Introduction

Social media has empowered individuals to actively disseminate information and share their opinion with broad audiences. This has enabled users to not just be passive receivers of information, but potentially influential information disseminators that can profoundly influence public opinion about a product or a brand (Zheng et al. 2020). When this information is negative or scandalous the impact upon a brand’s reputation can be significant (Hansen et al. 2018). The increasing use of social media and the pace that negative sentiment can spread means a focussed approach to understand how scandals can be shared and managed on social media is necessary.

We already know how influential social media users can benefit corporates through brand endorsement (Dhanesh and Duthler 2019; Silva et al. 2020). These social media influencers play a significant role in promoting brands and corporates’ relationship with the public (Dhanesh and Duthler 2019; Lin et al. 2018; Lu and Seah 2018; Ramadanty and Safitri 2019). Yet, less is known about how social media influential users can spread information regarding a corporate scandal and what impact the positive and negative information can have on brand reputation (Ao and Mak 2021). This research seeks to close this knowledge gap by examining the spread of scandals by influential social media users. We offer insights into how influential social media users drive the spread of a corporate scandal and how their behaviour may change depending on the context of the scandal and the nature of the organisation itself. That is, we explore potential differences if the scandal involves a value failure or a performance-based failure, as well as whether the scandal revolves around a for-profit or not-for-profit organisation. We also show the impact that an initial scandal being made public can lead to future scandals receiving significantly greater attention and becoming seen as a series of scandals.

Scandals can have wide-reaching consequences, such as decreased profits (An et al. 2018), damaged reputations (Coombs 2007), market share loss (Zheng et al. 2018), and may involve widespread victims who are emotionally affected by the incident and whose trust is violated (Bowen et al. 2018; Marcus and Goodman 1991). Scandals and brand attacks can spread within hours (Rauschnabel et al. 2016) and this spread is often outside the organisation’s control (Zhao et al. 2013). Despite the growing body of research on online scandals, their differential implications on for-profit and not-for-profit organisations are not yet fully understood (Adbi 2022). Because for-profit and non-profit are the two most common organisational forms utilised by researchers (Adbi 2022), it is important to expand the knowledge of whether these two types of organisation in scandal can attenuate or deteriorate the situation (Adbi 2022; Hornsey et al. 2021) and how influential social media users react to scandals associated with these two types of organisations. A review of the recent literature shows that little research has examined how influential social media users and their discourse impact online scandals associated with for-profit versus not-for-profit organisations. There is still a lack of understating of how influential users engage in online brand attacks and affect the users’ opinions depending on the nature of the organisation and types of scandals. Understanding how influential social media users react to scandals and how these reactions influence the transmission of information help managers to better respond to scandals (Li et al. 2020). This research seeks to close this knowledge gap by examining the spread of scandals by social media users and the manner in which organisations look to respond (if at all) to information regarding a scandal. We tested the way influential users and their approach towards that incident (reporting the incident with a neutral tone, attacking the brand by using e.g. anger-related words, or supporting the brand) affect the number of reposts following brand scandals.

Few studies have described how influential social media users may influence the public response to not-for-profit versus for-profit scandals based on real-time social media data. However, given the prominent role of social media users in influencing other users’ opinions and the different approaches they may have towards different types of scandals and organisations in scandals, it is important to understand how social media users can use their networks and influence to affect the spread of scandals. Understanding how influential social media users react to the scandals and how these reactions influence the transmission of information help the managers to better respond to the scandals and protect the brands reputations (Li et al. 2020). Therefore, this study fills these research gaps by answering the following questions:

  1. 1.

    How do influential social media users impact the spread of scandals via social media?

  2. 2.

    How do influential social media users’ participation and engagement approach in online scandal discussions vary based on the type of organisation in the scandal and type of scandal?

This study has several notable contributions. First, it links the concept of the spread of crisis information to the scandal typology based on performance and value and types of organisations to better explain online users’ sharing (retweeting) behaviour in the context of scandals. Second, this study contributes to the crisis communication literature by shedding light on the differences between for-profit and not-for-profit in online scandals. It shows that influential social media users are inclined to react differently to a brand crisis associated with for-profit versus not-for-profit organisations. Despite previous studies highlighting the role of the public press in the spread of scandals, this study shows that individual influential users have a contributing role in the spread of scandals, mainly in value-based scandals.

Theoretical Background

The Impact of Influential Social Media Users

Over the past five years, the number of individuals who seek news and information on social media and rely on influential social media users as a source of information has been drastically increasing (Leung and Chung 2019). Influential social media users are netizens with a high number of followers (Li et al. 2020; Wu and Shen 2015) who have the ability to influence other users (Li et al. 2020; Wu and Shen 2015). The concept of influential users is also cited as online influencers (Ao and Mak 2021); brand endorsers who shape public opinion about a brand or a product through their social media posts and online content (Ao and Mak 2021). Studies on social media influencers mainly focus on how brands can promote their brand and build relationships with the public through endorsements. Fewer studies focus on the role of influencers in crisis communication and management (Ao and Mak 2021).

