Public Relations Crisis and Social Media : An Investigation into Extant and Prospective Consumers ’ Perceptions through the Lens of Attribution Theory

Online social media has shifted the balance of power from businesses to consumers, with consumers now being able to share information almost unrestrictedly in real-time. As a result, an effort to suppress what may be considered as harmful information can easily backfire, causing the information to instantly spread through user ties. However, what is considered important in such occasions is the way extant and prospective customers may interpret such crises. In this paper, we discuss the case of a Greek Apple authorized service provider and the manner in which it responded to a customer’s complaint in a blog post. The company, instead of using social media to reach out to its customers, chose to move legally against the disgruntled customer, aiming to suppress the relevant blog post, which quickly resulted to the company’s viral defamation. Building upon attribution theory and employing a content analysis of user postings in social media, our study explores stakeholders’ perceptions regarding the company’s reaction and response strategy, seeking to investigate attributions of cause and responsibility. Our results show that, in the initial stages of a crisis, hostile behaviour or refraining from comments altogether, can lead to negative outcomes in relation to a company’s reputation.


Introduction
Social media's explosion has transformed several sectors, including those of e-marketing and e-commerce.As soon as individual users started embracing online social networks and social media, web 2.0 began attracting the interest of marketers for the purposes of marketing, building brand reputation and strengthening e-commerce efforts [43].
Analysts can now mine the information flowing across online social networks (OSNs) to launch cost-effective electronic word-of-mouth (e-WOM) strategies [40], handle customer complains in order to maximise customer satisfaction, while minimising the possibility of future complaining behaviour [9].In more detail, technological advances have enabled the mining of this otherwise dispersed information towards gaining insight into consumers' needs and opinions regarding products and services.As a result, irrespective of their size, many companies seek to integrate these new media into their processes and actively use blogs, OSNs, virtual worlds and dedicated forums to develop direct relationships or strengthen existing ones with customers.Additionally, marketers incorporate these applications into their PR strategy and treat them as platforms through which they are able to handle dissatisfied customers by responding to negative comments and criticisms and develop brand and reputation management strategies [10].
At the same time, the internet has been transformed into a user-driven, non-linear repository of information, where consumers create information themselves, whose presentation and propagation is beyond marketers' control [44].Web 2.0 in particular has transformed individual users into content creators and has given them control over the information they create and share.This has essentially shifted the power balance from companies to consumers, allowing potential customers to be better informed regarding a company's products and services and to inform others, by sharing their experience in forums and various social media.These actions, however, can be translated into actual sales, both offline and online.As consumers turn to online sources to look for information and form their purchasing intentions and decisions [39], and as dissatisfied customers have the opportunity to publicly express and share their opinions online, possibly harming the company's image and reputation, the various social media can be seen as platforms, which hold a potential marketing and/or public relations PR risk.Consequently, many companies choose to moderate user comments or refrain altogether from having an online presence.In reality however, a company can still be exposed to such risks, as consumers may criticize it in their own blogs and social media accounts, with the company being essentially unable to stop users' e-WOM.Therefore, it is obvious that e-WOM is no longer used exclusively by companies, but by individual users as well; while in the past a disgruntled customer could usually influence a handful of peers, today (s)he is able to influence the opinions of thousands [44].As a result, while web 2.0 can offer excellent opportunities for marketing purposes, it may also pose significant challenges for a company's online and offline reputation.
However, research on social media and their role within the context of business practices and public relations is still somewhat underdeveloped [48].On the one hand, studies within the field of crisis communications tend to be based on experimental designs; as such, participants are most often presented with fictitious stimuli or interviewed retrospectively, essentially requiring the reconstruction of events and reactions from memory [26].Nevertheless, it has been argued that the analysis of the public's spontaneous reactions due to real life events may be more helpful toward delineating perceptions regarding a specific type of crisis, the deployed communication strategy and the consumers' evaluation of the organisation's effort, and eventually, the impact on reputation and relationship with the public [49].On the other hand, the emergence of web 2.0 and that of the various social media has significantly altered the landscape of public relations.Their interactive and real-time nature have transformed them into valuable tools for the social construction and deconstruction of crises and organisations begun approaching them as more reliable tools for the building of relations; yet, the effect of social media on consumers' perceptions during crises are, too, less researched than that of classic media [48].
Most importantly however, such an investigation would be valuable and pertinent for the field of e-commerce as well.Shopping behaviour, and decision-making, both online and offline, regarding purchases have been shown to be influenced by e-WOM [8].Specifically, there is evidence that online social networks do drive an increasing volume of traffic to retail sites, which shows in turn that "highly influential customers directly affect other consumers' decision making" [32].As a result, companies, which conduct their activities online with the help of e-commerce, can use this influence to increase their sales, and build their customer base among others [40].At the same time, assuming that consumers are influenced by a negative e-WOM, the aforementioned companies can take advantage of the available information, which harms their reputation and, essentially, business, and seek to restore trust on their brand and their status, and overturn the effect of the negative e-WOM.
With these in mind, this paper discusses the case of CompanyX, (name withheld for anonymity purposes) a Greek Apple authorized reseller and service provider, and the manner with which it tried to intercept e-WOM, aiming to protect its reputation.It is an extended version of work published in [57] and we extend our previous work by considering attributions of cause and responsibility and investigating the public's evaluation of the company's reaction during and beyond the crisis.To do so, we adopt an audience-oriented approach and build upon the Situational Crisis Communication theory, by employing a content analysis of user postings in social media.Most importantly, this extended version seeks to explore and assess the company's communicational approach toward preserving its reputation.The main contributions of our study are twofold; on the one hand, it illustrates the strength

