-tested entertaining post of Snickers.
A total of 30 (mainly Dutch, 76.70%) respondents between the ages of 21 and 52 (Mage = 24.40; SDage = 5.53; 13 men) participated in the pre-test for which they were told to
be exposed to brands’ communication on Facebook. First, they were exposed to the pre-selected brands to indicate the attitudes (for the means and standard deviations for the participants’ attitudes towards the brands, see Appendix C, table 1). After eliminating the brands that scored below the scale’s midpoint (N = 3), 4 brands were randomly selected: Cornetto, Magnum, Milka & Oreo. In order to guarantee that those brands could be
considered as equivalent, paired samples t-tests were conducted that showed that the brands didn’t significantly differ from each other (ps > .568, see Appendix C, table 2). In addition, paired samples t-tests confirmed that the manipulation of the content was successful; the entertaining content (Mmilka_ent = 4.87, SD = 1.41; Msnickers_ent 1 = 5.57, SD = 1.17; Msnickers_ent 2
= 5.47, SD = 1.20) was perceived as significantly more entertaining than the informative content (Mmilka_inf = 3.07, SD = 1.36; Msnickers_inf = 2.80, SD = 1.35), t(29) = 5.45, p < .001,
95% CI [1.13, 2.48] (Milka); t(29) = 8.19, p < .001, 95% CI [2.08, 3.46] (Snickers1); t(29) = 9.78, p < .001, 95% CI [2.11, 3.22] (Snickers2). Thus, the manipulation of the content could be considered as successful and the stimuli for the main experiment were prepared in the same manner.
Stimulus material
Based on the pre-test, stimulus materials were developed that resemble real Facebook / brand community posts as closely as possible. For this purpose, the 4 brands selected based on the pre-test’s were randomly assigned to the conditions. Following, for each brand a screenshot of an existing Facebook post was taken where the text was replaced with the experimental message, for which the same procedure was followed as for the pre-test. Hence, this resulted in the posts for the Facebook condition. For the corporate community posts the same pictures and text was adapted to a corporate look and feel of the brands. Therefore, screenshots from the different websites were taken to construct posts with the corporate design.
Furthermore, 4 filler stimuli were constructed to distract participants from the real purpose of the study (see Appendix A). It was chosen for online articles about celebrities that are expected to rouse low involvement and arousal that could influence the results. The fillers contained between 87 and 366 words and were also shown in a randomized order.
Participants
Participants were gathered by convenience sampling; the link to the online survey (Qualtrics) was placed on Facebook groups of UvA students, Twitter and LinkedIn, resulting in the link being opened 192 times. Unfortunately, the dropout rate was quite high (43%), which resulted in a total of 110 participants who filled in the survey completely. The 110 participants,
whereof the majority was female (62.70%), had an average age of 24.73 years (SD = 5.61). The highest completed degree for most participants was the Bachelor’s Degree (63.60%). On average the respondents were online for 36.65 hours per week (SD = 20.34) whereof they spent 16.41 hours on social media (SD = 14.10). Moreover, participants considered them as active social media users to a moderate extent (M = 4.27, SD = 1.80). Furthermore, it was interesting to see that the majority of the participants were neither a follower of the shown brands on Facebook (94.50%) nor any community member (88.20%).
Measures
Reputation towards the brand was measured by means of 18 items based on the Reputation Quotient (Fombrun et al., 2000; Dijkmans et al., 2015). Participants were asked to what extent they agree to the statements (1 = not likely at all – 7 = very likely, 8 = I don’t know).
Unfortunately, it resulted that for 9 of the items a high amount of participants (between Nrep18
= 39 and Nrep11 = 65.75) wasn’t able to indicate their opinion about the brands. Because of
that it was decided to exclude those items for further analyses. By conducting factor analyses the 9 items were averaged to form three dimensions (Dijkmans et al., 2015), three items each: Dimension 1 consisting of the first three items regarding the emotional appeal, dimension 2 consisting of the three items regarding an organisation’s products and services and dimension
3 constructed by the three items concerning an organisations vision and leadership (EVdim1 =
2.61; R2
dim1 = .87; Cronbach’s alphadim1 = .92; EVdim2 = 2.33; R2dim2 = .78; Cronbach’s
alphadim2 = .85; EVdim3 = 2.48; R2dim3 = .83; Cronbach’s alphadim1 = .89). Those three
dimensions were averaged to form one scale (EV = 2.46; R2 = .82; Cronbach’s alpha = .89). Positive word-of-mouth was measured by means of three indicators; liking, commenting and sharing. The respondents indicated how likely it is for them to like, comment on and share the shown posts (1 = not likely at all – 7 = very likely). In order to measure credibility,
respondents were asked how credible they perceived the post to be by means of 5 semantic differentials (Zafer Erdogan, 2010). A principal component analysis (PCA) showed that the 5 items loaded on two components. It resulted that the three items that loaded on the first
component could be averaged to one scale (EV = 2.23; R2 = .76; Cronbach’s alpha = .82) – the other two items (reliable – unreliable; sincere – insincere) were excluded from further
analyses.
