Enablers of and constraints on employees’ information sharing on enterprise social media

Purpose – The purpose of this study is to investigate the various enablers of and constraints on employees’ information sharing on an enterprise social media platform. It draws on two theoretical perspectives, communication privacymanagement theory and the technology affordance framework, as well as on empirical data in an attempt to paint a comprehensive picture of the factors shaping employees’ decisions to share or not share information on enterprise social media. Design/methodology/approach – This qualitative field study is based on semi-structured interviews and enterprise social media review data from a large Nordic media organization. Findings – On an enterprise social media platform, privacy management principles shape employees’ information-sharing decisions in relation to personal privacy boundaries, professional boundaries and assumed risks, online safety concerns and perceived audience. Additionally, the technological affordances of visibility, awareness, persistence and searchability shape employees’ information sharing in varying and sometimes even contradictory ways. Finally, organizational factors, such as norms, tasks and media repertoires, are associated with employees’ information-sharing decisions. Together, these three dimensions, personal, technological and organizational, form a model of the enablers of and constraints on employees’ decisions to share information on enterprise social media. Originality/value – This study extends the understanding of different factors shaping employees’ decisions to share or not share information on enterprise social media. It extends the two applied theories by uniquely combining interpersonal privacy management principles with a technological affordance framework that focuses on the relationship between the user and the technology. This research also furthers the authors’ knowledge of what privacy management principles mean in the organizational context. This study shows connections between the two theories and extends the understanding of technology affordances as not only action possibilities but also constraining factors. Additionally, by revealing what kinds of factors encourage and inhibit information sharing on enterprise social media, the results of this study support organizations in their efforts to manage information sharing on enterprise social media systems. KeywordsKnowledgeworkers, Privacy, Qualitativemethod, Social media, Information sharing, Affordances, Interview, Intranets Paper type Research paper Introduction Companies are currently investing in social media both to strengthen their communication with external parties and to utilize social media internally. Internal enterprise social media (ESM) platforms are multidimensional web-based communication tools that allow Employees’ information sharing on ESM © Kaisa Laitinen and Anu Sivunen. Published by Emerald Group Publishing Limited. This article is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of this article (for both commercial and non-commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at http://creativecommons.org/licences/by/4.0/legalcode. We would like to acknowledge William Barley, Jeffrey Treem, Jennifer Gibbs, and the three anonymous reviewers for their valuable feedback on themanuscript. Their thoughts allowed us to refine the paper to its full potential. Funding: This research was supported by the Finnish Work Environment Fund (no. 116400) and The Academy of Finland (project no. 318416). The current issue and full text archive of this journal is available on Emerald Insight at: https://www.emerald.com/insight/0959-3845.htm Received 25 April 2019 Revised 22 October 2019


Introduction
Companies are currently investing in social media both to strengthen their communication with external parties and to utilize social media internally. Internal enterprise social media (ESM) platforms are multidimensional web-based communication tools that allow Employees' information sharing on ESM collaboration and information sharing both throughout organizations and in group settings . Examples of such platforms include Yammer, Jive and Workplace by Facebook. As these platforms grow more popular, there is a need for empirical studies looking at social media's role in the internal communication of organizations. As social media encourage sharing and participation (El Ouirdi, El Ouirdi;Segers;Henderickx, 2015;Leonardi et al., 2013;Majchrzak et al., 2013), the question of information-sharing practices becomes important. Internal information sharing is a crucial part of an organization's success and makes effective collaboration possible. By sharing information with their coworkers, members of an organization can distribute knowledge, complete tasks, contribute to innovation processes and build relationships (Chow and Chan, 2008;Leonardi et al., 2013;Lin, 2007;Kiezmann et al., 2011). According to Kiezmann et al. (2011, p. 245), the concept of sharing refers to "the extent to which users exchange, distribute, and receive content" on social media. In this study, we regard information sharing as posting or commenting on the organization's enterprise social media platform.
Forms of information sharing have recently gained researchers' attention in the context of personal social media platforms, such as Facebook and Twitter (Ashurin et al., 2018;Ledbetter et al., 2010;Spottswood and Hancock, 2017). However, ESM have unique characteristics compared to these personal social media platforms, as they have different goals for use as well as different audiences and users (members of an organization) (Ellison et al., 2015). Therefore, the motives for information sharing may also differ. Information sharing on ESM is especially crucial in large and distributed organizations, as it enables members to be aware of organizational matters as well as share and store knowledge to a cross-organizationally accessible platform.
In this paper, we study employees' information sharing on the ESM of a large Nordic media organization. The 3,000-employee-organization implemented a Google þ -based organization-wide enterprise social media system to develop the internal communication of the organization. In this study, we examine how employees' information sharing is shaped by various factors, i.e. the factors that enable employee information sharing as well as the key constraints that result in employees carefully considering before posting or deciding not to post at all on ESM. Drawing from two theoretical frameworks, communication privacy management (Petronio, 2002(Petronio, , 2013 and the technology affordance perspective (Treem and Leonardi, 2013;Rice et al., 2017), we argue that personal privacy management principles and technological affordances are critical in understanding information-sharing behavior on ESM. For example, public social media users have been found to calculate the benefits and risks, as well as the credibility of the platform, before sharing information on social media platforms (Ashurin et al., 2018). Additionally, the issues of control, surveillance and privacy have been found to play a role when individuals are using different kinds of open, interactive technologies (Humphreys, 2011). While privacy management theory focuses on intrapersonal and interpersonal processes by looking at information sharing on social media as something happening in the mind of the individual and in the relationship between the sharing individual and the perceived audience (Petronio, 2013), the technology affordance perspective considers the relationship between the user and the technology (Treem and Leonardi, 2013).
By bringing these two perspectives together, we contribute to the literature in four ways. First, our study extends the understanding of personal factors shaping employees' decisions to share or not share information on ESM by applying communication privacy management theory to the context of organizational social media (Petronio, 2013). All members of organization can be, and are often expected to be, active content creators on ESM. However, the decisions to share, withhold or modify shared information, especially information regarded as personal, are shaped by personal or professional privacy preferences. This study aims to further illuminate what kinds of privacy management considerations employees have in mind when planning to share information in this technology-mediated organizational setting.
Second, as ESM's unique characteristics (Ellison et al., 2015) may afford different kinds of information-sharing capabilities, strategies and constraints than public social media (Choi and Bazarova, 2015;Spottswood and Hancock, 2017), we aim to extend the literature on technological affordances by using it as a perspective to find technology use-related explanations for employees' information sharing or decisions to refrain from posting on ESM. Additionally, our study highlights the contextuality of affordances and contributes to theory development by showing how the same affordances in the same organization can be perceived both as enablers and as constraints in regard to information sharing and privacy management. Third, our study provides interesting new avenues for the theoretical viewpoints considered. This study extends the theories, for instance, by showing how privacy management principles are related to the affordances of visibility and persistence and what these principles mean in the organizational context.
Finally, we contribute to the literature by showing how organizational factors, such as organization-level structures and norms, emerge as a third important dimension of employees' information-sharing decisions. Our findings show that employees handle information differently depending on whether they perceive it as organizational or personal. Additionally, organizational goals and regulations shape the possibilities and expectations for information sharing. By combining these three dimensions (personal, technological and organizational), this study aims to paint a comprehensive picture of the factors behind employees' information-sharing decisions on ESM.
This study consists of four parts. First, there are three sections that illustrate the existing literature on the topic and draw from previous research to provide a comprehensive understanding of information sharing on ESM as well as the theoretical frameworks utilized in this paper. At the end of the literature review, we propose research questions for this study. In the second part, we describe the methodology used to obtain the results. The third part presents key findings of the study following the research questions. Finally, at the end of the paper, we discuss the theoretical and practical implications of this study as well as limitations and future research possibilities.
Information sharing on enterprise social media Modern organizations can choose from a vast selection of different kinds of communication technologies to be used for their internal communication. The development of ESM platforms can be derived from the breakthrough and popularity of public social media systems. While some organizations simply utilize publicly available sites as their own internal platforms, an increasing number of organizations use "in-house" sites that are private and often custom-made for the organization's needs . ESM systems provide employees with the possibility of communicating in various ways via one platform as well as an archive in which the members of the organization can view posted information anytime and anywhere Treem and Leonardi, 2013). ESM afford both visibility and persistence of information in a way that provides unique possibilities, such as increased social learning .
ESM, such as public social network sites, include many kinds of communication channels (e.g. chats, message boards and news feeds), which enable social interaction and information sharing across organizational units. Employees are not only users of these social media tools but also creators of content on ESM by means of updating statuses, commenting and maintaining their own social media profiles (Ellison et al., 2015). This is a core difference between ESM platforms and other organization-wide communication tools, such as traditional intranets. As all members of the organization are, or at least have the possibility to be, active content creators for the platform, the factors behind information-sharing decisions become especially crucial to understand.

