Corporate impression management on social media:
testing the effect of impression strategies, fit and dialogical
interaction on stakeholder relationship strength.
MSc Communication Science: Corporate Communication
Master’s Thesis
Christine Schneider (student 11107235) University of Amsterdam Thesis supervisor: Jim Slevin Graduate School of Communication
2 of 42 Abstract
With the advancement of media and communication technologies, institutions face the challenge of having to keep nurturing stakeholder relationships whilst adapting to increased media visibility and decreased control over image and information (Thompson, 1995). As a possible solution, past research has found use of corporate impression strategies on social media to favour stakeholder relationships (Men & Tsai, 2015; Kelleher, 2009). However, stakeholder relationships occur between different groups, and little is known on the benefits of adapting impression strategies according to fit of different groups for stronger stakeholder relationships (Bozzolan & Michelon, 2015). Moreover, dialogue on social media has been suggested to potentially foster positive relationships between organisations and their stakeholders (Kelleher, 2009; Macnamara & Zerfass, 2012), but there is lack of strategic employment of this. The purpose of this thesis is to investigate the effects of using corporate impression strategies in interactions on Facebook, fitting these to the type of stakeholder addressed, and communicating these in a dialogical manner, on stakeholder relationship strength. An experiment conducted on 178 social media users reveals a significant positive effect of using an ingratiation strategy, as well as fitting impressions to stakeholder type, and dialogical interaction. Strategies are proposed for practitioners to better manage impressions and relationships with stakeholders on social media through the use of ingratiation, fit and dialogue, and further research suggestions on corporate communication and stakeholder management are offered.
Introduction
Social media are increasingly used as interactive platforms for organisations to foster and manage relationships with their stakeholders online. However, in the age of digitalisation and media visibility, institutions have much less control over images and information flow they wish to project (Thompson, 1995), meaning they are more visible to stakeholders and more closely scrutinised than before (Tapscott, & Ticoll, 2003). To gain some control over
stakeholder perceptions online, organisations can use strategies in their interactions with them on social media. In fact, past research has found a positive effect of the use of corporate impression strategies on stakeholder relationships on social media, such as corporate character (Men & Tsai, 2015), and conversational human voice (Kelleher, 2009). However, stakeholder relationships occur between different groups, and relationship benefits of adapting impression strategies according to the variety of different groups has not yet been tested (Bozzolan & Michelon, 2015).
Moreover, dialogical social media employment has been suggested to potentially foster positive relationships between organisations and their stakeholders (Kelleher, 2009; Macnamara & Zerfass, 2012), and could bring them closer to realising the practice of ‘excellent’ public relations (Grunig, Grunig, & Dozier, 2002, in Macnamara & Zerfass, 2012).However, companies lack strategies for effectively managing stakeholder interactions and relationships on social media (Macnamara & Zerfass, 2012), including their dialogical use (Watkins & Lewis, 2014). Solutions should thus be provided for companies to be more strategic in their dialogical use of social media to positively manage stakeholder
4 of 42 The purpose of this thesis is to investigate the effect of impression management strategies on stakeholder relationships. Combining impression strategies with dialogical rather than one-way interaction form, and fitting these to stakeholder type, could reveal as effective methods for practitioners to be more strategic in their social media interactions, in a way which would help maintain positive stakeholder relationships. Furthermore, little research compares effects of impression management strategies across different stakeholder types on relationship outcomes. An attempt is made to answer the following research questions: What are the effects of using corporate impression strategies in social media interactions, and fitting these to the type of stakeholder addressed, on stakeholder relationship strength? Can dialogical interaction moderate these effects?
The study aims to help design appropriate strategies for practitioners to exercise some control over corporate impressions they generate when interacting with stakeholders on social media (in this study, the focus is on Facebook posts), to benefit relationships with them. This would allow practitioners to be more confident in engaging with stakeholders on social media, and to use these new communication platforms more strategically. On a research perspective, this study tackles gaps in corporate communication and stakeholder management research by comparing impression management strategies across stakeholder groups on social media, and analysing effects on stakeholder relationship outcomes.
Theoretical Background
Our current times are characterised by the extensive use of new media and communication technologies, including social media, “a group of internet-based applications that allow the creation and exchange of user generated content” (Kaplan and Haenlein, 2010, p. 61).
Organisations are increasingly using social media to facilitate intra and inter-organisational activities among different stakeholders, including cultivation and maintenance of positive stakeholder relationships (Kent, 2010).In the following theoretical section, the impact of social media on stakeholder relationships is discussed. Impression strategies are first suggested as a solution to gain control over stakeholder perceptions online and encourage positive relationships, after which the type of stakeholder and nature of relations is argued as important in devising impression strategies, and to finish, dialogical interaction is argued to further positively impact stakeholder relationships.
Impression management on social media in the age of visibility.
Social media are transforming how institutions are expected to nurture relations with stakeholders, by opening up opportunities for new forms of interaction. These conditions pose challenges for organisations, which Thompson terms as the “new age of visibility” (Thompson, 2005). According to him, we live in an age of high mediated visibility, where the increased flow of audio-visual content to communication networks enables a wide range of individuals to create and disseminate content once reserved for uni-lateral dissemination by the institutions in charge. This leads to an important loss in control of visibility: because it is difficult to control the flow of symbolic content within the Internet, it is harder for institutions to ensure that the images made available are those they wish to be circulated. Thus, because the information environment is more intensive and extensive, institutions struggle to control the images and information that constitute impressions in the public domain, as well as consequences of these impressions. Because stakeholders scrutinise companies more,
relationships are affected by this increased visibility, and companies in the digital age have to pay increasing attention to how they are perceived online to avoid negative consequences on relationships (Tapscott & Ticoll, 2003).
