Creating and destroying b2c stakeholder relationships on Facebook - a case study of four Finnish companies more

A paper presented at EUPRERA congress 2011 – Public Relations in a Time of Turbulence Leeds, UK, September 2011 Creating and destroying b2c stakeholder relationships on Facebook Salla-Maaria Laaksonen & Susanna Neiglick University of Helsinki salla.laaksonen at helsinki.fi Abstract The purpose of this paper is to study current practices of Finnish b2c organizations’ Facebook use in building stakeholder relationships. We utilize Hallidayan systemic-functional grammar, PR strategies proposed in van Ruler (2004) and reputation categories proposed in Weigelt & Camerer (1988) to analyze the Facebook Pages of four Finnish organizations. Our key findings are that marketing and customer services are the main areas to which Facebook Pages are utilized at the moment in the organizations studied. Although there is an attempt at dialogue, the amount of it is lesser than expected within the otherwise social arena. Difficult topics are approached with a tactique of non-response. Content is a key factor defining the amount and quality of dialogue. The study has practical implications in organizations planning to or currently running their Facebook Pages in Finland, and academical value in comparative analysis of Facebook Page practices. Keywords: Facebook, relationships, public relations, reputation Introduction PR professionals' web usage has been seen as effective in monitoring business environment (see e.g. González-Herrero & Smith 2008), and it is feasible to suggest that as the importance of social and ubiquitous media rises, the focus will increasingly be on engagement. Social media services give organizations and their public relations practitioners a unique opportunity to collect and disseminate information, monitor stakeholder opinions, and to engage in dialogue with their stakeholders. With over 500 million active users, 50% of which are daily visitors (Facebook 2011), the popularity of Facebook has made it an important tool in business communications as well as in web audience and e-participation research. In this article, we study the current practice of Finnish b2c organizations' Facebook use in building stakeholder relationships. We assess the communicative means utilized by four organizations organizations and their members, chart out their communication strategies, dialogue and participants involved in interaction. Our approach is based on the assumption that Facebook and other social media services have increased the visibility of previously undetected multifocal communication, and also made it more frequent, thus changing the communicative environment in a way that requires a novel approach to organizational communication practices. We build on the view that Internet users and organizations meet within virtual arenas and that these encounters give meanings to the organization (Aula & Mantere 2008). With the ascent of social media, the formation of reputation has also become an increasingly communal process. In this novel, often uncontrollable environment where reputational stories can equally well be built and mass-distributed by the public or other interested parties as by public relations professionals, the practices of reputation management and public relations face the challenge of keeping up with the flow as well as being reactive and proactive. We assume that in this new situation organizations need to communicate and engage bilaterally with their stakeholders (de Bussy & al. 2001), and argue that this is especially true in a usecontext where the stakeholders are communicating visibly with one another (see also van der Merwe & al. 2005). With our data-based analytical approach, we also wish to either confirm or reject our previous arguments that in digital publicity, reputation and reputational stories are increasingly built by the public, and that they would thus require a dialogic approach (Aula &. al 2010). Facebook is an interesting platform in this sense, since its pages are created firstly as an arena for virtual visibility for the organization, but added to this there is a more democratic networking layer, where all Facebook users have the option for commenting and sharing posts. It is up the organization to decide whether they wish to open their page’s wall for audience comments as well. All four organizations studied had taken advantage of this openness. We have to keep in mind though that we have no way to tracing possibly censored or deleted individual posts afterwards. Even though discussions and thought on interactive, bilateral and consensus-building communication with stakeholders are not new (see. e.g. Grunig & Grunig 1992), and the possibility of multifocal stakeholder communication has been brought up by researchers and practitioners alike, it is only in the recent few years that the situation has changed in practice. Usage of social media services and especially Facebook has grown significantly in developed countries during the past few years, and also continents such as Africa and South America are showing a rapid increase (Socialbakers 2011). Theoretical background and analysis methods On a text-analytical level, we are primarily interested in 1) charting out the communicative practices and roles of both the representatives of an organization, as well as other participants, and 2) in building a comparative, data-based framework for analyzing different strategies. Besides looking into actors and roles, we build on systemic-functional grammar, which divides textual meaning into three, co-existing levels; the ideational, the interpersonal and the textual level (Halliday & Matthiessen 2004). We looked into features such as proactivity, reactivity, ideologies expressed and text-level influencing methods used. To chart out interpersonal communication practices, we also tracked the level of informality as evidenced by Finnish morphological constructs, typographical, as well as stylistic features. Systemic-functional grammar and the three metafunctions Figure 1: The three metafunctions of language (Halliday & Matthiessen 2004). Halliday & Matthiessen (2004) define language through three metafunctions, illustrated in