Employees' information sharing on ESM
The factors shaping organizational members' information-sharing behavior and participation in information exchange can be divided into, for instance, environmental, personal, interpersonal and sociocultural factors (Matschke et al., 2014). Additionally, factors such as a lack of time or lack or trust have been found to constrain sharing behavior in organizations, while the willingness of employees to help others and monetary rewards can encourage sharing (Razmerita et al., 2016). In the context of information and communication technologies more broadly, Lin (2007) has found that individual, organizational and technological factors influence knowledge-sharing processes in organizations. This study extends the preexisting understanding by examining the different factors behind information sharing in the context of modern enterprise social media platforms and by utilizing two theoretical perspectives to further explain and clarify these factors and dimensions.
Five key differences can be found between ESM and public social network sites (Ellison et al., 2015). First, user behavior on public social network sites is influenced mainly by site norms, whereas corporate social media use is usually guided by company guidelines. Second, the users of ESM are an organization's employees. Third, public social network sites are designed to promote interactions among users, whereas in organizations, the design encourages interactions that serve company needs. Fourth, the audiences of public social network sites are a more general combination of various networks, while the audience of ESM is an organization's members. Fifth, while public social network sites are primarily used for social and interpersonal goals, enterprise social media are meant to be used to accomplish work-related goals. These differences suggest the need to study information sharing in organizational settings to discover the specific contextual affordances of ESM.
Information sharing on public social media has found to be motivated by, for instance, information-sharing needs, information archiving and storing, self-entertainment, boasting and keeping up with trends (Waters and Ackerman, 2011). However, since ESM are largely a workplace communication tool, the organizational context plays a more important role in employees' information sharing than on public social media platforms. Research shows that organizational culture, relational considerations and the calculus between the desire to obtain feedback and protect privacy (risk/benefit considerations) are associated with decisions to share or withhold information in face-to-face settings at work (Smith and Brunner, 2017). Therefore, in the examination of the different reasons behind employees' information sharing on enterprise social media, it is necessary not only to consider technological issues but also to take into account the contextual features of the organization as well as the communication privacy management principles of individual employees.

