Intra-organizational knowledge sharing, ambidexterity and firm performance: evaluating the role of knowledge quality Open Access

Organizational ambidexterity has become indispensable in today’s unpredictable and multi-dimensional business environment, allowing organizations to concurrently execute exploitative and explorative activities (Lis et al., 2018; Stelzl et al., 2020; Rojas-Córdova et al., 2023). Organizational ambidexterity is defined as the “ability of an organization to both explore and exploit – to compete in mature technologies and markets where efficiency, control, and incremental improvement are prized and to also compete in new technologies and markets where flexibility, autonomy, and experimentation are needed” (O'Reilly and Tushman, 2013, p. 1). While exploitation entails developing existing products, services and processes and exploring opportunities in mature markets, exploration is concerned with innovation in emerging markets (Eisenhardt et al., 2010).

Simultaneously managing the distinctive requirements of exploitation and exploration processes is a significant challenge for ambidextrous organizations (Ali et al., 2022; Turner et al., 2013), especially in developing economies (López-Zapata and Ramírez-Gómez, 2023). These organizations strive to balance (1) adaptability, which is necessary for prospecting new products, markets, technologies and industry trends, that is, exploration and (2) alignment around existing products/markets (i.e. exploitation). On the one hand, the focus on adaptability from a short-term perspective may inhibit the capacity of the organization to foresee emerging trends in the industry. On the other, adaptability overtaking alignment will likely jeopardize ongoing business operations (Birkinshaw and Gibson, 2004). Moreover, contemporary organizations have increased their focus on environmental protection issues, where ambidextrous green innovation practices tend to increase these entities’ green performance (Shehzad et al., 2023) and ultimately lead them to proposing and adapting adequate ambidextrous environmental strategies to better comprehend its important drivers (Xi et al., 2023).

While previous studies on ambidexterity have largely focused on its characteristics, antecedents and outcomes, recent research is increasingly building on organizational learning theory to address the dynamics of knowledge creation and utilization in organizations (Ali et al., 2022; Eraslan and Altindag, 2021; Lissillour and Rodriguez-Escobar, 2023; Zaim et al., 2019). Organizational learning ambidexterity refers to the ex ante strategic motives underlying ambidexterity, which consist of using and refining existing knowledge assets to exploit current products/markets and creating new knowledge assets to fill knowledge gaps and support explorative activities (Ali et al., 2022). Knowledge assets – also referred to as intellectual capital – encompass three categories: human, organizational and social capital (Ali et al., 2022; Lin et al., 2017).

In this sense, the centrality of the knowledge perspective is salient in organizational ambidexterity research. Numerous studies have incorporated the knowledge-based view (KBV) to further understand the dynamics of organizational ambidexterity and its capacity to influence organizational performance at different levels (Alshawabkeh et al., 2020; Amankwah-Amoah and Adomako, 2021; Muñoz-Pascual and Galende, 2020; Ramachandran et al., 2019; Shafique et al., 2022). A specific array of these studies focused on the interplay between two forms of knowledge – tacit and explicit – and ambidexterity within an extensive perspective of organizational outcomes such as performance, creativity, and innovation (Lin et al., 2017; Muñoz-Pascual and Galende, 2020). However, the role of knowledge quality in context of knowledge sharing and ambidexterity has not been sufficiently explored in the extant literature in spite of several studies alluding to the importance of the quality of knowledge flows in organizations (Chang and Chuang, 2011; Doronin et al., 2020; Ganguly et al., 2019; Hujala and Laihonen, 2021).

This paper contributes to ambidexterity research in multiple ways. First, we examine the role of knowledge quality, which has yet to be sufficiently explored in the literature connecting knowledge sharing with organizational ambidexterity. Second, by delineating tacit and explicit knowledge sharing within organizations, we provide more nuanced insights into how each may distinctly impact exploitative and explorative capabilities. Furthermore, we offer a more international context and greater validity to ambidexterity research by empirically exploring its link to performance of organizations in the United Arab Emirates (UAE). Additionally, we present several theoretical and practical implications related to the model based on empirical results.

Consistent with the premises of the social capital theory, proposed by Nahapiet and Ghoshal (1998), the circulation and integration of tacit and explicit knowledge forms require effective communication and interaction between individuals, which is largely influenced by the relationships between them (Hau et al., 2013; Lin et al., 2017). The influence of social capital is effective at the individual and team levels, both of which determine the extent of individual tacit and explicit knowledge sharing (Yu et al., 2013). Social capital, defined as individuals’ propensity to share knowledge and cooperate with peers and external actors in pursuit of innovative solutions, is thought to support the exploitation and exploration processes underlying organizational ambidexterity in a context of high absorptive capacity (Lee et al., 2021) or open innovation systems (Lazzarotti et al., 2017). Such a dynamic arguably relies on social capital’s potential to stimulate knowledge sharing in learning networks (Chumnangoon et al., 2023; Lefebvre et al., 2016). Social capital enhances employees’ intentions to engage in tacit and explicit knowledge sharing (Hau et al., 2013). The relational dimension of social capital affects tacit and explicit knowledge sharing between organizational members working in different teams (Santos et al., 2023). Another study found that two social capital dimensions – social relational and social cognitive capital – increase tacit knowledge sharing, which is further positively related to an organization’s innovation capability (Ganguly et al., 2019). Hence, scholars recognize that social capital significantly and positively affects organizational ambidexterity (Lee et al., 2021), particularly based on its capacity to shape tacit and explicit knowledge transfer.