Theoretical Background
Information sharing among users of social media and OSNs has been shown to be more influential and more effective than traditional marketing communications [40], [42], shaping consumer attitudes and behaviours [37], [54].Focusing specifically on the efficiency of e-WOM, it has been argued that a given company can leverage social media's low cost to reach a much broader customer base and spread more widely its marketing message [51], while it can compete more successfully for consumers' attention; e-WOM is immediate, accessible by more than the initially intended audience and, being in digital print, makes it permanent [28].Moreover, as others have noted e.g., [7], e-WOM has experienced significant growth and today it is strongly related to consumers' shopping behaviour both online and offline [8], [37].
However, as frequently showcased in the past, e-WOM may not be aligned with a company's marketing efforts.While users may share positive experiences, they may equally share negative experiences, causing the viral defamation of the company in question.At the same time, while some may actively seek others' opinions, others may be exposed to e-WOM without looking particularly for it, as web 2.0 allows one-to-many interactions, thus permitting a single user to reach a much larger audience [54].In this sense, and keeping in mind that the information drawn from "[s]ocial media is perceived by consumers as a more trustworthy source of information regarding products and services than corporate-sponsored communications" [39], the impact of e-WOM on companies can either be beneficial or harmful depending on the polarity of e-WOM.In other words, if one's experience with a product or service is positive, sharing it with others can result into creating a positive impact for the involved company or organisation and, essentially, building a good reputation for it.In contrast, if the experience is negative, e-WOM will discourage people from buying the particular products and/or services.Further to this, Madhavi and Akbar [38] point out that, while the motivational form may be interpreted into sales or a general good feeling towards the company, its discouraging form may transform into an active protest.
As a result, and because e-WOM exerts a strong influence over potential consumers and can disseminate very quickly through OSNs, it is of paramount importance to examine e-WOM through the lens of reputation and risk management.Specifically, as it is common knowledge that a positive reputation is integral for reducing the possibility of risk within the context of purchasing products and services [31], either online or offline, e-WOM is eventually tightly related to risk perceptions and consumers' trust on a given company.In more detail, assuming that a company has a bad reputation among its customers and that the latter transmit their negative experiences and opinions through social media, then it is very likely that prospective customers will choose to refrain from interacting with the given company as they will perceive a higher level of risk and consider the company to be untrustworthy [31].Yet, trust has been shown repeatedly to be an integral feature of e-commerce and for transacting with companies, and several studies have found that it builds over time, through various transactions and interactions with a specific party [20], [22].
What is important, however, is that in web 2.0, users are able to form their opinion regarding a product or an entire company quite easily.They can access others' experiences, and they can influence the opinions of others by communicating theirs.Moreover, when it comes to information that has been specifically communicated through user ties in order to advise about potential risks and the breach of trust and, while it is understandably a possible threat for a company, an effort to supress it or control it, can in most occasions lead to consumers' discontent [43].Often, the very action of trying to forcefully remove a piece of information, which is considered as harmful or defamatory, backfires and has the contrary consequences, by driving online users to spread more broadly the information through user ties all over the internet much like an epidemic, with offline consequences as well.In essence, web 2.0 has transformed the dynamics of reputation and crisis management, consumers' attitudes toward such efforts and their perceptions regarding an organisation under crisis.In the next section, we elaborate further on these and introduce our study's theoretical foundation.

Attribution Theory and the Situational Crisis Communication Approach
Several studies to date have analysed the various response strategies, which an organisation may adopt when faced with a crisis [12].Most of them approach crises as a threat to a company that can lead to reputational damages.This point is particularly important because a company's reputation is among its intangible assets, as it succeeds in attracting customers, investors, high-performance employees, and, as a result, helps towards creating and In this line of thought, it has been argued [14], [23] that Attribution Theory can prove to be a useful framework for the examination of companies' and public's reactions under such conditions.In more detail, the central tenant of Attribution Theory is that individuals feel the need to ascribe some responsibility for the various occurring events [56], by identifying the location of the cause of the crisis, characterising it as either internal or external to the organisation, and introducing the dimension of controllability, aiming to investigate whether the crisis is within the control of the organisation.As such, a given crisis is mapped "along a continuum of internal/ controllable-external/uncontrollable" [35].
Building upon Attribution Theory, Situational Crisis Communication Theory (SCCT) identifies the reactions of stakeholders in relation to various types of crisis and foresees the impact of the crisis and the organisation's response strategy on its reputation [14], [15], [49].In essence, SCCT can be seen as an extension of Attribution Theory which "prescribe[s] crisis response strategies designed to protect reputational assets" [15].Specifically, SCCT dictates a two-stage assessment, closely examining the initial crisis responsibility, the crisis history, and the organisation's prior reputation.During the first stage, one needs to look into the perceptions of stakeholders on whether the organisation is to blame for the crisis.If there is an increased attribution of responsibility to the organisation, several studies have shown that there is also a greater reputational threat [16], [17].The assessment takes place by investigating into how stakeholders frame the crisis, i.e., how they interpret the chain of events, and whether they consider it to be due to e.g., misconduct or a natural disaster [14].Within this context, SCCT identifies the victim, the accidental and the preventable type of crisis, depending on the stakeholders' attribution of organisational responsibility, ranging from low to high.Within the victim cluster, organisations are seen as unable to foresee or prevent a certain crisis, whereas within the preventable cluster, they are seen as highly responsible for the crisis [11].Naturally, as the level of responsibility increases, so does the negative impact on the organisation's reputation [13].In the second stage, two additional factors are examined; consistency and distinctiveness.Consistency refers to the history of the organisation regarding similar crises.As such, when consistency is high, i.e., similar crises have happened before, it is said that it experiences recurring issues.Distinctiveness refers to the organisation's prior reputation or relationship with the public and its stakeholders.In this instance, if distinctiveness is low, it is suggested that the organisation has a history of treating its stakeholders poorly.In short, high consistency or low distinctiveness can increase the reputation threat [14].
Following the two-stage assessment, SCCT posits that the organisation needs to consider the viable response strategies, which will help in repairing its reputation, and minimising or preventing altogether any negative outcomes.The matching between the crisis and the possible response strategy, within the SCCT framework, happens through the link of responsibility.Conceptually, this is rather intuitive because through responsibility it is suggested that the organisation needs to be held accountable to some extent for the crisis; in other words, the evaluation of the crisis, and its impact on reputation depends largely on the organisation's responsibility for it.Therefore, depending on the perceived level of responsibility, the organisation decides how it should respond to the situation, which in turn it formulates into a response strategy embodying the organisation's level of acceptance of the responsibility [15].With regards to the response strategies, extant research has shown that accepting responsibility can lead to positive reactions and lead to improved reputation [5].However, Schultz et al. point out that more recent studies suggest that expressions of sympathy can also help organisations restore their reputation.As such, being both apologetic and sympathetic can minimise reputational damages [48].Table 1 presents the various response strategies, matched with the type of crisis for which they are considered to be most appropriate and effective.As shown, more aggressive tactics are considered as more appropriate for those crises, in which the organisation is perceived to have little responsibility, or in other words, in which attributions are external.This has found support in extant studies, which have illustrated that as responsibility for a crisis increases internally within the organisation, the organisation in turn needs to be less defensive [12], [16].
To summarise, SCCT and Attribution Theory can be considered as particularly useful tools for crisis communication research.First, SCCT suggests that stakeholders are more likely to negatively evaluate an organisation, i.e., they tend to make unfavourable attributions (internal attributions), if the latter has failed in the past to meet their

