EXPLORING THE IMPACT OF ENTERPRISE SOCIAL MEDIA AFFORDANCES ON WORK PERFORMANCE AND KNOWLEDGE TRANSFER

Knowledge is the result of collective work. As an important component of human capital and intangible assets of the company, when properly processed, knowledge transfer can foster work performance. For a decade, enterprise social media affordances have been studied based on surveys of different platforms users, not allowing to submit the users to the same conditions, therefore, ignoring the effects of platform design on enterprise social media. To address this gap, this empirical research studies the effect of enterprise social media affordances on employee performance from the perspective of knowledge transfer, using a single platform. The analysis of survey data collected from 317 Malagasy employees using the company’s own intranet has shown that association, visibility, persistence and editability affordances of ESM foster knowledge acquisition and knowledge provision, which in turn promote employee performance. Findings show that the relationship between ESM affordances and work performance are essentially mediated by knowledge provision. This study provides the key variables for a sustainable performance of the co-creation of value in the knowledge transfer process of the technological innovation process, and the levers for action on these variables. These variables are identified from the intersection of the technological (ESM affordances), social (knowledge provision and knowledge acquisition) and organizational (task performance and contextual performance) dimensions of the knowledge transfer process.


INTRODUCTION
In recent years, social networks or social networking sites like Facebook, Twitter, Google+ and LinkedIn have gained popularity on the Internet (Statista, 2021). Millions of people have started to (1) create their public or semi-public profile on these platforms, (2) list the users with whom they are in contact, (3) view and browse their list of relationships and that of other system users (Aisenberg, 2016). Today, more than 2 billion people around the world use at least one social network and on average, global internet users spent 2 hours and 23 minutes on social media per day, though trends differed widely by country (Statista, 2021). While the general public is in favor of social networks, companies, for their part, face a dilemma: doing nothing, prohibiting it or on the contrary encouraging these practices within the professional sphere (Alahmad et al., 2018). However, in a context where large organizations have to deal with geographically dispersed work teams or different business units, social networks can, if they are used properly, add value (Behrendt et al., 2017). Moreover, social media platforms have rendered communication so easy that any physical kind of journey is regarded unnecessary to conduct businesses, saving both time and resources. Companies have thus seen online social media possibilities in the workplace and have since encouraged the use of sites such as Facebook and LinkedIn within their organization (Alahmad et al., 2018). Nonetheless, information and communications between employees remain outside the corporate firewall and control on these public social networks. Concerned about IT security, data confidentiality and wishing to maintain a certain control over the reputation conveyed externally, more and more companies are inclined to develop their own social network functionalities and to deploy them on their very own Intranet (Carter et al., 2018).The prevalence and practical value of ESM have attracted the attention of scholars. Previous studies have shown that ESM can promote job performance, knowledge sharing, and organizational learning (Ahmed et al., 2019;Ellison et al., 2015;Kane, 2014;Qi and Chau, 2017). In using ESM, employees are exposed to several communication partners that yield disparate knowledge and ideas, which may result in creative combinations and divergent thinking (Perry-Smith, 2006). Consequently, as a collaboration tool that enables real-time and bidirectional communications within organizations, ESM has the potential to stimulate new perspectives. It is important to note that the effective use of ESM in an organization promotes information flow and transfer in an effective manner, among the organization's employees who can rely on some of the technical features available, such as the publicly available contacts list, the ability to send instant messages to individuals and groups, and video meetings. The ability of employees to use these features also contribute towards the affordances for them to gain and also contribute new knowledge to others. In social settings where individuals have a difficulty in developing a working relationship so that knowledge can be shared, ESM is an effective tool at providing lubrication so that problems such as knowledge stickiness can be overcome (Leonardi and Meyer, 2015). Another school of thought propagated by Fulk and Yuan (2013) notes that challenges that arise in social interactions among the employees within the organization can be overcome or significantly mitigated by ESM affordances on knowledge sharing. These challenges, which could include location of knowledge, motivation of the team and development of the social capital can be mitigated through ESM affordances. Some researchers also have noted that where ESM tools are used in an organization then it is possible for the affordances such as persistence, editability, association and visibility to change significantly the communication process while also promoting knowledge sharing (Treem and Leonardi, 2013). Judging from the thoughts put forward by these researchers on the issue of ESM affordances it can be asserted that it provides a viable platform for efficient knowledge flow within an organization and drive the subsequent innovation. This research will therefore seek to determine the extent to which ESM affordances can drive variations of work performance; conceptual and task performance, through two patterns of knowledge transfer too; knowledge acquisition and knowledge provision. A theoretical model was developed conceptualizing ESM affordance as enablers of knowledge transfer and creative performance. This study aims to address the following two research questions: (1) How do ESM affordances promote employees' participation in knowledge transfer processes? (2) How does participation in knowledge transfer processes affect employees' contextual and task performance?

