Relationship building on social media : about the empowerment of users and companies’ control attempts

(1)

Relationship building on Social Media -

About the empowerment of users and companies’ control attempts

Marina Adam Student ID: 10968296

Master Thesis

Graduate School of Communication Master’s programme Communication Science

Supervisor: James Slevin Date of completion: 29-01-2016

(2)

Abstract

Social Media delivers high potential for Marketing and Public Relations in terms of

relationship building with customers but also bears reputational risks with regard to negative word-of-mouth. The power over the content is shifted towards the users and communication professionals fear to lose control over their brand messages. They are caught in a dilemma because personal relationships to the customers and a high level of engagement are the goals, but the insecurity about the empowerment of the users affects their communication. The present study investigates the effects of user power and control on behalf of the companies on the personalization of the conversation between companies and their followers and on

engagement. Additionally, it compares those relationships for two types of company structures, namely traditional companies and online-only brands. This paper delivers new guidelines for Corporate Communication professionals that want to make the relationship to their users more personal. Furthermore, it shows users how to enact their power in order to establish a personal conversation with the companies. The theoretical framework is grounded in the understanding of Corporate Communication in the sense of a constructing, constant and reciprocal development process. The study is methodologically based on a quantitative

content analysis including user-generated-content as well as marketer-generated-content. In the case users are empowered on Social Media and companies dare to lose control, their conversation can become more personal. A higher engagement in this context could not be proven and a moderation effect for the company structure was only evident for the effect of power on personalization on Facebook. So far, research on Social Media mostly focused on one specific platform. This study investigates three Social Media platforms simultaneously, namely Facebook, Twitter and Instagram. Furthermore, it explores the often claimed power shift towards the users as well as the executed control by professionals and their effects on personalization and engagement which, so far, has not been empirically tested in a

(3)

Relationship building on Social Media -

About the empowerment of users and companies’ control attempts Social Media is in the centre of attention of Marketing and PR today. As people altered their media use and spend increasingly time on Social Media, companies realized that the users communicate about them there, regardless of their presence (Kent, 2010). They entered the platforms in order to leverage the advantages of the direct access to their customers and to manage the conversation about their brands. This immediate interaction brings a company closer to its users, as there are no journalists necessary to transmit their message. Therefore, professionals use the platforms to strengthen their company’s brands by building relationships with the online users, aiming to increase their reputation and sales. Social Media engagement enhances the purchase behaviours of consumers (Goh, Heng, & Lin, 2013) and was shown to have a direct effect on customer loyalty and reputation (Brodie, Ilic, Juric, & Hollebeck, 2013; Cvijikj & Michahelles, 2013; Dijkmans, Kerkhof, &

Beukeboom, 2015). In order to encourage engagement, it is very important that a company is able to build a strong relationship to their customers by creating a personal interaction level (Kerkhof, Vonkeman, Beukeboom & Utz, 2011), as the willingness to engage depends on the emotional connection to a brand (Sashi, 2012).

What changes on Social Media with regard to traditional communication media for Marketing and Public Relations is the power shift towards the users, in particular the impact that users have by contributing to the development of a brand and stepping into direct interaction with it (Berthon, Pitt, Plangger, & Shapiro, 2012; Labrecque, vor dem Esche, Mathwick, Novak, & Hofacker, 2013). While companies plan their communication strategies, conversations also emerge unintendedly in the context of the interaction (Macnamara & Zerfass, 2012). Positive interaction can have benefits for the companies in terms of

recommendations via word-of-mouth but negative interaction can harm them substantially (Killian & McManus, 2015). Therefore, a lot of practitioners fear the loss of control over their

(4)

messages and do not dare to use Social Media to their full potential (Buchanan-Oliver & Fitzgerald, 2014). Especially traditional companies that build up a strong reputation, might still try to protect themselves and limit their transparency and openness towards their

followers. On the contrary, companies that are founded after the rise of Social Media, and that operate exclusively in e-commerce, are more likely to appreciate the possibilities Social Media offer them to get awareness for their business and interact with potential clients. In order to evaluate the effects that those developments have on the relationship building processes between companies and the online users, to investigate if the users indeed are empowered and the fear of the professionals of losing control is justified, those effects will be empirically tested and guidelines for the optimal communication strategy will be derived in this context. A content analysis will explore the effect of the power shift towards the users and the attempt to hold the control by the companies on the communication between the two actors on Social Media. The comparison will be made between traditional companies with physical stores and online-only brands, exclusively active online, by addressing the following research questions: How do user-power and brand-control influence the

relationship between companies and their stakeholders with regard to the personalization of the conversation and the level of engagement? Which impact has the company structure of traditional companies and online-only brands on those effects?

Theoretical framework

In order to set the stage for the following analysis, existing research on Corporate Communication within the context of Social Media will be discussed. Afterwards, a definition of Social Media will be given and the three platforms Twitter, Facebook and Instagram will be presented. Further, the use of the technologies will be explained from a marketer

perspective and the perspective of the online users, concluding in characteristics of their relationship building on Social Media. The power shift towards the user and the intent to control by the company will be discussed. To wrap up the theory section, the different

(5)

company types that are subjects to the study, namely online-only brands and traditional companies, are introduced.

Corporate Communication on Social Media

In current scientific literature, Corporate Communication on Social Media is

oftentimes understood as ideal with regard to relationship building and increasing sales, when it follows the strategy of an integrated communication (Killian & McManus, 2015). Hereby, the organisation’s aim is to communicate internally and externally with one voice and a clear message. In this approach, the company tries to submit an image to their stakeholders and protect the reputation from external influences. Social Media in this sense would be one additional tool that needs to be integrated in the communication strategy of the traditional tools for Marketing and PR (Cvijikj & Michahelles, 2013; Mangold & Faulds, 2009; Van Noort, Willemsen, Kerkhof, & Verhoeven, 2014). In earlier years, the perspective of efficient stakeholder management was prevalent, suggesting that managers should recognize varying stakeholder interests and respond to them in a cooperative way in order to legitimize the existence of a company (Donaldson & Preston, 1995). Even if former research acknowledged the potential of a strong relationship to the stakeholders, the focus was oftentimes on

leveraging effects for the company, downsizing the meaning of the individuals and still putting the company and its message in a clear focus of an asymmetrical sender-receiver relationship (Berger & Milkman, 2012; Liu-Thompkins, 2012). Although a mutual

satisfaction was the goal, this view is very organisation centred and neglects the power of the customers as it is nowadays present on Social Media.

Christensen & Cornelissen (2011) argued that this approach is outdated and a new perspective on Corporate Communication needs to be taken. A company is no longer a transmitter of communication but is subject of construction within this context.