Apology
Crisis manager indicates the organization takes full responsibility for the crisis and asks stakeholders for forgiveness In the next section, we present in detail the case of CompanyX, aiming to get a deeper understanding of the crisis, the causes and the eventual response on behalf of the company, before presenting our study's research method and our findings.

The CompanyX Effect Crisis: a 200,000€ Worth Blogpost
In this section, we first present the case of CompanyX and the exact events that brought up the public relations crisis.Then, we focus on the crisis situation itself and provide details on the company's reactions during it.We consider this information as necessary as it offers a rich description of our case study and because it helps in understanding the dynamics among the involved parties.

Chain of Events Leading to the Crisis
Mr Papadopoulos, a 35-year-old doctor, living in Athens, Greece, purchased on January 2010, a 27'' iMac computer from a large department store, which was an Apple authorized reseller.Nine months later, on September 2010 and while the computer was still under warranty, he noticed some grey spots on the screen, and attempted to clean the screen.However, the spots persisted.He therefore consulted Apple Inc.'s official website, aiming to find an authorized service provider and have the issue fixed.The website indicated three official service providers and Mr Papadopoulos decided to visit the one closer to home, which was CompanyX Ltd.
Therefore, on September 27, 2010, he took the computer to CompanyX.There, one of CompanyX's employees informed him that this was in fact a common problem with 27'' iMacs and ensured him that they could repair the screen.He then instructed Mr Papadopoulos to visit their premises a week later to collect the computer.A week later when Mr Papadopoulos visited the company, he was informed that the computer's screen had been cleaned and that the LCD panel had been replaced.He was also given a repair report with details on the items replaced and the repairs carried out (Table 2: BF17).After returning at his residence and switching on the iMac, Mr Papadopoulos realized that the issue hadn't been fixed; there was a grey shadow on the centre of the screen and traces of moisture.
Following some verbal conflict, the senior officer asked Mr Papadopoulos to submit his request in writing and fully signed.Immediately, Mr Papadopoulos wrote down his request, signed it and, in turn, requested that the senior officer also signed the document as proof of having received it.The senior officer initially refused to do so; yet he eventually did sign it, adding a written comment that he received the request without having read it.Next, the senior officer stated that CompanyX could no longer accept the computer for service on the grounds that it hadn't been purchased from them, and that responsible for replacing it was the Apple authorized reseller, i.e. the department store from which Mr Papadopoulos originally had purchased it, or the Greek Apple authorized distributor.Mr Papadopoulos replied that since CompanyX had already serviced the computer once, they were responsible for the defective new screen, and thus they had to replace it according to European legislation; yet, this request was also rejected.On October 9, 2010, he turned to the Greek Consumer Ombudsman (GCO), the independent authority that seeks the consensual resolution of consumer disputes.There, he stated the facts according to his interpretation, and he was informed that the Greek legislation was quite clear and on his side.He therefore returned to the Greek Apple authorized distributor, informing them on the GCO's opinion.Yet, the distributor claimed that Mr Papadopoulos was to be blamed in the first place for requesting a service from the authorized service provider, rather than requesting a replacement from the original reseller.
On that day, Mr Papadopoulos decided to share his consumer experience on the web.He posted his story on the official Apple Support communities, explaining the issue.His post did not receive much attention, but there were some Apple users claiming they had similar problems (Table 2: AW9).He also wrote a quite detailed post for the website to which he contributes until today, and which shared through twitter (Table 2: BF17).Two days later, he began a new thread in a technology-related forum (Table 2: BF9), where he was describing his experience.It should be noted that the website to which Mr Papadopoulos posted the story at that point wasn't particularly well known.In contrast, the forum where he initiated the new thread, being a specialized forum, where users discuss equipment, software, hardware, and seek advice, had been and is still attracting many readers.
On October 18, 2010, a CompanyX representative replied to Mr Papadopoulos' thread at the forum, describing the story from the company's viewpoint.Specifically, he stated that the company was not responsible for repairing, replacing or refunding the computer because it had not been purchased from them.He also informed the forum that the company's position was that Mr Papadopoulos should had directed his request to the department store from which he had originally bought the computer, and accused him for using harsh and emotive language against CompanyX and the Greek distributor.He also implied that the incident would be resolved through the judicial path (Table 2