Nevertheless, the social interaction underlying tacit or explicit knowledge sharing may be affected by the quality of the knowledge that supports ambidextrous activities in the organization. As per the social capital theory proposed by Nahapiet and Ghoshal (1998), effective knowledge sharing depends on recipients’ degree of satisfaction with the quality of the knowledge flows conveyed by their peers (Ganguly et al., 2019; Muhammed and Zaim, 2020). The theory also stipulates that employees’ tacit and explicit knowledge-sharing intentions are positively stimulated by the perceptions of reciprocity between individuals. Recipients of valuable knowledge feel “indebted” and strive to return knowledge of similar quality to the senders (Hau et al., 2013). A related research area is social networks within organizations that stipulate informal conduct and social relations among the organization’s members. These networks serve as a conduit for knowledge transfer (Blau and Scott, 2003; Caimo and Lomi, 2015).

However, as Chang and Chuang (2011) pointed out, the knowledge contributors in an organization tend to be more focused on increasing the quality of the knowledge shared, not just its quantity, which mainly depends on whether the members of a certain community of practice exhibit intensive interactions, trust each other, perceive that the knowledge exchange is fair and have a sense of belonging to a community. Ultimately, and in line with the social theory stipulated by Nahapiet and Ghoshal (1998), there are four distinct factors that affect the creation of intellectual capital: opportunity for exchange, anticipation of value, motivation to engage and the combination capability of the receiver. Hence, the successful creation of intellectual capital in an organization depends on knowledge-sharing quality (Doronin et al., 2020). Previous studies have investigated the quality of knowledge shared indirectly through effective development of intellectual capital. This study takes a more direct approach and measures the quality of knowledge shared by incorporating a distinct measure of knowledge quality and separating it from knowledge-sharing behavior (see Figure 1).

Although a number of studies have used the KBV of organizations in ambidexterity research, few (if any) have examined if firm’s knowledge quality impacts the relationship between tacit/explicit knowledge sharing and exploitation/exploration processes. Prior studies on knowledge quality were directed toward understanding its impacts on innovation in the context of external knowledge sourcing (Corral de Zubielqui et al., 2019; Demirkan et al., 2013) and on firms’ innovation capabilities and performance (Abdollahbeigi and Salehi, 2022; Ganguly et al., 2019). Building on the intersection between social capital theory and the KBV, we argue that exploring the role of knowledge quality is crucial to further comprehend the relationship between knowledge sharing, ambidexterity and firm performance. Therefore, this study aims to reveal the importance of knowledge quality in the relationship between knowledge sharing, organizational ambidexterity and performance. In particular, we suggest that knowledge sharing (both tacit and explicit knowledge sharing) impacts ambidexterity directly and through improved knowledge quality. Furthermore, organizational ambidexterity (through both exploitative and explorative capabilities) positively impacts organizational performance (see Figure 1).

The remainder of this paper is organized as follows. Section 3 draws on the existing knowledge-sharing and ambidexterity literature and explores the conceptual model presented in Figure 1 in greater detail to formalize the hypotheses. Specifically, we draw on the literature on the KBV of the firm and organizational learning theory to examine the role of knowledge quality in the relationship between knowledge sharing and ambidexterity and present a detailed model with the hypotheses being tested and then describe the study research methods. Section 4 describes the methodology, and Section 5 reveals the analysis results. Section 6 discusses the results and their implications on theory and practice within the limitations of this study and offers some suggestions for further research. Finally, Section 7 presents the major conclusions of this research.

Knowledge is a key element of companies’ sustained competitive advantage (Barney, 1991), which is why firms are perceived as “institutions for integrating knowledge”. During the 1990s, organizations were typically seen as machines for processing information and thus relied primarily on formal and quantifiable data, known procedures and relevant principles, otherwise labeled as “explicit” organizational knowledge. However, this view vastly disregarded the knowledge that is personal, hard to quantify and standardize and difficult to communicate. Thus, another form of information, regarded as “tacit” knowledge, was denoted as the key differentiating factor between successful and less successful organizations (Nonaka, 1991). In the light of this changing paradigm, the field of knowledge management emerged as a crucial topic in research and practice with two main objectives:

1. to assure that organizations act intelligently in all aspects of operations; and

2. to fully capitalize on knowledge-based assets (Wiig, 1997) through added value to their stakeholders.