Communication constitutes organisations in terms of constant adaption and reinvention of communication patterns in a dynamic interaction with the external environment and internal

(6)

members (Giddens, 1979). For this definition of Corporate Communication, the constructive effect of language on meaning creation is essential, formed in the social and institutional settings but respecting individual interpretations (Cornelissen, Christensen, & Kinuthia, 2012). Consequently, issues form the centre of communication nowadays, not the organisations (Luoma-aho & Vos, 2010). Not the balance between external and internal perception of the identity of a company is important but its ongoing development. The

characteristics of Social Media enable interconnected communication processes through their network structure, and their dynamic of action and reaction can result in collective sense making (Christensen & Cornelissen, 2011). The view of an organisation as holistic construct, that is able to control the total amount of messages, is not compatible with Social Media, where monitoring can help to observe the communication about the company, but control is lacking (Hennig-Thurau et al., 2010). If this shift of control is perceived as a share of control with the users online instead of a loss of control, it opens up for a transparent dialogue (Dekay, 2012). The use of Web 2.0 technologies can diminish control for marketers in the sense of ruling over the message, but enhance their opportunities in the sense of relationship-building, as they permit two-way conversations (Fournier & Avery, 2011; Hennig-Thurau et al., 2010; Luoma-aho & Vos, 2010; Mills & Plangger, 2015). Communication on Social Media is fast and can follow diverse paths which enables more personal interactions with the individual stakeholders (Christensen & Cornelissen, 2011).

Social Media in the context of the study

Generally speaking, Social Media are internet-based applications that make use of the interactive possibilities that are offered by the technology of the Web 2.0 (Macnamara & Zerfass, 2012). Social Media differ from traditional media in their interconnected social network structure with opportunities for dynamic interaction. Their characteristics enable a democratization of communication, as every individual can participate and share an opinion (Berthon et al., 2012); “a social medium is, by definition, multi-way, immediate, and

(7)

contingent” (Peters, Chen, Kaplan, Ognibeni, & Pauwels, 2013, p. 282). The platforms provide a wide range of tools for interactions between brands and their followers, offering the opportunity to share different types of content and implying a two-way-communication approach (Constantinides & Fountain, 2008). Openness for participation and interactivity have an effect on the Corporate Communication activities that are shifted to a dialogue or multi-player conversation instead of a controlled one-way communication process (Jahn & Kunz, 2012; Macnamara & Zerfass, 2012). Social Media allow for real-time interaction, supporting a more active conversation (Kent, 2010). On the contrary, they also enable an asynchronous communication (Liu & Shrum, 2002), whereby companies and users do not need to react immediately, as for example in a chat, but can solve a problem in advance to formulating an appropriate answer. However, especially in webcare, certain time frames should not be exceeded, to avoid disappointment of the other partner in the conversation who is waiting for a response (Van Noort & Willemsen, 2012). Conversations on Social Media are more personal, as anonymity is reduced, revealing habits and character traits of the customer and providing the company with a human voice (Kelleher, 2009). Social Media are seen as powerful tools for organizations in terms of PR and Marketing with respect to electronic word-of-mouth. Interaction on Social Media can satisfy customer needs, improve the relationship between users and a brand and thereby influence the conversion rate (Ahuja & Medury, 2010). On the downside, brand managers face the risk of losing control over the messages they want to transmit on Social Media (Bonsón & Ratkai, 2013; Hennig-Thurau et al., 2010; Peters et al., 2013).

Most studies on Social Media focus on one platform in their analysis but a comparison of several platforms can show differences and similarities between them and is able to help practitioners to choose their strategy accordingly to their business (Smith, Fischer, & Yongjian, 2012). In this study, three different types of Social Media will be investigated: Microblogs, Social Networks and Content Communities, represented by Twitter, Facebook

(8)

and Instagram. The different Social Media platforms offer a varying potential for content sharing and interaction. Consequently, companies use them differently with respect to their own structure and communication strategy, while the users are influenced by their needs and motivations (Java, Song, Finin, & Tseng, 2007; Muntinga, Moorman, & Smit, 2011; Peters et al., 2013).

Twitter. Microblogs enable their users to deliver short text messages in real-time (Jansen, Zhang, Sobel, & Chowdury, 2009). Twitter is currently the most popular Microblog that was launched in 2006 and has around 304 million monthly active users (Q2/2015) (Statista, 2015a). Its immediate short messages are limited to 140 signs, providing space for fast interaction. Twitter fulfils several functions for companies that can be grouped into information, community and action (Lovejoy & Saxton, 2012). Information means simple transmission of a message about facts and news of the company. Community implies dialogic conversations between a brand and its followers (positive or negative) as well as relationship building functions such as positive comments or praise. Action tweets aim to activate

followers in favour of the company by encouraging purchases, donations or advocacy

behaviours (Lovejoy & Saxton, 2012). Former research found that Twitter is often applied in a one-way-communication manner by companies (Waters & Jamal, 2011; Xifra & Grau, 2010) and is used for complaints by the followers (Peters et al., 2013).

Facebook. The most popular social network worldwide has 1.49 billion of active users per month (Q2/2015) (Statista, 2015b). Initially, Facebook was an online platform exclusively for persons, aiming to connect them to their acquaintances and friends. In the beginning phase of Facebook, fan pages for brands were created by users, without any direct influence of the brand but attracting high numbers of followers and exposing the willingness of users to interact with brands in their personal space (e.g.: Nutella, Coca-Cola) (Fournier & Avery, 2011). Companies got aware of that matter and took over the control of those pages or opened up their own brand pages. They had to be aware of the preconditions of the platform and

(9)

adapt to the personalized space by taking a human approach. By acting consciously, they can avoid being perceived as intrusive with the consequence of being ignored or even attacked (Fournier & Avery, 2011; Jahn & Kunz, 2012). On Facebook, engagement can take place as a reaction to marketer-generated-posts in the shape of comments, likes and shares or in a self-initiated way by posting user-generated-content on the wall of the company (Cvijikj & Michahelles, 2013) or writing a private message. Bortree & Seltzer (2009) investigated the volume of content provided by the company and the users and the social network of the company’s profile on Facebook. They found that a lot of companies do not use dialogic strategies on their profiles and therefore fail to use the potential of the medium properly.

Instagram. Content communities focus on sharing specific types of content (Kaplan & Haenlein, 2010) such as photos and short videos on Instagram. The platform has reached a number of more than 400 million users in 2015 (Instagram Blog, 2015). The main motivation of users for being active on Instagram is entertainment (Safko, 2012). A dialogue can take place in the form of shared hashtags from a brand and their users, through comments to the posted content or links to other user profiles. Despite its popularity, Instagram was scarcely subject to Web and Social Media research (Manikonda, Hu, & Kambhampati, 2014). While Facebook and Twitter are suitable for critique forwarded to the company, photo sharing platforms such as Instagram are mostly positive in the tonality of the posts (Smith et al., 2012).