Enterprise Social Media
Enterprise Social Media (ESM) Network is a group made up of natural and legal persons, united by a social networking device, within an organization. By extension, the term also refers to computerized networking tools that facilitate this shared and collaborative functioning within the company (Ennes, 2016). Members of a corporate social network are grouped together for professional purposes and are linked by an application platform that brings together the employees of a company, its customers, its shareholders or its partners (Thomas, 2015). Beyond communities, a corporate social network is based on the following concepts: • Profile or enhanced profile enhances the expertise of employees and their interests; • Monitoring wall of activity allows to have a consolidated vision of the activity of its ecosystem (colleagues, communities, exchange around a theme, etc.) An enterprise social network is equivalent to a "mainstream" social network. Members of a corporate social network are generally based on the flood inventory of address books, the exchange on projects, the sharing of documents between individuals is impossible in the same company (Agrawal, 2018). Under the term social media are gathered different sites and social features of the web and these include publishing contents such as articles, briefs, photographs, videos, PDF files, giving opinions or dialoguing with other users, contributing to participatory projects, aggregating information and then restoring it (Thomas, 2015).

Affordances
Affordance has attracted considerable attention in recent years, especially in the field of ESMPs (Cai et al., 2018;Leonardi, 2017;Ali-Hassan et al., 2015;Chaves et al., 2018). Although different studies have different definitions of affordance, it provides a new perspective to explain the impact of ESMPs on organizations and employees. The concept of affordance was raised by the psychologist Gibson in 1977 and was derived from the field of eco-psychology where affordance is the dynamic interaction between people and the materiality of the environment with which they come in contact and is considered to have an objective existence (Leonardi, 2017). Norman (1988) introduced this concept in the design field and suggested that there are two kinds of affordances when designers design an artefact: actual affordance and perceived affordance, of which the latter one was emphasized and viewed as dependent on actors' knowledge storage and experience. This concept was introduced into information systems by other researchers describing affordance as the functional and relational aspects that frame but not determine the opportunities for action in relation to an object (Chaves et al., 2018). Affordance is the user's perception of an object's ability and ESMPs afford users behaviours which are impossible to achieve with traditional communication and collaboration tools. Several affordances for knowledge sharing have been proposed in prior research.
Prior literature has identified several affordances. These include: reviewability, recombinability, and experimentation (Faraj et al., 2011); persistence, visibility, association, and editability (Treem and Leonardi, 2013); and network-informed associating, metavoicing, generative role-taking, and triggered attending (Majchrzak et al., 2013) and have been shown to have different impacts such as organizational knowledge collaboration, socialization, flow of knowledge, power relations, and knowledge sharing. In social commerce literature, researchers have examined the positive correlation between social commerce affordances and customers' purchase intention using empirical methods (Lin et al., 2019;Sun et al., 2019a). Leonardi (2012) identified four affordances: visibility, editability, persistence and association.

Knowledge Transfer
Knowledge transfer is regarded as an increasingly important process of knowledge management due to its ability to help the governing body to benefit from private knowledge. Knowledge transfer is defined as a process through which knowledge moves between a root and a recipient and where knowledge is given and practiced (Boh et al., 2016). In fact, knowledge can be transferred among people between levels in the hierarchy of the organization and between units and departments different in organizations, knowledge transfer is practical problems transferring knowledge from one part to another organization. Such as knowledge management, knowledge transfer, which aims to organize, create, capture or distribute knowledge and ensure its availability for future users (Calipha et al., 2018). Knowledge transfer is "the knowledge exchange process between knowledge providers and knowledge recipients" (Kim et al., 2011). It can be realized through communications and dialogues so that knowledge seekers can learn and apply the knowledge from another person (Cao et al., 2016). In organizations, employees are both acquirers and providers of knowledge. Knowledge acquisition and knowledge provision are two different ways to boost knowledge transfer and flow (He and Wei, 2009;Soda et al., 2019). Information technology (IT) can promote the exposure to diverse sources of knowledge and thus accelerate the knowledge transfer process (Alavi and Leidner, 2001). Harman and Brelade (2017) stated that knowledge transfer is dedicated to the transfer of explicit knowledge. Harman and Brelade (2017) discussed on the different view of authors in defining knowledge transfer when it comes to the application in service provision which related to IT service and outsourcing. The article in details discussed at the similar key point as other articles where different term definition given, but based on the same concept of knowledge transfer.

