The potential of social influence(rs) in organizational crisis ?
Kiona Stierman 10365982 Master’s Thesis Graduate School of Communication Master’s programme Communication Science Toni van der Meer 02-02-2018
1
Abstract
This study is aimed at uncovering the potential role of social influencers in the strategic approach towards crisis communication in organizational crisis. Through an experiment with 133 participants this study aims to test the mechanism of crisis communication in an online context and specifically aims to uncover what role social influencers play in this strategic communication approach. The study compares the denial- and apology crisis response strategies in their relation with reputation evaluation and how different sources (corporate vs. social influencer) affect this relationship. The findings indicate that source of the message moderates the relation between crisis response strategy and public reputation evaluation. Meaning that in case of the organization denying responsibility for the crisis messages could best be shared by social influencers. When on the other hand the organization aims to acknowledge responsibility for the crisis and apologize for it crisis messages can best be shared by the organization itself to ensure the most positive reputation evaluation.
Introduction
A crisis is an unexpected event that threatens to disrupt an organizations’ operations and poses threats to its existence (Coombs, 2007). It can negatively impact organizational performance, financial wellbeing and organizational reputation. Communication can be identified as a tool to manage information and create meaning towards the public (Coombs, 2015; Coombs & Holladay 2002). By communicating the organization can manage how people perceive the crisis and is therefore important to limit and repair damage caused by crisis situations.
The media landscape is changing towards a more social and interactive oriented approach towards communication. Social media networks that evolve around
user-manipulated content and interactive communication are used by most individuals daily (Liu, Austin & Jin, 2011; Van der Veer, Boekee, & Peters, 2017). This creates the possibility for the public to gather and share information in real time with everyone (Mangold & Faulds,
2
2009). Certain individuals on social media encompass a large audience and show leadership abilities. They can function as role models in this complex network of interactive
communication. These role models, also called social influencers, can be identified as individuals that show the ability to influence other individuals in a certain group more than others (Langner, Hennings & Wiedmann, 2013). Social influencers are viewed as role models that engage in a digital conversation with their public through which they can influence attitudes, emotions and behaviours (Booth & Matic, 2011).
This influence of social influencers can be of great potential. When organizational credibility is threatened by a crisis and its trustworthiness is questioned, which is a
fundamental attribute to providing believability and effectiveness of messages (Reynolds & Seeger, 2005). Social influencers could serve as a more credible source in times of crisis as they are independent from the organization and are therefore not threatened in their
credibility (Yang, Kang & Johnson, 2010). In this way they can reduce the public’s suspicion of manipulated and inaccurate information. Therefore the use of social influencers could be of great interest for organization to improve their crisis communication strategy. This will be highlighted in this study to understand todays communication and uncover the potential value of social influencers for organizations during a crisis.
To understand the underlying mechanism of crisis communication the following factors will be taken into account. Firstly crisis response strategies which are aimed at reducing and repairing reputational damage (Coombs, 2007). These strategies as described in the Situational Crisis Communication Theory (SCCT) will be studied to see if the social influencer potential differs amongst these strategies. The second factor underlying the mechanism of crisis communication is the source of the crisis response message. A social influencer can be identified as a third party source, as described by the Social-Mediated Crisis Communication model (SMCC), with the ability to rebuild organizational image (Jin, Liu & Austin, 2014; Zhu, Anagondahali & Zhang, 2017). This expected impact is the reason to study the differences between a corporate and social influencer source of the crisis
3
incorporated in this study to understand the mechanism behind the expected effects as they can determine the potential threat of the crisis and the effectiveness of the source.
It is scientifically important to study the potential value of social influencers as a third party source in organizational crisis. Since there is not much empirical research regarding this potential in this context, as current research is mostly focussed on marketing purposes. Researching the influence of source on public perception, attitude and organizational reputation evaluation in context of an organizational crisis will therefore be aimed at filling this gap in the research field. This experimental study will be guided by following research question:
“What role do social influencers play in the relation between crisis response strategy
and organizational reputation?”
To understand the effects the findings will be evaluated by means of the following follow up question:
“To what extent is the effect dependent on source credibility, -expertise and attribution
to the organization?”
The societal relevance of this study is twofold. Firstly the ability of social influencers to step in as role models, navigate the public sensemaking in a crisis and guide the forming of attitudes towards the organization could be of great relevance from an organizational perspective when loss of organizational credibility occurs. From a public perspective on the other hand it is important to research the potential of social influencers as the public should be aware in which way they are influenced by different sources. Which could, in the case of social influencers, not only relate to influence on consuming behaviour but also influence on formation of attitudes towards organizations and situations during crisis.
4
Theoretical framework
Organizational reputation and secondary crisis communication in times of crisis A crisis is defined as an unexpected event that threatens to disrupt an organization’s operations and poses both a financial and a reputational threat (Coombs, 2007). In times of crisis it is particularly important to protect organizational reputation as it can change the way stakeholders interact with the organization when threatened. Organizational reputation is defined as the public evaluation of stakeholders on how they perceive the organization and to what extent their expectations are met (Coombs, 2007; Coombs & Holladay, 2002). This is a highly valued asset for organizations as it provides the possibility to compare organizations to one another. This way, organizations can be viewed as superior to others and be ranked in order of relative standing with regard to perceived factors like status, image, desirability and favourableness (Deephouse & Carter, 2005). This asset is also important as it can positively effect organizations financially as well as improve customer trust and loyalty (Walsh, Beatty & Shiu, 2009; Bontis, Booker & Serenko, 2007). As it is important for
organizations to distinguish themselves from others in a positive fashion and gain and retain loyal customers reputation is a valuable asset that is to be protected in case of a crisis. Another variable that is of importance in the context of online crisis communication is secondary crisis communication. This is defined as consumer intentions to share, forward or react to organization’s crisis communication messages (Schultz, Utz & Göritz, 2011). Where reputation can be seen as the public evaluation of organizational behaviour, secondary crisis communication can be viewed as a way to act upon these attitudes based on how the
organization handles the crisis. It is important to study the intention for this behaviour as the online media environment provides easy access to a large audience. This poses a risk for the organization to lose control over the communication process as consumers can share their message in a negative context which could be harmful for the organization (Utz, Schultz & Glocka, 2013).