Furthermore, in order to increase internal validity, some control variables that might confound with the dependent variables were measured. Accordingly, socio-demographic variables were measured (i.e., gender, age, educational level and nationality) and respondents were asked to indicate their Internet and social media use per week (“How many days per week do you use the Internet/social media?”) and a day (“On the days you use the Internet, how many hours do you use the Internet/social media?”). Further, participants were asked to distribute 100% of their social media use over the three types described by Muntinga et al. (2011) (consuming, contributing & creating). Summing up the percentages of contributing and creating, a new variable was built; social media engager. Furthermore, respondents’ experience with communities was assessed (“Are you a member of any online communities”) as well as their relation with the brands in question (“Are you a customer/follower on social media of the brand?”).
Procedure
Participants were asked to take part in the online experiment via social media (no incentive was given). The study started with an informed consent form in which the procedure of the study was briefly explained and that it would take about 10 minutes (for the form and the whole questionnaire, see Appendix B). Participants were told that the study’s aim was to research their opinion towards online content and that they would be asked to indicate their honest opinion. After providing their consent, the participants were randomly assigned to one of the two questionnaires where they were shown the 4 stimuli. Between the stimuli, the fillers were shown. Finally, after answering some socio-demographic questions and providing information about their social media use and control variables, participants were asked to repeat the study’s aim to check whether they guessed the study’s actual research question. Finally, participants were thanked for their participation and provided with the researcher’s email-address for further questions or complaints. In order to ensure their anonymity there was no possibility to leave their contact details.
Results Prior analyses
With regard to the dependent measures, it is interesting to note that respondents reported overall a low likeability to engage in pWOM (Mliking = 2.68, SDliking = 1.39; Mcommenting = 1.58,
SDcommenting = 0.86; Msharing = 1.39, SDsharing = 0.76). This even though the reputation towards
(M = 4.94, SD = 0.80) and the credibility of (M = 4.73, SD = 0.78) the investigated brands was reported as fairly positive.
Furthermore, it was tested to what degree pWOM and credibility correlated with reputation. Therefore, single regression analyses were conducted with liking, commenting, sharing and credibility as independent variables and reputation as dependent variable for each. The regression models for liking, F (1, 108) = 19.42, p < .001, as well as credibility, F (1, 108) = 114.35, p < .001, were significant. The analysis showed that 15.20% of the variance in reputation could be predicted by people’s intention to like the presented post.
Thus, people’s intention to like a post is moderately strongly associated with the brand’s reputation, b* = 0.39, t = 4.41, p < .001. Furthermore, the analysis revealed that a post’s credibility predicted 51.40% of the variance in reputation. So, the association between a post’s credibility and the brand’s reputation was shown to be a strong one, b* = 0.72, t = 10.70, p < .001. However, the analysis showed no significant association between reputation and commenting, F (1, 108) = 3.03, p = .084, and sharing, F (1, 108) = 0.61, p = .437.
Randomization check and controls
In order to check whether the randomization to the two questionnaires was successful the distribution of gender, age, education, Internet use and social media use over the two between conditions was investigated. In order to check the distribution of age, Internet use and social media use, a MANOVA with platform as independent variable and age, Internet use and social media use as dependent variables was conducted. The multivariate test, F (3, 106) = 1.05, p = .374), as well as the univariate tests for age, F (1, 108) = 2.29, p = .133, Internet use, F (1, 108) = 0.74, p = .392, and social media use, F (1, 108) = 0.177, p = .676, were non-significant, which showed that no significant differences between the two
questionnaires existed. The distribution for gender and education was checked by conducting Chi-square tests for both variables combined with the between factor platform. No significant difference between the conditions was found with regard to the number of men, χ 2 (1, N = 110) = 0.70, p = .401, nor with regard to the number of community members, χ 2 (1, N = 110) = 0.25, p = .617. However, the analysis showed that there was a significant difference
between the two questionnaires with regard to the educational level, χ 2 (4, N = 110) = 13.03, p = .011. As a result of these analyses it can be concluded that the randomization, except for education, was successful.