Figure 1. First, they see it as a tool that constructs human experience through categories which turn into taxonomies, thus providing a theory of human experience. This aspect of language is the ideational metafunction, and divides into two components: the experiential (representation of perceived processes) and the logical (representation of the relationships of those processes) (1999; 2004). Second, they continue that simultaneously with the ideational metafunction, or “language as reflection”, language is also used as a means for enacting personal and social relationships; it is utilized when persuading and expressing attitudes towards addressees or issues discussed. This aspect of language is the interpersonal metafunction, “language as action”, both interactive and personal. Third and last, they define the textual metafunction, which relates to the way in which texts are construed. This third metafunction can be seen as the building block of both the ideational and the interpersonal metafunctions in that it helps construct sequences of discourse by creating cohesion and continuity. It also relates to our competence to form, evaluate and follow texts beyond sentence-level (2004). The Hallidayan perspective thus focuses on the structural, lexicographical, aspects of language as the means of realizing interaction and thought, organizing communication and experience through language. We utilize aspects of the textual metafunction to carry out our analysis. The ideational metafunction is present in our analysis through different theory-based categorizations, which represent the metalevel processes. Similarily, the interpersonal metafunction is accessed through the mapping for conversational roles and emotional content representing social relationships, positions and communicative attitudes. The textual level is our passage to these more sophisticated levels of communication and sensemaking. When coding the dataset, we look into the actors and roles within each page, discerning various organizational roles: initiators, responders, complainers. We also mapped the formality and informality of the posts evidenced by the usage of smilies, slang and punctuation errors. In our material, these lexicographical aspects illustrate and realize the ideational and interpersonal metafunctions of language, to which the analysis in this study thus mostly focuses. Public relations strategies To chart out strategic conversation methods used by the organizations we compare our classified data with a macro-level framework proposed in van Ruler (2004), of four different strategies of public relations: Information, Persuasion, Consensus-building and Dialogue. According to van Ruler (2004), one-way communication can be either informational or persuasive: either disseminating a flow of information and transmitting objective information or disseminating a flow of influence in a case where an attempt by a sender is made to produce a predefined attitudinal change in the receiver. Two-way communication can be either connotative or denotative: communication in this view is seen as an ongoing emotional process of co-creating meaning in a dialogic manner (dialogue) or as rational co-creation of new meaning (consensus-building). The two-by-two field depicting this approach is shown below in Figure 2. Figure 2. The communication grid according to van Ruler (2004). Creating an information strategy refers to providing information in order to help people form ideas, opinions and discussion. Most traditional public relations materials such as press releases fall into this category as they are initially made to inform. Persuasion strategy is the basis of advertising or propaganda, aiming to create a favorable basis for future relationships with the stakeholders and also targeted to change the opinions and behaviors of the receiver. Dialogue strategy is seeking consultation with the stakeholders, and is conducted by sending interactivity and informational messages on both sides. Consensus-building strategy deals with building bridges between the organization and the environment, and can be deployed when there are conflicting interest at stake, to cover a mutual agreement (ibid.). Van Woerkum & Aarts (2008) call consensus-building-like activity negotiation, which in our opinion covers the idea much better. Reputational interplay To cover the perspective of reputation, we traced the interplay between different types of reputation: company reputation, product reputation and corporate culture reputation (see Weigelt & Camerer 1988), combining these to issue reputation and stakeholder reputation proposed by Mahon & Wartick (2003). These categorizations imply that in fact a company as a totality can have multiple reputations: for example a good product reputation as a service provider but a bad corporate culture reputation as an employer or a bad company reputation as an responsible corporate citizen. Mahon & Wartick’s addition broadens the scope a bit more by implying that reputations of the stakeholders are also involved, and that the issue at hand can affect the course of the conversation. The writers remark that in each actual situation there is an interplay with all different types of reputation and reputation management practices should take this into consideration. Thus, practitioners need to assess their reputation in relation with the issue and process at hand, and in relation to the stakeholders present (ibid.). In practice, this is one way of categorizing what stakeholders are talking about on corporate Facebook pages and also what aspect of their reputation companies are aiming to build with their own updates. Even though reputation is a concept primarily referring to the evaluations of the publics, here we classified messages from both the organization and the participants as reputational talk, since the organizational messages can be seen as means of building reputation. However, we included only posts and comments clearly discussing one of the reputational types. Many actions of normal PR or organizational communication are building reputation, even if it's not explicitly and overtly mentioned in the data. Emotionality Our final interest deals with Facebook page conversations, stakeholders