Marketer-Generated-Content (MGC)

MGC is either initiated by the company or it is a response to user-generated-content and can be used for Marketing and PR purposes (Goh et al., 2013). As valuable content is one of the most important factors to attract fans to brand pages on Social Media (Jahn & Kunz, 2012), the focus is not only on the addressing of the right stakeholder groups, but more on finding the right issues and topics to engage with (Luoma-aho & Vos, 2010). Topics can

(10)

range from pure information over products and services to more general or entertaining content without a direct connection to the core business (Ahuja & Medury, 2010).

The gratifications users search on Social Media are information, entertainment and social connection (Heinonen, 2011; Muntinga et al., 2011). Thus, brands should deliver valuable content that satisfies those needs (Jahn & Kunz, 2012). Former research showed that communication that was initiated by a company was most effective when it was perceived as entertaining without obvious Marketing focus (Dekay, 2012). MGC as response to user postings contributes to reputation and relationship management. (Van Noort et al., 2014). While several studies explored webcare in respect to negative word-of-mouth, scarce attention has been dedicated to responses on positive customer feedback (Demmers, Van Dolen, & Weltevreden, 2013).

Control. The definition of control is multifaceted and embraces different concepts depending on the context. In general, control can be defined as „purposive influence toward a predetermined goal“ (Beniger, 1986, p. 7). Building on this definition, influence and purpose on Social Media can be derived. Companies attempt to influence their followers by the content they provide on the platforms. Accordingly, their purpose is to build a positive relationship with their customers in order to increase the sales and improve the reputation. Control can vary from an absolute level to a slight influence on an actor (Beniger, 1986), mirrored by the different levels of control in this analysis. The fear of losing control over their content by communication professionals has been evident in a study by Buchanan-Oliver and Fitzgerald (2014) but the actual dealing with control on Social Media and its effects has, so far, not been operationalized within empirical research on Social Media.

Companies can try to execute control by moderating the Social Media platforms whereby more moderation results in less openness and social exchange (Kent, 2010). Reasons for the resistance to an open Social Media approach are the feeling of being unprepared for the challenges and a lack of understanding of the optimal Social Media use. This might also

(11)

be due to a lack of established metrics, making the use of Social Media intuitive and leading to insecurity (Buchanan-Oliver & Fitzgerald, 2014). Social media platforms offer varying opportunities for engagement and its surveillance. Every platform that is subject to this study enables companies to follow back to their users’ profiles in order to gather information about them and observe their influential activities (Lovejoy, Waters, & Saxton, 2012). With concern to the content, a sender-focused one-way approach and a topic focus on information about the companies’ core business indicates their attempt for control. Additionally, Facebook enables companies to block the opportunity to review their company, to post on their wall and to use filters to exclude offending posts. As negative comments by users are oftentimes seen as a direct threat to the companies’ reputation, they can be ignored or censored (Dekay, 2012; Killian & McManus, 2015).

User-Generated-Content (UGC)

The most important innovation brought by Social Media to Corporate Communication is the possibility of the contribution of UGC (Berthon et al, 2012; Heinonen, 2011). UGC defines all content that is produced by non-commercial users on Social Media, that is published openly on a networking platform and can be considered creative in any way (Organisation for Economic Cooperation and Development (OECD), 2007). Users deal with UGC in three ways, similar to their interaction with MGC: First, they consume it in order to satisfy their needs for information and entertainment. Second, they participate by interacting with their social connections and third, they produce content for expression and self-actualization (Shao, 2009). Concerning the creation of UGC, users engage in varying degrees in dependence on their motivations. Empowerment is one motivation, where people act out of the conscious will to influence other followers (Muntinga et al., 2011). UGC has the potential of network spreading but exceeds the control of online marketers and shifts the power towards the consumer (Constantinides & Fountain, 2008; Labrecque et al., 2013). However, it can still be used for marketing purposes by companies, as it gives insights into the external perception

(12)

of the organisation that can be used for improvements of products and services (Van Noort et al., 2014).

Power. While a shift of power towards the user on Social Media is addressed in various research papers (Cova & Pace, 2006; Labrecque et al., 2013; Van Noort &

Willemsen, 2012), an operationalization of the concept for empirical research is missing so far. Social power can be defined “as asymmetric control over valued resources in social relations” (Magee & Galinsky, 2008). Further, Magee and Galinsky (2008) argued that those resources can have a positive or negative value which includes rewards or punishments for others. In the context of the study this means that users have the asymmetric control over their generated content, which can be positive or negative and therefore either help the company to improve their reputation and sales or harm the company through criticism and negative word-of-mouth. Handgraaf, Van Dijk, Vermunt, Wilke, & De Dreu (2008) stated that it is the mutual dependence of the actors that enables power in the first place.

On Social Media, there is no control over the message distribution, MGC as well as UGC can spread virally. While one might argue that the control over the message was an illusion earlier with traditional media as well because of the involvement of journalists as mediators, the degree of influence shifted towards the users increased strongly with the rise of Social Media, as content can now be evaluated, altered and commented by the public

immediately. Organizations and users can play a gatekeeping role by deciding if they want to interact with each other and to what extent they want to open-up and be transparent (Castells, 2011). For instance, users opt for notifications on Facebook via likes and an organisation can determine if they open up their wall for user comments. Furthermore, power can be executed from the participants of the network in terms of an agenda-setting function (Castells, 2011), for instance by users who set content links to other users’ profiles. Consumers get the power to influence brands by contributing content, distributing it and thereby shaping brand

(13)

also against companies. By posting negative comments, they challenge a company’s

reputation. Further, they can demand for webcare, as they post in a two-way-communication approach, asking for direct reaction on behalf of the company. In the perspective of a

marketer, UGC can help to reach Marketing goals through virality, generate new products through coproduction and identify areas for improvement through discussions (Berthon et al., 2012; Cvijikj & Michahelles, 2013; Sawhney, Verona, & Prandelli, 2005). Brands that are influenced by UGC can be expected to better fit the preferences of consumers (Leeflang, Verhoef, Dahlström, & Freundt, 2014). While this empowerment of the users seems to claim an asymmetric strength on the user’s side, it is not that simple. First, a lot of users only use Social Media passively and do not exploit their influential potential. Second, companies are empowered as well by Social Media, as they can leverage the user generated data for

marketing activities and learn from it for their Public Relations efforts. Third, the knowledge over their core business and responsibility as a corporate citizen is centred in the company and cannot be given away to the users. Compared to the traditional media, the power is shifted towards the user but on Social Media in general, the power lies within the communal interaction between a company and the users (Mills & Plangger, 2015). However, a power shift towards the users through UGC is evident and will be addressed hereafter.

