Enabling Interpersonal Dynamic Capability: Four Emerging Collaborative Practices in Globally Distributed Software Development

ABSTRACT Dynamic capabilities (DCs) are noted drivers of organizations’ competitive advantage in fast-changing, digitalizing operating environments. However, research has reached no consensus on the essence of DCs, and the focus has shifted from managerial excellence to employee involvement and participation. The interpersonal view of DCs is suggested as crucial in enabling organizations’ change capacity. At the same time, globally distributed work is a common operating model today but often hindered by tensions and collaborative challenges. In this qualitative single-case study, we explore the collaborative practices enabling interpersonal DC in globally distributed software research and development (R&D). By semi-structured interviews with senior key informants from a northern European site of a case organization, inductive and abductive analyses, and the practice perspective, we identified four emerging practices nurturing organizational dynamism in global software work: dialogical organizational development, constructive working culture, global open engagement, and facilitated shared learning. The findings highlight the role of employees and create a new understanding of the influence of collaborative practices in generating DCs in a global R&D working environment. Finally, the paper proposes a dynamic model for evaluating the development of such practices and suggests a stronger adoption of the practice perspective in further study of DCs.


Introduction
Globally distributed work (GDW) is a widely utilized model of organizing business process services, software and information systems (IS) projects, research, development, and innovation activities, and other such operations today (Brooks et al., 2020;Kotlarsky & Oshri, 2005;Levina & Vaast, 2008).The adoption of GDW operating models has been influenced by the developments where digital technologies, such as improved connectivity and the decreasing cost of data transfer continue to dilute organizational and functional boundaries globally (Bharadwaj et al., 2013).However, GDW settings are often complex, and various organizational, operational, and cultural tensions hinder effective global collaboration (Brooks et al., 2020;Cramton & Hinds, 2014;Ravishankar, 2015).
At the same time, influenced by the trends of digitalization organizational capability for continuous change, agility and transformation (Tallon et al., 2019;Vial, 2019;Wessel et al., 2021) is ever more important.As a response to these requirements, dynamic capabilities (DCs) are, by their very definition, understood as enabling organizations to be nimble and capable of change, that is, dynamic.The core argument is that companies seek to capture and offer new value to remain competitive amid fast changing technologies and uncertain environments by the DCs of sensing and seizing opportunities and threats, and transforming, renewing their business models, resource bases, and capabilities accordingly (Day & Schoemaker, 2016;Tallon et al., 2019;Teece et al., 2016;Vial, 2019).DCs have thus been of interest to IS, management, and organization scholars for a long time (e.g., Queiros et al., 2018;Salvato & Vassolo, 2018;Teece, 2007).Questions such as what constitutes DCs (Eisenhardt & Martin, 2000), how they interact with learning in organizations (Zollo & Winter, 2002), how to apply them for organizational agility and performance (Chakrabarti & Mukherjee, 2022;Queiros et al., 2018), and how to best study them (Wenzel et al., 2021) have been explored.
However, many unanswered questions on the acclaimed dynamism of DCs remain (Salvato & Vassolo, 2018;Wenzel et al., 2021).First, the paradoxical nature of DCs baffles theorists.The paradox here refers to DCs simultaneously being both understood as enablers of change and being tied to organizational routines: Routines are often seen as relatively slow to change and "pathdependent," that is, dependent on the organization's past developments, which appears counter-intuitive to dynamism (Salvato & Vassolo, 2018;Wenzel et al., 2021).Second, the scope of DCs has raised questions.While DCs were originally framed as managerial-level strategic capacities (Teece, 2007), calls have been made to understand their creation, enactment and nurturance encompassing the role of employees (Salvato & Vassolo, 2018;Wenzel et al., 2021;Wohlgemuth et al., 2019).Recently, an answer has been sought from interpersonal relationships, participation, and routines as enablers of change in organizations (Salvato & Vassolo, 2018;Wenzel et al., 2021;Wohlgemuth et al., 2019).In particular, collaborative social phenomena, such as interpersonal participation, appear focal in the creation of DCs, which are even seen as "effortful social accomplishment[s]" (Salvato & Vassolo, 2018, 1728, 1734).Finally, to the best of our knowledge, the formation of DCs in the context of GDW in the information technology (IT) field has largely remained outside the focus of research.
Practice theory revolves around "dynamics, relations, and enactment" in studying organizational phenomena in today's complex environments (Feldman & Orlikowski, 2011, p. 1240).The practice perspective has been found fitting, for example, in the research of the dynamics of organizational routines, stability and change in organizations (Wenzel et al., 2021), and the complexities of GDW (Levina & Vaast, 2008).In this paper, we aim to advance knowledge in the problem area of globally distributed software research and development (R&D) work by utilizing the conceptual framing of DCs as organizational collaborative phenomena and drawing from practice theoretical concepts to guide our analysis (Mathiassen, 2017).Our research question is this: how do collaborative practices, as enacted by management and employees, enable interpersonal dynamic capability in the context of global software R&D work?
We explore the research question by a qualitative, interpretive (Klein & Myers, 1999) single-case case study (Yin, 2018).The case company is a large, global operator in the industry of complex information and communication technology (ICT) products in the business-to-business sector.The data were collected by eight semi-structured interviews with senior professionals as key informants from a northern European site of the company.The data were analyzed thematically through inductive (Urquhart, 2013) and abductive approaches (Kennedy & Thornberg, 2018).As findings, we present and discuss four emerging, collaborative practices suggested to enable DCs at the interpersonal level of global software R&D work.The practices are dialogical organizational development, constructive working culture, global open engagement, and facilitated shared learning.
The study contributes to the understanding of the complexities of GDW in the context of software R&D (cf.Brooks et al., 2020;Cramton & Hinds, 2014;Levina & Vaast, 2008) and the DC theory by investigating phenomena close to an organization's practical life, beyond managerial roles and on the interpersonal level (cf.Salvato & Vassolo, 2018;Wenzel et al., 2021;Wohlgemuth et al., 2019).To complement the recent theorizing in the field (e.g., Salvato & Vassolo, 2018;Wenzel et al., 2021), our study attempts to advance the empirical and practical understanding of the functioning of DCs through the practice lens (Feldman & Orlikowski, 2011).
This study is a part of a research project involving three case-companies and in total multiple cycles of data collection.The overall aim of the project is to create a new understanding of the relationships of dynamic and operational capability development in interactions between the employees and management in digitalizing operating environments.This paper reports one part, a sub-study, of the project in order to reach the desired depth of discussion in this individual case (Sarker, 2021) and timely reporting of research findings for discussion, utilization, and further development (Ågerfalk et al., 2020).
The remainder of the paper is organized as follows.Next, the theoretical framing of dynamic capabilities is presented.The section introduces the debated nature of DCs, discusses them as collaborative social accomplishments, and connects the related areas of organizational learning, culture, and innovativeness to DC development.After that, we describe the case of globally distributed software work and the rationale of the research design.The latter part of the section considers the plural contexts (McLaren & Durepos, 2021) of GDW in reflection to the case.After that, we move to the qualitative method, including the application of practice theory and data collection and analysis.The findings are then presented in a narrative form illustrated by interview examples and summarizing figures.In the discussion section, a dynamic model for assessing the practices as well as theoretical and practical contributions are presented.We close the paper by addressing the limitations, making recommendations for further research, and providing concluding remarks.The key concepts will be presented in their respective sections, and Table A1 in the Appendix provides a brief overview.