and stakeholder emotion. Here we use the categorization proposed by Luoma-Aho (2009, 2010), who calls the most committed shareholders faith-holders and accordingly the most mistrusting hateholders. Both these groups have strong feelings towards the organization or it’s actions and are keen to express their feelings publicly and colorfully, for example by using social media. Drawing a bigger picture on human behavior, Luoma-Aho (2009) notes that the level of public involvement in organizational affairs is changing. Publics no longer only want to participate, they also show and express emotion. This is clearly visible on a variety of increasingly personal social media platforms. Further Luoma-Aho argues (2009, 1) that “we have moved into a time of emotional publics, where feelings toward organizations range from love to hate, and the different stakeholders have several ways of showing their emotion and recruiting others to join in and comment on their feelings”. In addition, emotional stakeholders should be regarded with high caution due to spreading word-of-mouth. Research shows that emotional content is being spread and is shared more likely both in offline and online social networks (see e.g. Berger & Milkman 2010, Hansen & al. 2011, Sweeney & al. 2005). Thus there is a need for an organization to recognize emotional stakeholder at an early stage. According to Luoma-Aho (2009), faith-holders are often found as surprise to the company, and companies are having hard time anticipating and profiting from them. Nevertheless, storytelling faith-holders are a handy resource for an organization, because they act as reputation ambassadors for free, and are also more convincing in the eyes of their peers. Non-marketing sources are given substantially more weight by the customers when evaluating information on product and services and making decisions (e.g. Herr & al. 1991). Thus communicative acts performed by the customers or page likers themselves are more likely to been seen as trustworthy or considered more thoroughly when forming opinions. In order to trace the faithholders and hateholders in this study, we coded expressions of positive emotional content (e.g. laughing, smiling, positive locations and events, and commonly positively regarded things such as beautiful beaches). Respectively, negative content was categorized containing negatively perceived things (e.g. crying, moaning, screaming, disgusting objects, fighting or arguing). After the analysis, we reflect through these emotional stakeholder categories and see if we can identify “faith-holders” and “hate-holders”, the most committed and most mistrusting participants. Data selection In order to study the current practice of Finnish b2c organizations' Facebook use in building stakeholder relationships, we chose four different Finnish corporations’ Facebook pages as our data. We chose to study the Facebook pages of Valio, Blue1, VR and Suomen Yleisradio (YLE). Our main focus was on b2c organizations, as we find that customer-centered marketing enabled by social media plays an especially prominent role within these companies. In addition, all of them are popular Facebook company pages as measured by the count of likers (the term Facebook uses for people following the updates of a page). More detailed information of the organizations studied is viewable below in Table 1. The dataset under analysis was coded using ATLAS.ti and spans March 2011. In total, it contains 142 wall posts of which organization-initiated were 69, as shown in Table 2 along with other initiator statistics of the pages studied. Coding categories were partly based on theories proposed earlier and partly grounded in the data. When coding the data we first tried to give our chosen categories to all wall posts, but soon we noticed that the conversations are far too complicated to be listed in such clear units: it is possible a conversation on a totally different topic is started below a wall post. Thus we ended up coding our categories somewhat freely, which lead us making an exploratory analysis with statistics that are not actually comparable with each other and not generalisable due to the small size of the sample. However, we choose to use the percentages to give a good overview of the data and to compare the cases. The interplay with different categories was studied using the Co-occurence Table Explorer tool provided by ATLAS.ti. For two codes, the cross table shows the count of their co-occurrence in all used documents and a normalized coefficient of probability. Blue1 Blue1, The second biggest airline company in Finland, flying to 29 different destinations within Europe. Blue1 started as a domestic low-cost airline but has been expanding and creating a broader brand since. However, the flight pricing is still following the conventions of lowcost airlines. Blue1 is owned by SAS Group. Valio Valio, a company owned by Finnish dairy farmers that secures milk production in Finland and the vitality of the nation's countryside by processing milk into tasty products that promote wellbeing (Valio 2011). Thus a corporation operating clearly on consumer business, producing aliment products. Valio has been actively promoting their presence in social media and their new social media strategy on 20102011. An issue regarding Valio’s connections with nuclear plant building in Finland has been lately brought up by ecological NGO’s, especially Greenpeace, and this has provoked a great deal of conversation on Valio’s social media sites. Likes: 42 230 http://www.facebo ok.com/blue1.fi VR VR (Valtion rautatiet), Finnish state owned railway transport company in Finland. VR is a monopolistic freight and passenger carrier in Finland. Their Facebook site VR - Yhteisellä matkalla (“travelling together”) focuses on commuter transport and customer service. The communication is usually (?) carried out by a character called “Kivimies Mikko” (Stone Man Mikko), referring to the two stone figures of Helsinki main railway station used frequently in VR marketing and advertising. Yleisradio Suomen Yleisradio (YLE) Finnish public broadcasting service company. Yle has been actively pursuing social media strategies since spring 2009. They are present in Facebook