Knowledge Acquisition
Prahlad and Hammel (2017) relying on the knowledge-based view had noted that competencies are seen as the basis for a company's ability to acquire competitive advantage. They had further observed that employees improved work action in relation to assigned responsibilities is not a function of tangible or extrinsic incentives or the conducive work environment rather, the todays worker characteristically acquire knowledge which constitute the asset that reengineer all work processes towards goals. As earlier noted, the acquisition component of the entire knowledge management process is fundamental as it precedes other activities in the entire knowledge management spectrum.

Knowledge Provision
Rendenick (2018) had considered knowledge spread among employees as being a significant practice especially within tacit classification of knowledge. The argument put forward here is that since knowledge is in built within the individuals, it will require germane organizational platform created to help in its acquisition. To achieve this Renderick (2018) believed that a continuous interaction platform that will help in knowledge sharing and transfer is important for knowledge acquisition and sharing. This is in addition to the authors thinking that the sources from which the individual acquire knowledge and share it is also imperative for quality knowledge that meets the goals of building intellectual capital that is sustainable for competitiveness.

Work Performance
Work performance is defined as the total expected value to the organization of the discrete behavioural episodes that an individual carries out over a standard period of time. This definition is a slightly revised version of the definition of performance presented in a previous publication in connection with a theory of individual differences in task and contextual performance (Kinicki and Kreitner, 2018). One important idea in this definition is that performance is a property of behaviour. In particular, it is an aggregated property of multiple, discrete behaviours that occur over some span of time. A second important idea is that the property of behaviour to which performance refers is its expected value to the organization (Anjun et al., 2018). Thus, the performance construct by this definition is a variable that distinguishes between sets of behaviours carried out by different individuals and between sets of behaviours carried out by the same individual at different times. The distinction is based on how much the sets of behaviours (in the aggregate) are likely to contribute to or detract from organizational effectiveness. In a word, variance in performance is variance in the expected organizational value of behaviour (Rasool et al., 2019). Performance is a multi-dimensional concept. On the most basic level, Borman and Motowidlo (2013) distinguish between task and contextual performance.

Task Performance
Task performance refers to an individual's proficiency with which he or she performs activities which contribute to the organization's 'technical core'. This contribution can be both direct (e.g., in the case of production workers), or indirect (e.g., in the case of managers or staff personnel). Task performance in itself is multi-dimensional. For example, among the eight performance components proposed by Quinn and Cahill (2018), there are five factors which refer to task performance: (1) job-specific task proficiency, (2) non-jobspecific task proficiency, (3) written and oral communication proficiency, (4) supervision-in the case of a supervisory or leadership position-and partly (5) management/administration. Each of these factors comprises a number of subfactors which may vary between different jobs (Zacher et al., 2018). For example, the management/administration factor comprises subdimensions such as (1) planning and organizing, (2) guiding, directing, and motivating subordinates and providing feedback, (3) training, coaching, and developing subordinates, (4) communication effectively and keeping others informed (Borman & Brush, 2013). In recent years, researchers paid attention to specific aspects of task performance. For example, innovation and customer-oriented behaviour become increasingly important as organizations put greater emphasis on customer service (Kunze et al., 2015).

Contextual Performance
Contextual performance refers to activities which do not contribute to the technical core but which support the organizational, social, and psychological environment in which organizational goals are pursued (Lorincova et al., 2018). Contextual performance includes not only behaviours such as helping co-workers or being a reliable member of the organization, but also making suggestions about how to improve work procedures. Researchers have developed a number of contextual performance concepts. On a very general level, one can differentiate between two types of contextual performance: behaviours which aim primarily at the smooth functioning of the organization as it is at the present moment, and proactive behaviours which aim at changing and improving work procedures and organizational processes (Di- Marco et al., 2018). The 'stabilizing' contextual performance behaviours include organizational citizenship behaviour with its five components altruism, conscientiousness, civic virtue, courtesy, and sportsmanship, some aspects of organizational spontaneity (e.g., helping co-workers, protecting the organization and of prosocial organizational behaviour (Koopmans et al., 2017;Johns and Al-Hajj, 2016). The more pro-active behaviours include personal initiative (Frese, Kring, Soose, & Zempel, 2016), voice (Van Dyne & LePine, 2018), taking charge (Morrison & Phelps, 2009). Thus, contextual performance is not a single set of uniform behaviours, but is in itself a multidimensional concept.