6 of 42 Consequently, managing stakeholder impressions while adapting to the loss of control of images and information posed by digital media is a key issue for public relations today. However, strategies can be used: impression management is defined as “the goal-directed activity of controlling information about a person, object, entity, idea or event.” (Connoly, Ahern & Broadway, 2007, p.343). To gain some control over stakeholder perceptions online, organisations can use strategies in social media interactions. In fact, research points to positive effects of corporate impression strategies on stakeholder relationships on social media: Kelleher (2009) found employing a conversational human voice and communicating relational commitment in public interactions to be linked to positive stakeholder relational outcomes, including trust, satisfaction, commitment, and control mutuality. Similarly, research from Men & Tsai (2015) suggests the effectiveness of a personification approach in social media communication to construct an agreeable corporate character, which could enhance public engagement, and induce intimate interpersonal interactions and community identification, in turn improving organization–public relationships.
However, there is a surprising lack of strategic social media use for managing impressions: the use of social media for organisational communication is still mostly experimental and ad hoc, rather than strategically planned with clear objectives (Macnamara & Zerfass, 2012). Therefore, this thesis aims to resolve this, by testing impression strategy effects on
stakeholder relationship strength in social media interactions between companies and stakeholders. A relevant study on impression strategies used on corporate websites
categorised five possible impression strategies: ingratiation (appearing likeable), competence (appearing intelligent), intimidation (appearing threatening), exemplification (appearing virtuous), and supplication (appearing weak)(Connolly-Ahern & Broadway, 2007). The corporate websites in the study found a trend to reflect competence and exemplification
strategies, but these were not implemented in application of two-way communication capabilities of the internet, and effects of these strategies on stakeholder relationship
outcomes were not measured. Both of these gaps are tackled here, by operationalising some of these strategies, integrating them into Facebook interactions between companies and stakeholders, and measuring relational outcomes. Consequently, these strategies are predicted to have a direct impact on the strength of stakeholder relationships.
H1: Impression strategies used in social media interactions have a direct effect on stakeholder relationship strength.
Now that stakeholder relationships have been argued to be affected by corporate impressions on social media, it is explained why identifying and targeting stakeholders could play an important role in this effect. Stakeholder theory is presented as a useful framework for understanding how different stakeholder categories play a role in relationship management, and could matter in choosing impression strategies to use.
Stakeholder relationship management and importance of stakeholder type.
A stakeholder refers to “any group or individual that can affect or is affected by the
achievement of an organisation’s objectives’’ (Freeman, 1984, in Fassin, 2009, p. 116). To conceptualise how stakeholder groups are positioned vis a vis institutions, several models have been developed. A prominent model by Freeman (1984) assumes that all stakeholders are equally important to the firm, thus the organisation makes no priorities of one set of interest and benefits over another (in Donaldson & Preston, 1995). It envisions corporations as a constellation of cooperative and competitive interests, and that relations between stakeholders and organisations have mutually beneficial value (Donaldson and Preston,
8 of 42 1995). However, Fassin (2009) contests that Freeman’s model is too simplistic, and does not account for the different levels, strengths of relationships, and power held by different groups. His triangular stakeholder-stakewatcher-stakekeeper model accounts this perspective (2009), and demonstrates the complexity of the relationships corporations have with their stakeholders. According to this, stakeholder groups have different degrees of dependencies and involvement in a firm, posing differences in managerial and legal interpretations of groups. Fassin sees that there are three categories of stakeholders, linked through a triangular relationship: stakeholders, who have a stake in the company (e.g. consumers), stakewatchers, who have an indirect claim and protect the interest of stakeholders (e.g. unions), and
stakekeepers, who regulate and impose external control on the firm, and have no stake (e.g. government). This model is useful on a managerial perspective as risks and opportunities can be devised in an adaptive manner according to the degree of responsibility of the firm toward each group.
Fassin’s triangular conception of stakeholder relations demonstrates that stakeholders vary significantly in the position and nature of the relations they have with organisations. His model supports that managing stakeholder relationships is best done by taking into account the specificity of the stakeholder type. Therefore, when employing impression strategies for relationship management, fitting impressions according to appropriateness with stakeholder type can be proposed as an important factor for success. Companies have already shown to differentiate their social media activities to address specific target groups (Krüger, Dang-Xuan, Schneider & Stieglitz, 2013), and to adapt their impression management techniques to stakeholder salience (Bozzolan & Michelon, 2015). However, effects of this on relationship outcomes have not yet been tested. Based on the assumptions of Fassin’s model, it is predicted that fitting impression strategies with stakeholder type could be more effective in establishing a strong relationship than using the same impression for all stakeholders.
H2: Fit of impression strategy with stakeholder type has a direct positive effect on stakeholder relationship strength.
Effects of dialogical social media employment on stakeholder relationships.
Now that fit of impression strategy has been proposed to affect strength of stakeholder relationships, it is next explained why dialogical use of social media could also do this. Excellence theory supports that dialogue between companies and stakeholders improves their relationship, but this is criticised. Also, there is a need to be more strategic in employing dialogue on social media for companies to reap from relationship benefits. Combining impression strategies with dialogical interaction is offered as a possible solution.
A useful perspective to understand the interactional feature of social media and what they entail for stakeholder relationships is Thompson’s ‘interactional theory’. In this angle, communication media are analysed “in relation to the forms of interaction which they make possible and of which they are part” (Thompson, 2005, p.32). Social media, like other communication media, actively affect relationships between individuals who, through the medium, “create new forms of action and interaction which have their own distinctive properties”(p. 32). Thompson makes a distinction between different types of computer-mediated interaction, which can be “computer-mediated interaction” or “computer-mediated quasi-interaction”. The first is dialogical, oriented to specific others and acts like face to face interaction (e.g. e-mail, real-time chat). The second is not, being more open-ended and oriented to different audiences (e.g. newsrooms, online bulletin boards). Because nature of the interaction differs between these communicative features, relationships between interactants are shaped in different ways.