However, it was also tested whether there were any variables that could confound the interested relationships. The Spearman’s correlation coefficients showed that there were no significant relationships between education and any of the dependent measures (see Appendix C, table 3). Furthermore, using Pearson’s correlation coefficient and Chi-square analyses, it
was tested whether the other measured variables correlated with the dependent variables liking, commenting, sharing, reputation, credibility and attitude. It was found that age significantly correlated with reputation, r = -.33, p < .001 credibility, r = -.26, p = .006 and attitude, r = -.25, p = .008. Also the participants’ self-reported active behaviour on social media was found to significantly correlate with sharing, r = .28, p = .003 commenting, r = .35, p < .001 as well as liking, r = .24, p = .011. Furthermore, liking was the only dependent measure significantly correlating with the participants’ social media use, r = .22, p = .022, and social media engager, r = .27, p = .004. Furthermore, to test whether either the
participants’ gender or the fact that a respondent is any community member may confound with the interested relationships, two ANOVAS (one with gender and the other with
community member as independent variable) with the above-mentioned dependent measures entered as dependent variables were conducted. The analyses showed that the fact whether a respondent was a community member or not did not have a significant impact on any of the dependent variables entered, Fs (1, 108) < 1.94, p > .167 (see Appendix C, table 4). Also gender was found to not significantly influence liking, F (1, 108) = 3.72, p = .056, η2 = .03, commenting, F (1, 108) = 1.80, p = .183, η2 = .02, sharing, F (1, 108) = 0.52, p = .471, η2 = .00, nor credibility, F (1, 108) = 1.13, p = .291, η2 = .01. However, a significant difference between women and men was found with regard to reputation, F (1, 108) = 6.30, p = .014, η2 = .06, and attitude, F (1, 108) = 8.58, p = .004, η2 = .07. As a result of those analyses the following control variables were entered in the analyses: age, gender, social media use, social media activity and social media engager. Moreover, it was entered whether participants follow any of the shown brands on Facebook and also which post they liked the most.
Main analyses
It was expected that pWOM, reputation and credibility would be higher for entertaining (compared to informative) content, generated by users (compared to marketers) and when placed on a corporate community (compared to Facebook). Furthermore, some interaction effects were anticipated as well, such as that the effects of the content and the source would
be stronger for corporate communities compared to Facebook groups. In order to test those expectations, the effects of the three factors on the dependent measures in interest were tested; pWOM (liking, commenting and sharing), reputation and credibility. Therefore, single
ANCOVAs for each dependent variables were conducted with content and source as
independent within-factors, platform as independent between-factor and the above-mentioned as dependent variables, controlling for age, gender, social media use, social media activity and social media engager.
PWOM. To test the effects on pWOM, the three pWOM indicators were tested; liking, commenting and sharing. For the effects on liking (the means can be seen in table 2), the test showed no significant main effects of content, F (1, 95) = 0.51, p = .476, η2 = .01, or source, F (1, 95) = 3.75, p = .056, η2 = .04. However, the analysis showed that the factor platform had a significant effect on liking, F (1, 95) = 5.43, p = .022, η2 = .05. As expected, people were more likely to “like” a corporate community post (M = 3.01, SD = 1.39) than a post on Facebook (M = 2.34, SD = 1.32). No significant interaction effects on liking were found (see Appendix C, table 5).
Table 2. Means for liking (Likert scale 1 – 7) over the 8 conditions (N = 110) Between-subject factor: Within-subject factors
Platform Content: informative Content: entertaining Overall
MGC UGC MGC UGC
Corporate community 2.26 (0.22) 2.83 (0.22) 3.11 (0.26) 3.55 (0.26) 3.01 (1.39) Facebook 2.03 (0.21) 2.11 (0.21) 2.51 (0.25) 2.79 (0.25) 2.34 (1.32)
The ANCOVA with commenting as dependent variable also showed that only the platform had a significant main effect, F (1, 99) = 14.33, p < .001, η2 = .13 (see Appendix C, table 5). Although in general very low, the respondents’ likeability to comment on corporate community posts (M = 1.89, SD = 1.07) was significantly higher than on Facebook posts (M = 1.27, SD = 0.44).
Table 3. Means for commenting (Likert-scale 1 – 7) for the three factors (N = 110) Between-subject factor: Within-subject factors
Platform Content Source Overall
informative entertaining MGC UGC
Corporate community 2.04 (1.24) 1.75 (1.16) 2.09 (1.25) 1.70 (1.06) 1.89 (1.07) Facebook 1.24 (0.55) 1.31 (0.63) 1.22 (0.54) 1.33 (0.61) 1.27 (0.44)
More interestingly, significant interaction effects on commenting were found between platform and content, F (1, 99) = 5.21, p = .025, η2 = .05, as well as between platform and source, F (1, 99) = 9.48, p = .003, η2 = .09 (for the means, see table 3). Whereas on Facebook there was no significant difference in respondents’ intentions to comment on informative (M = 1.24, SD = 0.55) or entertaining (M = 1.31, SD = 0.63) content, t(55) = -0.67, p = .504, 95% CI [-0.28, 0.14], in a corporate community people were more likely to comment on
informative (M = 2.04, SD = 1.24) than on entertaining posts (M = 1.75, SD = 1.16), t(53) = 1.87, p = .067, 95% CI [-0.02, 0.58] (see Figure 3).