The Crisis Situation
Understandably, upon receiving the lawsuit, Mr Papadopoulos was burdened with seeking legal help.He therefore turned to Mr Potiropoulos, a lawyer specialising in internet-related cases, who frequently acted as an advocate in cases relating to social media, the internet and freedom of speech.Furthermore, he was and still is a well-known blogger himself within the Greek blogosphere, a twitter user and a radio producer, discussing legal issues pertaining to the aforementioned matters.
A month later, on December 27, 2010, Mr Potiropoulos uploaded a post on his personal blog, which he later shared via Twitter, titled 'A 200,000€ damages claim against a disgruntled consumer -blogger' (Table 2: BF13).The post was briefly describing Mr Papadopoulos' consumer experience, without revealing the name of the company or that of Mr Papadopoulos.The lawyer was remarking that the company was claiming reputational damages under the provisions of press law, for which Greek courts had already ruled in the past that it doesn't apply to the internet, and that this would be the first time in Greece for a company to sue a customer for publicizing his consumer experience; as a result, this would immediately make it a landmark case.He was also arguing that the case had more to do with consumers' and bloggers' rights to freely express their opinion online rather than defamation or libel.
Regarding online presence, when Mr Potiropoulos took over the case, he had a little more than 1,300 Twitter followers.At the time, this could be considered as quite a large number, considering that, in late 2010, the specific OSN was not yet particularly popular among the Greek online community.Therefore, his tweet, and his blogpost were read, and possibly shared, by a large portion of the Greek Twitter community, causing the story to further propagate.In addition, as he was remarking that the case would inevitably touch upon consumers' and bloggers' rights and freedom of online speech, the case immediately turned into a popular discussion topic among Greek online users.
While it is presumed that Mr Potiropoulos was responsible for propagating the story within the Greek borders, a CNET article is most probably the one that helped the story spread beyond the Greek borders, while it has also been a source of information for several Twitter users (e.g., Table 3: Q13).CNET, being part of the Columbia Broadcasting System (CBS) Corporation, receives a lot of traffic and is among the first international technologyrelated websites that hosted the story.Among the sources of the story, the author of the article cited a blogpost (Table 2: BF1), maintained in English by a Greece-based blogger, and a forum thread (Table 2: BF6), also maintained in English.Before the particular article, and to the authors' knowledge, there were no other mainstream websites reporting the story.Immediately after it, several other outlets began reporting, commenting on and, therefore further propagating the story (Table 2).As a result, even though during the two previous months, the case was attracting the attention of the Greek online community, it was only after the CNET article that there were, by the end of January 2011, more than 2,000 posts on blogs, forums and news outlets, more than 10,000 comments in blogs, forums, Twitter and other mainstream social media, on both national and international level.Furthermore, more than 500,000 views of the aforementioned posts have been reported (Table 2: BF18), while the story has travelled around the world and has been read in German, English, Hebrew, Brazilian and Danish, among others.Additionally, the story was broadcasted on the national TV network (Table 2: NW4), while it was picked up by blogs and forums, international technologyrelated websites, Apple-focused websites, news websites and websites commenting on PR practices (Table 2).
On January 3, 2011, a day before the trial, the GCO invited the involved parties to meet at its premises on January 18, 2011 in order to jointly resolve the issue with an out-of-court settlement (Table 2: BF14).However, the Greek Apple Authorized Distributor failed to attend the meeting, claiming they did not receive the Ombudsman's invitation, while the participants failed to reach to an agreement (Table 2: BF19).Following the meeting, CompanyX issued an official announcement, stating that the company decided to withdraw the lawsuit, claiming that it wasn't their intention to stand against consumer rights and freedom of speech.Their announcement also acknowledged that Mr Papadopoulos didn't intent to damage the company's reputation, and that the lawsuit was an excessive measure toward defending their reputation and credibility.Nevertheless, they weren't going to take responsibility for replacing the defective computer or offer any compensation for Mr Papadopoulos' legal expenses (Table 2: AW8).Finally, on January 31, 2011, the reseller from which Mr Papadopoulos had originally purchased the iMac announced that they would replace the defective computer.
The same day, Mr Papadopoulos launched a petition urging Apple Inc. to investigate its after-sales support in Greece and establish satisfactory customer services.He announced his initiative through his blog and his twitter account and declared that once the petition gathered 1,000 signatures, he would personally visit Apple Inc. and file the petition himself.The following days, the petition gathered a significant number of signatures.The goal was reached in just over a month, while 1,077 people in total signed the petition (Figure 1).After the petition reached its goal, Mr Papadopoulos commented that Apple Inc. had indirectly refused to receive it, by remaining silent to his request to meet with a representative, despite his repeated efforts and the pressure exerted by reporters (Table 2: BF19).Specifically, CNET reported having contacted Apple Inc. for an official response and their perspective on the case; to date, however, there hasn't been any update on the issue (Table 2: TW3).The defective computer was finally replaced on March 16, 2011, jointly funded by the authorized distributor and the department store from which it was originally purchased (Table 2: TW15).
On February 9, 2011, a Greek member of the European Parliament brought the case to the European Commission's attention (Table 2: NW5).The member highlighted that the case could potentially become a milestone regarding practices against consumers, due to the lawsuit being an act against [their] freedom of expression.He further noted that within the Information Society, citizens use continuously OSNs in order to express personal opinions; therefore this case could also affect the right to publicly express one's opinion.On March 22, 2011, a representative of the Commission replied that the question did not provide sufficient information so that the Commission could decide whether the particular matter concerned the implementation of European Commission's law and thus could not further comment.