Theoretical framing by dynamic capabilities
In this section, we describe our understanding of the current knowledge on DCs.We more closely examine the "inter-personal (meso) level" DCs as the suggested driver of dynamism in organizations (Salvato & Vassolo, 2018), and integrate understanding from the areas of organizational culture, learning, innovativeness, and dialogic leadership.Overall, as a theoretical premise, we follow the emergent perspective where change emerges from "complex social interactions" and "the interaction of people and events" (Markus & Robey, 1988, p. 583, 588).The theoretical aspirations resemble the logical structure of process theory with a mixed level of analysis (Markus & Robey, 1988).

Dynamic capabilities as collaborative social accomplishments
DCs are a much studied and debated area by many IS, organizational, and management researchers (e.g., Eisenhardt & Martin, 2000;Peteraf et al., 2013;Queiros et al., 2018;Salvato & Vassolo, 2018;Wenzel et al., 2021).At the core of the DC theory is the view that a company capable of employing its DCs of sensing, seizing, and transforming in the right proportions is equipped to remain competitive in fast-changing, turbulent environments (Teece et al., 2016) that require the constant aligning and realigning of resources, capabilities, and value propositions (Marabelli & Galliers, 2017;Vial, 2019).
However, recent literature indicates that the field has yet to reach consensus on the essence of DCs (e.g., Wenzel et al., 2021).In particular, we find two points compelling.First, exploring the meso-level organizational influencers in DCs seems to offer avenues for creating new knowledge (Salvato & Vassolo, 2018).Second, the employees' role and a fruitful organizational climate in creating, pursuing, and nurturing DCs appear to require more attention in research  (Ghosh & Srivastava, 2022;Wenzel et al., 2021;Wohlgemuth et al., 2019).In these efforts, Salvato and Vassolo (2018) dissected the interpersonal, meso level of DCs in their multi-level framework, which we conceptually utilize in this paper, and which is visited next.
The meso-level DCs, dynamic interpersonal capabilities, revolve around the mechanisms of interpersonal participation and appear to be influenced by the organization's "quality of relationships and dialogue" (Salvato & Vassolo, 2018, pp. 1733-1734).If dialogue is "productive" it is said to result in solidarity and constructive opposition, and to yield intense participation and interpersonal relationships enabling change (Salvato & Vassolo, 2018, 1734, 1742).On the contrary, the authors argue, "unproductive" dialogue may lead to conformism and noninvolvement, hindering change.
In the framework, interpersonal relationships are a particularly important component.Employees' collective relationships and interaction characterized by "relational engagement and productive dialogue" are even identified as the locus of DC emergence (Salvato & Vassolo, 2018, 1736, 1737).Candor, inclusion, confirmation, and presentness in interaction are characters of relational engagement, which entails change-oriented "cooperative behavior" despite "conflicting viewpoints and opinions" (Berkovich, 2014as cited in Salvato & Vassolo, 2018, 1738).Thus, collaboration that builds on genuinely being interested in the other nurtures productive dialogue (Salvato & Vassolo, 2018).Dialogue is termed the "glue" connecting the different levels of individuals, teams, and organizations, enabling "shared consensus and commitment" even in demanding change initiatives (Salvato & Vassolo, 2018, 1739).Productive dialogue supports creating and transferring knowledge, improves cohesion among employees, and transforms learning from an individual effort into participation in mutual interactions (Salvato & Vassolo, 2018).
Cohesion helps increase the likelihood of employees engaging in change proposals (Salvato & Vassolo, 2018).It is noted as being critical in environments requiring a high level of work synchronization and recycling (Yli-Renko et al., 2001as cited in Salvato & Vassolo, 2018), such as in global software R&D.It bears a resemblance to the capability of "collective knowing" in distributed and complex work environments (Orlikowski, 2002).
Participation, then, is viewed as the "ultimate meso-level outcome" entailing the management and employees acting together toward organizational goals (Salvato & Vassolo, 2018, 1742) and warranting a research focus in the context of DCs (Wenzel et al., 2021; see also Wohlgemuth et al., 2019).Employees are recognized as important actors in the development of DCs (Salvato & Vassolo, 2018;Wenzel et al., 2021) and discovered positively to impact the DCs of an organization (Wohlgemuth et al., 2019).
Recent studies also indicate the importance of other collaborative aspects in the development of DCs.Wohlgemuth et al. (2019) addressed the framework (Salvato & Vassolo, 2018) by looking at the impact of managerial trust and informal control.Trust and informal control were discovered to facilitate employee participation, which in turn has a positive effect on the enactment of DCs (Wohlgemuth et al., 2019).Trust is noted as "trust by managers" (Wohlgemuth et al., 2019, p. 760), while trust and rapport among collaborators are found generally important for successful collaboration and knowledge sharing in distributed environments (Kotlarsky & Oshri, 2005; see also Orlikowski, 2002).Finally, the inclusion of employees within the DC framework may result in better informed decisions, faster adjustment, and a more efficient transformation process (Wohlgemuth et al., 2019).As Wohlgemuth et al. (2019) summarize, "involving employees help firms unleash the strategic value of dynamic capabilities" (p.768).