The concept of social media influence and influencers plays an important role in crisis communication (Lee and Abidin 2021). Some studies evaluated the negative aspects of the partnership with influencers, in backdoor advertising (Lee and Abidin 2021), deceptive advertisements (Abidin and Ots 2016), and scandal spillover caused by the paid influencers’ wrongdoing (Kintu and Ben-Slimane 2020). Other studies introduced the factors that determine the influential users in the time of crisis (Ma et al. 2021; Zhao et al. 2018). Some studies have also evaluated the involvement of influential social media users in spreading crisis information (Brummette and Fussell Sisco 2018; Rauschnabel et al. 2016; Zhu et al. 2017). For example, in the study of Chipotle’s E.coli crisis, Brummette and Fussell Sisco (2018) highlight that news media has a central role in the diffusion of information and framing of public opinion during the crisis. Their study indicates that while the number of crisis-related posts from the news was minimal, their influence on online users sharing behaviour was significant. Zhu et al. (2017) also highlight the leadership role of influential bloggers in leading public opinion, especially in a collective culture.

To better understand an online crisis, the social-mediated crisis communication model (SMCC) introduces three types of publics interacting during a crisis: influential social creators, followers, and inactive users (Liu et al. 2011). Influential social media users leverage the opinion of others by satisfying their emotional and informational needs (Zhu et al. 2017). This information will be consumed by the followers, in online media, and inactive users, in offline environments. While inactive users are not involved in online sharing behaviour, they may indirectly consume influential creators’ posts through word-of-mouth communication (B. F. Liu et al. 2011). Among the three publics, influential social media users play the most critical role in the spread of crisis information and crisis management (Ao and Mak 2021; Liu et al. 2011). They can induce other users to repost and comment on their posts (Li et al. 2020; Wu and Shen 2015). Ao and Mak (2021) also introduce two types of influential social media users engaging in online scandal information dissemination, primary and secondary influencers. Primary influencers are defined as famous users who are directly involved in the incident. Yet, secondary influencers tend to participate in crisis communication for their own personal purposes. These users who have a large number of followers can be famous people, organisations, or news outlasts (Ao and Mak 2021).

Studies on crisis communication examine general and influential users’ discourse during online crises. Li et al. (2020), for example, suggest that influential social media users with anger-related words trigger a higher number of reposts compared to other emotion-related posts. Kirkwood et al. (2019) also assert that social media users may use different discourses to show their resistance to organisations’ actions. They may directly question the organisation’s behaviour or express their disagreement by using humorous references and hyperboles. The study argues that in an online collaborative attack, humour and entertainment may intersect to ridicule and punish a brand. The participants may use political and cultural humour and hyperboles as resistance discourse to call for action, criticise, and ridicule a brand (Kirkwood et al. 2019). Finally, some studies investigated the impact of influential users’ support on the effectiveness of response strategy (Singh et al. 2020) and brand evaluation (Ma et al. 2021).

Although previous studies have highlighted the role of influential social media users in the spread of online scandals, there is little understanding of how different types of influential users and influential users’ engagement approaches influence the diffusion of news about a crisis or scandal. In analysing influential users’ engagement approach, we focus on how they interpret a scandal. We examine how influential users conveyed and clarified their positions through their discourse and engagement in online sharing behaviour (Lazar 2022).

Crises and Scandals

A crisis is an unexpected and non-routine incident that can seriously affect corporate performance and severely damage a company's reputation and stakeholders’ trust (Coombs 2007). Every organisation may make mistakes or wrong decisions during its operation. But not all mistakes and errors lead to a crisis, and not all crises have the same level of reputational damage (Karl Grebe 2013). Crises can be classified based on their origin, the level of attribution of responsibility, the degree of media exposure, and consequently the amount of reputational damage (Karl Grebe 2013). A crisis that involves transgressions with attribution of a high level of responsibility and vast public outrage and reaction can be classified as a scandalous incident. A scandal is defined as an incident that brings public shame with a high level of undeniable responsibility stemming from perpetrators’ misbehaviour or fault, which requires an immediate apology and preventive measures (Marcus and Goodman 1991). In scandals, victims tend to be a large number of people whose sense of fair play is violated (Bowen et al. 2018; Marcus and Goodman 1991). The victims of a scandal may be those directly affected by the event and secondary or indirect victims, who are affected indirectly and through their relationship with others. The secondary victims can even extend to an increasing number of people who observed or heard about the incident losing their sense of security and fair play (Condry 2010).