5
organization with its public. The main goal of communication in a crisis situation is containing and reducing harm to the organization by spreading information and aiming to meet public expectations (Seeger, 2006; Coombs, 2007; Coombs & Holladay, 2002). Guidelines to provide a strategic approach for organizations to communicate with their public are described in the Situational Crisis Communication Theory (SCCT), the most dominant approach for crisis communication in today’s literature (Coombs, 2007). This model poses ten different crisis response strategies accompanied by guidelines on effective use to optimally protect organizational reputation. Coombs (2007) describes three categories of primary response strategies; deny-, diminish- and rebuild strategies. This research will specifically focus on two crisis response strategies. Firstly (1) the denial strategy where it is stated that there is no crisis and every attribution and responsibility for the crisis is denied and (2) the apology strategy where full responsibility for the crisis is taken and stakeholders are asked to forgive the organization (Coombs, 2007). These specific strategies are chosen as they represent the complete opposite of the spectrum of possible responses and will therefore provide diversity and possibility for comparison in this research. As all crisis response strategies are designed to protect organizational assets, like reputation, both strategies are expected to limit damage to the organization. However a denial strategy is aimed at shaping crisis attribution by
convincing the public that the organization is not responsible for the crisis. Where an apology strategy is a positive reputational strategy that is aimed at strategically rebuilding trust and reputation (Coombs, 2007). An experimental study found that admitting responsibility and apologizing for the crisis was found to have the most positive effect on organizational reputation (Bradford & Garrett, 1995). Therefore it is expected that in case of a crisis reputation evaluation will be more positively influenced by an apology strategy in comparison to a denial strategy (H1). Regarding secondary crisis communication it was found that individuals are more likely to engage in secondary crisis communication when the message they have read was filled with indignation about the incident (Schultz, Utz & Göritz, 2011). It is therefore expected that the apology strategy will to lead to less harmful secondary crisis communication than the denial strategy (H2).
6
Crisis attribution
Attribution of crisis responsibility can be viewed as one of the underlying mechanisms of the relation between crisis response strategies and reputation evaluation. Crisis attribution forms the base of consumer brand evaluation (Klein & Dawar, 2004). In order to form crisis
attributions the public relies on information from the organization which makes crisis communication is indispensable. The attribution theory argues that individuals make attributions when negative and unexpected events, like a crisis, occur (Coombs, 2007). Attribution can therefore be seen as a way in which individuals make sense of a situation by blaming an actor for the negative situation at hand. In this way responsibility is attributed to an actor based on the intentionality of actions which influences feelings and behaviour towards the actor. When the actor is attributed as responsible individuals develop negative feelings and act upon negative images of the actor which intensifies the threat to the
organization (Coombs, 2007; Coombs & Holladay 1996). This means that public perception is highly relevant for the blame that the organization has to endure. Which is related to their image and reputation and influences the actions they can take in order to reduce the threat caused by the crisis.
The Situational Crisis Communication model provides guidelines for which strategies to use in which situation taking the effect of attribution in to account (Coombs, 2007). These guidelines and experimental research show that a denial strategy is stated to be most effective when the level of attribution to the organization is low (Claeys, Cauberghe & Vyncke, 2010). Whereas the use of an apology strategy is effective when the level of attribution is both low and high. Therefore it is expected that the relation between crisis response strategy and reputation evaluation is moderated by crisis attribution (H3).
Media environment
The media environment has changed(Keller, 2009). Where newspapers and news
programmes used to be the most used source of information. Today’s society has embraced social media. Social media encompasses a wide range of online media sources where
7
consumers can gather information in real time and communicate and share information with a wide range of other consumers and organizations (Mangold & Faulds, 2009). Social media evolve around user-manipulated content that facilitate interactive communication and content exchange among public and organizations (Liu et al., 2011). These platforms are viewed as efficient tools for communication as they provide a large audience in today’s society and are viewed as more dialogic, fast and an interactive tool to build public relations (Schultz et al., 2011). These platforms seem to provide many opportunities. However these platforms also pose a threat as it provides interactive communication between consumers and
organizations which leads to a loss of control for organizations because they do not control the context in which the information is shared (Utz, Schultz & Glocka, 2013).
Social media can be of specific importance during an organizational crisis as it can be used as a tool for communicating with the public (Veil, Buehner & Palenchar, 2011). By active users social media is regarded as a credible medium which could improve message acceptance (Schultz et al., 2011; Jin et al., 2014). Next to that social media provide the opportunity for organizations to understand and engage with their the public (Veil, Buehner & Palenchar, 2011). The opportunity to update information at any time to meet public
demands. Lastly social media provide possibilities for organizations to collaborate with other influential sources whom are trusted by the public to ensure consistent information.
Source
One of the possible influential sources an organization can collaborate with are social influencers, defined as third party endorsers that are able to shape public attitudes through sharing information and opinions online on blogs and social networks (Freberg, Graham, McGaughey & Freberg, 2011). They attain this great influence through creating social media content and sharing it with a substantial audience (Zhu et al., 2017). They can be viewed as role models in a specific social group with the ability to influence other individuals within that social group more than others (Langner et al., 2013). This ability can be due to different characteristics such as source expertise, tie strength to the social group and leadership
8
abilities. Social media influencers engage with the public in a digital conversation through which they can influence attitudes, emotions and behaviour of the public (Booth & Matic, 2011). The effectiveness of social influencers as a source in times of organizational crisis will therefore be highlighted in this research.
In the context of a crisis situation, communication to protect organizational assets is most commonly deployed through organizational channels by corporate sources to address the public directly, which is also referred to as ‘official crisis information’ (Jin et al. 2014). However the Social-Mediated Crisis Communication (SMCC) model describes that information about a crisis can also be shared by third parties, defined as any group or individual outside the organization that is involved in the crisis including the public and the media (Jin et al., 2014). The inspiration to look into social influencers as a possible source for organizational crisis communication was found in a case study regarding two food related crises in China. This study found that influential blogger support that was offered to one organization which helped shifting away blame from the organization and repairing their image (Zhu et al., 2017). This study aims to see if this finding can be supported by quantitative research, by researching how this effect relates to corporate crisis
communication. As well as finding out to what extend social influencers are effective as a source for crisis communication in the underlying mechanism of crisis communication. The effectiveness of social influencers compared to corporate sources is dependent on the public evaluation of the source. It is indicated that the public can identify with social influencers as they are similar to their social group and interact closely with them (Katz, 1957; Immink, 2017). Social influencers have the potential to provide great value for organizations that lose part of their credibility during a crisis. As the social influencer is not directly linked to the organization there is no threat to their credibility. Therefore they can step in as role models and navigate the public in making sense of the crisis and forming of attitudes regarding the organization (Jin et al., 2014). In line with this finding the SMCC predicts that social influencers affect crisis outcomes by their persuasive power on social media followers, inactives and traditional mass media. This power has been shown to lead to
9
an increased ability to guide public opinions and emotions within a great group of their followers. (Zhu et al., 2017). This is effective as the influencer provides information and opinions on the crisis that provide leadership and meet the emotional needs of consumers in time of crisis (Jin et al., 2014). Therefore it is expected that reputation will be evaluated more positive when the source of the message is a social-influencer than when the source of the message is corporate (H4).