10 of 42 Thompson’s ‘interactional theory’ underlines the relationship creating component of
communication technologies like social media. But what types of mediated-interaction occur through corporate social media use? And how does this affect relationships? A prominent view is that social media allows companies to implement dialogical interaction with their stakeholders, thus the practice of excellent PR. Excellence theory is a normative model assuming that public relations practice should apply two-way, symmetrical communication characterised by negotiation with stakeholders for resolving conflicts and reaching mutual understanding (Grunig, Grunig, & Dozier, 2002, in Macnamara & Zerfass, 2012). Instead of using persuasion via one-way and asymmetrical communication forms, stakeholder
relationships are managed through transparency and listening to public opinions in a dialogical manner.
Recently, increasing attention is drawn to social media as enabling the application of “excellence” in the digital age. Characteristics of social media environments are defined through openness for participation and interactivity involving dialogue, conversation, collaboration, and co-creativity, and due to diminished control, authenticity instead of pre-packaged content (Macnamara & Zerfass, 2012). Grunig believes that optimal use of social media can make public relations practice more “global, strategic, two-way and interactive, symmetrical or dialogical, and socially responsible” (Grunig, 2009, p.1).
Therefore, social media, when used appropriately, could allow organisations to cultivate ethical, authentic and positive stakeholder relationships through dialogue, characterised by engaged listening and active participation of publics to organisational decisions. However, there is resistance to the idea that social media allow dialogical communication and excellent stakeholder relationships. Thompson argues that most forms of mediated publicness created
by modern communication media are non-dialogical due to the differentiated roles of producer and recipient, which produce asymmetrical relationships (Thompson, 1995).
Producers produce messages for an indefinite range of potential recipients, who receive these messages in conditions which don’t allow direct discussion. Similarly, it is argued that public relations practitioners tend to “use new media in the old way” (Grunig, 2009), where
interactions reveal “symbolic participation instead of genuine interaction”, with purpose to satisfy the socio-emotional rather than build relationships (Kent, 2010). This reflects through common use of social media to disseminate marketing messages with techniques like viral marketing (Kent, 2010; Grunig, 2009).
Despite this criticism, empirical research favours arguments for positive effects of dialogical social media use on stakeholder relationships (Watkins & Lewis, 2014). In an experiment conducted by Colliander, Dahlén & Modig (2015), it was found that while dialogue lead to positive customer relationship outcomes (enhanced brand attitudes and purchase intentions), one-way communication did not. Although mediated-dialogue on social media can be viewed as non-dialogical, it signals impressions about companies to by-standing audiences with potential stakes, which could have positive impacts on stakeholder relationships (Colliander, Dahlén & Modig, 2015). Dialogue signals greater expense and care on behalf of the company than one-way interaction, encouraging better brand evaluations among social media users (Connoly, Ahern & Broadway, 2007, Colliander, Dahlén & Modig, 2015).
In sum, there is debate on benefits of dialogical interaction on social media for stakeholder relationships: dialogue may enhance relationships, but mediated dialogue on social media can also be interpreted as non-dialogical, producing asymmetrical relationships, a contrary
perspective to the excellence argument. Dialogical use of social media could also be
12 of 42 debate, by comparing effects of dialogue vs. one way communication on stakeholder
relationship strength. Based on previously outlined findings, a stronger relationship is predicted to result from dialogical interaction than one-way interaction.
H3: Dialogical interaction has a direct positive effect on stakeholder relationship strength.
Also, combining impression strategies with dialogical interaction could be an effective way to sustain positive relationships with stakeholders. As previously mentioned, companies lack strategies for effectively managing stakeholder interactions and relationships on social media (Macnamara & Zerfass, 2012), including their dialogical use (Watkins & Lewis, 2014). Communicating impression strategies in a dialogical manner could thus be proposed as a possible solution. This assumption is tested here: the combination between dialogical interaction and impression strategies is predicted to have a positive effect on stakeholder relationship strength.
H4: Dialogical interaction moderates the positive effect of impression strategies on stakeholder relationship strength.
Moreover, fitting impression strategies with stakeholder type in a dialogical manner could further positively impact stakeholder relationships. Research suggests that for organisations to best facilitate relationships through a dialogic process, communication is required to be tailored to specific recipients based upon relational needs (Bruning, Dial & Shirka, 2008). Apparently, the dialogical use of social media are yet to be used by practitioners, as they too often adopt a “one size fits all” strategy in relationship building activities (2007). There are however contesting examples: a study by Krüger, Dang-Xuan, Schneider & Stieglitz (2013) shows that the potentials of social media are exploited by firms to reach stakeholders and target groups, to interact with them in dialogue and to strengthen relationships. The authors
argue that social media can be used by firms to establish a dialogue with different stakeholder types, by enabling a bi-directional communication to be informed about stakeholder
expectations, so that conflicts of interest can be identified and harmonised through interactive communication. Based on these suggestions, it is predicted that combining dialogical
interaction and fit of impression strategy with stakeholder type will lead to stronger stakeholder relationships.
H5: Dialogical interaction moderates the positive effect of fit of impression strategy with stakeholder type on stakeholder relationship strength.