To define a scandal, Coombs and Tachkova (2019) also introduce the concept of “scansis” and argue that an organisation can be in crisis but not scandalised. “Scansis” is a situation characterised by inappropriate behaviour provoking strong moral outrage among the public (Coombs and Tachkova 2019). A crisis becomes a scandal when the organisation is considered highly responsible and accountable and an incident is regarded as an offensive transgression and unfairness (Karl Grebe 2013).

While not all crises are scandalous, they can escalate to a scandal if improperly handled. Where the response is interpreted as unethical and inappropriate, and people are suspicious that the organisation is lying, deceiving, misleading, avoiding responsibility, or even equivocating, the crisis can be exacerbated and turn into a scandal (Karl Grebe 2013). An example of this is the United Airline crisis; when in the eye of the public, United Airlines did not appropriately respond to the incident (Benoit 2018). In 2017, a video of a man forcibly removed from his seat and dragged off in the aisle in one of United Airlines' flights went viral on social media. While the incident caused deep emotional outrage across social media and was crying out for an apology, United Airlines attempted to downplay this hostile manner. In his initial response, the CEO relied on differentiation and mortification and did not sincerely apologise and promise corrective actions. This response strategy provoked public outrage and disgrace. People blamed the organisation not only for the incident but for the CEO’s inappropriate reaction to the incident (Benoit 2018). With the consideration of the “scansis” concept, the United Airlines incident and initial response provoked a high level of moral outrage among the public, making it a serious scandal for the brand (Jong and van der Linde 2022).

Studies have classified scandals into two main categories: product and performance-related scandals and value-related scandals (Brown and Dacin 1997; Guèvremont and Grohmann 2018; Liu et al. 2018; Pullig et al. 2006). While the former refers to the organisation’s ability to deliver functional benefits to the customers (Hansen et al. 2018), the latter is related to the social and ethical issues that questions organisation’s values and motives (Guèvremont and Grohmann 2018; Hegner et al. 2016; Pullig et al. 2006). Many studies focus on performance-based scandals (Kapoor et al. 2022; Ma et al. 2021; Thaler et al. 2018) and some studies examine the negative impacts of value-related scandals (Gabrielli et al. 2021; Guèvremont and Grohmann 2018; Sims 2009), but limited studies compared the differential implications of performance-based versus value-based scandals on public emotion and behaviour (Hegner et al. 2016; Liu et al. 2018).

Some studies argue that performance-based crises cause more significant threats than value-based ones (Hansen et al. 2018). Doherty et al. (2011), for example, compare two types of value and financial scandals and argue that in the scandals that are not involved in the abuse of power, people respond more negatively to financial scandals than to value-based scandals. Liu et al. (2018), on the other hand, assert that customers have stronger negative responses to values-related negative publicity than to performance-related one, and that “negative associations related to values are more harmful to a brand than those related to performance” (p. 138). The study also reveals that while corrective actions are an effective response strategy, the effectiveness of this response strategy is stronger in performance-based negative publicity than in value-based scandals (Liu et al. 2018). It is also argued that in an individualistic culture, people are more sensitive to value-based than performance-based scandals (Baghi and Gabrielli 2019).

The differential implication of types of scandal on brands might be moderated by the type of organisation in crisis (Zhao et al. 2018) (for-profit organisation vs. not-for-profit organisation) and the extent to which the scandal is related to the brand value. Companies extensively involved in benevolent activities are expected to be honest and have an ethical approach towards their stakeholders and society. Hence, for such organisations, a value-based scandal can cause more reputational damage than a product-based crisis (Hegner et al. 2016). In a performance-based organisation, likewise, it is expected that the scandals related to the main value of their products or services have severe negative impacts on their reputation (Pullig et al. 2006).

Types of Organisations in Scandal

The type of organisation is a factor that may amplify or weaken the public reaction to and the reputational damage of a crisis (Adbi 2022). The adverse impact of scandals on the organisations are not only due to their engagement in the transgressions but also is related to organisational core principles (Scurlock et al. 2020), different level of moral expectations from the organisations (Chapman et al. 2022), and the degree of media targeting (Adbi 2022).

When a crisis questions an organisation’s core principles, the event can impose substantial reputational damage on the brands (Dawar and Lei 2009; Scurlock et al. 2020). Baghi and Gabrielli (2019) find, for example, that when customers are more exposed to brand associations with corporate ability, they are more sensitive to performance-based crises. On the other hand, brands associated with corporate social responsibility are more sensitive to value-based crises (Baghi and Gabrielli 2019). A reputation for being good can also arouse high expectations from a brand, and in turn influence how people respond to future incidents. In the event of violations of these expectations, organisations that signal moral credentials may trigger greater anger and loss of trust. It is argued that “the more an entity’s behaviour deviates from expectations, the more profoundly negative the re-evaluations will be” (Chapman et al. 2022, p. 1259). These organisations range from for-profit organisations engaged in corporate social responsibility programmes to not-for-profit organisations driven by social missions.