through tens of different pages dedicated to different channels, individual programs and web services. We choose to examine here the main page of the company itself, where the page content is also less related to linking Yle’s own content elsewhere. Yle is also actively using other social media services, such as Twitter, YouTube, MySpace and blogs. Likes: 17 784 http://www.faceboo k.com/valionsivu Likes: 11 270 http://www.facebook.com /VRyhteisellamatkalla Likes: 4 798 http://www.facebook.co m/suomenyleisradio Table1: Organizations studied included Valio, Blue1, VR and Suomen Yleisradio (YLE). The strategies of initiating discussion We categorized all wall posts on each company page studied according to the actor initiating the conversation by making a Wall post. Top-level categories are the Organization and Others. The Others-category was further divided in Customers, Citizens and Other organizations. Customers were recognized due their post content that was clearly related to their recent or forthcoming interaction with the company. Citizens were ordinary people posting content related to company or other issues. Other organizations are other Facebook pages that are somehow affiliated with the company who owns the page. In the total dataset, post were divided surprisingly evenly between the two top-level categories of Organization (49%) and Others (51%). During March 2011, Yleisradio was the most active poster or initiator of discussion (31 versus 16), whereas in VR case most of the intiatives came from the users (mainly customers, 18 versus 7). Initiator Organization Others Customer Citizen Other Organization Employee Others Total 9 9 1 0 19 6 13 1 0 20 14 3 2 0 19 2 3 9 2 16 31 28 13 2 74 21% 20% 9% 1% 51% Blue1 20 Valio 11 VR 7 Yleisradio 31 Totals 69 % 49% All total 39 31 25 47 143 100% Table 2: Organizations initiated 49% of the discussions in the four Pages studied Formality and informality The formality and informality of the posts was studied by looking into the usage of smileys, slang and errors in punctuation and spelling. Unsurprisingly, the comments made by page fans were more informal than admin comments or initiated wall posts. Emotional states were expressed by smileys ranging from happy to sad and surprised and by the usage of exclamation marks and capital letters. It seems increasingly common to use smileys as sole sentence-enders in Finnish web writing, as evidenced by the co-efficient 0.30 between missing sentence-end marker and the use of smileys. From the companies studied, Valio’s textual approach in replying to comments was most formal, starting with a clear letter-type opening greeting and ending with a signature and explanation of where the responder comes from within the Valio organization. Valio’s representatives also used less smileys and other informality-signifying markers, such as multiple exclamation marks. Valio representatives mostly faced questions about products and responded them promptly, usually within the day the question was asked, with one exceptional topic: nuclear power, which we will cover more deeply later in this paper. Yleisradio representatives are less formal than Valio’s and undersign their messages with first names only and dropping opening greetings, whereas VR representatives normally begin with a greeting similar to Valio’s but is less formal in the letter-type ending, undersigning it with just “mikko”. Blue1 representatives do not indicate who speaks. When looking at the co-occurence tables produced by ATLAS.ti, it seems that in the content produced by the organization’s more informal speech is most often used in customer service. Most often informal speech is produced by the likers. Majority of the fan comments on pages are short opinion statements thus very informal; “Yes, a good product” “I watched the show and liked it” “Please go to Pula, it’s a nice city” - -or the opposite “didn’t try it”, “don’t much care about that”. Organizational functions and public relations strategies To study the communication practices used on Facebook we first classified all organizationwritten content in our data using different organizational functions of Communication, Marketing and Customer service. In practice, each wall post was coded to represent one of these three functions. In addition, from the functional point of view, we recognized a fourth distinct function of Dialogue, which might be placed between communication and customer service. When in dialogue, an organization is engaging in a discussion concerning it’s substance or service substance with its stakeholders, who are e.g. contradicting the organization’s actions. Dialogue can be started by the organization or by a question from a stakeholder. Thus we see that this form of conversation rises beyond customer service due to its more generalistic, conversational and negotiation-like nature. In the dialogical category of our data the organization is forced to explain its actions to its stakeholders. These categorizations were only applied to posts and comments sent by the organization (including roughly 50% of the total amount of wall posts as described in Table 3). When looking at organizational functions, Facebook pages are mostly used as customer service platforms (46% of total), and after that as marketing platforms (26%). Functions Customer Service Dialogue Marketing Communications Totals Blue1 18 3 16 5 42 Valio 18 6 7 14 45 VR 27 3 5 1 36 Yleisradio 22 7 19 12 60 Totals 85 19 47 32 183 % 46% 10% 26% 17% 100% Table 3: All wall posts divided in to categories of organizational functions. Next, we compared our classified data with a macro-level framework proposed in van Ruler (2004), of four different strategies of public relations: Information, Persuasion, Consensusbuilding and Dialogue, which were also coded to all wall posts and comments initiated by the organization. One wall post could be given more categories than one if multiple strategies were present. Since we were interested on the company behaviour, we choose not to classify the strategies used by the likers and commenters, though it probably would have offered some interesting insights to the actual dialogue. Persuasion is the most common strategy used in our sample in general (36% of