From the perspective of the two-step flow of communication theory it is stated that the influence of media on the public is not direct but flows through so called opinion leaders (Katz, 1957). Social influencers can be seen as opinion leaders. As opinion leaders are defined as experts in a specific field that interpret messages from mass media and translate this to messages they share with their personal network (Karlsen, 2015). They reach an audience where their message gains added value as these opinion leaders are influential in their social network (Katz, 1957; Feezell, 2017). These sources are valued and trusted regarding the specific field and are therefore expected to have stronger impact on the public opinion. In case of social media the public can search for or be exposed to messages from several opinion leaders from different fields which can influence their attitudes (Karlsen, 2015). Thus, as social influencers can be seen as opinion leaders that have the ability to shape a message to fit and spread their opinion to their personal network on a specific subject (Katz, 1957; Karlsen, 2015; Feezell, 2017) it is expected that crisis response
strategies shared by social influencers will lead to more positive reputation evaluation than strategies shared by a corporate source (H5).
Credibility and expertise
Social influencers have great potential of influencing the publics’ attitudes. Therefore they are expected to provide a positive effect on reputation, however it has been stated that social influencers utilize the most influence when their authority and credibility are both perceived as high (Jin et al., 2014; Immink, 2017).
10
which plays a big role in how the public evaluates the information from this source (Griffn, 1967; Nan, 2009). Source credibility is not found to influence retainment of factual
information, but was found related to the development of attitudes (Hovland & Weiss, 1951). This means that when a source is perceived to be credible it gains an important leadership role in forming attitudes and shaping public opinion through communication. This can in turn affect organizational reputation as this is based on public perception of and attitudes towards the organization(Coombs, 2007; Coombs & Holladay, 2002).On a critical note source credibility can be threatened when the public assumes a connection between the source and a profit making party (Nan, 2009). In this case a relation between the source and the
organization in crisis could be assumed, which would result in less influence on public opinion. Nonetheless the communication of social influencers is all about sharing opinions through leadership roles which is why a positive effect of source credibility on the
effectiveness of the source on organizational reputation is to be expected (Freberg et al., 2011). Corporate sources on the other hand are usually more focussed on sharing factual information which should not lead to a difference in effectiveness due to source credibility (Hovland & Weiss, 1951). However when a corporate communicator would take a leadership role and aim to navigate the public opinion this would still be more effective when the source is evaluated as credible. Therefore it is expected that the positive effect of source on reputation is mediated by source credibility (H6). Credibility also plays a role in affecting the amount of secondary crisis communication. As individuals are more likely to share information from a source that is perceived to be trustworthy (Utz et al., 2013) it is expected that a highly credible source will lead to more secondary crisis communication than a non-credible source (H7).
Another dimension of source credibility, that is treated as a separate variable in this research, is source expertise (Giffin, 1976). This is defined as the extent to which the source of a message is perceived to be capable of making correct assertions, which is evaluated by the source having the relevant skills (Homer & Kahle, 1990; Mun, Yoon, Davis & Lee, 2013). However it is argued that in todays’ online media environment a source’s expertise can also
11
be evaluated by numbers of followers and likes (Flanagin & Metzger, 2017). Either way expertise is evaluated by means of comparable characteristics of a source. It can play a big role in information acceptance and has a positive effect on perceived information quality of online information sources. Source expertise can therefore serve as a guide for individuals to form judgements of the quality of information they are provided with (Mun et al., 2013). Lastly expert sources were found to obtain more persuasive power than non-expert sources
(Homer & Kahle, 1990). It is therefore expected that an expert source will affect reputation evaluation more positive than a non-expert source (H8). A representation of all variables and their hypothesized relations are shown in Figure 1.
Figure 1. Conceptual model.
Method
Sample
The sample for this experimental research consists of 133 respondents (N) that were recruited through the social media platform Facebook by means of a snowball sample. As this research was focussed on online crisis communication the recruitment through
Facebook ensured only participants that use this social media platform. This research was focussed on participants between the age of 18 and 64 and with an average age of 28.50
12
(SD = 9.87) in this study. The biggest groups were found to be the age groups of 23 and 25 years old. Of the sample 60,2 % were found to be female participants. This experimental survey reached participants from 16 nationalities of which Dutch was the biggest group (73,7%) followed by German (7,5%).
Procedure
To test which crisis response strategy is most effective when distributed by a corporate- or social media influencer source, this experiment was set up as a 2 x 3 between subjects factorial design. With Crisis response strategy (denial and apology) and Source (corporate, social media expert, and social media non-expert) as between subjects variables.
To start the experiment the participants were asked to accept the conditions that were described in the informed consent based on the ‘Procedure of ethical testing of research at the department of communication science from the University of Amsterdam’. After which a description of the course of this research was shown to provide guidance. Followed by a description of ‘The BigBeefBurger shop’, the organization in this research. After which the participants were familiarized with the crisis by providing the Facebook message of ‘Sandra van Dijk’, a BigBeefBurger shop customer. This message contained a description of her experience of eating at this burger joint and how she got sick with food poisoning afterwards. A description about how this message spread over social media and the speculation it caused about the quality of food in this restaurant was added. The choice for a food related crisis was based on the case study of Zhu et al. (2017) that argued a positive relation between social influencer help and successful crisis communication in a food related crisis.
Hereafter the participants were randomly assigned to one of the six possible conditions, containing either a denial or apology crisis response strategy by either a
corporate source, a social influencer expert or a non-expert social influencer. The stimuli of these conditions were provided in the form of Facebook messages. Following this
13
intentions for secondary crisis communication were tested ending with some demographic questions about age, gender, nationality and level of education. For this experiment the chosen organization, crisis and all possible source were fictional to prevent the influence of existing attitudes towards any of the parties, and this was explained to the participants at the end of the survey. Transparency of all procedures in this research are provided to ensure replicability and reliability of this research.
Experimental manipulations
The participants were exposed to manipulated messages from three possible sources. The first source was Francien van Kampen the BigBeefBurger shops’ CEO describes as the shop owner that aims to make people enjoy tasty burgers which formed the ‘corporate condition’. The manipulation consisted of one message in the Facebook format as the participants would come across in their real timeline with a name and picture to improve external validity (Figure 1, Appendix). However only the single post was shown excluding the total feed, likes and reactions to ensure that participants were not influenced by other factors than the manipulated message, which could limit external validity.
The second source was Lianne Verhoek, the social influencer expert condition, described as a social influencer with a passion for food. Added was the description that Lianne responds to food related crisis more often and that her opinion is often highly valued and trusted by the public, to emphasize the manipulation of influencer expertise in this crisis. In this condition a biography was provided to ensure that the participants were well aware of the social influencers impact and goal (Figure 2, Appendix). The biography of Lianne
Verhoek provided her profile name ‘HappyandHealthy’ and her 78.800 Facebook likes and 77.900 Facebook followers. Added were follower statistics for other social media channels like 16.600 followers on YouTube and 51.500 followers on Instagram. Lastly this biography included a little description of her goal; to provide information how people can make great food at home and indispensable restaurant hotspots. Lastly this manipulation provided the Facebook message with her name and picture that participants would be able to come
14
across in a real life situation, without likes or comments to prevent this from influencing their opinions.