Not-for-profit organisations are regarded as moral and ethical entities that should not be corrupted by illegal persuasions. Therefore, the allegations of wrongdoing can cause widespread media coverage and reputational damage (Scurlock et al. 2020). Not-for-profit organisations are regarded as unique entities that work to sustain the fabric of society, a mission that cannot be achieved through government or for-profit activities (Hornsey et al. 2021). This sector benefits from its reputation as “purveyors of good”, enabling them to attract voluntary works and encourage donations. People may also be more willing to engage with their products over commercial organisations because of the high public trust in the sector (Hornsey et al. 2021). A question that needs to be answered is whether the not-for-profit sector also benefits from its reputation during negative events and whether its moral credential plays as a buffer that protects the brand in a scandal (Hornsey et al. 2021). The literature on the violation of these expectations mainly focusses on the for-profit sector engaged in corporate social responsibility, and little is known about the differential implications of a scandal on for-profit organisations versus not-for-profit organisations (Chapman et al. 2022).

A scandal can be a threat to all types of organisations and blemish the public trust in them. But this threat might be more detrimental to the not-for-profit sectors because the essence of their operation is based on public trust (Alhidari et al. 2018). The donors trust that not-for-profit organisations are ethical entities that are not abusing their power while serving vulnerable people and are not corrupted by misusing the donations. Because of the high expectations for this sector, scandals involving abuse of power or mismanagement of funds can arouse public anger and cause intense media coverage. The violations of expectations may incite donors to withdraw their funds and call for an investigation into other not-for-profit organisations (Scurlock et al. 2020).

While some studies argue that not-for-profits are more harshly punished in the allegation of transgression than commercial sectors (Chapman et al. 2022; Hornsey et al. 2021), Adbi (2022) asserts that for-profit brands may be more targeted by social control agents, such as government, activists, and media, causing greater financial losses following a scandal. The study argues that, following a depletion of trust, for-profit organisations are more likely to be under attack by the media and political stakeholders for engaging in unethical business practices. He argues that due to their commercial interest and likelihood of involvement in self-driven behaviour, for-profit organisations are often an easy target for popular media and suspension of unethical behaviour (Adbi 2022).

In sum, although the goal of crisis management in for-profit and not-for-profit organisations is maintaining brand image, considering the substantial differences between for-profit and not-for-profit sectors (Sepulcri et al. 2020), crisis management practices in for-profit organisations are not directly applicable to the not-for-profit sector (Sisco 2010). Not-for-profit organisations often have fewer resources for crisis communication than for-profit organisations. Thus, they need to be more creative in the allocation of limited resources on crisis communication and conveying crisis response to the public (Sisco 2010). Crisis management studies need to highlight the distinct characteristics of not-for-profit organisations. Therefore, in this study, we seek to explore the differences between the two types of organisations in scandals and suggest practical implications for crisis communication managers. The following sections outline the research method employed for addressing the research questions.

Methodology

This study collated multiple instances where scandals had been shared on Twitter. These scandals were analysed using content analysis methodology. Content analysis enables the researchers to understand the situation better and determine what could be done better in the future. Content analysis was considered appropriate for this research as it is a systematic method for observation, interpretation, and classification of different types of data and it provides meaningful information in logical categories and themes (Leung and Chung 2019). Content analysis enables the researcher to use varied concept-driven and data-driven categories in a single coding frame (Schreier 2012).

Given the exploratory nature of this research, we applied content analysis to multiple cases to gain a breadth of understanding by observing and analysing naturalistic behaviour online. In this instance, the evaluation of multiple cases on social media enables the researchers to better understand how scandals go viral on social media in the context of what occurs for different types of scandals and varied types of organisations rather than taking an in-depth analysis of a single case that may not be representative of other scandal diffusion patterns online.

Chosen Scandals for Analysis

The scandals were selected through an iterative process (Polkinghorne 2005). In the first step, the researchers searched for instances by using certain keywords related to crises and scandals (e.g. scandal, crisis, firestorm, backlash and non-profit, not-for-profit, profit, brand). When the primary list of scandals was completed, the news outlets’ articles on each scandal were evaluated to find new scandal-specific keywords. In the next round of search, the new keywords (e.g. fraud, fraudulent, scandalous, deceit, decisive) were used to complete the list of crises. The process was iterated until no additional keywords and cases were found (Polkinghorne 2005).

From here exclusion criteria were applied to screen the sample; more information is provided in Fig. 1. We included the crises that have occurred within the last five years (from 2016) and were related to worldwide brands with longevity. That is, if a scandal was related to a very small regional population or one that lasted only a few days, it was eliminated from the sample set. Any non-English cases were also excluded to ease the analysis process. We also decided to exclude the business platform and social media brands from our sample list. These businesses are likely to face crises caused by the actions of their users rather than the organisation itself.