all messages), but all strategies were well present, as can be seen from Table 4. Thus it seems that Facebook can easily be used as an effective platform for multiple models of communication. Communication GRID Consensus-building Dialogue Information Persuasion Totals Blue1 11 8 4 16 39 Valio 4 7 4 12 27 VR 7 2 4 4 17 Yleisradio 4 8 20 15 47 Totals 26 25 32 47 130 % 20% 19% 25% 36% 100% Table 4: All wall posts divided into categories of public relation strategies proposed by van Ruler (2004) By using the code co-occurences provided by ATLAS.ti, we took a look at the interplay between organizational functions and communication strategies. All frequencies by cases are shown in Table 5. The analysis shows that in customer service the commonly used strategy is consensusbuilding and that in marketing the commonly used strategy is persuasion. Secondly it seems that communications relies on information and partly on persuasion. Dialogue function, as expected, uses a dialogic strategy. Function / Strategy Consensus-building Dialogue Information Persuasion Customer Service 36 - 0.31 7 - 0.07 1 - 0.01 2 - 0.02 Dialogue 4 - 0.10 6 - 0.16 n/a 3 - 0.05 Marketing n/a 3 - 0.04 12 - 0.20 32 - 0.52 Communication 2 - 0.04 3 - 0.06 13 - 0.25 11 - 0.16 Table 5: Co-occurence probabilities with organization functions and public relations strategies. Dialogue In addition we marked the number of comments and likes to each wall post in the data and compared these to other categories. These results show firstly that 1) most likes go to the post initiated by the organization (normalized coefficient probability 0.56), while 2) posts initiated by the likers get only random likes as positive feedback from other likers (the normalized coefficient probability is 0.55 between zero likes and post initiated by others). This is most likely caused by a Facebook feature of showing only the organization’s own wall posts in users own news feeds. Thus the user would have to go to the actual page to see the postings made by other likers. Secondly the co-occurrences show that the function of marketing is receiving bigger amounts of likes in the data than other functions. The typical number of comments in the posts in all data studied ranged from 1 to 5. The longest thread in the data was 18 comments, and it was related to the city of Edinburgh, a marketingrelated initiation from Blue1. It’s notable that most of the longest conversations had to do with Valio’s nuclear power funding. It seems that VR has the most passive likers, who rarely comment or like the updates - even compared to Yleisradio who has significantly less fans. Comments (n): Blue1 Valio VR Yleisradio Totals % 0 1-5 6-10 11-18 12 20 5 2 8 21 3 5 7 19 0 0 18 21 5 1 45 81 13 8 31% 55% 9% 5% Table 6: Number of wall posts receiving n amount of comments. Reputational interplay To cover the perspective of reputation, we traced the interplay between different types of reputation in our data: company reputation, product reputation and corporate culture reputation (see Weigelt & Camerer 1988), combining these to issue reputation and stakeholder reputation proposed by Mahon & Wartick (2003). The code frequencies for all reputational types are listed in Table 7. The most discussed category of reputation is product reputation (60% of all reputation related content). To get a broader scale across our four different organizations, all discussions on corporate service reputation are classified as a part of product reputation. In fact, from the companies studied only Valio is clearly a product-orientated company. In Facebook and other social network service contexts, it’s important to note that also the fans and customers commenting on the page can have their own reputation, which affects the way a message is interpreted. Thus stakeholder reputation is referring to discussions where the reputation of the content produces other than the organization is at stake. Stakeholder reputation includes both their individual reputation and their reputation with regard to the given issue and to the process at hand (Mahon & Wartick 2003). On Valio’s page, for example, there’s one active person remarking on the Valio nuclear power case, and he clearly is getting a reputation of being a difficult person since other fans are commenting on his posts and asking him to stop. In frequencies, the Valio’s nuclear power case also shows in the high number of issue reputation related posts. VR case shows clear emphasis on content related to product reputation (or service reputation since VR is a train service provider). This probably is partly caused by a very difficult and snowy winter that caused a lot of technical problems for VR services. Issue was brought up extensively by traditional media and also by online audiences. During the winter, for example, there was a website dedicated to VR related failure stories people could submit. Two of the companies studied use Facebook actively as a recruiting tool (Blue1 twice and Valio trice). Thus we noticed that an extra category of employer reputation would have been useful. Here these activities are categorized under corporate culture reputation. Reputation types Product reputation Stakeholder reputation Company reputation Corporate culture reputation Issue reputation Totals Blue1 24 1 2 2 3 32 Valio 35 7 17 1 29 89 VR 19 0 2 0 0 21 Yleisradio 33 0 9 2 0 44 Totals 111 8 30 5 32 186 % 60% 4% 16% 3% 17% 100% Table 7: Different types of reputation in the data Discussion Observations by case Blue1 In the Blue1 case the issue reputation, thus the issue at hand, is travelling, which is a rather generally positively viewed topic. This shows clearly as it is easy for Blue1 to initiate conversations on their Facebook wall by just asking their fans’ opinions or travel experiences, or merely by mentioning a certain travel destination or a new route. Fans seem to react rather spontaneously, and write about their own travel plans especially if their flights are by Blue1. Blue1 supports these kind of conversations by actively using dialogic initiatives, often included in a more persuasive marketing message. Besides travelling in general, issues include certain destinations or e.g. weather. Product reputation discussions are about