The third and last source condition was formed by Sofia Veltman, in this case the social influencer that is not an expert in the field of food and restaurants. She is described as a social influencer with a passion for beauty and the biography provided shows her like and follower statistics that are exactly the same as in the social influencer expert condition described above. Only her profile name ‘Shine from the Inside’ and goal description differs from the expert condition. In this description it is stated that her profile is a place for beauty lovers on which she aims to show how inner beauty can be shown by using make-up, together with beauty product reviews. The manipulation was completed with exposure to the Facebook message from this source with her name and picture and without any likes or comments.
The second variable that was manipulated in this research was the crisis response strategy that was used to form the content of the message that the participants were exposed to. Two crisis response strategies were used which the participants could be randomly exposed to from either one of the three sources. On one hand the denial strategy messages contained statements like ‘A customer got sick which is very unfortunate’ and ‘The BigBeefBurger shop has nothing to do with this as we get all our products from safe and certified distributers’ for the Corporate condition. The denial response messages in both Social influencer conditions were identical except for the closing sentence and contained statements like ‘As you might have heard a customer of The BigBeefBurger shop felt sick with food poisoning’, ‘I do not believe this is caused by a burger from The BigBeefBurger shop’ and ‘I had a burger there often and never felt sick’. In the apology response message the content of the corporate condition contained statements like ‘It is very sad to hear a customer felt sick with food poisoning’, ‘We take full responsibility for the incident’, ‘We have contacted the customer and apologized personally’ and ‘We are reviewing all our products to ensure this incident never occurs again’. The apology response messages in the Social influencer conditions were identical except for the closing sentence and contained
15
statements like ‘It is very sad to hear a customer felt sick with food poisoning’, ‘However they have apologized to the customer personally and are reviewing all their products’ and ‘I still support The BigBeefBurger shop in their process and would still recommend their products’. The full descriptions of all the response messages for each condition can be found in Table 1 (Appendix).
A pre-test was conducted to test manipulations of strategy, source and expertise and assure internal validity in a sample of 24 participants (N). The manipulation of crisis response strategy was found to be successful with a 100% correct identification score assessed on a question with two answer options corresponding with the two strategies (λ= 1,00 p < 0,001). Secondly the manipulation of source was also found to be successful (λ= 0.92, p< 0,001). As both the corporate and social influencer expert condition were successfully identified in 100% of the cases and the non-expert social influencer were identified correctly in 87,5% of the cases, when assessed with a question with three answer options corresponding with the three source categories. Only the manipulation of source expertise, assessed with the question ‘do you feel like the source is an expert regarding this crisis providing the answer options yes and no, was not found to be successful (λ= 0.00, p= 1,00). As 87,5% of non-expert sources and only 50% of non-expert sources was identified correctly. Therefore source expertise was emphasized more clearly in the experiment which was controlled for with a second manipulation check. This second manipulation was found to be successful as the expertise level in the social influencer expert condition was evaluated significantly higher than the expertise level of the non-expert social influencer source (Mdifference = 0.94, p = 0.009).
Operationalization
This experimental questionnaire measured crisis attribution, source credibility, source expertise, organizational reputation and intention for secondary crisis communication for which existing scales were used to increase internal validity. Firstly blame for the crisis, referred to as crisis attribution, was measured on a three item scale from Klein and Dawar
16
(2004) which was indicated to be a reliable scale (α = 0.86) and is therefore used in this research to measure crisis attribution. In this scale the three items of responsibility,
accountability and whether or not the organization was perceived to be at fault determined the level of crisis attribution. These items were measured by asking for agreement to the following statements: ‘The BigBeefBurger shop is responsible for the customers illness’, ‘The BigBeefBurger shop should be held accountable for the consumers illness’ and ‘The incident of the ill consumer is the fault of The BigBeefBurger shop’ on a 7-point Likert scale ((1) ‘Strongly Disagree’ to (7) ‘Strongly Agree’). This scale was found to be reliable (α = 0.90) and a new construct for crisis attribution was constructed (M = 4.31, SD = 1.34), found to explain 83,1% of variance in the sample. Secondly, source credibility was measured on a scale where believability, accuracy, trustworthiness, bias and completeness of information were identified as determining factors (Metzger et al., 2003). For these factors agreement to the following statement was assessed: ‘The sender of this Facebook message is believable’ , where believability was changed for the other source credibility terms in order to create five statements measured on a 7-point Likert scale ((1) ‘Strongly Disagree’ to (7) ‘Strongly Agree’). This scale was found to be reliable (α = 0.82) with the exclusion of the item
regarding source bias, as this was mentioned to be misunderstood by multiple participants. Therefore the other four items were formed into the new construct of source credibility (M = 3.85, SD = 1.15) found to explain 52,2% of variance in the sample. Thirdly source expertise was assessed as a manipulation check on a 7-point Likert scale ((1) ‘Strongly Disagree’ to (7) ‘Strongly Agree’).on agreement to the statement that ‘The sender of the response message is an expert regarding the crisis’. Resulting in the source expertise measure (M = 2.95, SD = 1.60)
The dependent variable organizational reputation was measured on a five item scale as described by Coombs and Holladay (2002). This scale assessed the agreement on the following five statements ‘The organization is concerned with the wellbeing of its public’, ‘ The organization is basically dishonest’, ‘I do not trust the organization to tell the truth about the incident’, ‘Under most circumstances I would be likely to believe what the organization
17
says’ and ‘The organization is not concerned with the wellbeing of its public’, measured on a 7- point Likert scale ((1) ‘Strongly Disagree’ to (7) ‘Strongly Agree’).This scale was used as it was indicated to be a reliable scale (α = 0.87) to measure organizational reputation in the context of crisis. This scale was found to be reliable (α = 0.82). Therefore the five items were constructed into the reputation evaluation measure (M = 4.74, SD = 1.04), found to explain 58,0% of variation in the sample. Lastly the intention to participate in secondary crisis communication was assessed by three items (Schultz et al., 2011). Participants were asked how likely they were to ‘Share the message with other people’, ‘Tell their friends about the incident’ and ‘Leave a reaction to the message’, assessed on a 5-point Likert scale ((1) ‘Very Unlikely to (5) ‘Very Likely’). This scale was not found to be reliable (α = 0.66) and the items should therefore be handled as separate measures of sharing (M = 1.71, SD = 1.00), telling friends (M = 2.66, SD = 1.27) and leaving a reaction (M = 1.41, SD = 0.71). Exclusion of items would not increase reliability. As all items are intended to measure information sharing behaviour the choice has been made to use item one to represent secondary crisis
communication. This is the most fitting item to represent secondary crisis communication in this context as it focusses on sharing the message from the organization with other people which links closely to the ease of sharing on social platforms and the threat for organizations that their information is shared in another context.