Fig. 1
figure 1

The process of selection of samples

Full size image

To select the final scandal cases, we have run a group discussion among three authors. The purpose of the group discussion was to confirm the crises strongly associated with scandal and determine the type of scandal, whether the scandal is related to the organisation’s value and morality or is associated with its products and ability to satisfy customers’ needs. Finally, fourteen scandals that matched the criteria and resulted in considerable negative publicity on Twitter were identified. Table 1 shows the final samples.

Table 1 Summary of the brands, scandals, and other key information about the cases analysed in this study
Full size table

Coding Procedure

To develop the coding frame and ensure the consistency of the coding procedure, data analysis was initiated through a pilot analysis and group discussions. In this phase, Pepsi, Nike, and Australian Red Cross scandals were analysed to cover the two types of organisations and scandals. In this phase, the data were coded independently by two researchers. Following the coding, in the discussion meetings, the differences in coding were discussed and reconciled, and the coding frame and procedure were created. In the main phase, the remaining cases were randomly divided between the two researchers to be independently coded. After the completion of the coding stage, 20% of the data were randomly selected and multiply coded to determine the consistency of the coding and intercoder reliability. The intercoder reliability shows a 96% agreement rate, indicating the reliability of the coding procedure (Table 2).

Table 2 Samples of codebook for type of scandal
Full size table

Type of Influential Social Media Users and Their Engagement Approach

We define influential users as social media users with a high number of followers (Li et al. 2020; Wu and Shen 2015). We consider the users’ number of followers as an indicator that represents the popularity of the user (Cha et al. 2010; Zengin Alp and Gündüz Öğüdücü 2018).

The influential users were coded based on the reasons behind their fame. Following the coding procedure, the coders categorised the influential users into subcategories based on their self-discourse and their profiles’ information (e.g. Journalist, columnist, member of the parliament). If the subcategory covering the type of a given influential users has already been created, the user was placed in the respective subcategory. If not, a new subcategory was created to cover the new type of influential users. This process was continued until no additional new subcategory was found (Table 3).

Table 3 Categories of influencers
Full size table

The unit of analysis for identifying influential users’ engagement approach was the entirety of Tweets posted by users regarding a given scandal. In analysing influential users’ engagement approach, we focus on how they interpret a scandal. We examine how influential users conveyed and clarified their positions through their discourse and engagement in online sharing behaviour (Lazar 2022). The units of analysis were coded through a three-stage iterative process. We analysed the units of analysis at the word and sentence level to identify the concept and analytical code representing the meaning. In the second level, the codes were continuously compared and analysed to find the similarities between them and create wider categories of meaning. Finally, the codes were integrated and refined to develop the main themes. The coding process involved comparing and contrasting Tweets within and outside the subcategories to find the entire coding frame. The examples of codes are presented in Table 4.

Table 4 Samples of codebook for Users’ engagement approach
Full size table

Results

A total of 455 unique tweets were analysed across the 14 selected cases and categorised by (1) scandal type (performance based vs. value based), (2) types of influential users, and (3) influential users’ engagement approaches (attacking vs. defending vs. neutral). We observed that the role of these variables in the spread of scandal differs significantly between for-profit and not-for-profit organisations. In addition, we observed that facing the first scandal on Twitter, organisations are likely to experience a series of crises rooted in the first scandal. These crises may stem from inappropriate responses, similar mistake, or social change.

Influential Users and Their Role in Spread of Scandal

The primary result of the study demonstrates the differences between individual and corporate influential users. That is, online users’ engagements in the topic tend to be higher when they receive crisis information from individual influential users rather than news outlets or major corporations. For example, in the H&M scandal, The Weekend (a celebrity) tweet caused more than 109,000 retweets compared to CNN with 1,200 retweets. In United Airlines' second crisis, McNeil's (a blogger) tweet resulted in 85,000 retweets, yet The Washington Post’s tweet caused 866 retweets. This pattern existed even when the individual tweeters had a smaller initial following than the news corporation that also shared the scandal news. It was also observed that when a crisis happens, rival companies may seek to benefit from their competitor’s scandal. Examples of this are Wells Fargo’s scandal when its competitor, Aspiration, tweeted “how to close your Wells Fargo account”, the United Airlines’ scandal when Delta tweeted “you are welcomed with your leggings”, and Nike’s scandal when Puma tweeted “it would have never happened for Puma”. People’s reactions to rivals’ posts, however, varied considerably. Many users attacked Puma for taking advantage of the sportsman’s injury. Delta’s tweet, however, resulted in more than 35,900 retweets and numerous users positively reacted to the tweet.