Theoretical Model and Hypotheses
The model proposes that in this research the focus is on these four affordances of ESM which influence knowledge transfer (knowledge acquisition, knowledge provision), which also in turn influence work performance (task performance, contextual performance). In essence therefore, the proposition is that the relationship between ESM use and work performance is mediated by knowledge transfer.

ESM affordances and knowledge transfer
ESM affordance is defined as the potential or possibility for a particular action made possible by the characteristics or capabilities of ESM consistent with user goals (Faraj and Azad, 2012). In this study, four affordances-visibility, association, persistence and editability (Treem and Leonardi, 2013) -are expected to promote knowledge sharing and transfer (Ellison et al., 2015;Evans et al., 2017;Oostervink et al., 2016). Knowledge transfer is represented by two constructs: knowledge acquisition and knowledge provision. Knowledge acquisition is the act of seeking and collecting work-related information and know-how within organizations (Esmaeelinezhad and Afrazeh, 2018;Wang and Noe, 2010). Yuan et al. (2007) argued that employees acquire knowledge mainly from two kinds of sources within organizations: direct knowledge exchange with other individuals or knowledge extraction from knowledge repositories. Knowledge provision is the act of contributing tacit knowledge to other colleagues or compiling and storing existing knowledge into knowledge repositories (He and Wei, 2009). Prior studies have shown that individual motivation (e.g., enjoy helping and reputation) and cultural factors (e.g., innovativeness, openness, and fairness) tend to influence knowledge contribution (Bock et al., 2005;Wasko and Faraj, 2005;Yu et al., 2010).

Association affordance and knowledge transfer
Association refers to "the possibility of establishing connections between individuals or between individuals and content" (Pee, 2018) and it allows organization members to forge and maintain social ties as well as build social capital across disparate and geographically distant boundaries (Ellison et al., 2015). The newly developed weak ties and further-strengthened strong ties attained through association enable employees to connect with other colleagues and engage in direct knowledge exchange to acquire knowledge regardless of time and distance (Fox and McEwan, 2017 (Treem and Leonardi, 2013). Therefore, association affordance can enable employees to obtain knowledge or information from knowledge repositories on the platform or web links others share on ESM. Since ESM association provides a channel for employees to have bidirectional interactions with others regardless of distance, they can be both knowledge acquirers and knowledge contributors when engaged in interactions with others using ESM. Therefore, H1a. Association affordance is positively related to employees' knowledge acquisition behaviour. H1b. Association affordance is positively related to employees' knowledge provision behaviour.

Visibility affordance and knowledge transfer
Visibility refers to the possibility of enabling "employees' behaviours, knowledge, preferences, and communication networks that were once hard to see visible to others in the organization" (Treem and Leonardi, 2013) and it makes employees' tasks and social information widely known within the organization . Fulk and Yuan (2013) showed that visibility affordance could help with the challenge of locating expertise by reducing the transaction costs of time and effort when seeking knowledge. Leonardi (2015) argued that the consciousness of "who knows what" and "who knows whom" will increase through visibility, thus enabling employees to identify experts in relevant fields and acquire work-related knowledge from their colleagues. Visibility affordance can influence knowledge contribution since trust among employees becomes stronger with ESM use (Sun et al., 2019b). When employees confirm the identity of knowledge seekers through visibility, they tend to share or contribute knowledge based on reciprocity and trust. Visibility makes employees' job information and existing social networks more transparent and can enhance the identity authentication of members within the organization (Ellison et al., 2015), which may help reduce concerns about sensitive information and core knowledge capital draining out of the organization when contributing knowledge. Visibility also allows people with similar interests and values to build relationships, making them more willing to engage in the knowledge exchange process. It is closely related to the presentation of self by knowledge providers through their expertise and competence (Pee, 2018) and thus drives employees to manage their self-presentation through knowledge contribution to create a more favourable impression, which is helpful to gain access to key resources in organizations (Osch and Steinfield, 2018). Therefore, H2a. Visibility affordance is positively related to employees' knowledge acquisition behaviour. H2b. Visibility affordance is positively related to employees' knowledge provision behaviour.