Relationship between companies and users

In the last decade, Social Media changed the way of Corporate Communication. In earlier years, the interaction between companies and their customers was oftentimes mediated by the work of journalists and communication agencies. On Social Media, those gatekeeping functions are no longer evident and the relationship between companies and customers gets more direct. This development brings opportunities for the companies in terms of relationship building but requests a more thoughtful communication, as the interaction with the consumers here is immediate (Mills & Plangger, 2015; Van Noort et al., 2014). Furthermore, users produce their own content over brands which is publically visible and can threaten the

(14)

reputation of the companies. Those circumstances influence the way of relationship building between a company and its users.

The social interaction between individuals from inside the organisations with the external stakeholders on a micro level constitutes the organisational communication on a macro level (Christensen & Cornelissen, 2011). In this understanding, a company cannot be viewed as one single actor but needs to consider the individuals that comprise the company and communicate with the stakeholders under a corporate account. A personal communication approach is important as it can increase the credibility of a company and purchase intentions of the users (Kerkhof et al., 2011). Besides the positive promotional effects of word-of-mouth, if a company succeeds in establishing strong relationships to their customers it can count on them as advocates in difficult times (Dekay, 2012) and can encourage their engagement (Tsimonis & Dimitriadis, 2014).

Personalization. Social Media have the ability to make the interaction between

companies and online users more personal (Berthon et al., 2012; Goh et al., 2013; Tsimonis & Dimitriadis, 2014). Stakeholders tend to interpret the identity of a company in the metaphor of character traits of a person (Ahuja & Medury, 2010; Christensen & Cornelissen, 2011).

Comparable to human relationships, a regular interaction is needed in order to build a bond between an organization and its customers. Consequently, authenticity and transparency are important to stimulate a feeling of propinquity (Ahuja & Medury, 2010). In order to establish a strong relationship, conversational human voice and communicated relational commitment were found to be important contributors that improved trust, satisfaction, control mutuality and commitment (Berthon et al., 2012; Kelleher & Miller, 2006; Kerkhof et al., 2011; Van Noort et al., 2014). A personal message can be indicated by direct addressing of the counterpart, self-indication, informal language and invitational rhetoric (Van Noort et al., 2014), as well as sense of humour (Kelleher, 2009).

(15)

Engagement. Due to the complexity and multidimensionality of the concept, no clear exclusive definition of engagement exists (Brodie et al., 2013). Former research identified fan page engagement as important contribution for the relationship between a brand and the user (Brodie et al., 2013; Jahn & Kunz, 2012), emphasizing the need of using brand pages actively and creating a community. An investigation of the factors making content spread virally, pointed out that practical useful and surprising content has a higher probability to be shared. The same goes for valence, as positive content had higher viral potential than negative content (Berger & Milkman, 2012; De Vries, Gensler, & Leeflang, 2012). Different content types bear more or less potential for virality (Peters et al., 2013). Ahuja & Medury (2010) found that posts with relation building motives stimulated user engagement in a blog environment. A simple understanding of engagement on Social Media comprises the numerical outcomes of Likes, Shares/Retweets and Comments in relation to the total amount of fans of the brand page (Cvijikj & Michahelles, 2013).

Traditional companies vs. Online-only brands

The Theory of Structuration states that human actions are determined by social structures but also influence them in a reciprocal process (Giddens, 1979). Transferred to Social Media professionals, their interaction with the users will be influenced by the structure of their company but can also be changed throughout the communication process. On the one hand, differently structured organizations will use different communication approaches and they will be influenced by their interactions to varying degrees (Buchanan-Oliver &

Fitzgerald, 2014). On the other hand, Social Media enables consumers to influence the development of a brand and their input can change the communication structure of a company. Former studies suggested that a company’s size, the industry, its structure and culture are important factors when it comes to Social Media implementation (Baxter & Connolly, 2014). However, the link between an organization’s mission and its Social Media use requires further investigation (Lovejoy & Saxton, 2012). A comparison between

(16)

traditional companies and online-only brands is interesting, as they enter the Social Media conversations with different predispositions in their structure and their communication process might develop in distinctive dimensions. Traditional companies as defined in this analysis are multinational retailers, selling their products primarily offline in physical stores. They use their online activities as secondary sale channel and for brand building purposes. Traditional companies oftentimes use Social Media in the old understanding of Corporate Communication, trying to transmit their message to their stakeholders in a one-way

communication approach and neglecting the importance of interaction with the public for the development of their company (Grunig, 2009). Even if practitioners acknowledge the value of Social Media for two-way-communication processes, they fear the loss of control because of a lack of understanding or experience in this field (Buchanan-Oliver & Fitzgerald, 2014;

Fournier & Avery, 2011; Tsimonis & Dimitriadis, 2014). Online-only brands as defined in this study, entered the market with Social Media as given communication media and are retail companies that exclusively sell their products and services online (Li, Lu, & Talebian, 2014). They depend strongly on their online communication and can be expected to use Social Media in a more natural manner. Those companies will be more active as their core-business is on the web and their customers online are closer linked to Social Media (Kaplan & Haenlein, 2010; Kent, 2010).

Building on the theoretical framework presented above, the following hypothesis will be investigated:

Twitter

H1a: There is a positive effect of power on personalization.

H1b: The positive effect of power on personalization will be stronger on pages of traditional brands.

(17)

H1d: The negative effect of control on personalization will be stronger on pages of online-only brands.

H1e: There is a positive effect of power on engagement.

H1f: The positive effect of power on engagement will be stronger on pages of traditional brands.

H1g: There is a negative effect of control on engagement.

H1h: The negative effect of control on engagement will be stronger on pages of online-brands.

Facebook

H2c: There is a negative effect of control on personalization.

H2e There is a positive effect of power on engagement.

Instagram

(18)

H3e: There is a positive effect of power on engagement.

Figure 1. Model for the influence of power and control on the relationship between a

company and the Social Media users with a moderation effect of company structure Method section

In order to answer the research questions and to test the hypotheses stated above, a quantitative content analysis is considered the appropriate method because it reveals differences in the textual communication and makes them measurable and consequently comparable.