Method
Since our objective is to understand consumers' and online users' perceptions regarding the crisis, CompanyX's response strategy, its appropriateness in relation to these perceptions and the impact on the company's overall reputation, our study follows the interpretive tradition and approaches knowledge of reality as socially constructed.This permits us to embrace the perspective of the public and gain access to multiple interpretations of the examined concepts, while achieving a greater insight into stakeholders' rationale of the situation [45].Furthermore, we posit that adopting such a perspective can provide valuable insight into how individuals understand and react to an organizational crisis, as it is largely aligned with Vasquez's homo narrans perspective and the audience-oriented approach [53].
Our study evolves around an incident, which developed online and was then propagated through, and mainly thanks to, the various social media.Therefore, the empirical material of our case study comprises entirely of data collected from online communities (e.g., online social networks, blogs, forums), where users were posting and/or sharing comments, opinion pieces, blogposts and so forth in relation to the case, and from websites, online newspapers and magazines, and broadcast networks, where contributors, journalists and editors were reporting the case to their readership.The data collection period spans from October 9, 2010, i.e., the day Mr Papadopoulos published his consumer experience, to November 11, 2012, when data begun being saturated.The collection procedure entailed the use of three different approaches.First, we began monitoring significant social media accounts (i.e., the blogs and twitter accounts of all immediate involved parties).Second, we were conducting, on a daily basis, a simple web search, using CompanyX, Papadopoulos and Potiropoulos as keywords (both in English and Greek), so as to find relevant posts on the story, and third, we were doing a web search with the help of the Google Real-Time Search tool, which was still available at the time, using again CompanyX and Papadopoulos as our keywords.The added value of Google Real-Time Search tool, and the main difference between that and the simple web search, was that it allowed us to search in real time for posts and status updates from Twitter.In total, we analysed 53 blogposts and online articles along with their 610 comments, 3 forum threads, which generated 1371 posts, and 195 tweets, posted by 66 unique twitter users.Overall, the examined blogposts and online articles were shared in other outlets 1599 times, while the examined tweets received 608 retweets.A representative list of the online sources used can be found in Table 2.
The pool of data was investigated using a directed content analysis [27].The Attribution Theory and the SCCT approach dictated the coding scheme and the entire procedure was guided by extant studies in the particular field; initially, this helped us identify the central coding concepts, which later functioned as the main coding categories.In more detail, coding involved the identification of stakeholder interpretations of the crisis type, as well as interpretation from the behalf of the company.Therefore, the coding of data was focused on attributions of responsibility for the

Signatures per day
(Causal) attributions of responsibility to one or more of the involved parties During this stage, we also coded any references to the company's reputation and stakeholders' evaluations on its past behaviour, so as to estimate the weight of CompanyX's history on the overall assessment of the crisis.Finally, we coded stakeholders' reactions regarding the company's response strategy; this was done so as to evaluate whether the strategy was matched to the type of crisis and whether it was perceived as adequate and proper in stakeholders' perceptions.
The first author coded the collection of the empirical material and all second-order concepts (i.e., coder's interpretations) were discussed on each stage with the second and third authors for plausibility and validity.As this study is qualitative and interpretive in nature, rather than e.g., positivist, we did not proceed in calculating metrics such as intercoder reliability.In what follows, we present our study's findings and discuss them within the framework of Attribution Theory and SCCT.

Findings and Discussion
It must be noted that, according to Attribution Theory and SCCT [15], in order to evaluate the exact level of reputational threat and decide on the crisis response strategy, one needs to follow a two-step process; first, the responsibility attached to a crisis is examined, which helps in evaluating the initial crisis responsibility level and consumers' perceptions with regards to the organisation's reputation; second, one needs to focus on the organisation's crisis history and its past relationship reputation, as these two factors may intensify the overall crisis.These are discussed next, in relation to our study's findings.

Attributions of Responsibility
The case quickly resulted in being discussed across the Greek and international blogosphere and the most popular social media.What is important however is that there is little doubt as to the perceptions of external stakeholders (i.e., OSNs users, consumers etc.) regarding the attribution of responsibility and as a result, the type of crisis.As it will be shown next, there were strong attributions of crisis responsibility, with the consumers pinpointing the real or perceived failings of the company.
Examining first the concept of responsibility attribution, the overwhelming majority of Twitter users considered that CompanyX was responsible for the chain of events.This can be showcased through various tweets, which ranged from clear attribution statements, to considerations regarding the wider aspects of individuals' online presence.In more detail, online users were focused on exhibiting their support to Mr Papadopoulos by encouraging him to continue his efforts against CompanyX and the lawsuit (Table 3: Q1).Several of them however were highlighting specifically CompanyX's responsibility for the entire case (Table 3: Q3, Q4) and requesting CompanyX to withdraw the lawsuit (Table 3: Q2).In addition, most, if not all online articles and blogposts reporting on the story, were mentioning that not only the computer was defective in some way from the start, but also that CompanyX, the authorised service provider, failed to properly service it (e.g., Table 2: TW7, BF1).Therefore, these findings indicate that the public, including extant and prospective customers and reporters, was clear with regards to the aspect of responsibility, and was attributing it, in its extreme majority, on CompanyX.
Nevertheless, beyond the issue of the failed service, online users raised concerns regarding the wider aspects of online presence.These concerns can be adequately illustrated through users' postings, in which they were pinpointing the importance of the case for consumers' and bloggers' rights and freedom of speech.In more detail, once Mr Potiropoulos announced via his Twitter account that he undertook the case, he caught the attention of the community and users begun requesting more information.As a result, bloggers and Twitter users began worrying about how this case could find its extension into other scenarios (Table 3: Q5) and formally linked the incident to individuals' right to freedom of speech (Table 3: Q6).Several users were further highlighting that the initial blogpost prepared by Mr Papadopoulos, as well as the subsequent thread in the forum were not aggressive so as to warrant a lawsuit.In more detail, it was argued that his post [did] not seem to offer harsh or emotive language (Table 2: TW3), but rather that [h]is forum post was calm, no bad words... (Table 2: BF6).This finding further supports the argument that online users were in fact attaching the responsibility of the crisis on to the company, as they did not consider Mr Papadopoulos to be at fault in any way.
Moving on to the second step of the analysis and looking into the prior relationship reputation aspect, within the empirical material there is evidence that some commentators were indeed referring to the company's past behaviour and to their personal positive experiences as customers of CompanyX.For example, a returning customer of