Persistence affordance and knowledge transfer
Persistence refers to the notion that "information remains available to users and does not expire or disappear" (Treem and Leonardi, 2013). It enables users to search, browse, replay, annotate, visualize, restructure, and recontextualize the information previously shared on ESM. Treem and Leonardi (2013) argued that persistence can be conducive to sustaining knowledge and growing content, which implies that it can be a facilitator for employees to acquire knowledge they need from ESM repositories. The perceived value of knowledge can influence knowledge providers' level of contribution (Fulk et al., 2004). Persistence enables the contributed knowledge to exist on ESM for a long time, allowing their colleagues to view and reuse knowledge no matter when they need, which can help increase employees' perceived usefulness and value of the knowledge they shared. Knowledge that endures on ESM can free knowledge providers from repeated and ineffective assistance to others, and employees are hence inclined to contribute knowledge. Therefore, H3a. Persistence affordance is positively related to employees' knowledge acquisition behaviour. H3b. Persistence affordance is positively related to employees' knowledge provision behaviour.

Editability affordance and knowledge transfer
Editability refers to "the possibility of crafting and re-crafting a communicative act before or after it is viewed by others" (Pee, 2018). Therefore, collaborative knowledge transfer becomes possible within the organization because of editability. By enabling participants to compile and modify knowledge documents, editability enables the content to be easily revised, reshaped, and coordinated by knowledge providers themselves and other members within the organization, which further improves the quality and value of the knowledge contributed and prompts employees to seek and acquire knowledge through ESM (Borgatti and Cross, 2003). Time and effort expended for providing knowledge influence knowledge contribution (Fulk et al., 2004). Editability affordance allows employees to codify their tacit knowledge into explicit knowledge, and also modify or revise content progressively, thus reducing the time and effort to organize and compile knowledge from scratch. Complex knowledge will become easier to codify if the artefacts support high editability (Pee, 2018). Editability can enable employees to strategically select the way they share knowledge with others, and retain some control of the knowledge they contribute, thus reducing the risk and uncertainty of sharing knowledge. Therefore, H4a. Editability affordance is positively related to employees' knowledge acquisition behaviour. H4b. Editability affordance is positively related to employees' knowledge provision behaviour.

Knowledge transfer and work performance
Individuals' factual knowledge or expertise in their task fields is the raw material and foundation for creative performance (Amabile and Pratt, 2016). When using ESM, employees are not passive recipients of knowledge but rather users and communicators with their colleagues according to their needs. Therefore, employees who participate in knowledge acquisition can usually acquire more diversified knowledge and get comprehensively informed (Sigala and Chalkiti, 2015). In addition, employees can have access to more diverse and remote knowledge sources and holders when acquiring knowledge, which provide employees with disparate information, knowledge and perspectives, so that employees have the opportunity to make new combinations of ideas (Perry-Smith and Shalley, 2003). When employees contribute knowledge, their cognitive structure and metacognition of knowledge in a certain field will be further developed and improved in the process of verbalizing and transmitting knowledge to colleagues (Rhee and Choi, 2017), which can help facilitate the generation of employees' creative ideas. Employees can enrich their knowledge reserves through informational and knowledgeable returns obtained when providing task-related content requested by their co-workers (Cropanzano and Mitchell, 2005). Therefore,

H5. Knowledge acquisition is positively related to employees' Task performance. H6. Knowledge acquisition is positively related to employees' Contextual performance.
H7. Knowledge provision is positively related to employees' Task performance H8. Knowledge provision is positively related to employees' Contextual performance.

Data Collection
We used a questionnaire survey on an online survey platform to gather data for model testing. The questionnaire was answered by full-time employees of a large company in Madagascar who used STARCom, an internal ESM platform that supports message notification, telephone/video conference, email exchange, schedule arrangement, task management, file sharing, business discussion, business approval, and resource reservation among employees. A pilot test was first carried out with 30 employees of the company and the questionnaire was refined based on the feedback. Two reverse-coded questions were included to check whether respondents were serious about the questionnaire. Data collection lasted two weeks. In total, 425 users participated in the online survey. After eliminating responses based on the shortness of time to complete the survey and the contradictory answers to reverse-coded questions, a total of 317 valid questionnaires, with an effective response rate of 74.6%. Table 1 shows the demographics of the respondents.

Measurement Scales
Some of the measurement items were based on prior literature and measured association affordance using items adapted from Rice et al. (2017) and Pee (2018) while visibility, persistence and editability affordances were measured with items from Rice et al. (2017). We measured knowledge acquisition and knowledge provision with the items from Reinholt et al. (2011). All measurements were gathered using a 5-Likert scale, in which 1 indicated "strongly disagree" and 5 indicated "strongly agree". Consistent with prior studies (Chung et al., 2015;Yan et al., 2013), demographic variables (such as gender, age, education, job tenure) that may affect knowledge transfer were regarded as control variables in the theoretical model.