Analysis plan
The relation between the independent variable crisis response strategy and dependent variable reputation will be tested by means of an independent t-test, as the strategy variable is measured dichotomously and reputation evaluation as an interval variable. The possible moderation of crisis attribution on this relation will be tested hereafter using a multiple
regression analysis, which is fitting because crisis attribution and reputation are measured at interval level and strategy is dichotomous. Next the relation between the independent
dichotomous variable crisis response strategy and the independent interval variable
18
the relation between source as a dichotomous variable and interval variable reputation evaluation will be tested by using an independent t-test. After this the possible moderation of source expertise as an interval variable on this relation will be tested by means of a one-way ANOVA analysis. This is the fitting analysis as the independent variable source is in this case categorical. The moderation of source as a dichotomous variable on the relation between strategy and reputation will be tested with a multiple variance analysis. Next in order to ensure the possibility of a mediation analysis firstly the relation between source as a categorical variable and reputation. Secondly source and credibility as an interval variable are tested with single regression analyses. Analysis will be finished with a single regression analysis to test if source credibility would affect the intention for secondary crisis
communication.
Results
Hypothesis 1 states that reputation evaluation will be more positive after exposure to an apology strategy in comparison to exposure of a denial strategy. An independent sample t-test was conducted to t-test which strategy would have the most positive influence on
reputation evaluation. As both groups contain more than 30 observations (Denial n = 64 and Apology n = 69) and the data is normally distributed the assumptions for this analysis were met. The results indicate that the reputation evaluation is more positive when a message included an apology strategy (M = 5.03 , SD = 0.98) in comparison to messages including a denial strategy (M = 4.42, SD =1.02). This difference was found to be significant but small (t (131) = -3.58, p < 0.001, 95% CI = [-0.96, -0.28], d = 0.008 ), assuming variances are equal in both groups (F (131, 129.30) = 0.01, p = 0.939). This means that there is a significant difference between the reputation evaluation after exposure to the crisis response strategies of which apology strategy was found to have the most positive effect. Therefore hypothesis one can be supported.
19
attribution (H3). To test this moderation of crisis attribution a multiple regression analysis was conducted. Before the analysis the attribution variable was standardized and an interaction term was constructed. Before performing the analysis it is important to note that the following assumptions were met: continuity of reputation as the dependent variable, linearity,
homoscedasticity and the normal distribution of residuals. The assumption of multicollinearity was not met as the VIF scores of the standardized variable and interaction term were above 10 (10.85 and 10.65) proving there is multicollinearity in this regression. However as the VIF scores are only slightly higher than 10 the results will still be interpreted, with caution. The regression model that tests the effect of crisis response strategy and attribution on reputation evaluation was found to be significant (F (3,132) = 11.06, p < 0.001). This model shows that 20,5% of variance in reputation evaluation can be explained by crisis response strategy and reputation (R2 = 0.21). The results confirm the direct effect of crisis response strategy on reputation evaluation (b = 0.80, t = 4.65, p < 0.001, 95% CI [0.46, 1.13]) (b* = 0.38). The direct effect of attribution on reputation evaluation was also found to be significant (b = - 0.97, t = -3.62, p < 0.001, 95% CI [-1.0, -0.44]). The beta value shows that this effect is negative (b* = -0.93) which means that a higher level of crisis attribution effects reputation evaluation negatively. Lastly also the interaction term (strategy x attribution) was found to
have a significant effect on reputation evaluation (b = 0.46, t = 2.68, p = 0.008, 95% CI [ 0.12, 0.80]). This effect was found to be positive (b* =0.69), which means that attribution moderates the relation between crisis
1 2 3 4 5 6 Denial Apology R eput at ion Eva lua ti on Low Attribution High attribution
Figure 2. Moderation effect of attribution on the relation between crisis response strategy and reputation evaluation.
20
response strategy and reputation evaluation. Figure 2 shows that both crisis response strategies have greater positive effect on reputation evaluation when crisis attribution is low. Where low crisis attribution is defined as one standard deviation below average and high crisis attribution is defined as one standard deviation above average (M = 4.31, SD = 1.34). If however a denial strategy is used reputation evaluation is clearly higher in low crisis
attribution situations than when crisis attribution is evaluated as low. This means crisis attribution moderates the relation between crisis response strategy and reputation evaluation and hypothesis 3 can be supported.
To test if the apology strategy is related to lower intentions of harmful secondary crisis communication than a denial strategy (H2) an independent sample t-test was
conducted. As both conditional groups that are compared contain more than 30 observations (Denial n = 64 and Apology n = 69) and the data is normally distributed all assumptions for analysis are met. The results regarding the item ‘sharing the message’ indicates that
individuals that read a message including a denial strategy (M = 1.67, SD = 1.06) are slightly less likely to share the message with other people than individuals who read a message including an apology strategy (M = 1.75, SD = 0.95). However it is important to note that in both groups the likelihood of sharing the message has been found to be between ‘very unlikely’ and ‘unlikely’. Under the assumption that variances are equal in both groups (F (131, 126.70) = 0.54, p = 0.463) this effect has not been found to be significant (t (131) = -0.47, p = 0.638, 95% CI = [-0.43, -0.26]). As the relationship between strategy and secondary crisis communication was not found to be significant hypothesis 2 is rejected.