To conclude, the organisational implementation of social media for public relations practice and stakeholder communications poses challenges. One challenge is managing stakeholder impressions in an increasingly scrutinising environment. Another is knowing how to use social media’s potential to engage in dialogue with stakeholders without “using new media in old ways”. In this thesis, an attempt is made to resolve these challenges. First, to facilitate controlling organisational image on social media, impression management strategies are proposed as a solution, and tested for their effect on stakeholder relationships. Second, as the nature of the relationship could vary significantly based stakeholder type addressed in
communication, fit of impression strategy with stakeholder type is tested for its effect on stakeholder relationship strength. Third, as dialogical use of social media is widely argued to have positive outcomes for stakeholder relationships, this is tested, along with the effect of the combination between dialogue and impression strategies, and dialogue and fit with stakeholder type, on stakeholder relationship strength. The predicted relationships are summarised in the following model:
14 of 42 Methods
Study Design
To test the hypotheses, an online experiment via Qualtrics was conducted on a sample of 178 participants. Experimental design was employed, allowing to isolate factors to better explain correlations that would occur, an appropriate method since several factors are predicted to correlate with stakeholder relationship strength. Furthermore, an online design is most appropriate, relevant to the social media focus of the study (respondents can visualise the conditions in their appropriate context), and convenient for recruiting a large sample in a fast and cost-efficient manner. The experiment included a 3 by 2 by 2 factorial design, testing the effect of three variables on stakeholder relationship strength: impression strategy, fit of strategy with stakeholder type, and interaction type.
Sample Recruitment
Respondents were recruited over 3 weeks, and were required to possess a social media account to ensure their familiarity with social media interactions. For convenience,
participants were recruited via multiple social media platforms, including Facebook, Linked-In and Twitter. An announcement of the study containing a link to the survey was shared on several social network platforms, and individuals were contacted through private message on Facebook for recruitment. Average age of respondent was 24 (SD= 6.52, n=124), ranging from 18 to 58 years. Out of 124 people who disclosed gender, 65% were female. A greater part of respondents were French (17%), American (15%) and British (14%) (n=124), and had completed (or were completing) a bachelor’s degree (52%) or a master’s degree (38%) (n=124).
Experimental Stimuli
A fictitious brand named “Proud Pizza” was created, and a simulation of social media interactions through the Facebook chat simulation website “simitator.com”. Facebook is an appropriate social media choice, as one of the largest used online social networks in 2016 (eMarketer, 2016). Other platforms were left out of the study only for convenience. A pizza brand was chosen to represent a company revolved around a product which appeals to all publics, fictitious to avoid prejudice from an existing brand. Three Facebook interactions were displayed to every respondent, including a post about the brand from three stakeholder types: a potential consumer, a competing pizza brand (Domino’s), and an animal activist (Peta). The conditions were in the comment by “Proud Pizza” in response to each of the 3 posts, demonstrating 8 possible situations (2*(3+1)): a dialogical or one-way interaction (2), in combination with either one of the three impression strategies uniformly displayed for each response (3), or a display of these strategies adapted to fit with each of the three stakeholder types (1).
Operationalisation
Impression strategy: Impression strategy is the first manipulation of the experiment, understood as the self-presentation strategy employed in communication to influence stakeholder perceptions of the organisation. This independent variable, measured at the nominal level, consists of 3 items, taken from Jones’(1990) typology of self-presentation strategies (Chu & Choi, 2010): ingratiation, competence and exemplification. These were restricted to the three conditions due to the sample limitations of this study. Competence strategy characteristics include showing abilities, accomplishments and performance. Exemplification focusses on self-sacrifice and self-discipline, and ingratiation reflects the goal of being liked by others through statements of modesty, familiarity and humour
16 of 42 (Dominick, 1999, in Chu & Choi, 2010). These three elements were operationalised in the content of Proud Pizza’s Facebook comments (see appendices).
Interaction type: This is the second manipulation, reflecting wether the social media interaction is dialogical or one-way. Dialogical interaction is to be understood as a situation where organisations apply the “dialogic loop” principle (Rybalko and Seltzer, 2010). This means the company interacts by answering stakeholders on social media with a question or in a manner which encourages response. Thus, dialogue is operationalised in the sense that it opposes one-way interaction. This independent variable is dichotomous, consisting of 2 items: dialogical interaction and one-way interaction. These were also operationalised in Proud Pizza’s comments (see appendices).
Fit of impression strategy with stakeholder type: This is the third manipulation, understood as whether the impression strategy matches and reflects congruency with the type of
stakeholder addressed. This independent variable is dichotomous, including the two items adapted strategy and uniform strategy. To operationalise the first, a different impression strategy was applied to each stakeholder type based on congruency: ingratiation was matched with the “public” audience, competence with the “competitor” audience, and exemplification with the “activist” audience. For the uniform impression, each of the three strategies were applied to all three stakeholder types.
Stakeholder Relationship Strength: This is the dependent variable, reflecting the attributed strength and positivity of relationship respondents perceive the organisation to be involved in with their stakeholders. It is measured with the OPRs (organisation-public relationships) scale from Hon & Grunig (1999, in Men & Tsai, 2015). This includes 3 items: trust, satisfaction and commitment. Based on Hon & Grunig, trust refers to the level of confidence and willingness to open oneself to the organisation. Satisfaction reflects that benefits are perceived as outweighing the costs in being in the relationship with the organisation.
Commitment is the extent to which it is believed that the relationship is worth spending energy to maintain and promote. A 7-point Likert scale (strongly disagree/strongly agree) measured each item, where questions asked to what extent these attributes respondents associate with the organisation based on the Facebook posts. Respondents do not represent company stakeholders in this study, but their perception of OPRs is a valid indicator of how strong relationships between companies and stakeholders could be, because they are
perceived to be this way by social media audiences (respondents).
Procedure
A pilot study was first conducted on 20 respondents, to make sure the manipulations take effect appropriately. This included filling out questions asking whether the interaction is believed to be realistic, whether respondents perceive the impressions as same or different across all stakeholder types (fit), whether they perceive different impressions to be
appropriate according to the stakeholder type (validity of fit), whether respondents attribute the correct strategy to the one used in the manipulation (impression strategy), and whether they perceive it to be dialogical or not (interaction type). These same manipulation checks were conducted in the actual study.