Crisis Response Strategy
According to SCCT, "as the reputational threat and negative affect increases, crisis managers should utilize crisis response strategies with the requisite level of accepting crisis responsibility" [14].This suggests that, in the case of a preventable crisis the adopted response strategy should seek to diminish the threat and rebuild the reputation of the company, rather attempt to deny responsibility or the validity of the crisis (Table 1).As a result, in an effort to evaluate the response strategy of CompanyX and its appropriateness, it is necessary to first examine the reactions and the intentions of the extant and prospective consumers and online users in general, and second, to investigate the online presence of the company prior, during and beyond the crisis incident.
With regards to OSNs, admittedly, the entire case evolved mostly within Twitter.The sum of examined tweets and their retweets (195 tweets from 66 unique Twitter users and 608 retweets), as well as their trackbacks, i.e., citations of these tweets in other websites, blogs (e.g., Blogger, WordPress), and other third-party applications (e.g., Storify, acting as news outlets and/or aggregators), were posted in less than three months (28/12/2010 -18/03/2011), reaching approximately 115,000 Twitter users (Greek and international).Having in mind the quantity as well as the content, as shown in §5.1, of the textual material, it is evident that the public was quite spontaneous in their reactions.This was not unexpected, since previous studies have shown that online communities tend to be rather spontaneous in their reactions, specifically during crises [e.g., 49].
When the public got informed on the incident, online users begun requiring further details, as well as requiring that the name of the involved company gets revealed (Table 2: BF12).Despite that the consumer's lawyer chose not to provide additional details and reveal the name, online users eventually did manage to identify the company, through an online search, and discovering the consumer's forum thread and the subsequent posts and replies by CompanyX.They thus updated their Twitter statuses (e.g., Table 3: Q10), while some of them written blogposts, containing their perceptions regarding the case (e.g., Table 2: BF2).This resulted in other users picking up the story, further propagating it and, turning to the company for a reaction.
CompanyX maintained a blog since March 2011 and a Twitter account prior to the incident, both of which were active and whose administrator had been posting information on new merchandise, Apple news, computer tips and so forth.However, when faced with online users' questions, the company didn't use these social media in order to comment or interact with their readership.In more detail, CompanyX didn't engage with the public nor sought to be proactive in any way, e.g., by making available their side of the story or their future actions.Instead, they posted solely a few updates in Twitter, which were later deleted but several users succeed in capturing them in verbatim (Table 2: BF21) or through screenshots.These were the following (split across six different tweets): We have nothing to say to people who begin accusing them based on the writings of a blog, without being interested in discussing or at least listening to the other side of the truth.As far as it concerns our reputation or the company's policy, everything we accomplished so far, we accomplished it through work, not through libel and those who have been our customers and continue on trusting us for sales and service support, know the reality.We do not respond to every message, since the 160 character limit can hold only an accusation; surely not the truth.Even more so when one is accused by tens of people who simply read the writings of a blog, made up their minds and accused someone, all within 5 minutes.
In other words, while the sued customer was being quite vocal, online users were supporting him (e.g., Table 3: Q1, Q2), and several online news outlets were presenting the episode as an example of poor customer service and a failed communication strategy (e.g., Table 2: MW4), the company didn't engage at all with the public.Therefore, the company's online attitude, combined with its decision to sue the customer, is in severe contrast to the Situational Crisis Communication Theory, and characterised as a deny crisis response strategy (Table 1).
Moreover, the company attempted to shift the blame for the negative online comments and the overall criticism to the online community, further attacking their critics.As illustrated through previous studies [41], [49], such behaviour is quite often and may occur irrespective of the locus of the crisis, i.e., regardless of whether the origin of the crisis is interpreted as external or internal.Yet, when a crisis is perceived as originating from within the company, the various stakeholders are less likely to justify a defensive-focused response strategy and tend to receive it with scepticism and, in turn, evaluate even more negatively the investigated company.Indeed, CompanyX's attitude further enraged online users, and the company's denial to comment was interpreted as admittance of failure and, in some occasions, both as arrogance and ignorance (e.g., Table 3: Q11).In other words, the chosen response strategy led in severing ties with extant or prospective customers.
As far as the lawsuit itself is concerned, the Greek courts had already ruled that the press law, as formed at that time, could not find an application for the case of the internet (Table 2: BF13).Therefore, one could argue that the decision to file it aimed at suppressing the blogpost from proliferating.However, this in fact gave way to the propagation of the incident since it led the sued consumer to publicise the lawsuit, as well.