To test whether a social influencer source provides a more positive effect on reputation evaluation than a corporate source (H4) an independent sample t-test was executed. As both groups have more than 30 observations (Corporate source n = 41 and Social influencer source n = 92) and the data is normally distributed the assumptions for this analysis were met. The results show that the average reputation evaluation of individuals exposed to a message from a corporate source is slightly lower (M = 4.70, SD = 1.30) than the average reputation evaluation from individuals exposed to a message from a social
21
influencer source (M = 4.75, SD = 0.91). However this result has not been found to be significant (t (58.29) = -0.25, p = 0.800, 95% CI = [-0.50, -0.39]), under the assumption that variances of both groups were not equal (F (131, 58.29) = 9.73, p = 0.002). This means that hypothesis 4 is rejected as source has not been found to have a significant effect on
reputation evaluation. Source expertise was expected to affect reputation evaluation more positive when evaluated as high, as compared to a low expertise level (H8). Before testing this hypothesis a one-way ANOVA analysis was conducted to find out how the level of expertise from the social influencer sources compare to the expertise of a corporate source (Corporate source n = 41, Social influencer non expert source n = 43 and Social influencer expert source n = 49). The results show that expertise of a corporate source is evaluated as highest in expertise (M = 3.66,SD = 1.74), followed by the social influencer expertise source (M = 3.08,SD = 1.44) and the social influencer non expert source was evaluated as having the lowest level of expertise (M = 2.14,SD = 1.28). The assumption of equal variances was not met (F (2, 130) = 5.68, p = 0.004) but as the group sizes are similar the results will be interpreted. These differences were found to be significant (F (2, 130) = 11.13, p < 0.001). The post- hoc test provides the insight that a corporate source is found to be significantly higher in expertise than a social influencer non-expert source (Mdifference = 1.52, p < 0.001), but not differ significantly from the social influencer expert source (Mdifference = 0.58, p = 0.211). Also the social influencer expert source was found to have significantly higher expertise evaluation than a non-expert social influencer (Mdifference = 0.94, p = 0.009). To test hypothesis 8 a one-way ANOVA analysis was conducted with all three source variables of which the expertise was just determined. The reputation evaluation after exposure to a message from a social influencer expert source was most positive (M = 4.78 , SD = 0.95), followed by exposure to a non-expert social influencer source (M = 4.73 , SD = 0.87) and the least positive was the reputation evaluation after exposure to a message from a corporate source (M = 4.70 , SD = 1.30), The assumption of equal variances was not met (F (2, 130) = 4.88, p = 0.009) but group size was similar and therefore results will be interpreted. These group averages were however not found to be significantly different (F (2, 130) = 0.07, p =
22
0.930). These findings show that the source identified as highest in expertise (corporate) scores lowest in reputation when compared to the other sources. This shows that a high level of expertise is not related to a good reputation evaluation and the differences based on both source and expertise were not found to be significant, which means hypothesis 8 needs to be rejected.
To test whether the source of the message moderates the effect of crisis response strategy on reputation evaluation (H5) a multiple variance analysis was conducted. Before executing the analysis the following assumptions were controlled for: the dependent variable, reputation, is measured at interval level and the measures for both groups can be viewed as independent measures. As both groups are not equal in size (Corporate n = 41 and Social influencer n = 92) equality of variances was checked. The Levene’s test showed that the variances in both groups are not equal (F (3, 129) = 3.80, p = 0.012), however the results of this analysis will still be interpreted but will be handled and stated with caution. The results show that there is a significant weak but direct effect between crisis response strategies and reputation evaluation (F (1, 129) = 20.17, p < 0.001, η2 = 0.13), but no direct effect of source was confirmed (F (1, 129) = 0.03, p = 0.866). The results also show that the interaction term of strategy and source has significant weak effect on reputation evaluation (F (1, 129) = 8.44, p = 0.004, η2 = 0.06). The post-hoc multiple comparisons test provides the insight that when a denial strategy is used the reputation is more positively evaluated when the message came from a social influencer source (M = 4.60, SD = 0.80) than from a corporate source (M = 4.04, SD = 1.30). This difference was found to be significant (Mdifference = -0.56, p = 0.032). The apology strategy was found to have a more positive effect when exposed to by a
corporate source (M = 5.39, SD = 0.87) compared to a social influence source (M = 4.89, SD = 0.99). However this difference was not found to be significant (Mdifference = 0.50, p = 0.055). This means that source is found to moderate the effect of crisis response strategy on
reputation evaluation. This shows that response strategies can result in higher reputation evaluation when shared by a social influencer and hypothesis 5 is therefore supported. As described in the section regarding hypothesis 4 there is no direct relation found
23
between the source of the message and reputation evaluation tested in an independent sample t-test. Nonetheless the expectation for mediation of source credibility on this relation (H6) is still tested. To start the analysis the relation between source and reputation was re-evaluated using the new source variable (Corporate n = 41, Non-expert social influencer n = 43 and Expert social influencer n = 49) in a single regression analysis to ensure possibility of executing a mediation analysis. The results show that the assumptions of continuity of reputation, linearity, multicollinearity and normal distribution of residuals were met. The assumption of homoscedasticity was not met as source is a categorical variable. The regression model was not found to be significant (F (1,132) = 0.14, p = 0.708) (R2 = 0.001), and neither was the effect of source on reputation (b =0.04, t = 0.38, p = 0.708, 95% CI [-0.18, 0.26], b* = 0.03). Secondly the effect of source on source credibility was tested with a single regression analysis. The assumptions of continuity, linearity, multicollinearity and normal distribution of residuals were met. Only the assumption of homoscedasticity was not met as the source is a categorical variable. The regression model was not found to be significant (F (1,132) = 0.07, p = 0.786) (R2 = 0.001). The effect of source on source credibility was also not found to be significant (b =0.03, t = 0.27, p = 0.786, 95% CI [-0.21, 0.27], b* = 0.02). As the effect of source on reputation and source credibility were not found to be significant the chances of finding a mediation effect are nearly zero. Therefore the mediation analysis is not conducted and hypothesis 6 is rejected.
Lastly the expectations that a highly credible source would lead to more secondary crisis communication than a non-credible source (H7) was tested using a single regression analysis with item one ‘sharing the organizational crisis communication message’. This test was chosen as both dependent and independent values were measured at interval level. Firstly the effect of source credibility was measured after controlling for the following
assumptions. The relation was not found to be linear, the assumption of homoscedacity was met, no multicollinearity was found and the residuals were not found to be normally
distributed. As not all assumptions were met the analysis was not found reliable on the base on these criteria. The findings show that this regression model is not found to be significant
24
(F (1, 132) = 0.52, p = 0.474) and that based on this model only 0,4% of variance in secondary crisis communication can be explained by source credibility (R2 = 0.004). The results confirm that there is no relation found between source credibility and intentions regarding secondary crisis communication (b = 0.05, t = 0.72, p = 0.474, 95% CI [-0.10, 0.20]) (b* = 0.06) which means hypothesis 7 is rejected.
Conclusion and Discussion
The aim of this study was uncovering what factors influence organizational reputation in crisis situations. Specifically how an organization can influence the evaluation of reputation from the public through online crisis communication and what role social influencers can have in this strategic process.
This study found that the crisis response strategy that is used during crisis
communication, between the organization and public, is related to the reputation evaluation of the organization. The use of both crisis response strategies was found to be related to a positive reputation evaluation. However using an apology strategy resulted in the most positive effect on reputation evaluation in this online food related crisis. This finding is in line with the ‘Situational Crisis Communication Theory’ (Coombs, 2007). This theory states that the crisis response strategies are a strategic approach towards communication with the public to minimize the negative effects of a crisis and protect organizational reputation. The current study also showed that the level to which the organization is blamed for the crisis moderates the relation of crisis response strategy and reputation evaluation. Providing the insight that reputation evaluation is in all cases more positive when the level of blame the organization receives from their public is low. In case of using a denial strategy specifically, reputation evaluation was found strongly more positive when crisis attribution was evaluated as low compared to a high attribution situation. This confirms that attribution is an important factor for strategic crisis communication as it forms the base of customer brand evaluation and confirms that a denial strategy has the most positive effect when crisis attribution is
25
evaluated as low (Klein & Dawar, 2004; Coombs, 2007).