The roles of organizational learning, culture and innovativeness
First, organizational learning is understood as central to DCs (Ghosh & Srivastava, 2022;Salvato & Vassolo, 2018;Zollo & Winter, 2002).Organizational learning can be attributed to the mechanisms of "experiential" and "deliberate" learning, contributing to both the everyday organizational activities (operating routines) and the modification of those routines (DCs) (Zollo & Winter, 2002, p. 340).Learning to execute and modify the routines can also be connected to the "exploitation" and "exploration" sides of organizational learning as utilizing existing competences, and as creating new knowledge (Levinthal & March, 1993).
Experiential learning appears to have similarities with relational engagement and productive dialogue (Salvato & Vassolo, 2018).Collective reflective processes together with social, cognitive, and operational ones lie at the core of experiential learning in a work organization (Järvinen & Poikela, 2001).They "follow, influence and shape each other in a process of continuous learning" (Järvinen & Poikela, 2001, p. 286).In contrast to understanding individuals, groups, and organizations as different levels, they can be seen as "flows and processes" binding the different actors together into an "organizational entity" (Järvinen & Poikela, 2001, p. 286).
Next, organizational culture influences both organizations' learning and their ability to innovate, and it has been found to have a crucial role in building capability dynamism (Ghosh & Srivastava, 2022).This discovery has resulted in recommendations to invest in strengthening openness, participation, results orientation, trust, and "constructive dissent" in organizations (Ghosh & Srivastava, 2022, p. 967).It is further proposed that openness and participation contribute to "behavioral innovativeness," which could be supported by flat organizational designs, the encouragement of social interaction and information sharing by communication, collaboration, and coordination systems, as well as participation, inclusion, and diversity starting from the policy level.Overall, organizational culture can influence whether it is considered pro-or "countercultural" to share knowledge (Feldman & Orlikowski, 2011, p. 1247).
Further, productive dialogue (Salvato & Vassolo, 2018) appears philosophically rather close to the dialogic leadership paradigm.Dialogue supports active participation, commitment, and learning with the discovery and acceptance of new and different ideas, views, and opinions (Syvänen & Loppela, 2013;Ahonen & Pohjanheimo, 2000 as cited in Syvänen & Loppela, 2013).For example, in development dialogs, principles such as the democratic approach, reflective thinking as well as the aspects of learning, listening, voicing, and respecting have been found as requirements for successful organizational development processes (Syvänen & Loppela, 2013).Dialogic leadership may support the competitiveness of organizations, as it fosters the capacity for renewal, innovativeness, and skill and capability development (Syvänen & Tikkamäki, 2013).It has also been found that dialogic leadership can facilitate participation, knowing (cf.Orlikowski, 2002), interaction, and reflection all of which enable the potential for learning (Syvänen & Tikkamäki, 2013).
Moreover, admitting and learning from failure is promoted as fruitful for innovation (Danneels & Vestal, 2020).However, bringing up errors for discussion may be a delicate matter, and for it to be successful, an organizational climate allowing "constructive conflict" and frank discussion is required (Danneels & Vestal, 2020).Especially when the aim is to learn from mistakes for improved innovativeness, deliberate and collective reflection is important (Danneels & Vestal, 2020;cf. also;Järvinen & Poikela, 2001).
An important distinction has been made between tolerating and analyzing failure (Danneels & Vestal, 2020).The distinction can be illustrated by the difference between accepting failure "as an inevitable byproduct of taking a lot of initiatives" and "openly analyz[ing] past mistakes" (Danneels & Vestal, 2020, p. 16).Only tolerating failure appears not to benefit innovation.Instead, past failure should be explicitly analyzed, and the accurate and complete lessons learnt extracted, which requires a suitable organizational culture and managerial attention among other resources.These discoveries by Danneels and Vestal (2020) have a common note with the constructive dissent by Ghosh and Srivastava (2022), and constructive opposition by Salvato and Vassolo (2018).Based on the literature discussed here, it can be concluded that an organizational culture of openness and diversity of opinion benefits organizational dynamism and the capability for innovation.