Finally, we observed ordinary users with low number of followers can become an influential voice on a given topic. They can be impactful by proposing a creative and humorous perspective. They may express their disagreement with the brand and the way they responded to the event by using humorous references and hyperboles. For example, one user created a picture of an aircraft's seating area with a section for the fight club. Using humour, the user explained united airline is pleased to announce fight club (as its new seating area) on domestic flights. In some scandals, ordinary users who have posted the first videos of the incident had also a significant contributing role in the diffusion of crisis. In the Starbucks scandal, the first video about the incident was posted by a user with 11,000 followers. This video caused more than 179,700 retweets. In the Pepsi scandal, ‘Willy’ tweets with 5,000 followers and ‘Wallace’ with 395 followers resulted in 21,300 and 9,300 retweets, respectively.

Scandal Type and Type of Organisation

The results show that influential users’ participation in online scandal discussions is impacted by the type of scandal. Value-based scandals resulted in more negative tweets from influential users in terms of both number and tone. These tweets also received a considerably higher retweets than tweets on performance-based scandals. This observation was more prominent in for-profit scandals. That is, we observed that in for-profit crisis, value-based scandals cause more public anger and online attack than performance-based scandals. For example, in the Nike performance-based scandal, the sources of information normally narrated the scandal neutrally rather than blaming the brand for the shoe malfunction. Individual influential users also mainly expressed their concerns regarding Zion Williamson’s injury rather than Nike’s product fault, although the incident had been viewed by millions of people watching the basketball match. In addition, in the Pepsi, Starbucks, United Airlines, and H&M value-based scandals, the most viral tweets caused 144.000, 170,000, 178,000, and 120,000 retweets, respectively. In the Nike and Wells Fargo performance-based scandals, on the other hand, the most retweeted posts resulted in 15,000 and 10,000 retweets, in turn.

The results show that the spread of scandal information differs significantly based on the type of organisation involved in the scandal. We observed that influential users’ participation in online scandal discussion differs significantly between for-profit and not-for-profit organisation. The findings show that the number of famous people who attack not-for-profit organisations for their scandals is far less than the number of famous people who attack for-profit companies for their scandals. In the Australian Red Cross scandal, for example, the role of individual influential users who participated in the scandal was minor. It is noteworthy that, although celebrities, such as Ellen DeGeneres, Pink, and Dacre Montgomery, actively had been trying to raise money for the Australian Red Cross to help bush fire victims, when the scandal concerning the donation distribution for bushfire victims went viral, no celebrities were directly involved in the scandal on Twitter (Tables 5, 6).

Table 5 Descriptive analysis of (re)tweets for different type of scandals
Full size table
Table 6 Descriptive analysis of (re)tweets for different types of influencers
Full size table

Influential users’ Engagement Approach

Based on content analysis of tweets and results of the coding, we classified the influential users’ engagement approach into three clusters: attacking, defending, and neutral. It was observed that in the for-profit organisations’ scandals, the majority of individual influential users are attackers, and the impact of for-profit defenders, if any, is minor. For instance, when United Airlines faced the first scandal, Dana Loesch, a journalist with one million followers, tried to defend the company and convince people that company’s decision was appropriate, yet her tweet resulted in less than 270 retweets. On the other hand, the tweets of attackers contributed to thousands of retweets and sharing.

However, for the not-for-profit organisations’ scandals, the users’ engagement approach was more spread into attacking, defending, and neutral. When not-for-profit organisations are under fire, some people defended them and reminded people of the positive history and contributions of the not-for-profit organisations. It was observed that some famous people defended the not-for-profit organisations and encouraged people not to forget all the good that an organisation has done. For example, Simon Pegg and Tony Robinson in the Oxfam scandal and Piers Morgan in the Comic Relief scandal are high-profile influential users who defended the organisations. This type of defence was scarcely observed in for-profit organisations’ scandals.

We also observed that in organisational scandals, news outlets and magazines are often neutral, narrate the incident, and cover both the incident and organisations’ responses. Their tweets also receive significantly less retweets than the individual influential users (Table 7).

Table 7 Descriptive analysis of (re)tweets for different types of influencers
Full size table

Reputational Damage of “series of scandals”

In this study, we observed that six out of fourteen organisations (Johnson & Johnson, Wounded Warrior Project, Oxfam, Starbucks, Pepsi, and the Australian Red Cross) experienced a second or third wave of scandals on Twitter that stemmed from high engagement with an initial scandal. We called this phenomenon a “series of scandals”. After a substantial amount of attention garnered from the first crisis being shared online any other crisis or scandal draws more attention and establishes, perhaps falsely, a pattern of crises associated with a brand.