Communication privacy management and ESM
Communication privacy management (CPM) theory outlines how boundaries around private information are transformed when it is shared. According to the theory, people strategically coordinate their privacy management practices to suit the principles they hold for sharing private information in a particular situation (Petronio, 2002). The theory is not necessarily focused on the content that is revealed but rather on the process of disclosing and the underlying rules, boundaries and principles affecting the decision to share or withhold information . These privacy management practices are profoundly contextual; the decision to self-disclose or withhold information is evaluated depending on the situation and communication partners (Frampton and Child, 2013). Thus, it is necessary to constantly reevaluate privacy principles and privacy management practices to align them with the properties of different technological platforms.
Privacy rules around shared information and strategic coordination around how shared information should be handled by communication partners are always negotiated to avoid turbulence deriving from broken rules (Petronio, 2013). Therefore, the audience plays a ITP significant role not only before the decision to share private information is made but also after information sharing takes place and the ownership (Petronio, 2002) of private information is extended. Privacy rules are coordinated by considering both where and with whom the information is shared. The tensions between sharing and withholding information are constantly present as individuals evaluate both the situation and their communication partners and calculate possible benefits and risks (Ashurin et al., 2018;Petronio, 2013).
A vast body of literature has extended and tested the CPM framework (see, Petronio, 2013). Since this study utilizes CPM to understand information sharing in the ESM context, it is crucial to examine two distinct contexts where CPM has previously been applied: public social media and organizational communication.
The strategic choices for privacy management on public social media are guided by individual traits and background, as well as motivations and social media experience/s (Choi and Bazarova, 2015;Waters and Ackerman, 2011). Individuals' high need for control over private information can negatively affect information disclosures online (Benson et al., 2015). Recent studies have also shown that when using public social media, employees' privacy decisions are based on possibilities for privacy settings on social media, organizational privacy orientations and employees' satisfaction with their coworkers (Frampton and Child, 2013). Social media settings can also pose challenges for recognizing one's audience, as information is often shared with multiple audiences simultaneously (Vitak, 2012). This is also naturally relevant when talking about privacy management in larger organizations in the context of ESM, where the audience of the shared information can be the entire organization.
How can the rules of information ownership be negotiated when a full understanding of the audience is absent?
Privacy management in an ESM setting forces employees to consider not only where and with whom information is shared but also the underlying principles and norms of the organization, such as ESM guidelines or the communication culture. The organizational setting also brings up other aspects related to employee privacy management. Privacy in a workplace context is not simply about personal privacy issues; institutional and organizational privacy considerations also surface (Humphreys, 2011;Ball et al., 2012). Although the organizational context does raise these versatile considerations of what private information is in general, decisions to share or withhold information are always personal. Nevertheless, the consequences the organizational context creates for information ownership cannot be ignored.
As this study focuses on information sharing on ESM, it can contribute to theory by combining both CPM research on public social media and the CPM literature on the organizational context. Thus, this study extends privacy management theory (originally developed mainly for face-to-face contexts) by analyzing how sharers deal with privacy management challenges, such as audience recognition and information ownership, in technology-mediated organizational settings.
Technological affordances of enterprise social media Coming from the ecological viewpoint of seeing items in terms of not only their physiological properties but also how they are used and applied (Gibson, 1986), the affordance approach to technology allows social media to be defined from both a material and a social perspective. According to a review by Rice et al. (2017), the affordance perspective is one of the many frameworks that seek to balance technology features and users' perception and use. However, as this framework focuses on the relationship between users and technology, it is one of the rare approaches that allows for a middle path between technologically deterministic and socially constructivist stances (Leonardi;Vaast, 2017).

Employees' information sharing on ESM
Affordances can be defined as "relationships among action possibilities to which agents perceive they could apply a medium (or multiple media), within its potential features/ capabilities/constraints, relative to the agent's needs or purposes, within a given context" (Rice et al., 2017, p. 109). Affordances are the possibilities for action in a particular context. Technological affordances define users' perceptions of the technological platform and how it could be utilized.
Scholars have studied several technological affordances of ESM, such as visibility, editability, persistence and association (Treem and Leonardi, 2013;Oostervink et al., 2016), and sets of criteria for determining what counts as an affordance have been created . Visibility refers to users' ability to make their behaviors, knowledge, preferences and communication networks visible to other users. Persistence can be defined as how well communication remains in the same form after the sender has sent his or her message. Editability refers to the ability of users to craft and recraft their messages before they are viewed by others and to modify or revise the messages they have already sent (Rice et al., 2017). Association is defined as "establishing connections between individuals, between individuals and content, or between an actor and a presentation" (Treem and Leonardi, 2013, p. 162). Other studies list pervasiveness, searchability and self-presentation (DeVito et al., 2017) as key organizational media affordances (Rice et al., 2017).
The technological affordance perspective has been found to be a fitting framework for studying how modern social media platforms shape organizations and organizing. For instance, the affordance lens can be used to explain the diffusion of social media, the use of social media in organizations and the different organizing processes taking place on enterprise social media systems (Leonardi and Vaast, 2017). Affordances portray the perceived possibilities of ESM; therefore, affordances can be a relevant way to explain the technology-related dimension of employees' information sharing on ESM.
This review suggests three research questions based on the gaps in the current literature. Two first questions focus on the two theories presented, and the third question derives from the need to understand the role of the organization or workplace context in shaping employees' ESM use. The research questions are as follows: (1) How do employees' privacy management boundaries shape their information sharing on ESM?
(2) Which technological affordances can be applied to explain employees' information sharing on ESM?
(3) What organization-related explanations do employees give for their ESM information-sharing decisions?

Research site
To examine employees' information sharing on ESM, we conducted a qualitative field study at a large Nordic media organization, Kappa. In field settings, researchers are able to attain data that reflects the reality of actual organizations and communities (Tracy, 2013). In opposition to experimental settings, field research allows the exploration of different emergent communication processes as well as the complex perceptions of organizational members (Gibbs and Doerfel, 2014). Kappa has over 3,300 employees in over 20 locations nationwide. The organization consists of six different core units and four support function units. All organizational core units comprise several different departments that work with a range of television and radio productions, including news and local affairs programs, educational shows and dramas. Kappa was one of the first large organizations nationally to replace its traditional intranet ITP with a modern enterprise social media system as its main organization-wide communication platform. This change was a part of an internal development program, where the structures and habits of a relatively traditional organization were updated with the aim of facilitating more modern, democratic and mobile forms of work.
At the time of this study, the ESM platform had been in place at Kappa for a year and a half. The platform was based on Google þ technology; it consisted of a feed-based front page that displayed posts from groups that employees belonged to and provided access to services and group discussion sites. All Google services (e.g. Drive, Hangouts) were available to users. The ESM had multiple group discussion sites; although most were dedicated to work, some informal groups, such as exercise groups and a flea market group, also existed. The platform in question fit the definition of ESM by Leonardi et al. (2013) in three ways in particular. First, the platform was by all means versatile. It allowed both organization-wide posting and group work and functioned as a "homepage" for all other technological tools and systems. Second, the platform was open for all members of the organization to actively create content, and the organization hoped for active participation and communication from all employees. Third, the platform was web-based and therefore useable from different locations, including outside the office and while on the go through smartphones.
There were no strict organization-wide guidelines or policies in terms of what to post and what not to post on ESM. However, employees were encouraged and advised to post actively on the platform, and group leaders received written guidance on how to activate their group's ESM use. The guidelines had recommendations, such as "use the [ESM] community instead of email" and "be visible in your [ESM] community daily: post, comment and like". There were no ESM-specific privacy policies established in the organization. In the instruction material, the platform was described with the sentence, "Don't worry, [ESM] is safe and stable".