Furthermore, organizations’ actual capacity to share knowledge among their employees and to use it while performing critical operations represents the foundation of its long-term survival and growth (Haas and Hansen, 2007). Accordingly, this study focuses on revealing the nature of relationships among knowledge sharing (both explicit and tacit), knowledge quality, organizational ambidexterity and ultimately performance.

According to Castells (2010), the information age has given rise to a global, networked and interconnected society that is highly dependent on digital technologies. According to Castells, the information age originated in the 1970s with the development of microelectronics and new communication technologies that had widespread impacts on human communication, work and life. Before the information age, resources (e.g. capital, raw materials and human labor) were considered the main drivers of growth and development, and the creation and application of knowledge were seen as secondary in this process. However, as the information age flipped the paradigm, the demand for knowledgeable and creative individuals grew (Smith, 2001). Consequently, the need for managing all aspects of knowledge gained attention in the literature and in practice, giving rise to the field of knowledge management. Because intangible assets are the core of value creation in a knowledge-based economy (Janosevic and Dzenopoljac, 2013; Janosevic et al., 2012), capitalizing on related knowledge-based assets is crucial, made possible mainly through sharing practical knowledge within an organization and should be encouraged and nurtured effectively (Hau et al., 2013). Furthermore, some scholars argue that one of the most significant purposes of knowledge management is to encourage people to share knowledge with others (Dzenopoljac et al., 2018). In line with this, many organizations are continuously investing significant resources into knowledge management initiatives, coupled with various knowledge management systems that collect, store and distribute knowledge (Wang and Noe, 2010). According to seminal work by Nonaka (1991), the interplay between tacit and explicit knowledge creates value by transforming tacit to explicit, and explicit to tacit knowledge (see Figure 2). In ambidextrous organizations, these conversions gain higher importance due to the proven impact of tacit and explicit knowledge sharing on innovativeness (Tamer Cavusgil et al., 2003). Accordingly, we introduce the following general hypothesis:

One of the main ways knowledge sharing occurs is via the exchange of learning and resources through an organization’s network ties. These network interactions and exchanges play a significant role in information facilitation and transfer of knowledge, enabling knowledge sharing at the organizational (e.g. Law and Ngai, 2008) and individual levels (e.g. Kim and Yun, 2015; Muhammed et al., 2009). Additionally, there is evidence that knowledge sharing positively impacts organizations’ operational and financial performance (e.g. Wang et al., 2014), innovation performance (e.g. Sáenz et al., 2009; Yeşil et al., 2013) and individual employees’ work achievements (e.g. Du et al., 2007).

In terms of companies’ innovation performance, a study of firms in the IT industry in Poland and the USA revealed that an essential factor in the process of knowledge sharing is depicting to what extent “learning by doing” and “learning by interaction” affect the innovation of organizations’ processes, products or services. The results showed that learning by doing is more suitable for companies in the USA, while Polish companies tend to learn more through interactions (Kucharska and Erickson, 2023). However, previous studies focused primarily on the level of knowledge acquired or accumulated, not on the shared knowledge’s quality. In practice, it is vital to understand whether specific knowledge actually adds value to an organization or has a significant positive impact. In line with this, knowledge quality is perceived as the “acquisition of useful and innovative knowledge” (Soo et al., 2004, p. 3). When considering the quality of shared knowledge, high-quality knowledge is usually tacit, complex and highly asset-specific (Argote and Ingram, 2000; Han et al., 2018; Kogut and Zander, 1992). In this context, Kogut and Zander (1992) introduced the concept of combinative capabilities to illustrate the dynamism between knowledge transfer, knowledge creation and an organization’s learning and existing capabilities, which are updated with acquired and/or created knowledge.

Hujala and Laihonen (2021) reviewed the literature on the effect of knowledge management in the health and social care sectors and found that knowledge quality was an important factor in determining various capabilities of health-care-related organizations in several studies. Ganguly et al. (2020) found that knowledge quality had a significant effect on organizations’ innovation capabilities and argued that when implementing knowledge management frameworks, managers should explicitly consider how such systems can ensure that high-quality knowledge is available throughout the organization. Similarly, Ikonen (2020) indicated that knowledge quality was an important element of successful knowledge management in health care reform. Furthermore, studies have indicated that organizational teams’ knowledge quality plays a significant role in innovation and that higher knowledge quality positively affects organizational financial and innovation performance as a whole. In line with this, we posit that knowledge sharing is more significant for organizations when the quality of knowledge shared is higher (see Figure 2) and propose the following:

Knowledge sharing is typically defined as a process where people within or outside an organization mutually exchange their tacit and explicit knowledge to create new knowledge (Kamaşak and Bulutlar, 2010; Muhammed et al., 2009; Van Den Hooff and De Ridder, 2004). In line with this, knowledge, a crucial resource of any organization, is an enabling factor for improved results, including innovation performance. However, knowledge is not only held by individuals in an organization but also created by different interactions between employees, meaning that it is stored at both the individual and organizational levels (Kogut and Zander, 1992) and relies, in part, on interconnectedness and dependability. In the social context of organizations, knowledge is stored within the organization’s procedures, norms, rules and forms. This knowledge is built over time through interactions among individuals who contribute to the stock of knowledge via different socializing methods (March, 1991).