Study Population
Star is a beverage production company in Madagascar which was incorporated in 1953, and its stated vision is "To be the regional leader in the production and distribution of food drinks as part of a modern, efficient, profitable and civic organization. The company wants to ensure that there is constant availability and quality of its production at the best price, through an efficient distribution network to the service of its customers. The company also maintains twenty-five (25) brands and more than seventy references. Star operates seventeen (17) branches in Madagascar and more than twenty thousand (20000) points of sale for its products. Its operations in Madagascar have over 1800 staff members, and this research targeted the location at Antananarivo in Madagascar, where a total of 317 management and staff members participated in the research. The company operates its STARcom enterprise media sites which allows for comprehensive communication across different departments and locations in the company. This system was evaluated in terms of employee perceptions of its impact on their work performance, as well as the role knowledge transfer plays in that relationship.

Data Analysis and Results
The partial least squares (PLS) was used to evaluate the measurement and structural models. PLS is suitable for empirical research development as it is a mature structural equation modelling estimation method that can perform an accurate test of model fit. PLS is a widely used estimation method in empirical studies published on leading journals in IS research (Braojos et al., 2019). Compared to Amos or LISREL, PLS demands fewer rigid assumptions about population, distribution or scale of measurement (Chai and Kim, 2012). SPSS 24 was used to conduct single-factor test to detect common method bias, and correlation analysis to validate our hypotheses.

Measurement Model
In order to test the viability of the model a bootstrapping procedure which employed about 5000 subsamples and in doing so, it was determined that mediating variables such as age of respondents, gender, level of education and their length of time working with the organization did not show significant relation to their contextual or task performance. The results for the test for reliability and validity show that the values achieved for the Cronbach's alpha for each of the constructs were above the recommended threshold for suitability, with the range being 0.74 (PER) and 0.88 (EDI). The coefficients for the composite reliability (CR) achieved also ranged from 0.82 (PER) to (0.92) EDI, and this also met the minimum threshold for suitability. This indicates that overall, the measurements have good internal reliability, and looking at the average variance extracted (AVE) which was achieved for each of the constructs, they were all also above the acceptable suitability threshold of 0.5, ranging from 0.63 (KA) to 0.74 (VIS). The factor loadings which were also calculated for each of the constructs also showed that they were above the suitable threshold of 0.7 also, and these indicate adequate convergent validity. To show that there is also satisfactory discriminant validity achieved, it can be seen from the summary table that the AVE achieved for each construct is shown to be higher than the square of correlations they achieve with other constructs. The factor loadings shown in Appendices also indicates that the constructs achieved larger factor loadings that their cross loadings, and that is also a good indicator of the discriminant validity.

Common Method Bias
The issue of common method bias (CMB) often arises when the data collected using questionnaires is done at a similar time and there is no variability in the sourcing of the data using this method. To counter this phenomenon of CMB, it was gauged using the Harman's single-factor test which was done using the exploratory factor analysis. Based on an analysis of the results achieved it was determined that 10 factors had eigenvalues which were greater than 1.0, and the first factor actually was accountable for 39.3% of the variance, and this is lower than the acceptable limit of 40%.

Hypotheses testing
The results of the hypotheses testing are presented in table 4.4 and the results do show that the ESM affordances of Association (ASS), Visibility (VIS), Persistence (PER) and Editability (EDI) all achieved positive association with the elements of Knowledge Transfer; Knowledge Acquisition (KA) and Knowledge Provision (KP). This therefore supports H1a to H4b, and from the summary table, it can also be seen again that Knowledge Transfer has a positive impact on Work Performance. This is shown by the significant positive relationships of Knowledge Acquisition (KA) and Knowledge Provision (KP) to Contextual Performance (CP) and Task Performance (TP). From figure 4.1 below it can also be seen that the explained variances that exist with KA, KP, CP and TP are 61.2%, 47.7%, 75.1% and 72.4% respectively.
To further validate the results which were achieved through the hypothesis testing, it was necessary to also carry out the correlation testing of the dependent, independent and mediator variables which were used in the research. The Pearson Correlation Matrix in table 4.5 also shows that the variables identified in the hypotheses all retain significantly positive coefficients, which range from 0.60 to 0.75, therefore indicating further support for the hypotheses.