routes and prices, and service reputation conversations are affiliated with customer service contacts. Compared to for example the activist attention gathered by Valio, on Blue1 page there are no conversations on environmental issues related to flight travel. Blue1 uses their Facebook page mostly for marketing and for customer service. Faithholders or hateholders are not recognizable. Valio Valio’s Facebook page hosts several conversations on their products, often dialogically initiated by the company but several times also started by a customer asking a question. The companyinitiated questions bring up more responses and create longer conversation than comments by individual fans. Valio’s products are mainly different dairy products. The theme of milk in general and the good reputation of milk is very much present in the conversation. Milk as such is a very powerful brand in Finland, easily related to health issues and even to nationalistic values. The strong presence of milk as an ideology is even related to idealization of agricultural life: praising everything related to original natural production and old countryside professions such as dairy farming. Thus faithholders on Valio’s page are not necessarily related to the company itself, but to milk. For example in a post by a liker the earthquake victims in Japan are surmised to "need clean milk", so milk is used as a way to address reader's feelings. Several stakeholders are mentioned on Valio’s initiatives: MasterChef television program, which Valio sponsored during the spring (the events of the show are discussed as well as new shows advertised weekly); Elintarviketeollisuusliitto, whose competition included Valio’s product in the finals; poet Elina Karjalainen, dairy farms. This is related not only to marketing as arranged in the affiliation contracts, but also to using stakeholder reputation as a positive support in communications. Valio was an interesting case since during the research period the company was involved with a public discussion of funding new nuclear plants in Finland. Certain NGO’s, mainly Greenpeace, commanded a public social media attack on organizations involved in nuclear power plant building in Finland in late 2010 - May 2011. Since late November 2010, Valio has taken an approach of non-response in respect to the nuclear power activists and concerned citizens posting to Valio’s wall asking about Valio’s ownership in Fennovoima, a company holding a licence to build a nuclear power plant to Finland. The activists for and against nuclear power seem to have highjacked the Valio Wall and conversate among themselves on it. The conversation rely on just few commenters. Nevertheless, we assume this may have had influence to the fact that at some point Valio changed their landing tab (the first page a new visitor sees when entering the Page) to not display the Wall but to take the user to a tab, a feature that is no longer available on Facebook. The nuclear power conversation deals with company reputation and issue reputation: rather a company reputation related to an issue, but clearly mostly concerning individuals who are highly committed to the environmental value behind the issue at stake. Company reputation is at stake when likers are talking about the actions (or non-actions) made by Valio; issue reputation when argumenting the backgrounds of the writings. In these posts, Valio is often compared to Atria, another Finnish alimentary product company, thus we might say that competitor reputation is also present in the conversation. In Valio’s case, we can recognize some sort of hateholders as participants of the nuclear power conversation. However, there seem to be a lot of likers approaching the faithholder condition as well, and clearly the contradiction between Valio and activists is also encouraging them to stand up and express their faith towards the company. We might also argue that the faithholders and hateholders are actually issue-specific, relating to nuclear power. Valio seems to have a clear fan group related to milk production and another related to their products. There are a few enthusiastic likers who comment shortly on almost each product related posts initiated by Valio. They are not necessarily contributing much information in the discussion, merely commenting yes or no to product related questions initiated by the organization and thus participating the dialogue. VR VR page has the biggest amount of customer feedback posts, but less actual complaints than we would have expected after a difficult winter and an excessive amount of technical problems. In general, a negative voice is present in many of the conversations and there are only a few positively toned posts from the likers. Even though the company was a foregoer in Finnish social media scene, VR is the least active initiator of all the companies studied. They are replying customer to initiatives rather actively, but not punctually. Nevertheless, VR shows that dialogue in it’s lightest form on Facebook can be clicking the like button for customer posts. VR is trying to use participative means by for example collecting questions for an interview in their customer magazine and carrying out competitions, but often these initiatives are left without any feedback or evoke angry comments from people. This leaves VR in the mode of building consensus and mainly doing customer service through Facebook. Figure 3: VR participating in light dialogue. Yleisradio: Yleisradio was the heaviest linker of all the companies studied. Yleisradio has a clear advantage compared to other companies studied since they have the opportunity to link to their own products (i.e. TV shows), which are available online and free of charge. Thus Yleisradio is doing a lot of product advertising, especially by linking to their online video service, YLE Areena, and thus initiating discussions on product reputation. Fans are replying quite eagerly too and giving good comments on shows they like. Even though the product evokes plenty of dialogue between the fans, Yleisradio hardly participates in these conversations. From the comment-replies on Yleisradio’s Wall you can clearly see that Yleisradio has a team of people administering the Facebook