Innovation is one of the two ways through which organizational ambidexterity can be operationalized, the other being organizational learning (Ali et al., 2022), which is in line with the conclusion presented by Saleh et al. (2023) who also identified organizational learning as one of the important determinants for managing organizational ambidexterity. This implies that innovation represents one of the crucial outcomes of organizational ambidexterity. According to West and Farr (1990), innovation means an organization’s intentional decision to introduce and apply new products, processes, procedures, or ideas with a specific purpose to significantly improve the well-being of individuals, groups, organizations and society in general. The literature is abundant with evidence supporting the positive impact of knowledge management on innovation (Carneiro, 2000; du Plessis, 2007; Smith et al., 2005).

The interplay between various knowledge assets in organizations can successfully trigger the pursuit of different forms of ambidexterity, such as synchronous seeking after exploration and exploitation by employees or even asynchronous pursuit of ambidexterity within different time frames (Ali et al., 2022). At the individual level, the pursuit for exploration and exploitation is seen as the learning ambidexterity of individuals in an organization (Lin and Cheung, 2023). Organizational ambidexterity, on the other hand, is viewed as organization’s ability to successfully balance exploiting existing products through incremental innovation and exploring new opportunities to implement more radical innovation (Andriopoulos and Lewis, 2009). Indeed, their long-term success is closely related to their ability to efficiently and effectively conduct both activities, namely, exploiting current capabilities while exploring potential new competencies (March, 1991; Raisch et al., 2009).

Knowledge sharing should be viewed in various contexts to assess its relationship with innovation and ambidextrous behavior. For example, in the context of mergers and acquisitions, managers are continuously exposed to tacit knowledge sharing during negotiations, as well as devising and monitoring these transactions (Dzenopoljac et al., 2022). Han et al. (2018) found that high-quality overlapped knowledge positively affected the subsequent innovation performance. In contrast, the effect was negative for nonoverlapped knowledge, even that of high quality. Their results suggest that a certain amount of overlap facilitates the efficient exchange of complex knowledge, which is often tacit in nature, contributing to explorative capabilities such as innovation. In contrast, knowledge that has little or no overlap, even if it is of high quality, may not be easily integrated into the organization and may fail to advance innovation or even be detrimental to explorative capabilities. Explicit knowledge contained in databases, documents, and systems, can be considered nonoverlapping. Although it may not be effective in building an organization’s explorative capabilities, it can help improve organizations’ exploitative capabilities (see Figure 2). Tacit knowledge has also been associated with high-quality knowledge due to the fact that such knowledge is often complex and highly asset-specific (Argote and Ingram, 2000; Kogut and Zander, 1992).

In line with this, we propose the following:

Gibson and Birkinshaw (2004) explained organizational ambidexterity as an activity where an organization is engaged in both explorative and exploitative processes, which are considered vital for its long-term survival and success. Numerous studies have sought to ascertain whether organizational ambidexterity affects performance. Although exploration and exploitation represent two opposing frameworks of organizational learning, research has demonstrated that achieving a proper balance between these two approaches leads to increased performance (He and Wong, 2004). However, this conclusion is not universally unanimous in the literature. In their article, Tushman and O'Reilly (1996) advocated that companies must be ambidextrous (i.e. being successful in both incremental and revolutionary innovations) if they want to sustain success. Continuing on this positive note, several research studies have revealed a positive relationship between ambidexterity and performance. For example, Gibson and Birkinshaw (2004) assessed 4,195 managers and nonmanagers within 41 different business units during a three-year period and concluded that business units that exhibited ambidextrous behavior performed better than those that did not. Additionally, according to Chang et al. (2022), ambidexterity plays a mediating role between top management’s diversity and organizational units’ performance. Exploration and exploitation activities are also crucial prerequisites for performance of family firms (Hughes et al., 2018).

He and Wong (2004) examined ambidexterity in the context of technological innovation and inferred that companies that attain a balance between exploitative and explorative innovation strategies tend to achieve higher sales growth rates. Lubatkin et al. (2006) investigated the same relationship within small- and medium-sized enterprises (SMEs) and found that ambidexterity positively affects performance. Finally, Junni et al. (2013) provide a meta-analysis in the field of impact of organizational ambidexterity on performance, with several interesting conclusions. Specifically, they note that certain performance measures (e.g. growth) were more related to organizational ambidexterity, while others (e.g. profit) were not. Additionally, the combined measures of ambidexterity were more successful in capturing performance than balanced measures that sought the right ratio between exploitation and exploration (see Figure 2). Finally, the authors asserted that the relationship between ambidexterity and performance was industry-dependent and more emphasized in high technology and service companies than in the manufacturing sector.