Lastly the source (corporate- or social influencer) of the message was not found to have a direct effect on reputation evaluation, but to moderate the relation that was found between crisis response strategies and reputation evaluation. This confirms the influence of opinion leaders as described in the two-step flow theory (Katz, 1957; Karlsen, 2015; Feezell, 2017). This finding has to be handled with caution, but indicates that a crisis response message in which an organization aims to deny responsibility for the crisis provides the most positive effect when spread by a social influencer. Where a crisis response message
containing organizational acknowledgement and apologies can best be spread by a corporate source. In practice this means that organizations can consider using social influencers as a vehicle to spread a denial message. Since this was found to be the best approach to ensure a positive effect on reputation evaluation. Whereas the positive
associations of an apology strategy provides the most reputational benefits when shared by the organization itself. This could be the case as the denial of organizational responsibility has more distance to the organization when spread by a social influencer. Additionally, social influencers are seen as role models. Seeing them deal with the denial of responsibility from the organization in a positive matter could lead the public to follow this behaviour.
Supposedly this suggests that using social media to communicate with the public can, when the source is chosen appropriately, result in successful protection of organizational
reputation in times of crisis. It could be of great importance for crisis communication managers to be aware of this potential as it is overall challenging to limit harm to the organization by denying crisis responsibility.
However this study did not find confirmation for all relations as described in literature. Firstly the credibility of the source and used crisis response strategy were not found to be related to secondary crisis communication intentions. This could be explained by different factors. One reason might be that the level of indignation found in the messages was not clear or sufficient enough to activate the respondents to start participating in information sharing behaviour. This lack in effect of source credibility could also be explained by the fact
26
that all sources in this experiment used the same medium to spread their message which could explain the similar source credibility scores. The last reason could be that people were aware of this study not being a real life situation and therefore did not feel the urge or
anonymity to activate sharing behaviour. Secondly source was not found to have a direct effect on reputation evaluation and expertise was not found to affect this relation. This could be the case as there was no control over the extent to which the people were able to identify with the social influencer. This might have explained more about this relation as identification of the public is one of the biggest assets and forms of power for social influencers. Another aspect that might be of influence is that a social influencer might not be evaluated as a true ‘expert’ source as it is still regarding the opinion of an outsider to the organization. Lastly there was no mediation effect found of source credibility on the relation between source of the message and reputation evaluation. This could be explained by the idea that if the organization loses credibility due to the crisis and the social influencer would overall be evaluated as less credible. This would lead to both sources not being perceived as very credible and similar credibility evaluations would not result in different outcomes with regard to reputation evaluation.
On the basis of these findings the research question ‘What role do social influencers play in the relation between crisis response strategy and organizational reputation and to what extent is this effect dependent on source credibility, -expertise and attribution to the organization?’ can be answered by stating that social influencers could play a role as moderator between crisis response strategy and organizational reputation evaluation alongside the moderating effect of crisis attribution. This means that the content of a crisis response message is evaluated differently based on the source. Since the rise of social influencers the media environment has become more complex. As there are multiple sources providing crisis communication. This could arguably be a negative development with regard to crisis communication as this could lead to confusion of the public on what source to believe. On the other hand it could provide an opportunity for the public to gather information from different sources, from which they can choose a source they trust. The latter could be a
27
positive development for organizations as long as they manage consistency in communication amongst all sources.
To be able to state the potential of social influencers in this context as a fact it is important that more research is done regarding the role of social influencers as assets in times of organizational crisis. Especially as this study was done with a small sample and a specific crisis description. Therefore it is important to test if these findings hold up in a big sample, in different crisis situations and provide insight in to the possible generalization of these effects for the other crisis response strategies. Possibilities for further research would include manipulation of the full social network feed to mimic a real life situation more
precisely and evaluate if the findings would be different. Future research could also explore the full potential of social influencers and research their effectiveness across different social networks or study the effects of social influencers using video’s to spread the message to their public in the form of so called vlogs that are very popular nowadays.
Concluding it can be stated that there is much social influencer potential to explore in future research. This study indicates that social influencers could be used strategically by organizations to improve their crisis communication strategy, as social influencers can increase the protection of reputation during times of organizational crisis. This study has broadened the knowledge of social influencer impact and underlines the relevance for crisis communication managers to be aware of social influencer potential. This potential could be considered an asset to be included in the crisis communication strategy to protect
28
Literature
Claeys, A. S., Cauberghe, V., & Vyncke, P. (2010). Restoring reputations in times of crisis: An experimental study of the Situational Crisis Communication Theory and the moderating effects of locus of control. Public Relations Review, 36(3), 256-262. Coombs, W. T. (2007). Protecting organization reputations during a crisis: The development and application of situational crisis communication theory. Corporate Reputation Review, 10(3), 163-176.
Coombs, W. T. (2015). The value of communication during a crisis: Insights from strategic communication research. Business Horizons, 58(2), 141-148.
Coombs, W. T., & Holladay, S. J. (1996). Communication and attributions in a crisis: An experimental study in crisis communication. Journal of Public Relations Research, 8(4), 279-295.
Coombs, W. T., & Holladay, S. J. (2002). Helping crisis managers protect reputational assets: Initial tests of the situational crisis communication theory. Management Communication Quarterly, 16(2), 165-186.
Bontis, N., Booker, L. D., & Serenko, A. (2007). The mediating effect of organizational reputation on customer loyalty and service recommendation in the banking industry. Management Decision, 45(9), 1426-1445.
Booth, N., & Matic, J. A. (2011). Mapping and leveraging influencers in social media to shape corporate brand perceptions. Corporate Communications: An International Journal, 16(3), 184-191.
Bradford, J. L., & Garrett, D. E. (1995). The effectiveness of corporate communicative responses to accusations of unethical behavior. Journal of Business Ethics, 14(11), 875-892.
Deephouse, D. L., & Carter, S. M. (2005). An examination of differences between organizational legitimacy and organizational reputation. Journal of management Studies, 42(2), 329-360.
29
Feezell, J. T. (2017). Agenda setting through social media: The importance of incidental news exposure and social filtering in the digital era. Political Research Quarterly, 1- 13. doi: 10.1177/1065912917744895
Flanagin, A. J., & Metzger, M. J. (2017). Digital media and perceptions of source credibility in political communication. The Oxford Handbook of Political Communication. doi: 10.1093/9780199793471.013.65
Freberg, K., Graham, K., McGaughey, K., & Freberg, L. A. (2011). Who are the social media influencers? A study of public perceptions of personality. Public Relations Review, 37(1), 90-92.