page. They have a clear code of adding the respondent name at the end of the comment when replying to show more personality. This also allows the corporate voice and the employee voice to differ every once in a while. Some active YLE employees are also commenting and posting on page using their own personal profiles. This is not a clear practice for a random liker, who doesn't recognize their names and therefore not necessarily understands they are in fact representing YLE. It is difficult to differentiate between information and persuasion style in Yleisradio. Links to their own program content are presented in an informational way, but the objectivity of the content can be questionable since the introduction text is often formulated in a positive manner. However, a strictly factual information strategy is used with Fukushima news reporting and information of the TV news shows related - in these posts there is no copywriter-text around. All in all, Yleisradio uses a very free style when answering their customer feedback. This creates a feeling of a true conversation. A few times during the research period Yleisradio also opened their news production process to the likers. This offers a great channel for open views to corporate culture and work processes and thus builds corporate culture reputation and employer reputation. Company reputation often relates corporate responsibility issues since Yle is a public service broadcaster, commanded by the law to serve the Finnish nation. During our research period, Yle had a long discussion with a fan who begins by claiming that Yleisradio should have commercials, since that way they would have better content on offer. Page admins reply pertinently and explain issues with Finnish law, Yleisradio's history and research on media models. It seems that the inquirer is actually affected by the answers made. General observations We note that our sample is rather small to be a comprehensive analysis on the Facebook usage of Finnish companies. Our study shows, however, that at the moment in the organizations studied, Facebook pages are used more as a tool for customer service and marketing than as a tool for corporate communications or for having actual conversations. We met with less dialogue than expected. Often the case was that the likers or critiquers were left chatting with each other or that the organization might not follow up with the discussion it initiates. Most of the dialogic communication happened under customer-service-related posts. This study confirms that the content of the posts and the issues at hand have a great effect on the ensuing conversations - this shows when comparing the cases of Blue1 themed with holiday versus VR focusing on inland commuting. There is no reason why VR for example couldn’t engage their likers with train-related holiday themes, which in Blue1’s case spawned plenty of feedback. The amount of product reputation related discussion combined with the observation that most likes were given to marketing related content seem to confirm several earlier studies (see e.g. Morpace 2010) conducted by consulting groups telling that people follow company profiles on social media services in order to receive up-to-date product information and possible offers. The fact that most likes go to the post initiated by the organization while posts initiated by the likers get only random likes further confirms that customers following a company actually are expecting very practical and informational content let it be marketing or product-related tips. Many parts of the conversation fall outside the categories described here. Especially humorous content posted online is hard to precisely define to a certain category, even though it inevitably does have an effect on the corporate reputation. (e.g. on VR page 4 a person is asking what are the sanctions for driving the rails with a self-made trolley, and VR answers solemnly telling to ask the police since this kind of action is a crime). In the light of our study, the term hateholders seems to be a bit too harsh expression. Some fans are clearly hard to please and respond negatively, but they are not aggressive. Even the nuclear power opponents communicating on Valio's page are not being overly aggressive, rather questioning Valio's practises and demanding and answer they are not getting. Is non-response a good strategy in this case, then? The faith-holders are not getting much more attention in the conversation either, since none of the companies studied seems to communicate with positively commenting likers in a different way or note them in any ways. Conclusion and future research ideas How are reputations created and destroyed on Facebook Pages? In this paper, we assumed that organizations need to communicate and engage bilaterally with their stakeholders and argue that this is especially true in a use-context where the stakeholders are communicating visibly with one another. Is this so? Our data shows that building reputation on Facebook starts with stakeholder inclusion, regular posting to Page Wall and timely replies to questions asked. Via the great amount of stakeholder contribution on the pages we saw that stakeholders are clearly willing to use the opportunities of interaction offered them. Some of the companies studied are more active in this bilateral discussion, while some choose to maintain a more one-way style in their communication. Clearly a key issue in inspire conversations is the content offered by the initiative company. At it’s best, content can function as trigger to create versatile, dialogical stakeholder communication, and simultaneously offer important information and feedback for the company. But are all the conversationalists engaging in dialogue wanted? We saw that some examples of difficult stakeholder dialogue and organizational tactique of nonresponse, such as Valio’s in their nuclear power investment case and VR with their customer complaints, met with silence and no apparent retaliation at least as far as the venue of Facebook is concerned. We may only speculate with sales figures, as that data is not readily available for us. From the Valio case, we might further assume that non-response may be a feasible strategy when