The case of globally distributed software work and its contexts in the study
This section describes the setting of GDW, and the case explored by this study.While presenting the case, the rationale behind the research design choices is discussed.In the latter part of the section, the multiple contexts influencing the study are considered with reflection to the present case.
We understand GDW as a setting where teams are located in different countries and work toward a common goal (Bosch-Sijtsema et al., 2009) often relating to IT, business processes, or software R&D work.In this study context, the goal can, for example, be a piece software or code integrated into a larger project entity constructed by the input of multiple teams.When reflecting the present case against general global software development models, we highlight the categories of offshore outsourcing and internal offshoring (Prikladnicki & Audy, 2012).Prikladnicki and Audy (2012, p. 216) define that in the first, a company partners by contractual means with a service provider or a subcontractor located in another country.The second model utilizes wholly owned subsidiaries or company units likewise located abroad.In this paper, the research setting resembles the context of internal offshoring, where work is organized in a globally networked manner.The type and reach of team collaborations, team compositions, and work dispersion may change across time and across different projects.
Traditionally, the aim of distributing software work has related to reaching for global opportunities, such as skilled workforce, flexibility, competitive advantage, and cost reductions (Prikladnicki & Audy, 2012).Similarly, drivers related to financial motivators, competences and capabilities, quality improvements, and business transformation efforts have been identified in connection to IT outsourcing (ITO) (Obwegeser et al., 2020).Several variations in sourcing strategies exist (Oshri et al., 2015), and different forms of globally distributed and virtual work are common as enabled by technological development (Toiviainen et al., 2022).While the more traditional objectives of cost and efficiency gains are still recognized (Brooks et al., 2020), today the goals may involve areas such as strategic innovation (Oshri et al., 2015) and the distributed tasks include complex and high-end activities (Sayed & Agndal, 2022).Innovativeness and the innovation capacity of distributed teams can be seen as strong factors also in this case.
While capability-oriented research has been conducted in IT outsourcing (ITO) settings (e.g., Karimi-Alaghehband & Rivard, 2020;Koo et al., 2019), in the context of globally distributed teams it appears scarce.We also note that in outsourcing relationships, different regularities apply, as organizational boundaries are often more pronounced, collaborations are defined by contractual means (Obwegeser et al., 2020;Oshri et al., 2015), and organizational goals are based on each company's strategy.Thus, we understand ITO as a distinct field from GDW research, where we explore the topic from the perspective of global teams operating within a common organization under common goals.
Collaboration is noted in many ways as essential to the success of GDW, and similarly, many of the challenges seem to revolve around it.Recent work on globally distributed teams explored the "intrinsic and entrenched nature of tensions" stemming from the perceived gaps in knowledge, power, and identity among collaborating participants (Brooks et al., 2020).In addressing these tensions, both informal and formal organizational solutions supporting collaboration have been found necessary (Brooks et al., 2020).Clashes created by the differing expectations and conditions surrounding global collaborators are noted to create complex dynamics, which require constant and manifold adaption by teams (Cramton & Hinds, 2014).
Cultural assumptions and held beliefs of the other were found both helpful in navigating difficult situations and potentially hindering in the development of collaborative relationships (Ravishankar, 2015).The research body of knowledge notes factors -such as effective knowledge sharing and the quality of social ties (Kotlarsky & Oshri, 2005) and the role of management in spanning the various boundaries (Levina & Vaast, 2008) -that are important in enabling global collaboration.In a recent exploration of GDW tensions, the previous findings (cf.e.g., Brooks et al., 2020) were complemented by the work context-related nature of tensions among collaborating teams, and the utilization of unifying and tension attenuating discourses in promoting learningoriented everyday collaboration (Vartiainen, 2021).These findings provide an understanding to ground the present study and the new insights to which it contributes.