Figure 3. Interaction effect of content and platform on commenting (N = 110) 1 1.2 1.4 1.6 1.8 2 2.2 Informative Entertaining Corporate Facebook
Furthermore, as can be seen in Figure 4, respondents exposed to the corporate communities were significantly more likely to comment on marketer-generated (M = 2.09, SD = 1.25) than on user-generated content (M = 1.70, SD = 1.06), t(53) = 3.16, p = .003, 95% CI [0.14, 0.64]. However, on Facebook no significant difference between MGC and UGC was found:
Respondents’ likeability to comment on MGC (M = 1.22, SD = 0.54) didn’t significantly differ from the likeability to comment on UGC (M = 1.33, SD = 0.61), t(55) = -1.06, p = .293, 95% CI [-0.31, 0.10]. Thus, as expected, the influence of a messages content or source
depended on the platform it was posted on; the impacts of the content and source were stronger on corporate communities than on Facebook.
Figure 4. Interaction effect of source and platform on commenting (N = 110).
Another ANCOVA was conducted to test whether the investigated factors had an impact on people’s likeability to share the posts (the means can be found in table 4). This test showed that also for sharing no significant main effects of content, F (1, 99) = 0.00, p = .962, η2 = .00 , source, F (1, 99) = 0.02, p = .884, η2 = .00, nor platform, F (1, 99) = 1.97, p = .164, η2 = .02, were found, thus, not supporting this study’s expectations. Furthermore, also there seemed to exist no interaction effects of the investigated factors on people’s intentions to share the shown posts (see Appendix C, table 5).
1 1.2 1.4 1.6 1.8 2 2.2 MGC UGC Corporate Facebook
Table 4. Means for sharing (Likert-scale 1 – 7) over the 8 conditions (N = 110) Between-subject factor: Within-subject factors
Platform Content: informative Content: entertaining Overall
MGC UGC MGC UGC
Corporate community 1.31 (0.87) 1.35 (1.07) 1.76 (1.36) 1.69 (1.33) 1.53 (0.93) Facebook 1.18 (0.72) 1.21 (0.56) 1.39 (0.99) 1.25 (0.69) 1.26 (0.50)
Reputation. To test whether the researched factors had an effect on a brand’s
reputation, another ANCOVA was computed. Contradicting the expectations, no significant main effects on reputation of source, F (1, 98) = 0.50, p = .482, η2 = .01, and platform, F (1, 98) = 0.03, p = .864, η2 = .00, were found. However, the test showed a marginally significant effect of content on a brand’s reputation, F (1, 98) = 2.86, p = .094, η2 = .03. Furthermore,
also no significant interaction effects on reputation were found (see Appendix C, table 6). Because of those non-significant findings, it was also investigated whether the three factors had an impact on the single reputation dimensions (see Appendix C, table 7).
Table 5. Means for emotional appeal (Likert-scale 1 – 7) for the three factors (N = 110) Between-subject factor: Within-subject factors
Platform Content Source Overall
informative entertaining MGC UGC
Corporate community 5.12 (1.01) 5.07 (1.04) 5.04 (0.93) 5.15 (1.10) 5.10 (0.92) Facebook 4.78 (1.12) 4.88 (1.23) 4.93 (1.22) 4.74 (1.17) 4.83 (1.10)
Those ANCOVAs only showed a significant effect on the first dimension (the means are shown in table 5). Namely the significant interaction effect of source and platform on dimension 1 – emotional appeal (Dijkmans et al., 2015) – was found, F (1, 98) = 5.98, p = .016, η2 = .06. An independent samples t-test showed that for MGC there was no significant
(M = 4.93, SD = 1.22), t(102.48) = 0.52, p = .601, 95% CI [-0.30, 0.52] (see figure 5). However, for UGC there was a marginally significant difference between the platforms: As also can be seen in Figure 5, user-generated content on a corporate platform (M = 5.15, SD = 1.10) resulted in a significantly higher emotional appeal than when posted on Facebook (M = 4.74, SD = 1.17), t(108) = 1.93, p = .057, 95% CI [-0.01, 0.85]. Thus, it is especially UGC on corporate platforms that increases the emotional appeal towards the brand.
Figure 5. Interaction effect of source and platform on emotional appeal (N = 110)
Credibility. Because it was expected that the factor source not only impacts people’s intentions to engage in pWOM and the organisation’s reputation, but also its credibility, it was also tested whether those expectations can be supported. Thus, an ANCOVA was conducted to test the impact of the independent variables on a brand’s credibility (see table 6 for the means). As expected, the test showed a significant main effect of a content’s source on the brand’s credibility, F (1, 99) = 5.93, p = .017, η2 = .06. However, the means show that marketer-generated content (M = 4.79, SD = 0.84) was evaluated as significantly more credible than user-generated content (M = 4.66, SD = 0.83), which contradicted the
expectations. Furthermore, no significant main effects of content, F (1, 99) = 0.41, p = .523, 4.5 4.6 4.7 4.8 4.9 5 5.1 5.2 Corporate Facebook MGC UGC
η2 = .00, nor platform, F (1, 99) = 0.67, p = .415, η2 = .01, were shown. Moreover, also the
interaction effects resulted to be insignificant (see Appendix C, table 8).