Research subjects
First, with the help of our contact persons at Kappa, we identified and selected Kappa employees from several different units and work functions for an interview to obtain a broad sense of how the ESM platform was used in various parts of the organization. The interviewees were chosen using the following three criteria. First, we wanted to interview employees from different positions and organizational levels. Second, we wanted to study employees who worked somewhat remotely as well as those who worked close to each other. Third, we wanted employees who had relatively active ESM use and employees who had lower ESM use. These criteria allowed us to have a varied group of interviewees, which then led us to better understand the varying perceptions, needs and motivations that different kinds of employees in a large organization might have.
This interview round brought us 19 interviewees between the ages of 36 and 60 years. The age range of the interviewees represented the employees of Kappa well, as the average age of permanent employees in the organization was 49 years. The interviewees worked in different areas of television production, archiving and information processing and corporate training. They were varied in terms of what position and organizational level they occupied, whether they collaborated face-to-face or remotely with colleagues, and whether their work was more knowledge-based or involved only a little computer use each day. Furthermore, interviewees' ESM use varied from continuous use to use only a few times a month (see Table 1). These choices allowed us to have a data set that reflected the organization's structure.
Second, we had two sets of hour-long meetings with Kappa's ESM project group. The group consisted of three to four Kappa managers (depending on the meeting) who worked in the communication, strategic planning and organizational development departments of the organization and were responsible for the ESM platform. The two meetings were held at different points of the research process. The first meeting was held when the data collection was close to being finished, and the second meeting was held after the data were fully Employees' information sharing on ESM analyzed. These group discussions were held to (1) find out more about the organizational goals and strategies for the implementation of the ESM platform,(2) report on the data collection and reflect the data with the experts of the organization and (3) report early research findings. Through these project group meetings, we gained a better understanding of the contextual features of the organization and its information-sharing culture. Insights from these discussions were used to reflect the data collected from employee interviews.

Data collection
This study utilized two types of data: (1) employee interview transcripts and (2) photos, notes and transcribed descriptions of employees' ESM use and actual ESM posts observed during the interviews. This kind of data source triangulation (e.g. Bilandzic, 2008) was used to pursue versatility in the relatively small data set and allowed the statements made by the informants to be compared with their actual ESM behavior. Qualitative interviews were chosen to be the main data source based on the research objective: understanding the enablers of and constraints on employees' ESM information sharing. The observed ESM posts demonstrated and confirmed what the informants reported. The main dataset with 19 Kappa employees was collected through qualitative, semistructured interviews. The interviewees were contacted with the help of our connections at the organization. Before the actual interviews, we sent the informants an email invitation including information about the research and the interview setting, as well as a link to a short background questionnaire. This questionnaire included relevant demographic questions and questions about the frequency of the informant's ESM and social media use. This practice allowed us to save time in the actual interview. The first author conducted the interviews at the informants' workplaces and recorded the interviews with the informants' consent. The interviews lasted from 30 to 90 min. Most of the interviews were 60-90 min in length. In couple of occasions, we had to accommodate to the tight schedule of the informants which led to shorter interviews. We, however, utilized the same semi-structured interview protocol in all ITP of the interviews. For the shorter ones, the depth and amount of additional questions was not as extensive as in the longer ones, but all main themes were discussed with everybody.
The interview protocol included two parts: semi-structured interviews and ESM reviews. The semi-structured interviews included questions about work tasks, social media experiences, perceived technological capabilities and constraints, and ESM informationsharing goals, concerns and characteristics. After going through the semi-structured protocol, we moved to an ESM review, where the interviewees were asked to review the ESM site; they visited and talked about their own ESM profiles and showed some examples of their own posts. The ESM posts were photographed and/or written down by the researcher, and the discussion was recorded. All data, including the photographed posts, were transcribed, resulting in 323 pages of text with approximately 7,000 words per interview. The analysis was focused on the semi-structured interviews and the transcribed discussion part of the ESM review data, i.e. the part where the informant explained his or her goals in sharing or thoughts about the platform in general. The actual ESM posts were coded in terms of content and analyzed in terms of how well they portrayed the descriptions provided by the interviewees. The ESM posts did not generally bring new insights to the findings but rather illustrated and verified what the informants reported.

Data analysis
The analysis of the employee interview data was guided by abductive reasoning. This means that the analysis was carried out as a dialogical process between the empirical data and theory and that we constantly aimed at transparency and coherence in our interpretive inferences (Mantere and Ketokivi, 2013). The process included data-driven analysis as well as theory-guided analysis drawing from the aforementioned two theoretical frameworks: communication privacy management and technological affordances. As with the principles of iterative analysis, we conducted the analysis by systematically repeating and recursively alternating between the different phases (Tracy, 2013). Qualitative coding was conducted with the help of the Atlas.ti qualitative data analysis program. The analysis was initiated by the first author and then read through by the second author to achieve validity through peerdebriefing practices (Lincoln and Guba, 1985). Both researchers together discussed and solved the difficult cases to improve the validity of the results.
Analysis was conducted in three phases. To begin with, the first author analyzed the transcribed data to tease out first-round descriptive codes (see Tracy, 2013), a process guided by both the privacy management and technological affordance theoretical frameworks and the data. The first author started by inductively drawing out descriptive codes on any phrases or utterances regarding technology use, information sharing and ESM information sharing. This part of the first analysis round was fully data-driven. Immediately after this, the author went through all descriptive codes and continued the coding following the chosen theoretical frameworks, i.e. privacy management strategies and principles, as well as technological affordances. At the end of the first analysis round, the authors had initial theory-guided codes related to affordances and privacy management on ESM as well as datadriven codes related to employees' information sharing on ESM (see Table 2).
After the authors discussed the initial codes, the first author analyzed the data again, following the principles of the iterative analysis and constant comparative techniques (Strauss and Corbin, 1998). In this phase, the codes were reexamined, connections between them were made, and descriptive categories were developed into second-level analytical categories (Tracy, 2013). For example, individual descriptive codes such as "posting information that is relevant for work" and "time and other tasks constraining ESM use" were formed into a second-level category, "organizational tasks". Finally, after the authors discussed the analysis, the individual codes were integrated and compared, resulting in third-Employees' information sharing on ESM level conceptualizations fitting the emergent theoretical framework as well as the data (see Appendix). We were interested in discovering not only the characteristics of employees' information sharing on ESM but also their motives and strategies and the extent to which they were guided by technological affordances or privacy management goals compared with other factors, such as organizational norms. These findings were also categorized according to our emerging theoretical framework. In the end, two of the resulting categories were guided by the theoretical frameworks and one was formed from the data (see Figure 1). All data excerpts used in this article to illustrate the findings were translated to English, and all interviewees were identified with pseudonyms.