For example, in July 2018, Johnson & Johnson “was ordered to pay 4.69 billion to 22 women who had alleged the company talc-based product caused them to develop ovarian cancer” (“Johnson & Johnson told to pay $4.7 billion in baby powder case”, 2017). The company, however, denied both that its talc products cause cancer and that they ever contained asbestos (defensive response to a product-based scandal). This scandal was mainly covered by news media and magazines, whose tweets did receive significant retweets, and almost no individual influencers fuelled the debate. However, five months later, in December 2018, an article published by Reuters revealed that the company had been aware of the asbestos in its products for decades, so that the second stream of scandal, which resulted from the interpretation of an inappropriate and dishonest response, started. There were significant differences between the first and second streams regarding how the crisis went viral. In the first stream, neither news’ tweets fuelled the crisis, nor the individuals involved in it. However, although we observed that news tweets regarding the organisations’ scandal often do not receive many retweets, in the second stream Reuters, CNN, CNBC, and AJ + tweets resulted in thousands of retweets and caused significant debate. The role of bloggers also was significant in the second stream (Table 8). Table 9 also provides a summary of observations regarding scandal series.

Table 8 Johnson & Johnson first and second stream of scandal on product safety
Full size table
Table 9 Summaries of series of scandal
Full size table

Discussion

This study has several contributions to crisis literature. Firstly, this research suggests that while news outlets play a significant role in the spread of performance-based scandals, individual influential users overtake news outlets in the diffusion of value-based scandals. Despite having millions of followers, news outlets and magazine tweets are not shown to encourage sharing behaviour in value-based scandals compared to individual influential users’ tweets. Although news outlets and magazines actively cover the value-based scandals, their tweets often do not go as viral as individual influential users’ tweets, for several reasons. First, rather than attacking or blaming organisations, news outlets and magazines are often expected to narrate the incident and tell the story in a more neutral fashion, which might not be as provocative as individual influential users’ tweets. Second, online users may feel more connected with individual influencers who have attractive social personalities than with news outlets that officially cover the news. This may make online users more inclined to share and react to individual influencers' tweets. Third, news and magazine outlets do not focus on a specific topic but cover a range of topics, potentially diluting their follower base. Individual influencers, on the other hand, tend not to cover various topics while they are covering the scandal incident and all focus is on the scandal, creating a short-term body of work that draws more attention to the scandal. This finding is in line with Cha et al. (2010) that suggest when users limit their tweets on a specific topic, their influence tends to increase. On the other hand, the significant role of news outlets in performance-based scandals can be because of the low involvement of individual influencers in performance-based crisis coverage. That is, value-based scandals may be of greater concern to society, encouraging individual influencers and ordinary users’ reactions.

The results of this study demonstrate that individual influential social media users are more likely to tweet about a value-based scandal and these tweets subsequently often receive more retweets compared to tweets on performance-based scandals. In value-based scandals, both the number of individual influential users commenting on the incident and the retweets they receive are significantly higher. This study also concurs that the severity of moral misbehaviour positively influences the valence of negative expressions (Johnen et al. 2017). The result indicates that in value-based scandals, ordinary users can be significantly influential. The users who are pioneers in sharing reliable evidence about the incident, such as a video, or have a creative and humorous viewpoint towards the scandal can have extraordinary influence over a value-based scandal. In fact, everyday users can become opinion leaders and direct the flow of information over a specific matter (Xu and Pratt 2018).

Secondly, this study introduces three main categories for influential users’ engagement approach: attacking, defending, and neutral, and argues that influential users’ engagement approaches towards a scandal, represented in the valence of their tweets, influence their followers' participation in online scandal discussion. We observed that attackers’ tweets often trigger more engagement than neutral tweets. For example, news outlets and magazines which are likely to narrate the incident and cover both the incident and organisations’ responses in a neutral tone, receive significantly fewer retweets.

Thirdly, our study also highlights differences between varied types of organisations in scandal. It indicates that online influential users take different approach towards for-profit and not-for-profit organisations in scandal. When for-profit organisations are under fire, the number of individual influential users participating in the scandal and their engagement contributions are higher than when not-for-profit organisations are experiencing a scandal. The number of individual influential users who attack for-profit organisations for their mistakes is significantly higher than the number of those who attack not-for-profit organisations. Our study is consistent with Adbi (2022) that suggests for-profit organisations may experience greater reputational and trust damage from scandals because of greater media targeting. As a social mission-driven entity that runs on selfless principles, not-for-profit organisations may attract the support of the public, both general and influential users, following a scandal. We argue for-profit organisations are more easily suspected of engaging in malfeasance and opportunistic behaviour than not-for-profit organisations which are expected to run on selfless principles (Adbi 2022). For-profit organisations may also be more targeted by social control agents, such as government, activists, and media, even if not-for-profit organisations are also likely to participate in ethical transgression (Adbi 2022).

The result shows that for-profit organisations typically do not have influential users’ advocacy in the time of scandals. However, not-for-profit organisations can benefit from the online user’s support. As Scurlock et al. (2020) suggest, receiving celebrities’ endorsements is critical in organisational success. Hence, not-for-profit organisations are more likely to recover their brand reputation by receiving influential users’ support in the time of scandals.