Table 6. Means for credibility (Likert-scale 1 – 7) for the three factors (N = 110) Between-subject factor: Within-subject factors
Platform Content Source Overall
informative entertaining MGC UGC
Corporate community 4.85 (0.77) 4.78 (0.79) 4.85 (0.74) 4.77 (0.79) 4.81 (0.71) Facebook 4.64 (0.96) 4.65 (0.85) 4.73 (0.92) 4.56 (0.87) 4.64 (0.85) Overall 4.74 (0.87) 4.71 (0.82) 4.79 (0.84) 4.66 (0.83)
Discussion & Conclusion Discussion
This study tries to find an answer to the question what kind of content influences pWOM and the corporate reputation the most positively and whether that differs by the source of the message and the communication channel used. Thereby, it was expected that especially entertaining content, UGC and the content being posted on a corporate community would lead to the desired outcomes. However, the expectations are only partly confirmed: Whereas no effects on the sharing intentions are found, it is shown that a corporate community indeed leads to higher intentions to “like” and to comment on the posted content. With regard to people’s commenting intentions, it is further demonstrated that the effects of the message’s content and source depend on the platform; only in corporate communities significant
differences considering those factors are found. Specifically, on the one hand it’s informative content and on the other hand MGC that are more likely to be commented upon in corporate communities. Also, with regard to the corporate reputation, an interaction effect is found; the emotional appeal towards the organisation is shown to be significantly higher when UGC is posted on corporate communities instead of Facebook, whereas for MGC no differences between the platforms are found. Lastly, also a significant effect of the message’s source on
its credibility is shown; MGC is found to be perceived as more credible than UGC. The specific findings and probable explanations shall be discussed in the following.
Firstly, it shall be noted that the general intentions to engage in pWOM are very low. This corresponds with earlier research that has shown that lurking activities are more common (Men & Tsai, 2013). Thus, in general people tend more to passively consume an
organisation’s communication than to actively engage in pWOM and thereby shape the corporate reputation. However, this study shows that a post’s content nor its source have an impact on the corporate reputation or people’s intentions to engage in pWOM. Even though information and entertainment have been found to be important drivers for people to use SNSs (Muntinga et al., 2011; Tsai & Men, 2012; Lin & Lu, 2011) and to engage with organisations online (Men & Tsai, 2013; Muntinga et al., 2011; de Vries et al., 2012) these results suggest that it’s not as much the content’s characteristics that sparks off pWOM. With such low intentions to engage in pWOM in general, it seems to depend more on a person’s characteristics or satisfaction with the product (McQuail, 1983; Cheung & Lee, 2012).
Moreover, not even UGC can boost that engagement either – contrary to what earlier research suggests. Therefore, UGC seems to be an important factor for a higher perceived empathy, trustworthiness and credibility (Eagly et al., 1978), but not for engagement in pWOM.
The study does show, however, that the platform a post is posted on impacts people’s intentions to ‘like’ and comment on that content. In specific, people are more likely to engage in those types of pWOM when posted in corporate communities. This is in line with the expectations: It seems that brands can profit from the higher commitment and interest in the brand that people are expected to have compared to fan pages on SNSs, such as Facebook (Jahn & Kunz, 2012). Whereas on Facebook people probably join a fan page to show one’s self-identity to their network, in corporate communities the members are truly interested in the brand and are therefore more likely to engage in pWOM. Yet, it is worth noting that this expectation is found to be true even though only a minority of the respondents was any community member. This suggests that the construction of the posts was successful and that the layout of the constructed posts was able to create a feeling and ambience of the brands.
Furthermore, with regard to the commenting intentions it can be said that the role of the content or a message’s source mostly depends on the context it’s posted in. While on Facebook, there aren’t any differences between the source nor the content types, in corporate communities there are such differences found. In corporate communities – contradicting this study’s expectations – informative content is more likely commented upon than entertaining one and therefore stimulates people more to shape the corporate reputation. It may be that higher interest in the brand that results in people being more eager to comment on
informative, rather than entertaining content. Knowing that other community members are interested in the brand too, that interest combined with the need for social interaction (Muntinga et al., 2011) might have led people to comment upon informative content. Furthermore, the results suggest that corporate communities create a setting where the
marketer isn’t perceived as a pure seller of the brand; in fact, people show higher intentions to comment on marketer-generated than on user-generated content and thereby co-shaping the corporate reputation. The theory so far suggests that UGC provokes greater empathy and trustworthiness (Eagly et al., 1978). However, due to the higher and genuine interest in the brand of corporate community members, it appears to be of higher relevance to them to hear what the marketer, thus the brand, is communicating. So, people are less motivated to engage in a conversation with other users and are rather commenting on the brand’s posts. This may be because people evaluate marketer-generated content as more credible than user-generated content – independent of the platform it is posted on. This might even indicate that people ascribe anthropomorphic characteristics to the brand and therefore have a higher emotional appeal towards the brand (Rauschnabel & Ahuvia, 2014).