Findings
The employees' information sharing on ESM was both encouraged and constrained by privacy management principles, technological affordances and organizational factors. The findings suggest that the employees' incentives to share information and their perceived inhibitions toward information sharing on ESM can be traced to the following three dimensions: personal, technological and organizational (see Table 3).
Personal dimension: privacy management principles and information sharing Privacy management shaped the information-sharing behavior of the employees through considerations and coordination on four levels: personal privacy boundaries, professional boundaries and risks, online safety and perceived audiences (see Table 3).
First, employees at Kappa described their personal privacy boundaries as relatively strict; they did not want to share too much about themselves with their coworkers. In this case, information-sharing behavior was guided by the strict boundaries around the ownership of private information, which led to a lack of personal disclosures in the work environment. Employees did not generally use the ESM to share the ownership of personal information. This separation of "work self" and "private self" is illustrated in the following excerpt: This kind of strict separation between work and private selves clearly led the information shared on the ESM to be mainly issue-centered. Thus, the social possibilities of the social media platform, such as getting to know coworkers or developing relationships, were muted. Limiting personal disclosures on ESM seemed to be a common privacy management strategy, as the following excerpt portrays: You can include something lighter in a comment, but it also must have some relevant content. This is not like Facebook, where you can just write like that and be more communal. Here, everything is so issue-related that you do not really focus on relationships. (IN7) Second, the professional boundaries and potential risks related to professional roles shaped information sharing by making employees either consider the content carefully or abstain from posting altogether. This kind of privacy calculus was especially common among the employees in the leadership roles. In the following example, IN1 ponders how her role in the organization shapes her information-sharing behavior and coordination around private information: Third, online safety was also a factor considered by employees. They described the ESM as "vast" and too closely connected to the platform developer (Google). They were unsure if their private information was being distributed outside the organization and how much of it was recorded by the platform developer, a trend illustrated in the following excerpt: "Then of course I think about the fact that Google is a worldwide business and you cannot be sure who reads all these things we post here" (IN4). This example shows that employees' personal privacy management principles were strict on ESM because they were concerned about information sharing on the Internet. The fact that Google operated the ESM platform increased the privacy concerns of Kappa employees. Interestingly, however, the employees did not generally mention privacy settings as a part of their privacy management practices, even though the ESM platform did allow the privacy settings to be adjusted. The main privacy management strategies were withholding of private information and careful consideration of content. IN4 described the lack of a need for privacy settings as follows:

Employees' information sharing on ESM
Fourth, the perceived audiences were related to decisions to share or withhold information on the ESM. The ESM platform included different kinds of groups. Information sharing was perceived differently depending on the group context. Work-related groups tended to have more activity and more audience-focused content, material essential for the group members, than the general feed. These groups were often relatively small and comprise team members. Recreational groups were perceived to have more personal information shared than professional groups, especially in terms of one's own opinions and views. One of the interviewees illustrated this in the following way: Quite a lot of our groups are purely focused on the profession and work, and communication is very much related to work and work tasks. But then there are these groups like Discussion or some recreational groups . . . . I think that in those groups the discussion can be freer and more provocative.
In these discussion-type groups you can very well bring out your own views and question issues. (IN1) Sometimes the nature and breadth of the audience also posed potential risks and concerns for employee information sharing. These risks included sharing excessive or unsuitable content from a respected professional position, sharing too much private information to too large an audience, being misunderstood, or posting strong personal opinions that could lead to professional consequences. These are illustrated in the following two examples: If people do not realize how big of a group gets the information. If there is something that is meant to be confidential information and then suddenly too many people get access to it. (IN1) Well, one thing is that you kind of like build a picture about yourself [by posting] and you can never know how others understand or interpret the thing you have written. In that sense, I would be very careful, because people understand things so differently. (IN19) Privacy management principles were found to shape information-sharing behavior by shaping the content of the shared messages. Employees took into consideration their personal privacy boundaries, the responsibilities attached to professional roles, the realities of online safety and a general uncertainty about the audiences within the ESM.

Technological dimension: technological affordances as enablers and constraints
Technological affordances were used to explain both the behavior and attitudes toward the ESM. Among the various affordances that ESM provide, those of visibility, awareness, persistence and searchability were found to be the most important with regard to employees' information-sharing decisions on ESM. First, the affordance of visibility appeared to explain information sharing in two contradictory ways. Visibility was seen as an enabler when sharing work-related information because that kind of information was generally published on the ESM to reach as large an audience as possible. However, visibility was also seen as a constraint on information sharing, especially in relation to personal information. Most employees stated that information visibility plays a significant role when they are deciding what to post. Employees were often not sure how visible the information they shared was and were thus apprehensive about posting anything personal. The following excerpt illustrates visibility as a constraint: Similarly, like if I have a closed group on Facebook, I pretty much know who sees the posts . . . but then on our platform, I kind of do n't know who, in the end, sees my posts. (IN11) Second, the affordance of awareness-enabled information sharing as the ESM platform was perceived as allowing employees to follow and make connections with different parts of the organization. The ESM was described as a tool for knowing what was happening in the ITP organization and, understood in this way, it was perceived positively. Employees said that the ESM allowed them to "get the hang of" or "be aware" of issues taking place in other parts of the organization; because this aspect was important to them, it also increased their own willingness to share.
Third, the affordance of persistence was seen as an important factor in work-related information sharing and information availability. However, information remaining on the platform for a long time increased apprehension toward personal information sharing. In the next excerpt, IN17 describes how she prefers chat over the ESM group because of the persistence of information: Those [chat groups] are easier to just write something for everyone to see because on the [ESM] groups, the information spreads wider and stays available for a longer time. (IN17) The chat system allowed the employees to control the persistence of information, thereby lowering the bar for more informal content. Additionally, the overall amount of information on the ESM platform was perceived to be high because of both the platform design and information persistence. This heavily shaped employees' information-sharing decisions. As all new content was seen to be adding to an already excessive amount of information, it was expected that all additional information shared on the ESM should be useful to the organization. This is illustrated in the following example where an employee describes the consequences of information persistence related to the amount of information: If somebody shares their holiday pictures, well, feel free, who am I to decide for them? But I still think it would be good to concentrate on work-related content. Because nowadays the amount of information is basically limitless, so if that is made even wider and wider, nobody is going to read that anymore. (IN6) Fourth, the affordance of searchability proved to be crucial in explaining employees' information sharing on the platform. Finding information related to their work was described as one of the most important ways employees used the ESM. However, the search feature of the platform was perceived as confusing and poorly designed, as evidenced by the following example: As much as it should be a good search engine now, it actually is not because it does not understand the conjugations of [employees' native languages] or anything like that. If I do not remember if some word in the title of the document was written like this or like that, then I will have to do a lot of work [to find the document]. (IN4) The fact that the affordance of searchability was perceived as being of high importance while the actual search feature did not meet the expectations of the users seemed to clearly discourage information sharing on the ESM platform.