For the experiment, participants filled out a survey which first introduced them to the Proud Pizza organisation as a pizza company who recently won the Guinness World records for largest pizza. The stimulus was then introduced through a randomly allocated screenshot of the three Facebook posts and comments. These contained 1 of the 8 manipulations. The survey followed, including the manipulation check, and measure of the independent variable, stakeholder relationship strength, based on attributions from the Facebook posts. Questions included items taken from Hon & Grunig (1999, in Men & Tsai, 2015), adapted to the mock case. Sample questions were then asked (age, gender, nationality, education level), and the
18 of 42 survey ended on a statement which thanked the participants and debriefed the true purpose of the study and the fictitious nature of the brand and interactions.
Results
Scale Reliability
Stakeholder Relationship Strength: PCA showed that the 6 indicators for trust loaded on one single factor (EV = 3.87, R2 = 64.42, Cronbach’s alpha = .88, M= 27.0, SD= 6.90), which is what also occurred for the 4 indicators of commitment (EV = 2.74, R2 = 68.42, Cronbach’s alpha = .85, M= 18.27, SD= 4.88), and the 5 indicators of satisfaction (EV = 3.78, R2 = 75.53, Cronbach’s alpha = .92, M= 23.03, SD= 6.08). A scale for “Relationship Strength” including the 3 items (15 indicators in total) was computed, which proved very strongly reliable: Cronbach’s alpha= .95, M= 68.39, SD= 16.53.
Manipulation Check
A manipulation check was performed by testing whether the experiment conditions matched up with the responses to questions about which interaction type, impression strategy and fit with stakeholder condition participants thought to have seen. This was done by generating correlation tables measuring association between the condition variables and the responses they generated. Unfortunately, these show that out of the three independent variables, only “interaction type” had a strong manipulation effect, whilst all other conditions had either weak or no effects. The bivariate table with asymmetric effect measures (a directional relationship is assumed for the effect of condition on participant’s responses) indicated that the relationship between interaction type (dialogue vs. one way) and respondent indication is significant and strong, Goodman & Kruskall’s tau= .643, p<.001 (see Table 1 in appendices for percentage of participants who answered the condition questions correctly). However for
the fit condition, the association effect is very weak, Goodman & Kruskall’s tau= .061, p=.004. Additionally, a bivariate table was generated for each impression strategy, indicating weak manipulation effects of the ingratiation condition, Goodman & Kruskall’s tau= .249, p<.001, and the competence condition, Goodman & Kruskall’s tau= .128, p=.003, and insignificant effects of the exemplification condition, Goodman & Kruskall’s tau=.018, p=.454. In sum, these indications show that the manipulation has worked well enough for “interaction type”, but not well enough to allow to perform scientifically reliable analyses of the other independent variables. However, the findings are still interpretable as most
manipulations were significant, and the “exemplification” item was withdrawn from analyses for lack of significant manipulation effect. It is also useful to add that to check for valid design of the “fit” condition, a question asked whether the the impressions were considered appropriate for the type of stakeholder addressed, to which 92% of respondents in the “adaptive” condition answered “yes”.
H1
The first hypothesis, predicting a direct effect of impression strategies on stakeholder relationship strength, was tested with a multiple regression analysis. This included the conditions “Ingratiation” and “Competence” as independent dummy variables in the model (“Exemption” is excluded for lack of manipulation effect), and “stakeholder relationship strength” as dependent. The multiple regression reveals that the model as a whole is significant, F(2,64)=4.72, p=.012, but the strength of the prediction is small: 10% of the variance in stakeholder relationship strength can be predicted by these variables, adj.R2=.101. However, only ingratiation proved significant as a predictor, b= 0.94, b*= 0.38, t= 2.81, p=.007, 95%CI [0.27, 1.61], significantly predicting stakeholder relationship strength while holding competence, b= 0.11, b*= 0.05, t= 0.33, p=.739, 95%CI [-0.54, 0.76], constant. Per unit increase in use of ingratiation strategy, respondent-attributed strength of stakeholder relationship increases 0.94 units. This result partially supports the hypothesis, as an effect of
20 of 42 the ingratiation strategy is found, however the scientific reliability of this result should be questioned as manipulation effects of impression strategies were found to be weak as stated earlier.
H2
An independent samples T-test was used for hypothesis 2b, which predicts that adapting impression strategies to stakeholder types are linked to stronger stakeholder relationships than using a uniform strategy across all types. This included “Fit” (uniform vs. adaptive) as independent variable and “stakeholder relationship strength” as dependent. The T-Test showed that there is a small, significant difference between the means of participants in the “uniform” condition (n=67) and the “adaptive” condition (n=66), t(131)= 1.84, p=.034, 95% CI[-0.72, 0.01], d=0.32. Those who were exposed to an “adaptive/fit” condition (M= 4.69, SD= 0.99) attributed a higher positive stakeholder relationship strength than those exposed to the “uniform/non-fit” condition (M= 4.35, SD= 1.16). Thus, the hypothesis is supported, taking again in mind the weak manipulation effect of the fit condition.
H3
A higher score for stakeholder relationship strength was predicted for those exposed to a dialogical interaction than for those exposed to a one-way interaction. To test this, an independent samples T-test was produced, with “Interaction type” as independent variable. Results indicated that there is a small, significant difference between the two groups, t(131)= 1.77, p=.040, 95% CI[-0.70, 0.04], d=0.31. Those exposed to a “dialogical interaction” condition (n= 65) (M= 4.69, SD=1.10) appointed a stronger stakeholder relationship with the brand than those exposed to a “one-way interaction” condition (n= 68) (M= 4.36, SD= 1.06). The hypothesis is thus supported.