approaching a difficult topic, although this goes against our claim that a bilateral approach is generally the best option. However, Valio engaged in bilateral conversation on the topic before turning to the non-response strategy starting from 24th November 2010. Why the change? An interview with Valio’s representatives and a detailed look to their user statistics might be useful in illuminating the circumstances. During the composition of this article, we also noted a clear need to study the impressions of the likers of each page. Multi-user communities and online social networks that rely on users to contribute content all share one property: most of the users don't participate very much, they simply lurk in the background and majority of the content is produced by a small minority (see 90-9-1 rule or participation unequality law, Nielsen 2006). It is a very important to ask what is happening in the heads of the people who just read the content. In the future, it would be interesting to conduct a study with a more quantitative approach and for example take timestamps, like and comment numbers as variables to see what kind of posts create most feedback and during which times of the day. This data is already available for Facebook Page Admins. The limited access to this data is unfortunate, but we will continue to work on this analysing frame to build it as a schema for studying any online interaction that is based on a newsfeed structure similar to Facebook’s. It would also be interesting to compare reputational ratings given to the companies and especially stakeholder impressions and highlight them against the communication strategies utilized on Facebook pages. References Aula, P. & Mantere, S (2008). Strategic Reputation Management: Towards a Company of Good. Routledge, London and New York. Aula, P., & Laaksonen, S-M., & Neiglick, S., (2010). Reputational risks and the rising power of digital publicity. Paper presented at EUPRERA Congress 2010, University of Jyväskylä, Finland Berger, J., & Milkman, K. (2010). Social transmission, emotion, and the virality of online content. Wharton Research Paper. Available online http://marketing.wharton.upenn.edu/documents/research/Virality.pdf Blue1. Short information on the company. Retrieved May 17th 2011. http://www.blue1.com/en/fi/blue1/yrityksesta/ de Bussy, N. M., Watson, R. T., Pitt, L. F., & Ewing, M. T. (2001). Stakeholder communication management on the Internet: An integrated matrix for the identification of opportunities, Journal of Communication Management, Vol. 5 Issue 2, pp.138 – 146. Facebook 2011. Press Room, Statistics. Visited January 14th 2011. http://www.facebook.com/press/info.php?statistics González-Herrero, A., & Smith S. (2008). Crisis Communications Management on the Web: How Internet-Based Technologies are Changing the Way Public Relations Professionals Handle Business Crises. Journal of Contingencies and Crisis Management, 16: 143–153. Grunig, James E. & Grunig, Larissa A. (1992): Models of Public Relations and Communication. In Grunig, James E. (ed.): Excellence in Public Relations and Communication Management. Lawrence Erlbaum Associates, New Jersey, pp. 285–325. Halliday, M. & Matthiessen, C. (2004). An introduction to functional grammar, 3rd edn. Arnold, London. Halliday, M. & Matthiessen, C. (1999). Construing experience through meaning: a language-based approach to cognition, Continuum International Publishing group. Hansen, L. K., Arvidsson, A., Nielsen, F. \AA, Colleoni, E., & Etter, M. (2011). Good Friends, Bad News-Affect and Virality in Twitter. Arxiv preprint arXiv:1101.0510. Herr, P. M., Kardes, F. R., & Kim, J. (1991). Effects of word-of-mouth and product-attribute information on persuasion: An accessibility-diagnosticity perspective. The Journal of Consumer Research, 17(4), 454–462. Luoma-aho, V. (2010). Emotional Stakeholders: A threat to organizational legitimacy? Paper presented at the ICA Conference 2010. Luoma-aho, V. (2009). Love, hate and surviving stakeholder emotions. that Matters to the Practice, 323. A paper presented at the Internation Public Relations Research Conference, Miami, Florida, March 11th - 14th 2009. Available online http://www.instituteforpr.org/wpcontent/uploads/Love_Hate.pdf Mahon, J. F., & Wartick, S. L. (2003). Dealing with stakeholders: How reputation, credibility and framing influence the game. Corporate Reputation Review, 6(1), 19–35. van der Merwe, R., Pitt, L. F. and Abratt, R. (2005). Stakeholder Strength: PR Survival Strategies in the Internet Age, vol. 50, no. 1 (S), pp. 39-48. Morpace (2010). Is Facebook an effective marketing tool for retailers to influence consumer purchase decisions? Morpace Omnibus Report April 2010. Retrieved June 1st 2011. http://www.morpace.com/Omnibus-Reports/Omnibus%20ReportFacebooks%20Impact%20on%20Retailers.pdf Nielsen, J. (2006). Participation Inequality: Encouraging More Users to Contribute. Alertbox, October 9, 2006. Retrieved May 30th 2011. http://www.useit.com/alertbox/participation_inequality.html Ruler, B. van (2004). The communication grid: an introduction of a model of four communication strategies. Public Relations Review, 30(2), 123-143. Socialbakers (2011). The rise of Asia and Africa on Facebook, statistics by continent! Socialbakers. Retrieved June 1, 2011, from http://www.socialbakers.com/blog/116-the-rise-ofasia-and-africa-on-facebook-statistics-by-continent/ Sweeney, J. C., Soutar, G. N., & Mazzarol, T. (2005). The Differences Between Positive And Negative Word-Of-Mouth Emotion As A Differentiator? Proceedings of the ANZMAC 2005 Conference: Broadening the Boundaries (p. 331–337). Valio 2011. Company Information. Retrieved March 30th 2011. http://www.valio.fi/portal/page/portal/valiocom/Company_information Woerkum, C. van, & Aarts, N. (2008). Staying connected: The communication between organizations and their environment. Corporate Communications: An International Journal, 13(2), 197-211. doi:10.1108/13563280810869613 Weigelt, K. and Camerer, C. (1988). Reputation and corporate strategy: A review of recent theory and applications. Strategic Management Journal, 9: 443–454. Worcester, R. M. (1972). Corporate image research. Consumer Market Research Handbook. London: McGraw Hill, 505–518.