Case description and rationale
The case company is a large global organization operating in the field of complex business-to-business ICT products.The company is headquartered in Northern Europe, and it operates in more than 100 countries employing tens of thousands of employees.It has a revenue of over 20 billion EUR.The company exercises advanced, broadly utilized software R&D processes and has a high level of expertise in their portfolio of products in different life cycle stages.The company is also experienced at operating in GDW models.
The key informants of the study as the selected interviewees (cf.Kumar et al., 1993) include eight senior professionals from different work roles in software R&D and related functions in a northern European site of the company.The main selection criterion was that the informants had expertise in the different aspects of global software R&D work, such as software development, innovation, competence management, team management, testing, and quality control in a globally distributed working environment.In other words, their roles were "closely associated with the phenomena under study" (Kumar et al., 1993(Kumar et al., , 1635)).The selection of the key informants was coordinated together with the case organization representatives, who were knowledgeable about each participant's competence area and their openness to communicate about the research topic (Kumar et al., 1993).The aim of the multiple key informants was to obtain rich qualitative insight of the phenomena under investigation (Walsham, 1995, 79-80) while minimizing singleinformant bias (Kumar et al., 1993).The number of participants was aligned with the broader multi-case research project with the aim of a balanced number of interviews in the different cases (Sarker, 2021).
Due to the limited availability of participant locations within the timeframe of the research and the travel restrictions in place due to COVID-19 pandemic, it was decided that this study would focus on one location only.While the scope of the study is limited in this regard, we follow the views of Ågerfalk et al. (2020) in our aim for the timely reporting of research findings and their prompt opening for scientific discourse.Additionally, the present scope makes way for the sought-for "thick description[s]" of the interpretive research tradition (Walsham, 1995, p. 75) and "actionable insights" (Grover & Niederman, 2021, 1774), which are expected to provide value "in the future in other organizations and contexts" (Walsham, 1995, p. 79).

Consideration of the multiple GDW contexts in reflection to the case
In our attempt to address the research context, we identify multiple relevant contexts within the complex environment of GDW (Davison & Martinsons, 2016;McLaren & Durepos, 2021).While "to a point, [we are] using context as a container," we aim to explicitly acknowledge the plurality and fluidity of different contexts rather than relying on assumptions (McLaren & Durepos, 2021, pp. 81-82).Essentially, GDW environments are known to be influenced by and create "multiple and overlapping boundaries" (Levina & Vaast, 2008, p. 307) fueled by factors, such as cross-cultural differences (Cramton & Hinds, 2014) and various other tensions (Brooks et al., 2020).These boundaries and tensions may also be reflected in our findings, for example, when later discussing openness and engagement in global communication.Similarly, in connection to organizational development initiatives, the perspective of the informant is likely to influence how the initiatives are perceived.In these regards, the principle of multiple informants is essential.
Moreover, when considering the present case of study (cf.Davison & Martinsons, 2016), it should be emphasized that the findings concern senior professionals' perspectives in northern Europe.Findings from the perspective of junior engineers in east Asia, for example, may portray a differing set of practices and activities (cf.Hosack, 2021;Palvia et al., 2020).Therefore, the present study opens avenues for the further exploration of such contexts.Moreover, the networked structure of the case organization probably shows, for example, in discussing the development efforts across the different levels of the organization.We can also identify the context of a global organization with advanced competence in its field and a relatively flat corporate culture.These notes are likely to show in the findings relating to working culture, which would likely appear differently in a small start-up or in a more traditional industry, such as manufacturing.
In acknowledgment of the critique raised of often failing to consider and differentiate between the different contexts of a research environment (Davison & Martinsons, 2016;McLaren & Durepos, 2021), we note the handling of the contexts here is inevitably limited.For example, we choose to omit the societal developments from the early days of GDW (cf.e.g., Kotlarsky & Oshri, 2005;Orlikowski, 2002) from the analysis.Thus, we remain within the boundaries of the organization as it stands today as part of its industry, which could be described as rapidly changing (Day & Schoemaker, 2016), and exclude its historical development from the analysis.However, we note that by this case of GDW, we study a context of an experienced organization both in the area of software R&D and operating in a globally distributed manner.It is also an environment where the organizational members are used to changing work settings.These final notes are expected to be reflected in the findings related to organizational learning, working culture, and the style of global communication.These considerations aim to illustrate the importance of acknowledging the context of study in its reporting (Davison & Martinsons, 2016).

The qualitative method
The study was conducted as a qualitative, interpretive (Klein & Myers, 1999;Walsham, 1995) single-case study (Yin, 2018).This section first describes the application of the practice perspective as an analytical tool within the study, and then accounts for the qualitative data collection and analysis method in detail.