H4
A moderation analysis was performed with the use of Hayes’ PROCESS macro to test hypothesis 4, which suggests that dialogical interaction moderates the positive effect of
impression strategies on stakeholder relationship strength. This included “stakeholder
relationship strength" as outcome variable, the “Ingratiation” strategy as independent variable (chosen as it is the only significant “impression strategy” predictor of relationship strength) and “dialogical interaction” as moderation variable. Overall, the regression model is
significant, F(3,63)= 6.01, p=.001. Interaction type, use of ingratiation strategy, and their combination all explain 22% of the variance in stakeholder relationship strength, R2= .223. Moderation analysis showed a non-significant moderation of the effect of ingratiation
strategy use on relationship strength by dialogical interaction type, R2 change = .016, F(1,63) = 1.30, p= .258. Both dialogical interaction, b= 0.85, SE= 0.31, p=.008, and use of the
ingratiation strategy, b=1.20, SE= 0.38, p=.003, have a direct positive effect on stakeholder relationships strength, although the interaction between the two has no effect, b=-0.63, SE= 0.55, p= .258. The results thus do not support the hypothesis that use of dialogue moderates the positive effect of using impression strategies (ingratiation) on stakeholder relationship strength, but they do show that when controlling for dialogue, ingratiation has a significant positive effect on stakeholder relationship (and vice versa).
H5
A moderation analysis was performed with the use of Hayes’ PROCESS macro to test hypothesis 5, which suggests that dialogical interaction moderates the positive effect of fitting impression strategies to stakeholder type on stakeholder relationship strength. This included “stakeholder relationship strength" as outcome variable, “fit” as independent variable and “dialogical interaction” as moderation variable. Overall, the regression model is significant, F(3,129)= 3.12, p=.028. Interaction type, fit and their combination all explain 7% of the variance in stakeholder relationship strength, R2= .068. Moderation analysis showed a non-significant moderation of the effect of fit on relationship strength by dialogical
interaction, R2 change = .019, F(1,129) = 2.59, p= .110. Both fit, b= 0.64, SE= 0.26, p=.015, and dialogical interaction, b=0.63, SE= 0.26, p=.017, have a positive effect on stakeholder
22 of 42 relationship strength, although the interaction between the two has no effect, b=-0.60, SE= 0.37, p= .110. The results thus do not support the hypothesis that use of dialogue moderates the positive effect of fit on stakeholder relationship strength, but they do show that when controlling for dialogue, fit has a significant positive effect on stakeholder relationships (and vice versa).
Conclusion
In this thesis, an attempt was made to determine whether impression strategies could provide a solution to the visibility problem, and to allow more strategic employment of dialogue on social media to benefit stakeholder relationships. To do this, it was asked whether stakeholder relationships could become stronger by using corporate impression strategies in interactions on Facebook, fitting these to the type of stakeholder addressed, and communicating these in a dialogical manner. Findings support all hypotheses, excluding those concerning moderation effects of dialogical interaction.
The ingratiation strategy was found to have a positive effect on stakeholder relationship strength, while the competence strategy was found not to have a significant effect. It was unfortunately impossible to make any claims as for the exemplification strategy due to non-significant manipulation effects. This however suggests that not only could employing a specific impression strategy in social media interactions help strengthen stakeholder relationships, but the choice of the impression could matter as well: an ingratiation impression is suggested to be more effective than a competence impression for a positive impact. This resonates with findings on the link between conversational human voice and positive stakeholder relational outcomes (Kelleher, 2009). To explain, in this thesis, an
ingratiation strategy was operationalised by appearing as likeable and friendly through the use of humour and a smiley emoticon (see appendix). This adds human characteristics to the communication, comparable to conversational human voice. Results also support previous claims on the positive effect of a personification approach (Men & Tsai, 2015) on stakeholder relational outcomes and organisation-public relationships. Furthermore, the study on which the impression strategies were based on found a general trend to reflect competence and exemplification strategies rather than ingratiation (Connolly-Ahern & Broadway, 2007) in corporate communications with stakeholders. It may be thus advisable for companies to focus on using ingratiation rather than competence strategies if they want to strengthen stakeholder relationships.
Results also found a significant positive effect of fit on stakeholder relationship strength, in the way that adapting strategies according to stakeholder type could lead to stronger
stakeholder relationships than uniformly employing them across all types. Adapting impression management techniques to stakeholder salience has already been shown as a practice, but evidence of effects on relationship outcomes are scarce. The result uncovers this gap, and suggests importance of carefully designing communication strategies that reflect appropriate impressions for the type of stakeholder adressed. As Fassin’s stakeholder model points to the variability in position and nature of the relations different stakeholder groups have with companies, the results further suggest that this variability should be accounted for by adapting impression strategies in social media communications to the congruency of specific stakeholder types to foster strong relationships. Although this result lacks scientific relevance (weak manipulation effect), it can be considered as an interesting proposition for a relatively untamed research area, and it supports recent developments in stakeholder theory which underline importance of considering stakeholder position for positive relationships.
24 of 42 Moreover, results suggest that practitioners should aim to implement impression strategies in a dialogical manner, as a positive effect on stakeholder relationship strength was found. This is also most scientifically reliable out of all of the results, due to significant manipulation effects. This supports the current trend of research on social media which supports dialogue to positively impact relationships between companies and stakeholders (Colliander, Dahlén & Modig, 2015; Watkins & Lewis, 2014). It is particularly relevant to findings on mediated dialogue on social media, suggested to send out positive signals to audiences that lead to positive relationship outcomes (Colliander, Dahlén & Modig, 2015). This thesis shows that the simple addition of a question at the end of a comment, or anything that encourages a dialogical loop, can have a small, but existing positive impact on the way stakeholders make attributions about a company’s commitment, trustworthiness, and the satisfaction they could feel in engaging in a relationship with them. Although a weak effect was found, it is a
promising result which could, if further demonstrated, close the debate on the positive impact of dialogue in social media on stakeholder relationships.