Sample

For this study, the Social Media presence of six retail companies was analyzed, three traditional companies, namely Jack & Jones, Manhattan Cosmetics and Wilkinson Sword as well as three online-only brands: Trashness, Stowaway and Dollar Shave Club. They pairwise

MGC UGC Relationship brand-user Personalization Engagement power + - control Company structure

(19)

belonged to the same sector: two cosmetic companies (Manhattan cosmetics & Stowaway), two clothing companies (Jack & Jones & Trashness) and two hygienic product companies (Wilkinson Sword & Dollar Shave Club). The sample consisted of content derived from Facebook, Twitter and Instagram. In order to make sure that the official brand pages were analyzed, the link to the platforms were identified over the corporate website of the

companies and via cross links from indicated platforms there. The sample units were posts dating backwards from the 30th of November 2015. For every Social Media platform, the units of data collection were 20 posts generated by each company (MGC) and 20 posts generated by its followers (UGC). The coding units were extracted manually and coded simultaneously in order to ensure a proper interpretation, corresponding to the context of the post. In case of identical posts by the users, for instance due to a sweepstake, the posts were skipped and the following unique post was coded. The total amount of coded data units for all companies taken together was 720 posts (360 UGC/360 MGC).

Codebook

As three Social Media platforms were analyzed, the Codebook was divided in three parts with respective adaptions for Facebook, Twitter and Instagram. The complete Codebook can be found in the Appendix.

Procedure

In order to test the codebook, 10 MGC posts and 10 UGC posts were randomly selected for each company structure and for every platform (Twitter, Facebook, Instagram), resulting in 60 data units. The codebook was adapted and finalized afterwards.1

Categories and Measures

In the following, the main variables for the analysis, that are operationalized in the codebook, are described and further discussed.

(20)

Follower base on brand pages. With respect to the network structure, the follower bases of the different Social Media platforms were measured by the number of fans for the specific brand pages (Peters et al., 2013). This variable helped to set the engagement in relation to the amount of fans, avoiding a biased perception of it (Cvijikj & Michahelles, 2013).

MGC. Marketer-generated-content are posts that are generated by a company and publicly shared on their Social Media account. This can be a Facebook post, a Tweet or a post on Instagram.

UGC. User-generated-content can be a Facebook post directed to the company on its wall, a Tweet that is directed to the company’s Twitter account through @companyname or a picture/video on Instagram linked to the company via #companyname.

Control. This concept is referring to content provided by the marketer. In the context of this study it is comprised of five indicators for Facebook and three indicators for Twitter and Instagram. First, the following back behaviour shows if a company tries to keep the control over its content by observing more carefully who their followers are and how they interact. Furthermore, the following back function on Twitter enables a company to send private messages. Second, the one-way direction of the message reveals the intent of the company to transmit its message without asking for any feedback, neglecting the

opportunities of a dialogue. Third, the topic of the post indicates the control of the company as product related topics are more controllable than entertaining topics, exceeding their direct expertise and encouraging more creative user responses. The additional dimensions for control on Facebook consist of the functions of opening up for reviews by users and opening up for comments on the wall, as they encourage an opinion sharing of their users that exceeds their influence. All categories were coded on a binary level and then computed into one variable for control, reaching from 0 indicating no control intents to 3 (from 0 to 5 for Facebook) for a high executed control.

(21)

Power. The concept of power relates to content characteristics of UGC and consists of three factors. First, the negativity of the post evaluates the content tonality and expresses how much pressure the users put on the company. Second, if content is linked to user profiles the power over the content is shifted to the followers and exceeds the control of the marketers. Third, a two-way-communication approach indicates a power shift towards the users, as it demands a direct and immediate reaction of the company. The three categories were coded as binary variables and afterwards computed into one continuous variable with 0 indicating no power reaching to 3 meaning a high amount of user power.

Relationship company-user

Personalization. The factors of a personal message as explained by Van Noort et al. (2014) were combined with the concept of conversational human voice as defined by Kelleher (2009). First, the direct addressing of the user/company by the posting person with his/her name and the use of second-person-pronouns such as “you” and “your” indicate one aspect of personalization. The second aspect is self-indication of the user or company, expressed through first-person pronouns such as “I” or “we” and/or initials/name of the posting person. The third factor is informal speech which uses abbreviations, emoticons and imitates spoken language. Fourth, invitational rhetoric that tries to engage the receiver of the message and suggests an interaction was considered. In the fifth place, sense of humour that indicates a conversational human voice completed the concept of message personalization. Those five dimensions were added up into one variable, where 0 means no personalization of the message, following a continuum to 5 which means a highly personalized message. The concept of message personalization is applicable to UGC as well as to MGC.

Engagement. For the measurement of engagement, the indicators were adapted for the different Social Media platforms and put into relation to the total amount of fans (nf) of a brand page (Cvijikj & Michahelles, 2013). For Twitter, engagement consisted of likes/nf, comments/nf, and retweets/nf. For Facebook, the likes/nf, comments/nf and shares/nf build

(22)

the variable of engagement. On Instagram likes/nf and comments/nf indicated engagement. The numbers were added up, resulting in one variable for each platform separately.

Results

This section reports the results of the content analysis for the three Social Media platforms Twitter, Facebook and Instagram applied to the hypotheses presented in the theoretical section. Linear regressions were used to show direct effects of the independent variables power and control on the dependent variables personalization and engagement. In addition, a moderation of the company structure (traditional vs. online-only-brands) was tested for every relationship. Traditional companies, being active on whole sale and retail via physical stores without exclusively offering their products online, are in this sample: Manhattan Cosmetics, Jack & Jones and Wilkinson Sword. On the opposite, companies considered online-only-brands sell their products only on their own online shop without any whole sale options or physical stores, are in this sample: Stowaway, Trashness and Dollar Shave Club. The total number of research units consisting of posts was N=720, for Twitter (n=240), Facebook (n=240) and Instagram (n=240). The content was each time divided in two subgroups for user-generated-posts (n=120) and marketer-generated-posts (n=120).

Twitter

H1a: There is a positive effect of power on personalization.

A linear regression was carried out. The regression model with the effect of power as independent variable on personalization as dependant variable was significant, F (2, 117)

=11.708, p <.001. The variable power explained 16.7 per cent of variance in personalization (R² =.167). Higher levels of power lead to higher personalization of the message (b* =.349, p <.001). H1a could be approved. The effect was moderate for both company types taken

together which means that a higher user power in general lead to a higher personalization of the message on Twitter.

(23)

H1b: The positive effect of power on personalization will be stronger on pages of traditional brands.

A moderation analysis was conducted, exploring if the company structure had an effect on the relationship between power and personalization. The model was significant, F (3, 116)

=8.118, p <.001. The variables power and company structure taken together explained 17.5

per cent of variance in personalization (R² =.175). Anyways, the moderation effect of company structure was not significant (p =.291). Therefore, H1b was rejected. H1c: There is a negative effect of control on personalization.