Organizational dimension: organization-related enablers and constraints
In addition to privacy principles and technological affordances, we found three organizationrelated factors shaping employees' information sharing on the ESM: (1) organizational norms, (2) organizational tasks and (3) organizational media repertoires (see Table 3).
ESM are an extension of the organization; thus, organizational norms shape employee attitudes regarding what is suitable to post. Employees explained their information-sharing habits by stating, "Nobody else shares either" or "That is what others also do." This reflects the way employees interpreted not only the other organizational members' ways of using the platform but also the institutional logics for the use of the ESM. These interpretations formed the organizational-level norm on how the ESM should be used. This norm was evident in the ESM posts as well as in the employees' explanations of their posts. The organizational norms in the use of the ESM are also apparent in the descriptions of expected ESM content, as the following excerpt illustrates: Organizational norms were reflected in the expectation of how information sharing on ESM should look. Task-related, formal content was preferred over "cat videos," as illustrated in the following excerpt: Of course, the fact that this is this kind of a workplace's inner channel means that you need to act appropriately. If somebody would post cat videos, I would be like "whatever," but in the end, it would be away from the work-related feed. (IN6) This excerpt shows how personal or informal information was sometimes seen as a threat to sharing and receiving work-related information, and since the platform was organizational, work-related posts were preferred. However, this caused the social and networking possibilities of the social media system to fade to the background. Organizational norms were also visible in employees' ESM posts. Most of these posts were related to employees' experiences in training courses, opinions about work-related matters, out-of-office messages and greetings from work-related trips. Occasionally, employees also posted slightly more personal information, such as photographs from birthday celebrations of coworkers. All this information was, however, described by the employees as strictly professional. Sharing information concerning anything other than work-related topics was seen as unappealing because employees are on the ESM as their "professional selves." In the following excerpt, this is illustrated by IN15: [On the ESM] people play with these kinds of work roles. I do not think that anybody goes there to be personal, it is always about work. It is the work role you talk about and through which you act [on the ESM]. (IN15) This was also evident when actual ESM posts were examined. The following two posts from Kappa's ESM platform portray how the preferred content was very much workrelated, providing guidance and information about work or the rules of the ESM community: Now, all our department's cars and their reservations have also been moved to the digital reservation system. Calendars have been taken down from the break room, and cars can now be reserved only from the digital system. Ask for help to make a reservation from Matt or Ann, if needed. (An ESM post published by IN12) Welcome to the Video community. In this community you will find current information and tips about Kappa's video contents (TV programs, work materials, web clips, and programs) and their content descriptions and archiving. In this community, you can also ask and have discussions about content descriptions and archiving-related matters. (An ESM post published by IN1).
These types of normative posts further shaped other members' posts in the same community and on the ESM in general. Most of the group posts (excluding recreational groups) mainly consisted of work-related content aimed at informing others. Other larger categories of posts consisted of questions and help-seeking on work-related matters.
Second, organizational tasks were another strong encouraging factor behind information sharing on ESM. Interviewees reported that the posted information had to be somehow useful or relevant to the audience or other team members. Typically, this utility was explained as something that needed to be shared as part of accomplishing a work task. For example, in a remote team, employees often used ESM to present an issue that they wanted the whole team to address. In addition, employees in leadership positions felt that they were expected to share ITP information as part of their work tasks. These task-related expectations were often present when employees reflected on their decisions to post something on the ESM. They described these motivations as, "It is part of my job" and "I need to do that to inform others about my area of responsibility".
Another task-related, but constraining, aspect that was related to employees' information sharing was the sheer number of work tasks the employees had. This "busyness" took time away from posting on the ESM. An employee responsible for the finding and purchasing of props for TV and movie sets describes this time constraint as follows: I could make a better profile and write about my own skills and competence and such, but that would need time. Basically, there would have to be a shift without anything else, to book time just for that. (IN11) Finally, organizational media repertoires were also related to information sharing on the ESM. This was evidenced in two ways. First, lack of competence in the use of the organizationprovided ESM prevented employees from sharing information there and motivated them to choose other, even nonorganizational (e.g. public social media), platforms to share information instead. One informant describes the challenges of the ESM as follows: "If I first have to learn to post before actually posting, I am usually going to choose some other, quicker and easier, way to inform others" (IN8). This lack of experience with ESM guided the employees to use other, more established technologies to communicate the relevant issue. Many employees also reported that they would need time to study the versatile possibilities of the platform and that their daily work does not allow time to be spent on this kind of learning.
Second, employees felt that some of the other platforms in organization-wide use were more familiar and natural in terms of sharing information with others. Workgroup chat (Google Hangouts), email and face-to-face interaction were mentioned as great ways to share more versatile, and even personal, information at Kappa. This preference was justified, for example, by the ability to communicate only with a small group of people one knows: [The ESM] is not a very natural place to me. For me, it is more natural to communicate with email to those people I know and to whom I want to direct my information. (IN11) The existence of other organizational communication technologies proved to be an organizational structure that restricted the information sharing on ESM and guided it, instead, to these other channels. Additionally, the organization did not have a clear enough protocol on where information was expected to be shared, which allowed individual employees to lean on their personal preferences.