However, dialogical interaction was not found to moderate the positive effect of use of ingratiation strategy, neither fit of impression strategy to stakeholder type, although each of these variables show to increase relationship strength on their own. As a possible explanation, a dialogical approach could strengthen relationships for different reasons than impression strategies and fit. Ingratiation could signal that the company is friendly, playful, or has a sense of humour. Fitting impressions could signal efforts to answer each interactant in a personable way, and to tend to each interactant in a manner most relevant to the type of relation they have. These signals could then turn into attributions of trust, satisfaction and commitment in relationships. Dialogue can be understood to act differently, as positive signals from dialogue are far more implicit than appearing friendly or appearing to make an effort, which could explain the absence of a combination effect. Furthermore, this result
means that striving to combine a dialogical approach with an adaptive one in the way it was done in this study seems unnecessary. Although literature encourages practitioners to target stakeholders on social media and address their specific needs through dialogue (Bruning, Dial & Shirka, 2007; Krüger, Dang-Xuan, Schneider & Stieglitz, 2013), the study shows that encouraging a “dialogical loop” and fitting impression strategies with their type is not an effective way to do so.
To sum up, this study makes several interesting suggestions. First, a significant positive effect of using an ingratiation impression strategy on stakeholder relationship strength was found, suggesting that designing social media interactions reflecting an ingratiation impression as opposed to competence or exemplification could more effectively encourage strong
stakeholder relationships in a social media context. Second, the study suggests that fitting impressions to stakeholder type, as opposed to using a uniform impression for all
stakeholders, also encourages strong stakeholder relationships. Third, a significant effect of dialogical use of social media on stakeholder relationship strength was found, suggesting a possible signalling effect of using dialogical interactions for positive stakeholder attributions of companies. These suggestions allow to conclude that focussing on managing impressions in a dialogical and adaptive manner could help practitioners handle social media strategically, and could help bring long term benefits through maintenance of strong relationships with stakeholders online, addressing some of the issues posed by increasing media visibility and importance of social media for corporate communication and public relations. No moderation effects were however found between dialogue and use of an ingratiation impression, or dialogue and fit, meaning that although dialogue is important, it seems not to play a role in reinforcing the positive impact of impression strategies and their fit on stakeholder
26 of 42 Discussion
This study aims to propose appropriate strategies for practitioners to be more confident in using social media to interact and nurture relationships with stakeholders online, in a time characterised by increased online visibility and corporate social media use for public relations activities. It also aims to uncover research gaps by testing effects of fit of communication strategies to stakeholder type on relationship outcomes. Implications of the results are discussed.
Impression strategies used in communication on social media, as a way to positively manage and reinforce stakeholder relationships, are suggested in this thesis as a possible solution to the visibility problem. With today’s high media visibility, companies need to be increasingly careful with the way they present themselves online, and need to communicate strategically to positively impact stakeholder attributions, and in turn, relationships. The results suggest that ingratiation can be an effective way to bond with stakeholders online. Practitioners should aim to present themselves in a personable, charismatic, friendly, and when
appropriate, informal manner on social media, to obtain positive attributions and potential to build strong relationships with stakeholders online. Further research could explore other ways to apply an ingratiation strategy on various social media platforms (other than Facebook), as well as comparing this impression with others left out in this study (e.g. supplication,
intimidation), and testing effects on other outcome variables, like purchase intention or reputation.
Furthermore, type of stakeholder addressed revealed as important in this thesis when
regardless of the stakeholder addressed. However, adapting strategies to stakeholder types were found to be more effective than keeping the same impression for all. In this study, ingratiation was matched with potential consumers, competence with competitors, and exemplification with activists. It is suggested for practitioners to reflect about the suitability and congruency of their communication strategies with the type of corporate relationship at hand with stakeholders on social media, meaning considering responsibility of the company towards the stakeholder, and the degree of impact they have on the company. Additionally, importance of fit of communication strategies with stakeholder groups is largely undermined in stakeholder management research. It is encouraged to further research relational outcomes of fit beyond impression strategies, by applying fit to all types of communication strategies (e.g. crisis management).
Dialogical interaction showed to positively impact stakeholder relationships, supporting much of the literature on benefits of a dialogical approach. This is particularly in line with the findings from Colliander, Dahlén & Modig (2015), who found dialogue to lead to positive customer relationship outcomes as opposed to one-way interaction. Excellence theory further posits that dialogue allows authentic and positive relationships between companies and stakeholders. Despite Thompson’s belief that mediated dialogue is a form of non-dialogue, and criticism of social media dialogue as another persuasive PR strategy, the results point to a positive relationship impact of social media dialogue for Facebook. Practitioners are
encouraged to go beyond practicing a one-way interaction form when participating in conversation topics on social media, by establishing and inviting a dialogical exchange. Although a disclaimer should be made when interpreting results: practitioners should not consider use of dialogue as a mere copy-paste tool in their communication strategies to trick stakeholders into liking their company. Stakeholders value genuine interaction (Grunig, 2009), and dialogical interaction could signify genuineness, but the encouragement for
28 of 42 further exchange should be as natural and honest as possible, devoid of short-term marketing incentives. It is suggested to avoid repeating the same communication for every audience member, being personable, and thinking creatively when it comes to engaging in dialogue with stakeholders on social media.
Dialogue showed to be important, but results do not support that it plays a moderating role in the positive impact of impression strategies and their fit on stakeholder relationships. The absence of a combination effect means that there is room for further study. Combining fit with dialogue did not show as reliable in aiding practitioners use less of a “one size fits all” dialogical communication strategy on social media for relationship building, and solutions should be further pursued. For practitioners, it is still proposed for dialogue and fitting impressions to be used in combination, as both approaches reveal to have benefits on their own.
Several limitations are worth highlighting in regards to this study. First, the experiment could have benefitted from a control group. The results could have been more conclusive if effects of impression strategies were compared with effects of lack of impression strategies. This was missing from the design, as it was challenging to operationalise a situation where no impression strategies are used: participants would have perceived an impression to be in effect regardless of whether this is considered strategic or not. It is recommended for future research to consider designing a control group, perhaps based on similar relevant
experiments.