A linear regression was executed. The regression model with the effect of control as independent variables on personalization as dependant variable was significant, F (2, 117)

=8.12, p <.001. The variable control explained 12.2 per cent of variance in personalization (R² =.122). H1c could be approved. Higher levels of control lead to less personalization of the

message (b* =-.378, p <.001). The effect could be considered moderate for online-brands and traditional companies together. Therefore, less control executed by both company types lead to a higher personalization of the conversation with their followers on Twitter.

H1d: The negative effect of control on personalization will be stronger on pages of online-brands.

A moderation analysis was conducted with the independent variables control, company structure and the interaction effect of those two. The overall model was significant, F (3, 116)

=6.676, p <.001. The variables explained 14.7 per cent of variance in the dependent variable

personalization (R² =.147). However, the moderation effect of company structure was not significant (p =.066). Therefore, H1d was rejected. The company structure had no effect on the relationship of control and personalization.

H1e: There is a positive effect of power on engagement.

A linear regression was carried out with power as independent variable and engagement as dependant variable whereas engagement was not normally distributed. The regression model

(24)

was not significant (p =.557) and could therefore not be considered for further interpretation. Hypothesis H1e must be rejected, a direct influence of the user power on their engagement on Twitter could not be approved.

H1f: The positive effect of power on engagement will be stronger on pages of traditional brands.

The moderation analysis revealed a significant model F (3,116) =3.457, p =.019. The variables power and company structure explained 8.2 per cent of the variance in engagement

(R² =0.082). Engagement was not normally distributed. The interaction term of power and

company structure was not significant (p =.524). Thus, Hypothesis H1f was rejected, a moderation effect of the company structure on the effect of power on engagement on Twitter needed to be neglected.

H1g: There is a negative effect of control on engagement.

A linear regression was conducted but showed no significant results for the model (p =.12). Hypothesis H1g was therefore rejected, on Twitter, a negative effect of executed control by the company on the users’ engagement could not be shown.

H1h: The negative effect of control on engagement will be stronger on pages of online-brands.

The moderation analysis presented a non-significant result for the testing of the model (p

=.118). Hypothesis H1h was rejected, no influence of the company structure on the effect of

control on engagement could be proven on Twitter. Facebook

A linear regression was executed. The regression model testing the effect of power on

personalization was significant, F (2, 117) =4.639, p =.012. The variable power explained 7.3 per cent of variance in personalization (R² =.073). Therefore, H2a was approved, higher

(25)

levels of power lead to more personalization of the conversation on Facebook (b* =.271; p

=.03). The effect could be described as weak-to-moderate.

H2b: The positive effect of power on personalization will be stronger on pages of traditional brands.

A moderation analysis was conducted and the model was significant, F (3, 116) =4.501, p

=.005. The variables power and company structure explained 10.4 per cent of variance in

personalization (R² =.104). The moderation effect of company structure on the direct effect of power on personalization was significant (p <.005); b* =-.300 and showed a moderate effect of the interaction term. Thus, Hypothesis H2b could be approved. As the interaction graph (figure 2) shows, there was indeed a moderating effect of company structure on the positive relationship between power and personalization, which was stronger for traditional brands than for online-only brands.

Figure 2. Moderation effect of company type on the relationship between power and

personalization 1 1,5 2 2,5 3 3,5 4 4,5 5

Low Power High Power

P er son ali zat ion Tradition al Online

(26)

H2c: There is a negative effect of control on personalization.

A linear regression showed that the regression model with the effect of control on

personalization was significant, F (2, 117) =15.041, p <.001. The variable control explained 20.5 per cent of variance in personalization (R² =.205). Higher levels of control lead to less personalization of the message on Facebook (b* =-. 433, p <.001). The effect could be described as moderate and Hypothesis H2c was approved.

A moderation analysis for the effect of company structure on the relation between control and personalization was conducted. The overall model was significant, F (3, 116)= 10.057, p

<.001. The variables control and company structure explained 20.6 per cent of variance in

personalization (R² =.206). However, the moderation effect of company structure was not significant (p =.0601). Hypothesis H2d was therefore rejected.

H2e: There is a positive effect of power on engagement.

The linear regression model with the effect of power on engagement, which was not normally distributed, was not significant (p =.651). Therefore, Hypothesis H2e was rejected.

The overall model of the moderation analysis testing a moderation effect of control on the relationship of power on engagement showed significant results, F (3,116) =5.381, (p =.02). However, the interaction effect of company structure and power was not significant (p =.688) and Hypothesis H2f was rejected.

A linear regression for the effect of control on engagement showed no significant results for the model (p =.669). Hypothesis H2g was rejected.

(27)

The overall model of the moderation analysis with control and company structure as

independent variables and engagement, being not normally distributed, as dependent variable was significant F (3,116)=3.527, (p =.02). However, the interaction effect of company structure and control was not significant (p =.072). Hypothesis H2h was rejected, no

moderation effect of company structure on the relation between control and engagement could be proven on Facebook.

Instagram

The linear regression model testing the effect of power on personalization was significant, F

(2, 117) =12.707, p<.001. The variable power explained 17.8 per cent of variance in

personalization (R² =.178). Higher levels of power lead to more personalization of the message on Instagram (b* =.36, p <.001), the effect was moderate and H3a was approved. H3b: The positive effect of power on personalization will be stronger on pages of traditional brands.

A moderation analysis revealed a significant model with F (3, 116) =8.762, p <.001. The variables power and company structure explained 18.5 per cent of variance in personalization

(R² =.185), but the moderation effect of company structure was not significant on Instagram

(p =.346). Hypothesis H3b was rejected.

H3c: There is a negative effect of control on personalization.

A linear regression was executed and showed that the regression model with the effect of control on personalization was significant for Instagram, F (2, 97) =19.555, p <.001.

The variable control explained 28.7 per cent of variance in personalization (R² =.287). Higher levels of control lead to less personalization of the message (b* =-.538, p <.001). The effect could be considered moderate-to-strong and Hypothesis H3c could be approved.

(28)

A moderation analysis was carried out and the model was significant, F (3, 96)= 13.465, p

<.001. The variables control and company structure explained 29.6 per cent of variance in

personalization of the conversation on Instagram (R² =.296). There was no significant moderation effect of company structure (p =.275). Thus, Hypothesis H3d must be rejected. H3e: There is a positive effect of power on engagement.

A linear regression was conducted. The regression model with the effect of power on

engagement (not normally distributed) on Instagram was not significant (p =.57). Hypothesis H3e was rejected.

A moderation analysis presented a significant model with F (3, 116) =4.407, p =.006. The variables power and company structure explained 7.9 per cent of variance of engagement (not normally distributed) (R² =.079), but the moderation effect of company structure was not significant (p =.338). Hypothesis H3f was rejected.