Contrary to the above-mentioned research studies, there are empirical efforts that showcase counterintuitive conclusions. For example, Atuahene-Gima (2005) showed that the relationship between exploitation and exploration is inversely proportioned in the sense that when organizations are successful in exploiting their current competencies, they will have success with radical innovation only with a low level of exploration and vice versa. After investigating alliance formations, Lin et al. (2007) reached a conclusion that challenges the ambidexterity hypothesis, namely that the success of companies that pursue both exploitation and exploration when forming alliances is contingent on their own organizational characteristics and external conditions and is not always beneficial. They claim that companies that only pursue ambidexterity in alliances without considering other factors (i.e. firm size and level of environment uncertainty) may experience a decrease in performance. As the literature generally favors the positive impact of organizational ambidexterity on performance, we propose the following:

We used a cross-sectional research survey design to collect the data from a wide range of public and private organizations within the UAE. To obtain a comprehensive and unbiased perspective on the level of knowledge sharing within organizations, we targeted individuals from multiple levels of each organization. A Web-based questionnaire in the English language covering the basic demographic data and items related to the constructs of this study was distributed to 700 individuals working in various organizations in the UAE. The contacts were based on an initial convenience sample identified by the researchers and expanded based on snowballing method until a sufficient number of responses was obtained (Baxter et al., 2015). This approach was used due to two main reasons. A convenience sampling approach combined with snowballing facilitates rapid collection of data (Parker et al., 2019). Furthermore, this approach is more practical and cost-effective in the absence of reliable mailing lists from which a true random sample can be drawn. While there are limitations related to generalizability when using this approach several precautions were taken to minimize bias and to obtain a representative sample of the UAE industry as suggested by Zickar and Keith (2023). Data were collected over a period of two weeks during the month of July 2022 with two follow-up reminders, which ultimately yielded 363 usable responses (52%). The data collection was stopped once the responses received were deemed sufficient for model testing using structural equation modeling (SEM). A minimum sample size of 200 observations or 5–10 observations for each estimated parameter is recommended for SEM analysis (Bentler and Chou, 1987; Wolf et al, 2013). An evaluation of the differences between the means of the constructs from early and late responders did not indicate any significant differences, suggesting that nonresponse bias was not an issue (Armstrong and Overton, 1977). The organizations covered by the survey included both private (64.1%) and public (33.4%) sectors (excluding 11 missing data) mostly representing service-based organizations. The sample was heavily biased toward the service sector since only 7% of the organizations were related to manufacturing, construction, and oil and gas production. Within the service sector, a wide range of organizations were covered, including government, education, health care and banking/financial institutions; real estate and aviation and transportation agencies; retailers; and others. This is representative of the general make-up of the industries in the UAE, where 94% of all companies operating in UAE are SMEs, with 89% in the service, wholesale, and retail sectors (The UAE Government Portal, 2023). Evaluation of the differences between means of the key constructs between private and public sector groups was not statistically significant, and hence it was not necessary to differentiate them in the subsequent analysis.

As indicated earlier, we targeted individuals from multiple levels of each organization to obtain a more comprehensive picture of knowledge sharing within organizations. Previous studies have shown that both peer knowledge sharing that occurs within the hierarchical levels of organizations (Dzenopoljac et al., 2018; Muhammed and Zaim, 2020) and vertical knowledge sharing (De Long and Fahey, 2000; Ipe, 2003) are crucial for the successful knowledge use within organizations. While 31% of the participants in this study held nonmanagerial positions, the rest had some level of managerial responsibility. Among those who did, 16% were top management or held C-level positions. A total of 65% of respondents were female, and 80% had an undergraduate degree or higher. Approximately 44% were below 30 years of age, and nearly 5% were 50 years or older. Most participants (87%) had more than a year of experience, with 32% having more than 10 years of experience in their job. The majority (88%) had also been with their current organization for more than a year, with most having a tenure between one and five years (39%).

To test the proposed hypotheses, whenever possible, we used existing measures of knowledge sharing, knowledge quality, ambidexterity and organizational performance and adapted them for the current context when necessary. For knowledge sharing, we used distinct measures that differentiate tacit and explicit knowledge sharing. Similarly, because ambidextrous organizations are involved in both exploitation and exploration, we distinctly measured these two capabilities. Table 1 presents the constructs, their definitions and the related research. All items were measured on a five-point Likert scale ranging from (1) strongly disagree to (5) strongly agree. A five-point Likert scale was used throughout the questionnaire since many of the existing measures used in this study were developed using this scaling factor. Earlier studies have shown that the reliability and validity of a five-point scale and a seven-point scale are comparable and can be rescaled to be used equivalently (Dawes, 2008).

For knowledge sharing, we adapted the tacit and explicit knowledge-sharing measures used by Wang et al. (2014). Tacit knowledge sharing was measured based on whether the employees in the organization share and seek knowledge derived from experiences and expertise. The items used for explicit knowledge sharing included whether employees in the organization shared externalized knowledge such as that found in reports and official documents. Knowledge quality was measured based on the scale originally developed by Chiu et al. (2006) and Wasko and Faraj (2000). It included six reflective items measuring various aspects of the knowledge quality, such as ease of understanding, reliability, accuracy, and completeness. A later implementation of this scale by Chang and Chuang (2011) used a shorter scale with four items. This study used this more parsimonious measure of knowledge quality.