The Impact of the Crisis Response Strategy on CompanyX's Reputation
As shown through the findings thus far, despite that the various stakeholders interpreted the crisis as nesting within the preventable cluster, CompanyX adopted a combination of attack and deny response strategies.This can be characterised as inconsistent to what extant literature suggests as appropriate action.The result of this inconsistency can be witnessed through stakeholders' irritation, and, in some occasions even anger.Most importantly, this inconsistency prevented the company from regaining the public's trust.As evidenced (e.g., Table 2: BF2, Table 3: Q4, Q5), several users expressed concerns to deal with the company in the future, fearing of being treated similarly in case they, too, felt the need to complain regarding their consumer experience.Such concerns are in fact central to the definition of trust, which entails both parties' "intention to accept vulnerability" [46], and which, similar to reputation, is a form of social capital [33].
Trust however, is essential for every company.It has been argued that trust is significantly affected by reputation and therefore an injured reputation can have a long lasting effect on trust [47].Admittedly, it is quite difficult to restore both trust and reputation, and previous studies have shown that issuing an apology can help towards it by essentially initiating a successful post-crisis communication campaign [41].In essence, an apology suggests that the company in question takes the responsibility and regrets for its actions that led up to the crisis, and that it deals with its any shortcomings [24].In this regards, CompanyX did issue a relevant announcement, which was published through its official website and twitter account, in which they were apologizing for the events, and stating that the lawsuit was in fact an excessive measure toward defending their reputation.The company was also apologising for what took place as a result of the lawsuit, but not for what had happened before it, i.e., the perceived as poor service from the behalf of the consumer.However, since the locus of the crisis was considered by the various stakeholders to be internal, this apology-centred approach, which was combined with an overall denying responsibility strategy, was not particularly effective in restoring trust or repairing the company's reputation, especially because the company operates with the service sector [41].
Furthermore, with regards to reputation, it should first be noted that, in web 2.0, reputation is both more ephemeral and more enduring; it is ephemeral by being more vulnerable to easily accessible negative comments, and more enduring since any (negative) information, which is both retrievable and persistent, can be easily reproduced.In addition, traditional legal means, such as lawsuits, which are often used for printed media, can no longer help companies repair their reputation [1]; In this situation, even if the lawsuit had reached the court, and CompanyX had managed to secure the blogpost's retraction and the compensation they were filling for, their reputation would remain damaged.
Second, on a more abstract level, and keeping in mind that the lawsuit was filed claiming reputation damages, it should be noted that these require the injury of a relational interest, which encompasses extant and prospective relations and one's public image [6].Specifically, reputation itself is socially constructed through the interactions developed between the involved parties, it is defined as "a set of attributes inferred from the firm's past actions" [55], and depends upon the relationship between the company, which seeks to maintain its esteem, and its prospective and/or extant customers [6].As a result, and as far as the sued consumer is concerned, the company's reputation had already been affected due to his perceived consumer experience, prior to the lawsuit and regardless of the blogpost.When the company turned legally against him, their reputation got permanently damaged, because their action may be seen as an attempt to attack the accuser, since the company sought to silence the consumer, by seeking to suppress the perceived as defamatory information.
Further on the damages to the company's reputation and drawing from our study's findings, one could argue that Mr Papadopoulos could be indeed held partly responsible for damaging CompanyX's public image on the online world; indeed, some stakeholders argued for CompanyX (e.g., Table 3: Q9).Nevertheless, as evidenced through the majority of the reactions, the public held the company itself responsible for dissolving any extant relations and for impeding prospective ones with those requesting a comment or its viewpoint on the matters.In more detail, the company entered into a situation in which they treated the various stakeholders poorly by accusing them as a whole for the negative online criticism and for being biased against them.As a result, the reputational injury was further exacerbated, since CompanyX chose to severe ties with the online users, and any extant or prospective customers.