Post Hoc Testing
To test the effect of the mediation factors, knowledge acquisition and knowledge provision between the ESM affordances and work performance, bootstrapping procedures were used. To carry this out in an effective manner, the direct and indirect effects that the intermediary paths had, along with their significance were examined so that a better understanding of the mediation effects and type could be achieved. From the summary of mediation results in table 4.6 it can be seen that knowledge acquisition had no discernible mediation effect between the ESM affordances and either contextual or task performance. It is evident that the indirect effect at CI 95% for association affordance on contextual performance (0.004, 0.051) and task performance (0.004, 0.053) through knowledge provision, and the direct effect (0.038, 0.225) and (0.035, 0.230) are all positively significant. From these results, it can be concluded that there is a complementary mediation of knowledge provision on the relationships between association affordance and contextual performance, or task performance. Furthermore, the visibility affordance (95% CI = 0.008, 0.044), persistence affordance (95% CI = 0.0010, 0.036) and editability affordance (95% CI = 0.0011, 0.026) have indirect effects on contextual performance. In the same vein the visibility affordance (95% CI = 0.009, 0.073), persistence affordance (95% CI = 0.0010, 0.030) and editability affordance (95% CI = 0.0011, 0.064) have indirect effects on task performance. However, the summary also shows that there is no direct effect on contextual or task performance which is significant at the 95% CI, therefore showing that the indirect-only mediation of knowledge provision for the relationships that exist between visibility, persistence and editability on contextual and task performance.

Key Findings
This research was carried out to determine the impact that Enterprise Social Media (ESM) would have on work performance, within the framework of knowledge transfer. Within the model that was adopted for this research, ESM affordances were selected, namely, association, visibility, persistence and editability, and these were portrayed as enabling knowledge transfer with contextual and task performance as the outcomes. Based on the analysis of the data collected from respondents in Star and the testing of the hypotheses which were advanced, it was determined that the ESM affordances have significant effects on the factors on knowledge transfer selected, being knowledge acquisition and knowledge provision. This finding aligns with conventional knowledge from previous researchers who concluded that ESM use can ease knowledge transfer within an organization. The data analysis also showed that in more specific terms, the association affordance is seen to have a more direct impact on knowledge acquisition than on knowledge provision within an organization. The effect of this therefore is that the features within the ESM tools available to employees within the organization, that promote the association affordance through direct messages, group messages and public contact list searches will aid the employees to develop certain communication ties which give them access to knowledge sources. This will in effect reduce the associated challenges with development of social capital when users are seeking knowledge. The findings made from this research also show that knowledge acquisition and knowledge provision within an organization have a positive association with work performance. This essentially means therefore that employees in an organization, having relevant knowledge to apply towards the work that they do, and to be able to share with others can improve upon performance significantly. Employees are able to share and gain knowledge from one another and effectiveness in contextual and task performance can be stimulated, and this finding also reinforces the need for knowledge in an organization to promote effective and efficient work output. The research also considered the extent that knowledge transfer (knowledge acquisition and knowledge provision) in an organization can mediate the impact of ESM affordances on work performance. The findings also made from the data analyzed showed that in terms of knowledge acquisition on the ESM affordances and work performance (contextual and task performance) there was no mediating effect identified. Referencing the demographic data from the respondents, they more than 50% indicated that they relied on the STARCom tool at Star to solve some problems that they encountered, though this was not on a regular basis, and therefore aligns with this assertion that challenges often encountered may be routine, and not specifically related to their contextual or task performance as employees of the company. It may not be entirely efficient or effective for the employees of the company to rely significantly on the ESM affordances to solve these problems, but rather, as complementary to their own efforts at solving problems due to the fact that they only accessed the ESM tool at specific times or only when they had notifications. This therefore means that the knowledge that the respondents acquired through the ESM affordances may not have a significant impact on improving their contextual or task performance since time constraints and performance targets may lead the workers to be more focused on completing the task at hand using ESM. There was also a complementary mediation of knowledge provision on the relationship between the ESM affordances and work performance, and therefore this also indicates that the association affordance has a more direct influence on work performance, and more indirectly through the influence it has on knowledge provision. Knowledge provision fully mediated the relationships of visibility, persistence and editability affordances with contextual and task performance. This suggests that knowledge provision effectively delivers the positive impact of visibility, persistence and editability affordances of ESM on contextual and task performance.