We measured items related to ambidexterity based on the exploitative and explorative capability of Sanal et al. (2013), which are similar to the exploitative and explorative orientation and strategy presented by Clauss et al. (2021) and Sirén et al. (2012), respectively. The explorative capability items measured whether an organization was more involved in developing new ideas, technologies and methods than its competitors. Some items focused on improving organizations’ existing processes. We adapted organizational performance from measures used by Wang et al. (2014) related to organizations’ operational and financial performance. The measure included five items evaluating the various elements of organizational performance – profitability, sales growth, customer satisfaction, innovation and overall performance – and was similar to the measures used by Noruzy et al. (2013) and Shafique et al. (2022). Appendix contains a detailed list of items for each construct.

Common method bias is a potential issue in studies that use the same source to measure both independent and dependent variables. Because this study used a single informant to assess all variables used in this study, we conducted Harman’s one-factor test (Podsakoff et al., 2003) to determine whether common method bias would be an issue before further analysis. A principal component factor analysis yielded four factors with eigenvalues greater than one. The first factor accounted for only 18.3% of the variance. A confirmatory factor analysis (CFA) with all six constructs accounted for 73.5% of the variance. Because a single factor did not emerge and no one factor accounted for most of the variance, we deemed that common method bias was not a significant concern.

To further test the structural relationships and the associated hypotheses, we assessed the measurement model for convergent and divergent validity. We used SEM using linear structural relations (LISREL) to assess the measurement model by performing a CFA. Standardized factor loadings were significant at the 1% level and ranged from 0.64 to 0.92, which is greater than the recommended 0.60. The averages of variances (AVEs) for all constructs were above the recommended 0.50 value and ranged from 0.62 to 0.73. The composite reliabilities (CRs) were also well above the recommended value of 0.80, ranging from 0.86 to 0.93. Table 2 shows the convergent validity parameters for each construct, which support the convergent validity of the measures used in the study. Table 3 presents the correlations and the square root of the AVEs. Fornell and Larcker (1981) recommended that the correlations between the constructs be smaller than the square root of their AVEs (shown in the diagonal) as evidence of discriminate validity. The results show that all correlations were smaller than the corresponding square root of the AVEs, except for the correlations between tacit and explicit knowledge-sharing and the correlations between exploitative and explorative capabilities. The relatively high correlation between these constructs may be attributable to the fact that these are the subdimensions of knowledge-sharing and ambidexterity constructs, respectively. Furthermore, the absolute, incremental, and parsimonious fit measures of the measurement model (Table 4) indicate good model-data fit for us to test the structural model to assess the proposed hypotheses.

The SEM of the proposed model was tested in LISREL using the covariance matrix of the indicators. For brevity, Figure 3 shows the trimmed structural model with only the latent variables and their standardized path coefficients. The goodness of fit statistics showed a good overall model-data fit (Chi-square/df = 1.58, RMSEA = 0.040, GFI = 0.92, AGFI = 0.90, NFI = 0.98, CFI =0.99). All proposed paths were significant at p < 0.01 and provided support for the hypotheses, except for the relationships from explicit knowledge sharing to knowledge quality (β = 0.06, p > 0.10) and explorative capability (β = 0.10, p > 0.10).

The results indicate that explicit and tacit knowledge sharing impacted both exploitative and explorative capabilities through knowledge quality. As hypothesized, tacit knowledge sharing and explicit knowledge sharing were significantly correlated (H1), suggesting that organizations that share greater levels of tacit knowledge also share greater levels of explicit knowledge and vice versa. Higher levels of tacit knowledge sharing in organizations contributed to higher levels of their both explorative and exploitative capabilities (H4a and H4b). The direct impact of explicit knowledge sharing was significant only on the exploitative capability dimension of ambidexterity (H5b) and had no significant impact directly or through knowledge quality on explorative capability (H5a). Results suggest that tacit knowledge sharing in organizations has greater value in helping organizations become ambidextrous than explicit knowledge sharing. Although knowledge quality was hypothesized to meditate the impact of knowledge sharing on exploitative and explorative capability, knowledge quality only partially mediated the relationship between tacit knowledge sharing and both dimensions of ambidexterity. While the relationship between tacit knowledge sharing to knowledge quality (H2) was significant, the relationship from explicit knowledge sharing to knowledge quality was not significant (H3) and hence knowledge quality could not be considered to be mediating the relationship between explicit knowledge sharing and the two dimensions of ambidexterity. However, knowledge quality is viewed as an important construct in the research model as it had a significant direct impact on both the dimensions of ambidexterity (H6 and H7). The significant correlation between exploitative and explorative capability (H8) indicates that the organizations in this study that have a high (low) level of exploitative capability also tend to have a high (low) level of explorative capability. Enhancements in both exploitative capabilities and explorative capabilities significantly impact the organizations’ performance (H9a and H9b).