However, whilst marketers can use the potential of corporate communities to get people to engage in pWOM, it is also relevant to consider which platform to choose in order to increase the brand’s reputation. The results show that it is particularly UGC on corporate communities that increases the emotional appeal towards the brand. Hence, even though UGC is not found to foster pWOM, when posted in corporate communities it leads people to have more positive feelings towards the brand. Seeing that other people also engage with the brand
seems to elicit more favourable feelings – but no higher intentions to engage in pWOM, which suggests that the link between managing pWOM and the corporate reputation isn’t that clear. Additionally, it shall be discussed that only effects on that one dimension (emotional appeal) are found and not on the whole corporate reputation. In line with this study’s
expectations, the short messages that were shown were only able to impact the respondents’ feelings towards the brands but could not influence a thorough evaluation about the brand (Petty & Cacioppo, 1986). This shows that pWOM is somewhat easier to manage than an organisation’s reputation, but due to people’s low intentions to do so it’s certainly not key for a successful reputation management. That indicates that exactly blurring lines regarding the perspective on and use of pWOM result in a weaker link between pWOM and the reputation and manifests the importance of a well-elaborated combination of the tools to manage the corporate reputation (i.e. issue management, agenda setting, framing).
Limitations. Nevertheless, this study faces some limitations that should be discussed subsequently. Firstly, the vague operationalization of content might be the cause why there aren’t any differences found regarding pWOM and reputation between the content types. In this study informative content is operationalized as posts that contain information about the brand or its products, whereas entertaining content is one that doesn’t explicitly refer to the organisation or the product and is aimed at being funnier. However, of course also brand- or related information can be funny and entertaining, whereas brand- or product-unrelated content can be informative. Even though the pre-test showed that the
operationalization was successful and the entertaining content was perceived as more entertaining than the informative content, it might be the case that that information of the informative posts was perceived as “fun facts” and therefore not as strikingly less
entertaining. Future research probably should not see informative and entertaining as two ends of one scale, but as two separate scales; the extent of informative content and
entertaining content. With such an approach, experiments could be conducted with the factors ”extent of information” and “extent of entertainment”.
Secondly, the validity of the construct reputation and the scale that is used in this study is questionable. The high number of people that couldn’t make any utterances about half of the items of the whole reputation scale makes the usefulness of the reputation scale debatable. It seems that people don’t assess an organisation’s workplace environment, social and environmental responsibility and financial performance in their daily lives. Even though the corporate reputation was assessed based on one short post, it can be reasoned that the whole reputation scale by Fombrun and colleagues (2000) is accurate for objectively assessing an organisation’s reputation (e.g. by content analysis), but doesn’t reflect the reputation constructed in people’s minds. Furthermore, it must be recognized that the temporal dimension of reputation (Gotsi & Wilson, 2001) has been neglected in this study. On one hand, it complicates the interpretation of the results; no clear statements about the causal relationship can be made. Future research could improve that by gathering longitudinal data. On the other hand, it is interesting to note that the mental representation of an
organisation can only be constructed based on information that was gathered over time, which of course wasn’t provided in this study. However, since the researched brands are in fact existing brands, people should have longitudinal information about them, hich indicates that those FMCG brands don’t communicate about their workplace environment, nor their social and environmental responsibility or their financial performance.
Thirdly, also the question about the generalizability arises. Participants’ general intention to engage in pWOM is found to be very low. This brings up the question whether the sample is representative for today’s social media users. With a moderate Internet and social media use per week and also the evaluation of their social media activity to a moderate extent, the present sample can be assumed as representative for the average social media user. It could be speculated that the low engagement in pWOM is the result of the experimental setting where people had more time to think about their actions and probably didn’t behave, as they would normally do. Specifically, the question arises why people show such low intentions to share posts. Since sharing a post doesn’t cost that much more time than
there are other motives that drive the sharing intentions or the posts weren’t exciting enough to share – more vivid and interactive posts are more likely to be shared than this study’s posts containing only a picture and text (De Vries et al., 2012). Moreover, it might also be the fact that a shared post appears on one’s Facebook wall as if it was an own post. The higher intentions to ‘like’ and comment on posts suggests that the motives of integration and social integration are stronger than to show one’s personal identity. Even though posts also appear on one’s wall by liking or commenting on it, it seems that people evaluate on that more explicitly when sharing a post, which might demand a higher commitment for a brand – which probably wasn’t the case with the investigated brands. Moreover, the focus on only one branche (i.e. FMCG, chocolate in specific) also adds to the question of generalizability. Thus, future research could improve that by conducting a field experiment and including brands of different branches.
Conclusion
In this study, pWOM is approached from a corporate perspective and it was researched how a message’s characteristics can be used to manage pWOM as well as the corporate reputation. The investigated factors are shown to partly influence pWOM behaviour, but the effects on reputation weren’t that clear, which indicates that reputation, as an overarching concept, is complex to manage. Conducting long-time studies and simultaneously acknowledging and separating the blurred perspectives, could not only contribute to a better understanding of people’s evaluation of an organisations’ reputation, but also the potential of pWOM management to build long-term relationships.