Another limitation is the design of the dialogical interaction condition. This was done simply by including a question at the end of the message from the brand addressed to each
dialogue was chosen for convenience. It is however very different from the type of dialogue outlined in the literature, where a proper discussion and deliberation occurs between
company and stakeholder on topics of concern to both parties. In the experiment, dialogical interaction was simplistically suggested as an invitation to continue the interaction. Further research should explore more valid ways of operationalising dialogue on social media.
Other limitations include weak manipulation effects: more reliable measurements for the variables in this study should be developed (only dialogue had a strong manipulation effect), perhaps by testing these more thoroughly. The study could have also benefited from being conducted on a larger and more diverse sample (participants hardly stretch out beyond the author’s network), as well as on other social media platforms than Facebook, including Twitter, LinkedIn, Instagram and others, for stronger external validity.
30 of 42 Bibliography
Bozzolan, S., Cho, C. H., & Michelon, G. (2015). Impression management and organizational audiences: The Fiat group case. Journal of Business Ethics, 126(1), 143-165.
Bruning, S. D., Dials, M., & Shirka, A. (2008). Using dialogue to build organization– public relationships, engage publics, and positively affect organizational outcomes. Public Relations Review, 34(1), 25-31.
Chu, S. C., & Choi, S. M. (2010). Social capital and self-presentation on social networking sites: a comparative study of Chinese and American young generations. Chinese Journal of Communication, 3(4), 402-420.
Colliander, J., Dahlén, M., & Modig, E. (2015). Twitter for two: Investigating the effects of dialogue with customers in social media. International Journal of Advertising, 34(2), 181-194.
Connolly-Ahern, C., & Broadway, S. C. (2007). The importance of appearing competent: An analysis of corporate impression management strategies on the World Wide Web. Public Relations Review, 33(3), 343-345.
Donaldson, T., & Preston, L. E. (1995). The stakeholder theory of the corporation: Concepts, evidence, and implications. Academy of management Review, 20(1), 65-91.
EMarketer (2016, April 8). Facebook Remains the Largest Social Network in Most Major Markets [Web article post]. Retrieved from
https://www.emarketer.com/Article/Facebook-Remains-Largest-Social-Network-Most-Major-Markets/1013798
Fassin, Y. (2009). The stakeholder model refined. Journal of business ethics, 84(1), 113-135.
Grunig, J. E. (2009). Paradigms of global public relations in an age of digitalisation. PRism, 6(2), 1-19.
Kaplan, A. M., & Haenlein, M. (2010). Users of the world, unite! The challenges and opportunities of Social Media. Business horizons, 53(1), 59-68.
Kelleher, T. (2009). Conversational voice, communicated commitment, and public relations outcomes in interactive online communication. Journal of communication, 59(1), 172-188.
Kent, M. L. (2010). Directions in social media for professionals and scholars. The Sage handbook of public relations, 2, 643-656.
Krüger, N., Dang-Xuan, L., Schneider, A. M., & Stieglitz, S. (2013). Usage of Social Media for External Stakeholder Relationship Management-A Study of German Companies and International Non-government Organizations. Advanced Information Networking and Applications Workshops (WAINA), 2013 27th International Conference on (pp. 1479-1482). IEEE.
Macnamara, J., & Zerfass, A. (2012). Social media communication in organizations: The challenges of balancing openness, strategy, and management. International Journal of Strategic Communication, 6(4), 287-308.
Men, L. R., & Tsai, W. H. S. (2015). Infusing social media with humanity: Corporate character, public engagement, and relational outcomes. Public Relations Review, 41(3), 395-403.
32 of 42 Rybalko, S. and Seltzer, T. (2010) Dialogic communication in 140 characters or less: How Fortune 500 companies engage stakeholders using Twitter. Public Relations Review, 36(4), 336-341.
Tapscott, D., & Ticoll, D. (2003). The naked corporation: How the age of transparency will revolutionize business. Simon and Schuster.
Thompson, J. B. (1995). The media and modernity: A social theory of the media. Stanford University Press.
Thompson, J. B. (2005). The new visibility. Theory, Culture & Society, 22(6), 31-51.
Watkins, B., & Lewis, R. (2014). Initiating dialogue on social media: An investigation of athletes’ use of dialogic principles and structural features of Twitter. Public Relations Review, 40(5), 853-855.
Appendices
Word count of text body: 7465 Table 1
Manipulation Check Results
n Goodman & Kruskal’s tau percentage of correct responses p Interaction Type 149 .643 <.001 Dialogical 71 89 One-way 78 91.2 Impression Strategy 84 Ingratiation 28 .249 92.9 <.001 Competence 29 .128 58.8 .003 Exemplification 27 .018 26.7 .454 Fit 148 .061 .004 Adaptive 74 52.9 Uniform 74 71.4 Table 2
Differences between Interaction Type and Fit Group ratings of Stakeholder Relationship Strength n M SD df t p Cohen’s d Fit 133 131 1.84 .034 0.32 Uniform 67 4.35 1.16 Adaptive 66 4.69 0.99 Interaction Type 133 131 1.77 .040 0.31 Dialogical 65 4.69 1.10 One-Way 68 4.36 1.06
34 of 42
Table 3
Regression Model predicting Effect of Ingratiation and Competence on Stakeholder Relationship Strength b b* 95% CI Lower 95% CI Upper t F p R2 Model 4.72 .012 0.10 Constant 4.02 3.55 4.49 17.15 .000 Ingratiation 0.94 0.38 0.27 1.61 2.81 .007 Competence 0.11 0.05 -0.54 0.76 0.33 .739
Experimental Conditions
36 of 42 Ingratiation and One-way
38 of 42 Competence and One-way
40 of 42 Exemplification and One-way
42 of 42 Adaptive (Fit) and One-way