Application of the practice perspective
Studying "messy" areas of organizational life, such as compositions of competences and capabilities requires "get[ing] in deep into understanding the social processes of organization and the human and cognitive dimension" (Peppard et al., 2014, p. 5).In seeking an understanding of how DCs work and are accomplished in practice, it is recommended that research should by qualitative methods aim for "sociologically-informed, practice-based" knowledge (Wenzel et al., 2021, p. 400).Attention should be directed to actors and change potential "beyond managerial intent" (Wenzel et al., 2021, p. 400).
The practice perspective has been successfully applied to both theorizing around DCs (Wenzel et al., 2021) and empirically in the context of GDW (Levina & Vaast, 2008) as well as in IS strategizing and strategy-as-practice research (Marabelli & Galliers, 2017;Peppard et al., 2014).Therefore, the lens offered by the practice theory appears fit for exploring the software R&D case.Practice itself can be defined in many ways, and here we adopt a broad definition of practices being "configurations of actions which carry a specific meaning" (Nicolini, 2012, p. 10).
It appears that "practicing" (Feldman & Orlikowski, 2011) the practice perspective can be understood differently influencing its application.First, we understand the practice perspective to imply that practices and the practitioner view are the most relevant sources of knowledge (Peppard et al., 2014) in organizational research.Further, it seems that actions within practices actually form the reality (Feldman & Orlikowski, 2011, p. 1241).In other words, all would be tied to actors and their activities.In a sense, this view appears to go beyond the ontological notion that world and phenomena are socially constructed (cf.McLaren & Durepos, 2021) and to convey that reality is constructed by actions in social relations.
The second understanding involves the "strong programme" of practice theory with a methodological argument that taking the practice approach requires studying practice "as it happens" (Nicolini, 2012, p. 14; see also Feldman & Orlikowski, 2011, p. 1249).While Peppard et al. (2014) acknowledge the suitability of methods such as grounded theory and action research over quantitative ones, Nicolini (2012) calls for observational methods in practice research.Optimally, research should be conducted deep in organizations and with a longitudinal approach (Peppard et al., 2014).Even though our method is somewhat limited in these regards, we believe the rich qualitative data (cf.Salvato & Vassolo, 2018) warrants exploration into the "complexity," "ambiguity," and "the everyday realities of organizational life" (Feldman & Orlikowski, 2011, p. 1249) in studying DCs, also ridden with complexities and ambiguities.
Finally, according to Feldman and Orlikowski (2011), practice theory encompasses three approaches forming a triad of foci in practice research (pp. 1240-1241).The investigator may choose the emphasis in which one or more of them are used.First, the empirical approach focuses on everyday activity in organizations answering "the what" question.People's actions and practices are seen as central to organizational outcomes.Second, the theoretical approach explicitly applies the practice theory in the aim to explain "the dynamics of everyday activity."Thus, this side addresses "the how" question focusing on the generation and operation of activities.Finally, the philosophical approach concerns "the why" and grounds on the premise that everyday activity creates -brings "into being" -the social world.
Drawing from these understandings, Figure 1 describes the analytical lens formed based on the DC and practice perspectives.In essence, the DC theory is here utilized to guide what phenomena to seek, and the practice perspective is used to identify the correct "units of analysis" (Nicolini, 2012, p. 9) within the case.With these lenses, we aim to contribute to both theoretical generalization and practical relevance (Feldman & Orlikowski, 2011;Peppard et al., 2014;Walsham, 1995) in the problem area (Mathiassen, 2017) of globally distributed software R&D work.

Data collection and analysis
The data were collected in August 2020 by eight individual, qualitative, semistructured interviews (Walsham, 1995) with the key informants.Due to the COVID-19 pandemic restrictions in 2020, the interviews were conducted remotely via Microsoft Teams.Before conducting the interviews, several interactions of planning took place with the organization.
The recording of the interviews was done by researcher notes and as audio with the participants' permission (Walsham, 1995).Audio recording was used in seven out of eight interviews.The interview duration was from approximately 72 minutes to 101 minutes, which excludes the introduction of the research at the beginning of the session.While addressing the interview questions, discussion on views and topics also around the interview themes was encouraged (Walsham, 1995).The overall objective was to understand the factors and phenomena of importance in global software R&D work, such as effective collaboration, capability development, and potential challenges.
The analysis process was informed by interpretive and grounded theory methods (Corbin & Strauss, 1990;Klein & Myers, 1999;Urquhart, 2013;Walsham, 1995).Atlas.tiqualitative data analysis software was utilized as the main tool of the analysis with the support of Microsoft Visio software, as is described next.The audio material was first transcribed verbatim and after that the transcribed material was read thoroughly.Factors and phenomena perceived important and/or challenging and/or supportive in relation to globally distributed software R&D work were identified.These factors, phenomena, and their relations were first mapped into a Microsoft Visio drawing.This resulted in a complex map of well-functioning and desired states, challenges, risks, and their relations within the case.
After gaining this overall picture of the material, consciously also leaving room for "intuitive paths" (Yin, 1993, p. 5 as cited in Sarker, 2021, p. 251) of inductive discovery, the analyst returned to the transcribed material and systematically coded it by utilizing Atlas.ti.In the process, the data were first coded at a low level and then the codes were thematically categorized.The thematic categories resembled areas of the organizational life.For example, the initial category of organizational learning and competence development contained codes, such as "mutual, frequent knowledge and learnings sharing among teams," and "emphasizing and encouraging open sharing of mistakes for learning."Another initial category of effective collaboration contained codes such as "motivating and engaging people toward a common goal, vision," and "acknowledgment of succeeding or failing as one team." Both the Visio drawing and the Atlas.ticoding and categorization were conducted with an inductive approach (Kennedy & Thornberg, 2018;Urquhart, 2013).As the result of these analyses, a report with an explication of the identified challenging and supportive phenomena was prepared for the company to validate the findings and utilize them in practice.After delivering the report, e-mail exchange was carried out with the company representatives to comment on and discuss the report.Elaborative remarks were made, but no conflicting views on the results were identified.
The overall feel of these rich data was that areas such as operating together, dialogue for understanding, and mutual learning were prominent.Also, collaboration in different compositions of software R&D work, the relationship of organizational and local practices, and questions of learning and competence management appeared topical.Inspired by this impression, we returned to the data and started abductively (Kennedy & Thornberg, 2018) to explore how the data would respond to calls to pay more attention to meso-level issues, such as participation, productive dialogue (Salvato & Vassolo, 2018), and the employees' role (Wohlgemuth et al., 2019) in nurturing and building organizational DCs.It soon became apparent that "the social world" of organizational actors, in a way the social enactment of the organizational reality (Feldman & Orlikowski, 2011, p. 1241), such as communicating globally or engaging in the organizational development through improvement activities, was deeply entangled with the findings.At this point, the analytical lens of Figure 1 presented in the previous section started to form.
In this phase, the analyst systematically reviewed the list of codes by category and selected those that most directly related to the key concepts of relational engagement, productive dialogue, cohesion, participation, and interpersonal relationships (Salvato & Vassolo, 2018).In other words, the focus at this stage was on collaborative, socially influenced practices.Based on the previous coding work, the identified codes were first divided into enablers and challengers.To portray the level of detail in the data, of the 491 codes in total 137 were identified as resembling enablers, and 58 as resembling challengers, amounting to 195 codes.The total number of codes is high for two reasons: First, the analyst chose not to remove duplicate codes from the material but instead handle them as bundles.Second, the analyst wanted to break down the data, which could then be aggregated back into more meaningful and connected entities.This was to ensure an adequate understanding of the interview material from a complex organization had been reached.
The selected 195 codes were further aggregated into the emerging practices and activities with descriptions of related goals and challenges as identified based on the analysis.The findings were then organized utilizing the analytical lens of Figure 1 and sent for a final review by the case organization.The elaborative comments from this review are incorporated as notes within the findings, which are presented next.