The model of the linear regression with the effect of control on engagement (not normally distributed), was not significant for Instagram (p=.313). Hypothesis H3g was therefore rejected.

A moderation analysis was conducted, exploring a moderation of company structure on the effect of control on engagement, with the dependent variable not being normally distributed. The model was significant, F (3, 96) =11.739, p <.001. The variables explained 26.8 per cent

(29)

of variance of engagement (R² =.268) but the moderation effect of company structure was not significant (p =.492). Hypothesis H3h was rejected.

Conclusion

Social Media alter the relationship building between a company and its customers. As users gain the power to create content and react on content provided by the companies, the pressure for the communication professionals rises (Mills & Plangger, 2015). They fear to lose control over their message and oftentimes lack the knowledge or experience on how to communicate with their followers on Social Media (Buchanan-Oliver & Fitzgerald, 2014). The purpose of the study was to explore how interaction intents on Social Media by companies and users alter their relationship. More specifically, the power shift towards the users which arises with the possibilities of UGC and the intent to still keep the control over the content by the marketers was investigated. The influence of those two concepts on the relationship in terms of the personalization of the conversation and the level of engagement was compared for traditional brands and for online-only brands on Twitter, Facebook and Instagram. The three companies for the traditional sample and for the online-only sample were chosen respectively from the same sectors, two cosmetic companies (Manhattan cosmetics & Stowaway), two clothing companies (Jack & Jones & Trashness) and two hygienic product companies (Wilkinson Sword & Dollar Shave Club). While the traditional companies sell their products primarily over physical stores, the online-only brands, sell their products exclusively online.

Former research addressed an empowerment of the users on Social Media and the fear of losing control on Social Media by professionals, but a quantitative content analysis to empirically test the assumptions had not been done, so far. A strength of this study was the investigation of three different Social Media platforms simultaneously that vary in their characteristics and interaction possibilities while previous research on Social Media often focused on one specific platform. The Hypotheses of this study generally assumed a positive effect of user power on personalization and on engagement, regardless of the company type,

(30)

and for all three Social Media platforms. Further, a negative effect of executed control by the companies on personalization and engagement was anticipated. The comparison for company structure came into play for a moderation that was expected for both effects, the impact of power on personalization and engagement was supposed to be stronger for traditional brands and the impact of control on those outcomes was supposed to be stronger for online-only brands.

The results of the content analysis showed that the users’ power over their generated content had indeed a positive effect on the personalization of the overall conversational messages and that the companies’ control had a negative effect on the personalization level of the posts. The respective hypotheses could be supported on every analysed Social Media platform. The effects on Twitter can be considered moderate for both influences of power and control on the personalization of the message, the effect of power on Facebook is weak-to-moderate and is weak-to-moderate for control. On Instagram, a weak-to-moderate effect for power on personalization and a moderate-to-strong effect of control on personalization is evident. Motivations and expectations for the use of Social Media might play an important role as the highest positive effect of power and negative effect of control were evident on Instagram. This platform is normally very positive in tonality and criticism is more remarkable and an active power or control intent is seems inappropriate, resulting in stronger effects than on Twitter or Facebook.

With regard to the impact of power and control on engagement, no significant results could be reported for any of the analysed Social Media platforms. The variable engagement was not normally distributed and could not be transformed accordingly. This might be due to the low levels of engagement in general that were observable for both, MGC as well as UGC. The amount of engagement expressed in Likes, Retweets and Comments on Twitter, Likes, Shares and Comments on Facebook and Likes and Comments on Instagram, was very low when related to the number of followers/fans of the respective companies. That indicates a

(31)

high number of dead likes or lurkers, meaning that users follow the brand pages in a very passive way, consuming their MGC but not responding to it (Jahn & Kunz, 2012).

Concerning the difference for traditional companies and brands operating exclusively in e-commerce, the expected difference could not be approved in the analysis, except for one case: The effect of power on personalization of the message was stronger for traditional companies than for online-only brands on Facebook. That shows that a higher power of users on

Facebook can generate a higher personalization of the conversation with the company and this relationship is even stronger for traditional companies. Traditional brands might take the power shift more seriously, being highly concerned about a negative impact on their

reputation. Therefore, compared to online-only brands, they adapt the conversation with their users more strongly, once they feel their pressure. The characteristics of Facebook provide no textual limitations like Twitter and are more applicable for criticism than Instagram. As it is a Social Network that was initially build up for interpersonal relationships, companies partly enter the private space of the users (Fournier & Avery, 2011; Jahn & Kunz, 2012). It seems intuitive that the users’ power is very strong on this platform and that an online-only-brand can be perceived as less intrusive which can explain the stronger effect of power on

personalization for traditional brands.

The study is able to deliver guidelines for communication professionals and users that aim to foster their common relationship on Social Media. It shows that professionals do not need to fear a loss of control as a lower degree of control leads to more personal conversations with their users. Furthermore, users can strengthen the relationships to brands by enacting their power because criticism and the articulation of their demands consequently lead to more personal conversations. Those correlations hold true for both company types, traditional as well as online-only companies.

(32)

Discussion

The opening of Social Media platforms for brand building processes brings a number of opportunities but also bears a number of threats for the interaction of companies with their stakeholders. Benefits of Social Media are the possibility to increase the purchase behaviours of consumers (Goh et al., 2013) by generating dialogue and taking a direct impact on

customer loyalty and reputation (Cvijikj & Michahelles, 2013). While positive word-of-mouth spread by the users promotes the company without any additional efforts, negative word-of-mouth can harm the brands (Killian & McManus, 2015). Recent theories on Corporate

Communication stated that there is a power shift towards the users on Social Media, resulting in a fear of losing control over the message on behalf of the company (Bonsón & Ratkai, 2013; Castells, 2011; Hennig-Thurau et al., 2010; Killian & McManus, 2015; Peters et al., 2013). As the characteristics of Social Media reduce the anonymity of communication, the conversation between a company and its customers gets more personal and individual (Kelleher, 2009), fostering their relationship. As a higher level of power lead indeed to more personalization of the message on every Social Media platform, the power shift towards the users seems to have a positive outcome for their relationship with the companies, as the conversation gets more personal. Therefore, there would be no reasons for companies to fear this development. Further, a lower level of control by the companies lead to a higher level of personalization as well, putting emphasis on the fact that companies do not need to be afraid of giving away some of their control. With the regard to engagement, former research

detected a lot of positive outcomes for Corporate Communication (Brodie et al., 2013; Cvijikj & Michahelles, 2013; Dijkmans et al., 2015). As users who are empowered can be expected to be more active on Social Media (Muntinga et al, 2011), a higher engagement was expected for a rising degree of user power, but could not be approved. The same holds true for a lower degree of companies’ control, which had no positive effect on the users’ engagement. Former research indicated that different company structures affect the Corporate Communication

(33)

strategy (Buchanan-Oliver & Fitzgerald, 2014). A difference for traditional companies being very concerned about protecting their reputation, limiting their transparency and openness in comparison to brands operating exclusively in e-commerce was expected, but could only be approved for one case in the analysis: a higher level of power lead to a higher degree of personalization and this effect was stronger for traditional companies than for online-only brands on Facebook.