Furthermore, this study’s findings provide some guidelines for reputation managers. Firstly, it got apparent that organisations can benefit from the higher commitment to the brand in corporate communities compared to Facebook fan pages. It seems that on Facebook people join those fan pages to profile their personal identity and don’t intend to engage in pWOM or to engage with the brand. Secondly, this study shows that when choosing a corporate
that informative content is more likely to be commented upon and also that pWOM is triggered when the content is posted by the marketers and not by community members. Thirdly, the results suggest that it is also important to motivate users to get active themselves on corporate communities, since this increases people’s emotional appeal towards the brand.
In conclusion however, the research question can only be answered with certain reserves. The message’s content doesn’t influence pWOM and the corporate reputation as much as expected and largely depends on the platform it is posted on. It shows that there is no black or white regarding which content to post where and by whom. To build up and
strengthen an organisation’s reputation, it might be advisable to be present on SNS’s as well as corporate platform – bearing each platform’s advantages in mind: SNSs might be used to increase the brand awareness and maintain the reputation. Corporate communities, in turn, might serve as the platform to engage with the true fans of the brand.
References
Abbratt, R. & Klein, N. (2012). Corporate Identity, corporate branding and corporate reputations. European Journal of Marketing, 46, 7/8, 1048 – 1063.
Algesheimer, R., Borle, S., Dholakia, U. M., & Singh, S. S. (2010). The impact of customer community participation on customer behaviors: An empirical investigation. Marketing Science, 29(4), 756-769.
Argenti, P. A., & Druckenmiller, B. (2004). Reputation and the corporate brand. Corporate Reputation Review, 6(4), 368-374.
Awad, N. F., & Ragowsky, A. (2008). Establishing trust in electronic commerce through online word of mouth: An examination across genders. Journal of Management Information Systems, 24(4), 101-121.
Bickart, B., & Schindler, R. M. (2001). Internet forums as influential sources of consumer information. Journal of Interactive Marketing, 15(3), 31-40.
Boyd, D. M. & Ellison, N. B. (2007). Social network sites: Definition, history, and scholarship. Journal of Computer-Mediated Communication, 13(1), 210-230. Brønn, P. S. (2010). Reputation, communication, and the corporate brand. In Heath, R.L.
(Ed.), Handbook of Public Relations (pp. 307-20). London: Sage.
Brown, J., Broderick, A. J., & Lee, N. (2007). Word of mouth communication within online communities: Conceptualizing the online social network. Journal of Interactive Marketing, 21(3), 2-20.
Bryman, A. (2012). Social research methods (4th ed.). Oxford, NY: Oxford University Press. Buttle, F. A. (1998). Word of mouth: Understanding and managing referral marketing.
Journal of Strategic Marketing, 6(3), 241–254.
Cheung, C. M., & Lee, M. K. (2012). What drives consumers to spread electronic word of mouth in online consumer-opinion platforms. Decision Support Systems, 53(1), 218-225.
Chevalier, J. A., & Mayzlin, D. (2006). The effect of word of mouth on sales: Online book reviews. Journal of marketing research, 43(3), 345-354.
Cho, S., Huh, J., & Faber, R. J. (2014). The influence of sender trust and advertiser trust on multistage effects of viral advertising, Journal of Advertising, 43(1), 100-114.
Coombs, W. T. (2007). Protecting organization reputations during a crisis: The development and application of situational crisis communication theory. Corporate Reputation Review, 10(3), 163-176.
Cvijikj, I.P. & Michahelles, F. (2013). Online engagement factors on Facebook brand pages. Social Network Analysis and Mining, 3(4), 843-861.
Dekay, S. (2012). How large companies react to negative Facebook comments. Corporate Communications: An International Journal, 17(3), 289-299.
De Matos, C. A., & Vargos Rossi, C. A. (2008). Word-of-mouth communications in marketing: a meta-analytic review of the antecedents and moderators. Journal of the Academy of Marketing Science, 36(4), 578-596.
De Vries, L., Gensler, S., & Leeflang, P. S. H. (2012). Popularity of brand posts on brand fan pages: An investigation of the effects of social media marketing. Journal of Interactive Marketing, 26(2), 83-91.
Dijkmans, C., Kerkhof, P., & Beukeboom, C. J. (2015). A stage to engage: Social media use and corporate reputation. Tourism Management, 47, 58-67.
Eagly, A. H., Wood, W., & Chaiken, S. (1978). Causal inferences about communicators and their effect on opinion change. Journal of Personality and Social Psychology, 36(4), 424.
East, R., Hammond, K., & Wright, M. (2007). The relative incidence of positive and negative word of mouth: A multi-category study. International Journal of Research in
Marketing, 24(2), 175-184.