Theoretical Implications of Research
In terms of the theoretical implications of this research, the first noted is that this adds to existing knowledge on ESM, work performance and knowledge sharing. Though there has been prior research that has been focused on the effect of ESM on performance of employees as well as the level of knowledge sharing, the specific focus on work performance in terms of the contextual and task perspectives have not been explored. This is especially relevant as it was conducted with companies operating in Madagascar in mind, where there is essentially very limited research on this focus. This gap in knowledge was therefore addressed in this research through developing an understanding of how ESM affordances, being association, visibility, persistence and editability, can influence the work performance of employees through an empirical examination of the theoretical model which was developed within the framework of knowledge transfer. Through this research the knowledge gained is that the ESM affordances have significant effects on knowledge acquisition and knowledge provision. In more specific terms of the ESM impact on work performance, the research has also added new knowledge to the affordance's element. It cannot be denied that previous researchers have carried out research on ESM affordances, with some going more broader than those captured in this research, however in terms of the empirical perspectives, the data is limited. Through a conceptual framework that relied on the affordance's perspective, the research sought to determine how ESM can impact the performance of a worker through knowledge acquisition and knowledge provision. The research concluded that the ESM affordances significantly impacted on knowledge transfer and work performance of employees, and this reinforces the need for organizations to adopt social media and other digital technologies to effectively and efficiently improve on employee performance. The research also allows for an analysis of knowledge transfer from two unique perspectives, knowledge acquisition and knowledge provision, and how they can impact conceptual and task performance. These variables have limited existing literature on the relationship that exists between them, and this gap is also closed to some extent through the findings in this research. On the knowledge management front, knowledge transfer activities positively impact the knowledge creation process, which can also impact on how employees are able to effectively and efficiently undertake their assigned work.

Practical Implications of Research
In a practical sense, this research shows that ESM affordances within an organization can induce employees to engage in the knowledge transfer process, and therefore for employees in Star, it is important that they take advantage of association, visibility, persistence and editability to gain relevant knowledge to improve on their work performance. To gain knowledge from experts who may also have access to the ESM tools or software available in the organization, employees can consider the affordances such as visibility of the contacts lists and also contact cards, as well as the association affordance of direct messaging and group messaging which can allow interactions with such experts to gain relevant knowledge. Where there is already substantial existing knowledge within the organizations ESM network that can meet the knowledge needs of the employees then the persistence and affordability affordances can allow them to access this information. The research also showed that knowledge transfer could play a mediation role between ESM affordances and work performance, so the employees must actively work to contribute new knowledge and also absorb any new knowledge they may come across. This is relevant for the employees because the knowledge gained can stimulate their potentials for more effective and efficient contextual and task performance. The designers of the ESM tools and software for organizations must also ensure that the affordances identified in this research, association, visibility, persistence and editability must be included in features that are available to users so that they can actively seek ways to contribute and also absorb new knowledge to enhance their work performance.

Limitations and Future Research
There were some limitations encountered in this research that may have had an impact on the conclusions which were drawn. The data collected for this research was done through questionnaires which were completed by the respondents, so all elements of ESM affordances impact, knowledge transfer and work performance were self-reported, and did not rely on other sources of data that could verify them. This therefore could include significant personal biases of respondents into the data mix and influence the conclusions. Therefore, future research can rely on multiple data sources which can less impact of respondent biases. Secondly, the research was carried out on one company operating in Madagascar, and the findings will therefore be characteristic of the specific company and may not broadly represent the situation in the whole country or outside of the country. To ensure generalization, future research can be carried out with a cross-national context so that different national and cultural perspectives can be compared to derive relevant conclusions. The number of respondents who participated in this research was limited to just 317, and this is in stark contrast to the over 1800 staff members who have access to the STARCom platform. The conclusions drawn may therefore not be representative of the views or positions of the entire organization when taken more broadly. Finally, this research only focused on four affordances to determine the impact of ESM on work performance. Future research can also consider other ESM affordances and how they will influence work performance within the knowledge transfer framework.

CONCLUSION
This study examined the impact of ESM affordances on employee work performance from the perspective of knowledge transfer. The findings show that association, visibility, persistence and editability affordances of ESM can promote knowledge flows through two distinct ways of knowledge transfer (knowledge acquisition and knowledge provision). Knowledge acquisition and knowledge provision have positive effects on employee contextual and task performance. Through the mediating effects analysis, it was found that knowledge provision partially mediated the relationship between association affordance and work performance, and fully mediated the relationships between visibility, persistence and editability affordances with work performance. However, knowledge acquisition has no mediating effect on these relationships. This research extends our understanding of social technologies and user creativity in the context of organizations and provides guidance for enterprise practice.