Practical implications

The effects of power and control on personalization could be approved for Twitter, Facebook and Instagram. As personalization is an important factor for a close relationship with their customers, professionals should encourage the empowerment of the users. When the users are criticizing the companies, request interaction from them and spread content over the companies, they challenge the professionals which consequently adapt their

communication and make it more personal in order to stabilize the relationship. On the other side, users might sense that companies trust them and take them seriously and equally employ a more personal tone in the conversation. Companies can execute a high level of control by observing their followers closely, limiting the possibility of user reviews and their comments as well as avoiding dialogue by the dedication to information topics and a

one-way-communication approach. Those practices had negative effects on the personalization of the conversation and consequently on the relationship between the companies and their users. Therefore, professionals should not fear to lose control over their brand messages but should acknowledge the positive effects (Berthon et al., 2012). Especially on Facebook, professionals from traditional companies should dare to accept the power shift of the conversations towards the users and less control on their behalf, as it pays off in a more personal conversation, even stronger than for online-only brands. If they open up their brand pages for reviews and comments, opt for entertaining topics and two-way communication and show trust by not following their users, a strong personal relationship can be established. The approved positive

(34)

effects of power on personalization and negative effects of control on personalization were mostly weak-to-moderate. However, companies should not opt for completely giving in to the customers. In order to stay in charge of their brand, they need to find a compromise between keeping control over their message and acknowledging the empowerment of the users (Cova & Pace, 2006; Muntinga et al., 2011).

Limitations and future research

For this analysis, the variable power consisted of factors that were built upon theoretical research on Corporate Communication and Social Media, but were not tested in this combination in any quantitative analysis before. Therefore, future research could extent the scale for measurement of user power on Social Media. User characteristics and

motivations could be evaluated in the context of a survey, as they can influence the degree of power, and could help to identify the most important drivers of power. Similarly, the concept of control of the company was derived from theory but could be further established by combining the content analysis with a survey, in order to assess how the professionals perceive the power shift towards the users, which aspects they consider important and how they articulate their fear of losing control. The variables could be expanded with regard to the outcome of the survey and could be able to predict personalization and engagement to a better extent.

Another suggestion for further research is the inclusion of comments into the analysis. While this research focussed on the personalization of the message send by the company and the users, the interaction in the comments following the post could be further investigated. Private messages could not be analysed on any platform and the degree of personalization in this area kept unexplored. As engagement was only measured by explicit numbers of the provided indices by the platforms such as Likes, Shares and Comments for Facebook, it was very static. The content of the comments and other indices of perceptions of the users which could show a more dynamic view of the variable engagement were not assessed. Furthermore,

(35)

all numerical metrics were weighed equally, not considering that a higher involvement is required for commenting on a post than simply sharing it or pushing the like button, indicating different levels of actual engagement. Therefore, future research could try to identify more valid measurements for the concept of engagement, reflecting the dynamics of Social Media, instead of counting static indicators.

Concerning the company structure, future research should test a direct effect of company structure on the outcomes personalization and engagement, as the underlying regression models for the moderation analyses on the different Social Media platforms showed potential indicators in this direction. Furthermore, research could concentrate on one specific sector in order to give more detailed recommendations for the professionals. As all of the investigated companies operated internationally, no differentiation for their country of origin was made. Further research could focus on specific countries of origin and make a comparison in order to find out if they have an impact on the Social Media approach or if the online world crosses the country borders and equalizes the cultural differences (Berthon et al., 2012). With regard to the timely scope of the study, no specific time frame was set but a fixed amount of posts was examined. Therefore, the development of the communication process could not be observed for a longer period of time and the posting frequency of the companies and users could not be assessed.

Despite the presented limitations, the study was able to clearly show the influences of power and control on the level of personalization of the conversation which is an important indicator for the relationship between a company and its users. The study made a first

approach in investigating three different Social Media platforms simultaneously and is able to give recommendations for professionals who want to communicate more personal with their audience. Companies do not have to fear a loss of control on Social Media, as a lower amount of control improves the relationship to their users. Although they can feel pressured by the power of the users, this ultimately also leads to more efforts to improve the relationship with

(36)

positive results for their business. A significantly different approach of online-only brands and traditional companies could only be shown for user power on Facebook. Therefore, a diverse approach on Social Media by various company types leaves room for further investigation.

References

Ahuja, V., & Medury, Y. (2010). Corporate blogs as e-CRM tools – Building consumer engagement through content management. Journal Of Database Marketing &

Customer Strategy Management, 17(2), 91-105. http://dx.doi.org/10.1057/dbm.2010.8

Baxter, G.J., & M. Connolly, T.M. (2014). Implementing Web 2.0 tools in organisations: feasibility of a systematic approach. The Learning Organization, 21(1), 6-25. http://dx.doi.org/10.1108/tlo-11-2012-0069

Beniger, J. R. (1986). The Control Revolution: Technological and Economic Origins of the

Information Society. Cambridge, Mass.: Harvard University Press.

http://doi.org/10.2307/2069810

Berger, J., & Milkman, K. L. (2012). What Makes Online Content Viral?. Journal Of

Marketing Research, 49(2), 192-205. http://dx.doi.org/10.1509/jmr.10.0353

Berthon, P. R., Pitt, L. F., Plangger, K., & Shapiro, D. (2012). Marketing meets Web 2.0, Social Media, and creative consumers: Implications for international marketing strategy. Business Horizons, 55(3), 261–271.

http://dx.doi.org/10.1016/j.bushor.2012.01.007

Bonsón, E., & Ratkai, M. (2013). A set of metrics to assess stakeholder engagement and social legitimacy on a corporate Facebook page. Online Information Review, 37(5), 787-803. http://dx.doi.org/10.1108/oir-03-2012-0054

Bortree, D. S., & Seltzer, T. (2009). Dialogic strategies and outcomes: An analysis of environmental advocacy groups' Facebook profiles. Public Relations Review, 35(3), 317‐319. http://dx.doi.org/10.1016/j.pubrev.2009.05.002