Contributions to literature and theory
Our study shows that employees' information sharing on ESM was shaped by three dimensions: personal (privacy management), technological (affordances) and organizational (norms and practices). To form an even more comprehensive picture of the factors shaping employees' information sharing, we present a model (see Figure 1) of the relationships between the three dimensions. Additionally, the figure includes information on which categories seem to work as enablers of (plus) and which work as constraints on (minus) employees' ESM information sharing. Next, we discuss these associations and contributions further.
We implemented two theories to guide the analysis and explain our findings: communication privacy management and technological affordances. Through the analysis, we are able to make associations between some of the resulting categories linked to these theories. Two key associations were found: (1) there is a relationship between the personal Employees' information sharing on ESM and technological dimensions through employees' perceptions of audience and the affordances of visibility and awareness, and (2) the personal and technological dimensions are also linked through personal privacy boundaries that are connected to the perceived affordances of visibility and persistence. First, ESM can be seen as a technology of accountability (Treem, 2015). The term is used to describe how ESM increase the accountability of employees by making communication visible to all members of the organization. This sense of accountability can make employees reluctant to use social media in an organizational context and could therefore play a role in their information sharing. Our findings support this perspective, as employees expressed apprehensions about information sharing based on who might see their posts, i.e. the visibility of the information. Therefore, there is a connection between the employee's perception of the possible audience and the affordances of visibility. On the one hand, the visibility of the information may lead to increased awareness of organizational activities, and if the audience is suitable for the content, this is very desirable and encourages sharing. On the other hand, if an employee questions the audience and the content he/she is planning to post, the affordance of visibility can act as a constraint on using ESM and encourage the use of other organizational communication media.
Second, personal privacy boundaries are connected to the perceived affordances of visibility and persistence of information in a constraining manner. As the employees tended to have relatively strict privacy boundaries at work and preferred to have a separate "work self", the possibility of vast information visibility and long persistence of the posts could create second thoughts before posting anything on the ESM. Therefore, private information was shared face-to-face or via more focused communication media (such as chat or email).
Information visibility and persistence were perceived as both constraints on and enablers of information-sharing behavior. This suggests the interesting dual-sided nature of affordances. Even though visibility was perceived as enabling important information sharing with coworkers, it was also perceived as a constraint. While it has been noted that visibility is not necessarily always desirable and sometimes employees use technologies to be invisible (see Gibbs et al., 2013), visibility as a constraint on information sharing on ESM is a relatively new discovery. Interestingly, the constraint tended to rise from uncertainty about and even a slight distrust of the platform. Employees also highlighted the importance of knowing the audience. Without knowing who was in the audience, visibility became a source Model of the dimensions that motivate and constraint employees' information sharing on ESM ITP of uncertainty and thus a constraint. The previous affordance literature has only rarely considered the versatility of a perceived technological affordance. Kim (2018) has recognized how different status groups perceive and use the affordance of visibility in vastly different, sometimes contradictory, ways. Additionally, Hutchby (2001) mentions that there needs to be more recognition of technological affordances as constraints. Our findings contribute to this previous affordance literature by showing that affordances can be either enablers or constraints, depending on the communicative needs, e.g. organizational tasks, and on personal, privacy-related factors.
The affordance of searchability also had a significant role in shaping attitudes toward information sharing on the technological platform. We found that when an affordance is perceived as highly important but simultaneously does not meet the expectations of the users, it could lead to negative attitudes. The search function on Kappa's ESM platform had multiple limitations that caused user frustration. For some users, frustration arose when the expectation of a Google-based platform having a great search engine was not met in practice. The challenges of searchability were also connected to information persistence. As the platform filled up with messages and information, the constraining role of the poorly working search feature became increasingly emphasized.
Previously, CPM theory (Petronio, 2013) has been used to study public social media and organizational communication, although not necessarily combining the two and examining communication in the ESM context. Based on our findings, information sharing on ESM and its motivations seem to be different from public social networking sites and from face-to-face organizational settings. Privacy boundaries and coordination around information sharing are important both on public social media and on ESM; however, on ESM, the workplace context and professional roles are more important in defining what is suitable, expected and possible to share from a personal privacy management perspective. In our study, the fact that the ESM platform was based on a product of a large, well-known IT company (Google) also constrained the information-sharing behavior of employees from a privacy perspective. Thus, our study contributes to the current literature on CPM by highlighting how the workplace context creates different expectations regarding certain professional roles and privacy management principles; these expectations further enable or constrain informationsharing behaviors. Our findings also highlight the role of technological affordances in privacy management that make information sharing very different on an ESM platform compared to, for example, face-to-face organizational settings.
The kind of information perceived as private differs between individuals, and individuals also manage boundaries around private information differently (Petronio, 2013). However, compared to the results of studies focusing on users of public social network sites, privacy boundaries on ESM were relatively strict across the board. This could be explained by the organizational context, because social influences in the workplace, such as organizational norms and expectations, shape communication in organizational social media settings (Ellison et al., 2015;Frampton and Child, 2013) by framing what is suitable and accepted. In the same way that social learning affects sharing behavior on public social media (Ashurin et al., 2018), it may be that the sharing behavior of other employees shapes individual sharing decisions, thus reproducing a "sharing culture" typical of the ESM platform of that organization.
This kind of organizational culture was also present in employees' ESM posts and in their reluctance to share anything more personal on the platform. Consequently, the culture and the expected form of use of the platform diminished the use of the social possibilities of the ESM platform almost entirely. Since these social dimensions were hardly used, the potential to, for instance, get to know other employees through the platform and extend and develop relational networks over the ESM was also restricted.
Our findings also mirror the key findings of the study by Oostervink et al. (2016) concerning information management in organizational communication platforms. They Employees' information sharing on ESM noted how, "informed by the corporate logics, users were aware of their role as employee, 'my boss pays me to work,' and realized they had to prioritize what information to spend time on" (p.171). This role of complex institutional structures was also evident in our findings, as employees' information-sharing behavior was shaped by their organizational roles and tasks, as well as the expectations they faced coming from the organization. This relationship between the affordance of visibility and organizational expectations (see Figure 1) indicates that organizational logics, such as roles and responsibilities, shape information sharing on ESM and the willingness to post on ESM.
Overall, our findings reflect the model presented in the study by Lin (2007) of the factors behind information sharing in organizations. In their work, the three dimensions similar to our findings were regarded as enabling factors. Our work extends this model in two ways. First, this study takes into account not only enabling factors but also constraining factors that shape employees' information sharing. Second, our work combines interpersonal and technological theories to further explain the dimensions found by the analysis. These theories have not been combined previously.