Conclusions
Social media have become platforms of unlimited connectivity among consumers, where every brand, product, or service "can be discussed, argued over, researched, and bought as part of a vast conversation among the people interested in it" [50].As a result, the impact of consumers has changed tremendously; they can connect and support each other instantly, ubiquitously, effortlessly and effectively, while one is in position to reach out to a thousand others and her/his experience may have a serious effect on a company [52].Further more, since word-of-mouth ranks among the most important information source[s] when a consumer is making a purchase decision, the influence exerted by other consumers is especially important.This influence is even more significant when the product or service is either intangibleand thus it is difficult to evaluate it prior to purchase [36], or a high involvement one (as is a high-priced desktop computer and its timely and satisfactory service), when the consumer goes through a more detailed decision process, looking for all pertinent information and considering several evaluation criteria [4].
What is remarkable in the case of CompanyX is that, instead of using social media to reach out to its customers in order to handle a possible reputation tragedy, the company chose to suppress the information and refuse further interaction with its extant and potential customer base.This relates to the main contributions of the particular study; the very decision of filling a lawsuit significantly differentiates this case from similar ones, during which consumers Next, with regards to the comparison between the case of CompanyX and that of the Love Parade crisis, again we consider that there are several differences.To begin with, the Love Parade was a techno music festival, held in Duisburg, Germany in 2010, during which there were several fatally wounded people, as a result of panic.After the incident, the festival's organizers, the involved local authorities and the police began shifting responsibility for the various security failures.In other words, all the involved parties adopted the same crisis response strategy, by denying control over the situation and blaming others.As a result, the general public attributed responsibility to all of the involved parties, but primarily to the local authorities and the police, and to a lesser extent to the festival's organizers.When it comes to the case of CompanyX, however, the result regarding responsibility attribution is significantly different, despite the fact that a blame game also took place.CompanyX, the authorized distributor and the authorized reseller similarly accused each other, and attempted to shift the responsibility both for servicing the computer and for the course of events.In addition, all parties chose to refrain from commenting and interacting with the public in any way, while there was a wide disparity in the company's adopted behavior, as CompanyX endeavored to attack their accusers and deny responsibility altogether.As a result, CompanyX attracted all the criticisms and the responsibility for the situation was not distributed among the parties.In the case of the Love Parade, Schwartz [49] argues that, potentially, because the festival organizers refrained from making frequent statements and issuing announcements, the public criticized them much less.Nevertheless, this did not happen with CompanyX.
As a result, this study enriches extant literature by highlighting the importance of an appropriate response strategy, and by discussing the possible outcomes when the involved parties fail to understand the social media landscape.CompanyX ignored the manner and the extent to which social media platforms can affect their reputation.In addition, as illustrated through this case, the company reacted in haste, ignoring reputation management policies and choosing excessive measures toward handling the incident.As stakeholders attributed the crisis internally, thus placing it within the preventable cluster, it was expected that the involved parties, and CompanyX in particular, would adopt a response strategy, which would allow them to quickly and effectively rebuild their reputation.Drawing from past cases, this could be achieved by either issuing immediately an apology, offering some form of compensation, like United Airlines, and eventually, by engaging with their audience.In contrast, the company on the one hand abstained from interacting with online users and from commenting regarding the incident, while on the other hand, chose to attack, first offline through the lawsuit against Mr Papadopoulos, and second online, by suggesting that all critics were biased.Therefore, the chain of events and the adopted behaviour by CompanyX caused e-WOM to have an important, negative impact on the company's reputation.We consider this investigation as particularly important for crisis management and social media researchers and practitioners alike, since the proliferation of OSNs and social media leave little room for a company to react without having considered carefully all available options in the face of reputational damages.
Beyond the issues pertaining to e-marketing, public relations and social media, our findings highlight important implications for the field of e-commerce as well.Prior studies have adequately illustrated that it is typical for prospective customers and consumers to seek reviews, both positive and negative, in order to decrease the decision-making time and the perceived risk, and develop informed buying intentions [2], [25].Furthermore, Kuan and Bock argue that trust generates "through the social relations and networks of an individual, especially when he or she has had no prior interaction with the trustee" [34].Within this context, and as illustrated through the presented case, several individuals propagating the story and supporting the sued consumer or simply observing the chain of events were not extant customers of either the physical or the online store of CompanyX.As a result, they did not have prior knowledge of the company's services and products, other than that available through e-WOM, which in this case was particularly negative.In short, this, combined with the transmi[ssion] of a cold-shoulder message [21] from the behalf of CompanyX and the damaged reputation, which unquestionably affected the stakeholders' trust, acted against building or restoring consumers' trust and the formulation of purchase intentions in general.
Even today, the information on the incident is still retrievable and persistent.More than a year later, on January 14, 2012, a simple web search on CompanyX returned among the top ten search results, five listings referring to the lawsuit, while only three related directly to the company.Two years later, at the time of finalising our study, the same web search returned among the top ten search results, two listings referring to the case.In short, anyone looking up information on CompanyX can find readily available information on the particular incident, as a result of a persisting Therefore, we consider that there were several avenues the company could have taken and which could assist toward minimizing the persistent echo effect of the incident.For example, Bernoff and Li [3] have examined a number of brick and mortar and brick and click companies hit by a PR crisis, and it appears that, in their majority, they managed to restore their relationship with customers by actually interacting with them and inviting them to actively participate in dedicated blogs and social media platforms, aiming to collect insight on their products and/or services.While such business decisions may prove to be too resource-consuming for small-to medium-sized companies, the main rationale remains the same; a company's participation and proactive approach to user comments on social media platforms can be a valuable step toward minimizing public concerns.Such approaches are even more valuable when crises occur essentially within the online world, as they cannot afford not to connect with online communities during incidents of a PR crisis.As online users share information with or without the company's consent, it is up to the company in question to decide whether this will occur with its participation, by communicating online with customers, or without it, ignoring social media altogether [30].Both choices have a significant effect; yet not engaging with one's audience can only lead to self-isolation and as others have noted, there are already many online outlets that uncover companies' failures, by being openly critical [52].This has resulted in many companies building an active collaboration with bloggers and consumers, as for example Dell Inc. who succeeded in developing an online community, with whom the company engages toward resolving any technical failures, thus minimizing customer support costs and any negative impact on its reputation [3].Consequently, specifically for crisis management, and contrary to how CompanyX handled OSNs, web 2.0 offers the opportunity to promptly update any stakeholders with both sides of the story, which can be much more effective than an even successful effort to suppress or retract a defamatory statement [1].In more detail, CompanyX's twitter updates suggested that they considered Twitter as an ineffective communication tool and the majority of online users as biased.Nevertheless, having an already active blog and the official website at their disposal, the company could have directed users there, and, in turn, the latter could form an overall impression of the situation, forming indeed an unbiased opinion.
Concluding, in an age when extant and prospective consumers continuously reach out to each other for support and information, a company cannot afford to refrain from web 2.0 or ignore its principles.Developing a direct dialogue with online users can help a company be proactive in the case of a PR crisis, by adopting an empathetic approach.At the same time, a firm can leverage its online social presence to smooth out the echo effect deriving from a crisis and use e-WOM to its advantage.Even though online presence may give room for negative comments, refraining altogether, controlling or suppressing offensive information can lead to additional damages.As a result, companies need to accept loss of control over their brand and actively collaborate with consumers.Most importantly, when crises occur, companies need to carefully examine all available options and design their response; wrong PR tactics can leave behind an echo effect, damaging a company's status permanently as information nowadays is persistent, retrievable and ubiquitous.

Table 2 :
List of online sources

Table 3 :
List of representative tweets

Table 4
summarises our findings by juxtaposing the recommendations of Attribution Theory and the outcomes of the case.

Table 4 :
Type of crisis, response strategy and their impact on CompanyX's reputation