Four Emerging Practices Enabling Dynamic Interpersonal Capability
The findings are here presented in a narrative form organized under the four emerging collaborative practices, which we suggest help enable the dynamic interpersonal capability in globally distributed software R&D work.The practices and their descriptions are, in the order of discussion: • Dialogical Organizational Development: Organizational development in dialogue and with acknowledgment of the needs of the different organizational levels Figure 2, • Constructive Working Culture: Fostering a supportive and open organizational culture, where mistakes are utilized for learning and improvement Figure 3 • Global Open Engagement: Nurturing openness and engagement in global communication and collaboration Figure 4, and • Facilitated Shared Learning: Organizational learning and idea cultivation through the management's facilitation and mutual sharing of experiences gained Figure 5.
According to the analysis, these four practices represent focal areas of the case organization's everyday life.In our reporting of the findings by practices and activities, we are inspired by the seminal paper of Orlikowski (2002), which qualitatively "explore[d] a possible explanation" of constituents of effective distributed organizing discovering the role of "collective knowing" (Orlikowski, 2002, p. 249).We are further encouraged by the recent qualitative study on "practical examples of challenges, barriers and enablers" of organizational change (Bojesson & Fundin, 2021).
The findings sections go hand in hand, and they are not intended as exhaustive (cf.Orlikowski, 2002).Instead, we find them illustrative of the power of collaborative practices, where both the management and employees make a difference by acting in an organization (Feldman & Orlikowski, 2011;Salvato & Vassolo, 2018;Wohlgemuth et al., 2019).Next, the findings related to dialogical organizational development are presented and grounded with examples from the interviews.