Giffin, K. (1967). The contribution of studies of source credibility to a theory of interpersonal trust in the communication process. Psychological Bulletin, 68(2), 104-120.
Homer, P. M., & Kahle, L. R. (1990). Source expertise, time of source identification, and involvement in persuasion: An elaborative processing perspective. Journal of Advertising, 19(1), 30-39.
Hovland, C. I., & Weiss, W. (1951). The influence of source credibility on communication effectiveness. Public Opinion Quarterly, 15(4), 635-650.
Immink, M. (2017, April 1st). Influencer marketing onder de loep: van geld tot geloofwaardigheid. Consulted on December 8th 2017, from
https://www.frankwatching.com/archive/2017/04/01/influencer-marketing-onder-de- loep-van-geld-tot-geloofwaardigheid/
Jin, Y., Liu, B. F., & Austin, L. L. (2014). Examining the role of social media in effective crisis management: The effects of crisis origin, information form, and source on publics’ crisis responses. Communication Research, 41(1), 74-94.
Karlsen, R. (2015). Followers are opinion leaders: The role of people in the flow of political communication on and beyond social networking sites. European Journal of
Communication, 30(3), 301-318.
Katz, E. (1957). The two-step flow of communication: An up-to-date report on an hypothesis. Public Opinion Quarterly, 21(1), 61-78
30
Keller, K. L. (2009). Building strong brands in a modern marketing communications environment. Journal of Marketing Communications, 15(2-3), 139-155.
Klein, J., & Dawar, N. (2004). Corporate social responsibility and consumers' attributions and brand evaluations in a product–harm crisis. International Journal of Research in Marketing, 21(3), 203-217.
Langner, S., Hennigs, N., & Wiedmann, K. P. (2013). Social persuasion: targeting social identities through social influencers. Journal of Consumer Marketing, 30(1), 31-49. Liu, B. F., Austin, L., & Jin, Y. (2011). How publics respond to crisis communication strategies: The interplay of information form and source. Public Relations Review, 37(4), 345-353.
Metzger, M. J., Flanagin, A. J., & Zwarun, L. (2003). College student web use, perceptions of information credibility, and verification behaviour. Computers & Education, 41(3), 271-290.
Mangold, W. G., & Faulds, D. J. (2009). Social media: The new hybrid element of the promotion mix. Business Horizons, 52(4), 357-365.
Mun, Y. Y., Yoon, J. J., Davis, J. M., & Lee, T. (2013). Untangling the antecedents of initial trust in web-based health information: The roles of argument quality, source
expertise, and user perceptions of information quality and risk. Decision Support Systems, 55(1), 284-295.
Nan, X. (2009). The influence of source credibility on attitude certainty: Exploring the
moderating effects of timing of source identification and individual need for cognition. Psychology & Marketing, 26(4), 321-332.
Reynolds, B., & Seeger M. W. (2005). Crisis and emergency risk communication as an integrative model. Journal of Health Communication, 10(1), 43-55.
Seeger, M. W. (2006). Best practices in crisis communication: An expert panel process. Journal of Applied Communication Research, 34(3), 232-244.
Schultz, F., Utz, S., & Göritz, A. (2011). Is the medium the message? Perceptions of and reactions to crisis communication via twitter, blogs and traditional media. Public
31
Relations Review, 37(1), 20-27.
Utz, S., Schultz, F., & Glocka, S. (2013). Crisis communication online: How medium, crisis type and emotions affected public reactions in the Fukushima Daiichi nuclear disaster. Public Relations Review, 39(1), 40-46.
Van der Veer, N., Boekee, S., & Peters, O. (2017, January 23rd ). Nationale social media onderzoek 2017. Consulted on December 11th 2017, from
https://www.marketingonline.nl/sites/default/files/Newcom_- _Nationale_Social_Media_Onderzoek_2017.pdf
Veil, S. R., Buehner, T., & Palenchar, M. J. (2011). A work‐in‐process literature review: Incorporating social media in risk and crisis communication. Journal of Contingencies and Crisis Management, 19(2), 110-122.
Walsh, G., Beatty, S. E., & Shiu, E. M. (2009). The customer-based corporate reputation scale: replication and short form. Journal of Business Research, 62(10), 924-930. Yang, S. U., Kang, M., & Johnson, P. (2010). Effects of narratives, openness to dialogic communication, and credibility on engagement in crisis communication through organizational blogs. Communication Research, 37(4), 473-497.
Zhu, L., Anagondahalli, D., & Zhang, A. (2017). Social media and culture in crisis
communication: McDonald’s and KFC crises management in China. Public Relations Review, 43(3), 487-492.
32
Appendix
Corporate Source Social Influencer Expert Source Social Influencer Non-expert Source Denial Strategy As you might haveheard a customer felt sick with food poisoning, which is very unfortunate. However the
BigBeefBurger shop has nothing to do with this food poisoning as we get all the products to make our burgers from safe and certified distributers.
CEO Francien van Kampen
As you might have heard a customer from The
BigBeefBurger shop felt sick with food poisoning which is very unfortunate. However I do not believe this is caused by a burger from The BigBeefBurger shop as they only use safe and certified products to make their burgers. Also I have had a burger there often ad never felt sick.
Stay Happy Healthy, Lianne
As you might have heard a customer from The
BigBeefBurger shop felt sick with food poisoning which is very unfortunate. However I do not believe this is caused by a burger from The BigBeefBurger shop as they only use safe and certified products to make their burgers. Also I have had a burger there often ad never felt sick.
Keep Shining, Sofia
Apology Strategy It is very sad to hear that a customer felt sick with food poisoning after eating a burger in our restaurant. The
It is very sad to hear that a customer of The BigBeefBurger shop felt sick with food poisoning after eating a burger at
It is very sad to hear that a customer of The BigBeefBurger shop felt sick with food poisoning after eating a burger at
33
BigBeefBurger shop takes full
responsibility for this incident. By means of setting things straight we have contacted the customer and apologized personally. Next to that we are
reviewing all our products to ensure this incident never occurs again.
CEO Francien can Kampen
the restaurant. However they have apologized to the customer personally and are reviewing all their products. Therefore I still support The
BigBeefBurger shop in their process and would still
recommend their products.
Stay Happy Healthy, Lianne
the restaurant. However they have apologized to the customer personally and are reviewing all their products. Therefore I still support The
BigBeefBurger shop in their process and would still
recommend their products.
Keep Shining, Sofia
34
Figure 1. Example of manipulated Facebook message (Denial x corporate condition).