Finally, we observed that organisations experiencing a scandal may encounter a ‘series of scandals’ that reminds people of the first scandal. The findings show that after a substantial amount of attention garnered from the first crisis being shared online any other crisis or scandal may draw more attention and establishes, perhaps falsely, a pattern of crises associated with a brand. Despite the traditional crisis communication models that consider the crises to be static, these observations are consistent with Coombs (2017); Mak and Song (2019), suggesting that a crisis on social media tends to be dynamic with multiple peaks and troughs and triggering events.

Managerial and Theoretical Implications

Based on the results of this study, influential users may attack or defend the organisation or take a neutral position towards a scandal. Organisations could communicate with the influential users to help redirect their negative approach towards the organisation and change their perspective to positive or neutral. However, the efficacy of this approach, especially in the for-profit sector, is uncertain with an individual influencer unlikely to feel a need to change based on an organisation’s demands and a powerful influencer with a strong following unlikely to change for fear of being associated with the scandal itself and impacting their own following. Although for-profit organisations often do not benefit from the influencer’s advocacy in the scandal, communication with the influential users may mitigate the possibility of their further attack. However, influential users' advocacy is a valuable advantage for not-for-profit organisations facing a crisis or scandal. Influential users’ support for not-for-profit organisations can be a critical means to mitigate the attacks and maintain the brand reputation. Hence, both not-for-profit and for-profit organisations should (1) identify the most influential users who comment on the crisis and (2) provide a direct and immediate response to them and ask them to stand with the company while trying to find the reasons behind the scandal.

Organisations are also encouraged to provide clear information about the incident and an appropriate response through the news outlets and magazines. News outlets often have large number of followers and audiences and often narrate the incident neutrally. By providing clear information a balanced approach may be achieved via social media. An organisation can benefit from the news outlets large audiences to share the crisis response with their stakeholder while lessening the risk of causing a strong negative debate and attack without their voice and perspective being included. While this study shows that news outlets’ scandal coverage does not disseminate as deeply as the individual influential users' tweets, news outlets’ large number of followers results in broad message exposure (Jianqiang et al. 2017).

From a theoretical perspective, this research supplements situational crisis communication studies and online scandal research by identifying the new concepts of type of influential users, influencer’s engagement approach, and type of organisation as situational factors in a crisis. The situational crisis communication theory assists managers in better managing crises by determining the type of crisis being faced and introducing appropriate response strategies (Coombs 2010; Utz et al. 2013). The results of the current study, however, indicate that situational factors beyond the organisational control stemming from the essence of social media may influence people’s perceptions and attribution of responsibility. This exploratory research confirms the critical role of social media influential users and their engagement approaches in scandal spread and indicates that influential social media users’ scandal coverage can influence the other users’ attitudes towards a scandal and sharing behaviour. We also highlight the role of users’ engagement approach in the spread of scandal and indicate that attacking tweets encourage more sharing behaviour compared to defending and neutral scandal-related tweets.

This study also introduces the type of organisation in crisis as a factor that influences crisis situation. While previous crisis management studies have mainly focussed solely on either for-profit or not-for-profit organisations’ crisis communication, this research compares the two types of organisations (for-profit versus not-for-profit) in scandal. This study shows that type of organisation in crisis has an influence on the public’s perception of the incident. Hence, the study incorporates the concept of the type of organisation into crisis management as a factor that influences crisis situation.

Finally, this exploratory research introduces the ‘scandal series phenomenon’ to online scandal research. This study suggests companies would be better to be open, honest, and provide an appropriate response to the first scandal and avoid a series of scandals, wherever possible. However, if a further scandal were to arise on social media, managers would be advised to be prepared for the subsequent scandal to garner even greater attention than the first.

Limitations and Future Research

While producing interesting findings, this is important to consider that this study is initial exploratory research. We content analysed the secondary data from Twitter across multiple cases to explore the online scandal phenomenon. Although the findings of this qualitative research can provide the basis for quantitative research to examine the causal relationship between the constructs, the findings might not be applicable to all populations and all organisations for some reasons. First, our sample is big global brands’ scandals that cause a serious debate on social media and English tweets and cases. So, we should be careful while generalising the findings of this study to the brands that are not in line with our case selection criteria. Second, the results of this study cannot find the extent to which the constructs, including the types of scandals, types of organisations, types of influential users, and influential users’ engagement approaches influence the sharing behaviour or user’s perception of brands and scandals. Therefore, we encourage future research to develop quantitative and experimental research to examine the impact of this study constructs on, e.g. user’s attitudes towards the brand, sharing behaviour, reputation, brand trust, and attribution of responsibility. Future research that evaluates the impact of the size of the organisation on scandal situations can broaden our understanding of the phenomenon. Finally, we encourage future research to investigate the online scandal phenomenon by including other social media channels such as YouTube, Instagram, and Facebook.