Engel, J. E., Blackwell, R. D., & Kegerreis, R. J. (1969). How information is used to adopt an innovation. Journal of Advertising Research, 9(4), 3-8.
Enginkaya, E., & Yılmaz, H. (2014). What Drives Consumers to Interact with Brands through Social Media? A Motivation Scale Development Study. Procedia-Social and
Behavioral Sciences, 148, 219-226.
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(4), 241-255.
Fombrun, C. J., & van Riel, C. B. M. (1997). The reputational landscape. Corporate Reputation Review, 1(1/2), 5-13.
Fombrun, C. J., & van Riel, C. B. M. (2004). Fame & fortune: How successful companies build winning reputations. Upper Saddle River, NJ: Prentice Hall.
Fournier, S., & Avery, J. (2011). The uninvited brand. Business Horizons, 54(3), 193-207. Gauri, D. K., Bhatnagar, A., & Rao, R. (2008). Role of word of mouth in online store
loyalty. Communications of the ACM, 51(3), 89-91.
Godes, D., & Mayzlin, D. (2004). Using online conversations to study word of mouth communications. Marketing Science, 23(4): 545–560.
Goh, K-Y., Heng, C-S., & Lin, Z. (2013). Social Media Brand Community and Consumer Behavior: Quantifying the Relative Impact of User- and Marketer-Generated Content. Information Systems Research, 24(1), 88-107.
Gotsi, M., & Wilson, A. (2001). Corporate reputation: seeking a definition. Corporate Communications, 6(1), 24-30.
Gruen, T. W., Osmonbekov, T., & Czaplewski, A. J. (2006). eWOM: The impact of customer-to-customer online know-how exchange on customer value and loyalty. Journal of Business Research, 59(4), 449-456.
Grunig, J. E. (2009). Paradigms of global public relations in an age of digitalisation. PRism, 6(2), 1-19.
Grunig, J. E., & Grunig, L. A. (1992). Models of public relations and communication. In J. E. Grunig (Ed.), Excellence in public relations and communication management (pp. 285-325). Hillsdale, NJ: LEA.
Hall, R. (1992). The strategic analysis of intangible resources. Strategic Management Journal, 13(2), 135-144.
Hass, R.G. (1981). Effects of Source Characteristics on Cognitive Responses and Persuasion. In R.E. Petty, T.M. Ostrom, & T.C. Brock (Eds.), Cognitive Re- sponses in Persuasion (pp. 141–172). Hillsdale, NJ: Lawrence Erlbaum Associates.
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(1), 38-52.
Hennig-Thurau, T., Malthouse, E. C., Friege, C., Gensler, S., Lobschat, L., Rangaswamy, A., & Skiera, B. (2010). The impact of new media on customer relationships. Journal of Service Research, 13(3), 311-330.
Jahn, B. & Kunz, W. (2012). How to transform consumers into fans of your brand. Journal of Service Management, 23(3), 344-361.
Kaplan, A. M., & Haenlein, M. (2012). The Britney Spears universe: Social media and viral marketing at its best. Business Horizons, 55(1), 27-31.
Katz, E., Gurevith, M., & Haas, H. (1973). On the use of the mass media for important things. American Sociological Review, 38(2), 164–181.
Katz, E. & Lazarsfeld, P. F. (1955). Personal influence. New York, NY: Free Press.
Kimmel, A. J., & Kitchen, P. J. (2014). WOM and social media: Presaging future directions for research and practice. Journal of Marketing Communications, 20(1-2), 5-20. King, R. A., Racherla, P., & Bush, V. (2014). What we know and don’t know about online
word-of-mouth: A review and synthesis of the literature. Journal of Interactive Marketing, 28(3), 167-183.
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
Behavior, 27(3), 1152-1161.
MacKenzie, S. B., & Lutz, R. J. (1989). An empirical examination of the structural
antecedents of attitude toward the ad in an advertising pretesting context. The Journal of Marketing, 53(2), 48-65.
Macnamara, J. (2010). ‘Emergent’ media and public communication: understanding the changing mediascape. Public Communication Review, 1(2), 3-17.
McQuail, D. (1983). Mass Communication Theory. London: Sage Publications.
Men, L. R., & Tsai, W. H. S. (2013). Beyond liking or following: Understanding public engagement on social networking sites in China. Public Relations Review, 39(1), 13-22. Meuter, M. L., McCabe, D. B., & Curran, J. M. (2013). Electronic Word-of-Mouth Versus
Interpersonal Word-of-Mouth: Are All Forms of Word-of-Mouth Equally Influential?. Services Marketing Quarterly, 34(3), 240-256.
Muntinga, D. G., Moorman, M., & Smit, E. G. (2011). Introducing COBRAs. Exploring motivations for brand-related social media use. International Journal of Advertising, 30(1), 13-46.
Nambisan, S., & Baron, R. A. (2007). Interactions in virtual customer environments: Implications for product support and customer relationship management. Journal of Interactive Marketing, 21(2), 42-62.