While the linkage between knowledge sharing and ambidexterity has been firmly established in the literature (Fu et al., 2018; Kamaşak and Bulutlar, 2010; Kurniawan et al., 2020), a key objective of this research was in uncovering the importance and role of knowledge quality in this relationship. In the theory section, we proposed that knowledge quality was an important mediator between knowledge sharing and ambidexterity. To uncover the nuances in this relationship we examined the tacit and explicit dimensions of knowledge sharing. The results indicate that tacit knowledge sharing has a positive impact on knowledge quality, but the relationship from explicit knowledge sharing to knowledge quality is not significant. This finding may highlight an important characteristic that may have been overlooked in the knowledge management literature. Due to the nature of tacit knowledge being an artifact situated in people’s minds, sharing tacit knowledge facilitates a better understanding of the organizational knowledge for others and hence improves the overall knowledge quality in the organization. An improvement in the quality of knowledge shared within the organizations subsequently impacts the organizations’ exploitative and explorative capability. This is an important observation that needs to be explored further and to be considered in the theory building related to knowledge sharing. In this sense, the social capital perspective should inform any theoretical conception since knowledge quality depends on the degree of social interaction in the workplace (Fullwood et al., 2019). Previous research particularly emphasized the significant contribution of relational and cognitive social capital in developing high-quality knowledge sharing (Ganguly et al., 2020).

Our findings align with the existing theories, which suggest that tacit knowledge has a greater impact on the organizations’ competitiveness (Ganguly et al., 2019; Nonaka, 1994; Stenmark, 2000). However, the nonsignificant relationship between explicit knowledge sharing and knowledge quality, indicates that explicit knowledge may be expected to inherently have a certain amount of quality by definition of being externalized knowledge (compared to information or data), and a greater level of explicit knowledge sharing (which often involves sharing documents, policies and databases) may not necessarily contribute to an improvement in the quality of knowledge present in the organization. This contradicts the widely held view that tacit knowledge needs to be converted to explicit knowledge for it to become beneficial to the organization (Herschel et al., 2001; Nonaka, 1994), and may point to the limitations in trying to make all tacit knowledge explicit. These findings have implications for researchers exploring methods and effectiveness of converting tacit knowledge to explicit. If sharing of explicit knowledge do not contribute to the knowledge quality in organizations, future research may explore other factors that may act as a catalyst that will enable externalized knowledge to play a more prominent role in improving the knowledge quality in organizations. Perhaps, such externalized knowledge needs to be internalized back again to tacit knowledge by other organizational participants for it to be more beneficial for the organization.

Although explicit knowledge sharing had a significant role in improving the exploitative capability of the organizations, tacit knowledge sharing seems to have a more prominent role in improving the ambidexterity of the organizations in this study directly and through improving the organizational knowledge quality. However, this needs to be examined within today’s climate of information overload where excessive and constant information sharing could lead to a negative consequence where the organizational actors do not get sufficient opportunity to process that information for organizational gains (Filippov and Iastrebova, 2010; Fourné et al., 2019; Lingo, 2023; Roetzel, 2019). Future studies should take this dimension into consideration.

Path coefficients from exploitative capabilities and explorative capabilities to organizational performance indicate that both dimensions contributed to organizational performance at approximately the same level in the organizations covered by this study. Prior studies have indicated that certain industries and sectors may choose to focus on a more exploitative strategy or an explorative strategy, while some contemporary organizations strive to be ambidextrous and focus on both (Andriopoulos and Lewis, 2009; Peng et al., 2021; Raisch et al., 2009). This aspect needs to be explored further to understand the situations where organizations prefer a certain strategy over the other and in which situations organizations prefer to be ambidextrous.

By definition, tacit knowledge is knowledge that is inimitable and often contains rich context that may not be economical or even feasible to be codified for the purpose of sharing it (often for organization-wide sharing of such knowledge). The development of expert systems is one application of codification of tacit knowledge from experts in certain fields. Evidently, the development of such expert systems is a laborious task and is usually confined to a narrow field of application (do Rosário et al., 2015; Kambhampati, 2021). However, these very qualities of tacit knowledge, which are usually embedded in human minds and organizational culture, might be what provides a superior advantage in becoming ambidextrous organizations when organizational participants share it. Managers considering moving their organizations to be ambidextrous organization should consider these facts and explore ways to get their employees to share their tacit knowledge and provide incentives and structure to facilitate such knowledge sharing.

In this study, explicit knowledge sharing had a significant impact on only the exploitative dimension of ambidexterity. Its impact on the explorative capability of the organization was not significant suggesting that explicit knowledge, which is often codified in organizational documents, policies, and standard operating procedures, while helpful in improving the exploitative capabilities of the organization, such knowledge does not necessarily contribute to explorative capability. When employees share information that is externalized within organizations, it can help improve the organizations’ effectiveness and efficiencies related to their current operations. However, if they are to be innovative in exploring new products, solutions and markets, sharing such externalized information is not sufficient. Organizations looking to build such explorative capabilities should encourage their employees to share their tacit knowledge that is often held in their minds comprising of their understanding of specific organizational situations and contexts.