Contributions to practice
The findings of this study can be used in different stages of ESM implementation. The findings indicate that if stronger information sharing is expected, privacy management principles, technological affordances and organizational norms and practices need to be considered. From a managerial perspective, these results demand two kinds of considerations. Both of these considerations reflect the context of our study, in which a new internal communication platform was adopted in a traditional organization that previously relied on vertical communication practices and where the employees had not been producers of the internal organizational communication content.
First, when implementing the ESM platform, the purpose of the platform should not only be carefully thought out but also be communicated clearly to the employees. If the managerial goals for ESM are in contradiction with employees' perceptions of the affordances of the platform, employees are likely to choose alternative technologies to use. Additionally, if the purpose is not communicated clearly, employees are more likely to consider the platform to be something "extra" rather than part of their daily communication media repertoire. It is important not only that the purpose be communicated clearly but also that clear guidelines and policies that shape the communication norms be available. If there are not clear enough guidelines on the content that is suitable to post on ESM, employees may be more uncertain and worry about content-related issues, such as privacy. This is especially important to consider in traditional organizations, such as Kappa, that have implemented ESM after a long history of bulletin board-style intranet systems mostly used for sharing information from the top down. If employees think such use is, and should be, the only form of intranet use, they are more likely to consider social intranets that rely on participation to be an optional or extra feature of the workplace. Additionally, as employees' work time is limited and task-related organizational factors are crucial, any kind of uncertainty can increase the possibility of employees choosing some other easier channel to share information.
Second, if an organization wants to utilize ESM to boost the amount of information sharing or the number of employees who actively post on the platform, managers need to consider the personal and organizational factors that guide employees' decisions to share or refrain from posting altogether. For example, as visibility factors in this study proved to be critical in regards to privacy, organizations should make sure that the ESM platform allows users to control information visibility. This would mean, for instance, the ability to control the audience for each post and to increase the awareness of group memberships by making group members clearly visible for all users. From an organizational perspective, it is important that ITP employees' task-and position-related restrictions are taken into account. If there is a need to increase the number of ESM posts, such activities should be well integrated into employees' work tasks. Even allowing employees some time to explore the ESM and its functions or update their profiles could increase the utilization of the platform. Again, these issues are highlighted if the ESM system is implemented in an organization where the employees have not previously acted as content creators for the internal communication systems.
Additionally, we found that the managerial hopes for versatile and active ESM use were not always met with equally active use by the employees. Additionally, while managers also tended to consider privacy issues, especially from the perspective of possible professional consequences, they were often more active in their ESM use than regular employees. This was mainly due to the fact that they were more encouraged to use the platform, and as team leaders, they had more reason to do so. The fact that the upper management of the organization was also present on the platform might have played a role in terms of how employees evaluated the content they were willing to post. This was reflected by one of the interviewees, who said that employees would not want to be Facebook friends with upper management and that they always consider what is suitable to post on the ESM because of the work-focused context.
While the constraints presented in the findings suggest considerations for management to take into account, we also recommend being aware of the enabling factors that support ESM use. It seems that these enabling factors were mostly connected to work tasks and employees' need for information from or collaboration with other members of the organization, in one form or another. Therefore, it seems that our findings support the view that ESM, when used for information sharing, promotes organizational knowledge sharing and collaboration (e.g. Gibbs et al., 2015;Leonardi et al., 2013;Lin, 2007).
Limitations and future research possibilities As our data are based on only one organization, organizational and national culture may play a role in shaping the findings. Organizational culture introduces interesting dimensions in regards not only to employees' decisions to share or withhold information but also to the content of ESM posts. One indicator of the unique organizational culture at Kappa was the repeatedly expressed desire to keep interactions, especially on the ESM, very much work-related; personal information was even occasionally seen as a threat to sharing and receiving work-related information. Additionally, the professional roles at Kappa shaped information sharing significantly. This may also be an aspect of the organizational culture in which supervisors are expected to behave in a certain way. This causes the employees in such roles to approach the content of their posts with particular care.
In a discussion of organizational culture, the role of national culture must be considered. Employees of Kappa showed a willingness to keep work and private lives separate, even to the extent of conceiving of themselves differently at work (work self) and at home (leisure-time self). While this may be a manifestation of organizational culture, it can also be explained by the national culture of the employees of this Nordic organization. In Nordic countries, people tend to have a high appreciation of the separation of work and leisure time (e.g. Wieland, 2011), which may also shape our findings.
The data in this study consist of employee interviews that were conducted with employees between the ages of 36 and 60 years. This data sample leaves out some of the younger employees at the organization and may raise questions about those employees' unique social media experience compared to the older generation. However, this data sample was collected in a collaboration with the management of the organization, and they have confirmed it fairly reflects the age structure of the organization. Additionally, although age-related factors associated with information-sharing behavior on ESM could be a fruitful focus for future studies, previous research has already shown that the effect of employees' age on ESM use is Employees' information sharing on ESM not always clear-cut. For instance, it should not be taken for granted that the younger generations adapt better to social media platforms at work; sometimes the case is even the opposite . For some time, there has been a need to study technology-mediated communication processes outside the category of knowledge workers (see, e.g. Gilson et al., 2015;Rhoads, 2010). Although our informants worked in a media organization, not all of them would necessarily be described as knowledge workers. Some participants worked on the practical side of television production, for example, crafting set pieces and organizing props. These individuals were an interesting addition to our data, as computer access was limited for some of the team members and their daily work tasks included little computer-based work. There is, however, still a need for further research with more versatile data sets.
Additionally, the dual-sided nature of the affordances of visibility and persistence, that is, how affordances are simultaneously perceived as enablers and constraints, could be potential topics for further empirical research in the future. Similarly, the variance in informationsharing cultures, such as what is considered information worth sharing on ESM in different organizations and national cultures, is a topic that warrants further research. Overall, the model we present in this study should be examined in different contexts and different organizations to achieve a greater understanding of the matter.

Conclusions
The factors shaping employees' information sharing on ESM can be categorized into three dimensions: personal, technological and organizational. Personal privacy management principles shape employees' sharing behavior in four ways. Employees consider their personal privacy boundaries, professional boundaries and risks, online safety and perceived audience. The technological affordances of visibility, awareness, persistence and searchability are the most important affordances in the ESM context that shape employees' information sharing in varying, sometimes even contradictory, ways. Additionally, organizational norms and organizational tasks, as well as how other organizational media repertoires are utilized, play a role in information-sharing behavior on ESM. Our results show that these are the three key dimensions behind employees' decisions to share or not share information on ESM. This study extends the ESM literature and combines the theoretical perspectives of communication privacy management and technological affordances. Further research is required to better clarify how the personal, technological and organizational dimensions are present in other organizational, cultural and technological contexts.