We also find that greater levels of tacit knowledge sharing improve the knowledge quality, possibly, because the participants gain a better understanding and insight related to the knowledge being shared in that process. Furthermore, a high correlation between explicit and tacit knowledge sharing indicates that in organizations where high explicit knowledge sharing occurs, a high level of tacit knowledge sharing also occurs, and vice versa, pointing to a possible presence or absence of a knowledge-sharing culture. Although explicit knowledge sharing does not have a direct impact on the explorative capability of the organization, higher levels of explicit knowledge sharing could lead to building a knowledge-sharing culture where employees are more willing to share their tacit knowledge as well and may indirectly contribute to building explorative capability.

Managers should note that higher levels of tacit knowledge sharing within the organizations can improve the effectiveness of organizational operations related to the current strategies and priorities contributing to its exploitative capability. Results indicate that it can also provide organizations with key capabilities related to exploration, such as developing new products, processes, and technologies. Rather than expending considerable resources in making most of organization’s tacit knowledge explicit, organizations may consider using those resources for enabling the sharing of tacit knowledge in organizations while selectively focusing on codifying tacit knowledge. While encouraging knowledge sharing in organizations managers should also be aware of the negative effects of information overload that could potentially have an inverse effect on knowledge quality and organizational capability (Filippov and Iastrebova, 2010; Roetzel, 2019).

Policymakers should also be attentive to enhancing social capital elements that include social climate as well as internal and external relations. These dimensions support knowledge sharing by creating cooperative and equitable work conditions (de Frutos-Belizón et al., 2019). A parallel implication emphasizes the role of social capital in terms of intellectual capital creation, through the interplay of three components, namely, social interaction, trust and shared vision (Barrutia and Echebarria, 2022).

As with other similar studies in the field that use a cross-sectional survey research design, the results of this study should be interpreted within its limitations. First, because of the cross-sectional survey design, and the use of convenience sample used to reach the respondents, the generalizability of this study is limited. The study was targeted to reach organizations operating in UAE. Even though convenience sampling limits the generalizability, the distribution of the organizational profile presented in the results section indicates the distribution of firms in the sample closely represents the distribution of the overall sectors of firms in UAE. Furthermore, Dubai being a melting pot of organizations from East and West, there is no reason to believe that it may be different in organization from other parts of the world. However, this needs to be tested further in future studies. Second, the firms in the sample were mostly from the service sector due to service sector being the leading sector in UAE. This also limits the findings to firms mostly in the service industry.

Even though we collected the data regarding the organizations from private and public sectors, no significant differences were observed between them on key constructs, hence they were combined in our analysis. Future studies may include other variables commonly used as controls such as firm age and firm size because prior research has suggested that organizations’ demographics may have a significant association with firm performance (Jansen et al., 2012). Future studies may also focus on specific industries and explore if these relationships hold differently in different industries and within private and public sector organizations. Furthermore, even though tacit knowledge sharing and explicit knowledge sharing are highly correlated, future studies should explore organizational factors that facilitate sharing of specific type of knowledge.

In the present study, we examined to what extent knowledge sharing affects organizations’ ambidexterity and their performance while observing the role of knowledge quality. Knowledge sharing was assessed through the tacit and explicit dimensions, while ambidexterity was analyzed through exploitative and explorative capabilities. In the literature, it is evident that KBV is used in organizational ambidexterity research but there is an evident scarcity of studies that investigated whether a firm’s knowledge quality impacts the relationship between tacit/explicit knowledge sharing and exploitation/exploration processes. Data for this empirical research was collected from a cross-sectional sample of participants, working in a wide range of public and private organizations within the UAE. To obtain a comprehensive and unbiased perspective on the level of knowledge sharing within organizations, we targeted individuals from multiple levels of each organization. The presented results point to the conclusion that tacit knowledge sharing exhibits significant direct impact on organizational ambidexterity, namely, on both the exploitative and explorative capabilities, and it indirectly impacts both dimensions through knowledge quality. Conversely, explicit knowledge sharing does not have a significant impact on knowledge quality, and it affects only the exploitative dimension of ambidexterity. Both exploitative and explorative capabilities significantly impact organizational performance. In line with this, the research contributes to the ambidexterity field in multiple ways. The literature is scarce in terms of research studies that investigate the role of knowledge quality in the relationship between knowledge sharing and organizational ambidexterity. Apart from this, the present research provides a more nuanced insight, into how tacit and explicit knowledge sharing within organizations may distinctly impact exploitative and explorative capabilities. Finally, the current study offers a more international context and greater validity to ambidexterity research by empirically exploring its link to performance of organizations in the Middle East, more specifically in the fast-growing country of the UAE.

Table A1