Dialogical organizational development
According to the analysis, common organizational practices, processes, tools, and policies enable effective collaboration among people from different backgrounds; provide a common language and terminology for a shared understanding of roles, tasks, and requirements; and facilitate product and software quality in a large organization.Thus, they could be characterized as a desired backbone of global software R&D work, as is illustrated in the below excerpt.
[I]f we don't understand each other at all -then it is very difficult to start a conversation and dialogue -.So, as these issues are complex, -this is a large organization, and because of that, it is important that we would have a common operations model.It is easier to communicate.It is not only the language, or even the time difference -.However, it is not always easy to fit organization-level (process) changes into the diverse sub-organizations, units, and teams.Simultaneously understanding the common ways of operating and having room for area-and teamspecific variation (cf.Wenzel et al., 2021) appears desirable.Thus, understanding the impact of changes across the chain and at different process levels seems important.Our data indicate that this understanding can be gained by reaching across organizational levels and carefully analyzing suggested process changes to ensure their feasibility at all levels.In doing so, it is important to listen to the feedback from all the levels of the process, which the below sample exemplifies.
It is important to listen to -the lowest level in the process.-[W]ill some change cause a problem, or can it be taken without problems -? [P]roblems can be caused there, if it is just stated that everybody will implement this -without listening to the feedback -.
Situational consideration of teams' circumstances in evaluating different improvement requirements is valued.Thus, finding a balance and dialogue between organization-and team-level practices by mutual conversation seems key.It entails making it possible to adjust the practices and processes as perceived adequate, understanding the organizational policies and their impact, and learning the existing practices before introducing changes.This ties in with mindfulness in taking improvement action.Considering the impact of improvement actions on the implementing teams is focal, and it includes an understanding of the consequences, reasons, and potential problems of the actions and their implementation.Gaining such an understanding requires time, knowledgeability, and the analysis of metrics, information from the team, and visibility to the feasibility of improvements in practice.
Further, all employees' active participation in ideating improvements is reflective of productive dialogue (cf.Salvato & Vassolo, 2018), and the finding by Wohlgemuth et al. (2019) that inclusion of employees may result in better decisions and faster change.Indeed, being able to influence things, such as performance metrics, is considered a motivational factor, which could also increase participation and prevent noninvolvement (cf.Salvato & Vassolo, 2018).As noted by an informant, "at least I'm motivated by being able to influence things."Otherwise, a risk of changes being perceived to come from the outside, without clarity of the need in practice, arises.Importantly, improvement activities can be utilized for challenging the teams' thinking in a constructive way and preventing them from becoming locked in how things have previously been done, as is remarked next.
[I]n a way it also challenges it.So that you don't go too much into "this is how it has always been done, and here this is how it is done," and when you have that in a right quantity, it is good.
In sum, the interview material illustrates how dialogue influences organizational practice and process development efforts.Dialogue supports participation and helps the making of better-informed decisions by directing the operation in a way that the risk of unintended hindering consequences in different sub-organizations is diminished.As importantly noted by the case company while reviewing the findings, balance is indeed essential, since too much team-specific diversion from harmonized processes could tip the scales and hamper the goal of the common operations model.Dialogue seems equally important as a means for the management to affect how teams perceive change without losing their sense of being able to influence things, even if the requirement originates from the organizational level.
Figure 2 compiles the findings from the practice of dialogical organizational development as organizational development in dialogue and with acknowledgment of the needs of different organizational levels.The figure explicates the goals (the why approach), the practice contributing to the goals (the what approach), and the activities and actors contributing to the practice (the how approach) based on the analysis.

Constructive working culture
Based on the analysis, creating, adopting, and enacting a common organizational culture with shared values appears desirable in a global, complex and distributed organization.It is acknowledged that a diverse organization will likely have diverse cultures, which points toward inclusion (cf.Salvato & Vassolo, 2018): "There are different cultures and acting in different cultures is a little bit different, but still a company can have common values and a common operations model." The management's role in building a supportive organizational culture can manifest as an overall visibility and emphasis of the culture and values within the organization, including management communication.It can also show as a drive for building common organizational values, fostering the openness of the culture, and aiming to fix problematic areas.
However, it is noted that gaps in internalizing and absorbing the common culture may hinder fruitful collaboration and integrating organizational cultures across global teams may pose a challenge.Site politics, weak adoption of the common culture, and conflicts caused by cultural or power differences are identified as risks.For example, a culture of blaming could drive toward creating overly safe plans to buffer against failure, which can be seen as harmful for nurturing the dynamic capacity of teams.As pointed out, blaming "for sure does not improve the operation, and at the worst it could be so that some of the -issues are then hidden and filtered out -." On the contrary, it is apparent that the enactment of an organizational culture where making mistakes is allowed is valued as supportive of open collaboration and innovativeness.As the next excerpt notes, acknowledging that mistakes are part of the innovation process, and that they could be learned from rather than feared, opens way for creating new things.
[One can] minimize mistakes, but -if you want to innovate, usually you need to try something you have never tried before, and then there is a possibility to make a mistake.That is how the new is created and making mistakes must be allowed.
A perceived supportive organizational culture is here identified as casual, flat, and open featuring a good working spirit.Acting according to organizational values, even under pressure and during difficult times, is appreciated.As described, "corporate culture -generally happy that -you can say openlywe do things at least relatively openly.It has always been an advantage." Therefore, the influence of the organizational culture on innovativeness and collaboration is reflected particularly through enacting common values, transparency, and a mutual blame-free culture that has room also for making mistakes.Figure 3 summarizes the findings from the practice of constructive working culture as fostering a supportive and open organizational culture where mistakes are utilized for learning and improvement.The figure includes the goals, practice, and activity and actors in a similar manner as in the previous section.

Global open engagement
The theme of openness flows over from the organizational culture to the communication and collaboration practices of the global organization.We identify this as the goal of an open, communicative and collaborative way of working around the R&D pipeline, including goal setting and issue resolution across cultural, team and organizational boundaries.The clarity and communication of goal setting, early inclusion of experts in problem solving, and the uncovering and improving of bottlenecks throughout the process are examples of such a practice.
Challenges may occur, if there are breaks in communication and limited openness or visibility in inter-team and inter-organizational collaboration.As summarized, "[i]n communication, it can be called a challenge, if you don't get that conversation.If you don't become understood -."These could also be situations where mishaps have occurred in cross-cultural communication.Further, limited openness may cause misconceptions of a team's performance or schedule, leading to bottlenecks in the process and conflicts in collaboration.An informant explains that "if there is no openness in terms ofinformation sharing, then your starting point is -an incorrect assumption -." We suggest these challenges influence participation and productive dialogue (cf.Salvato & Vassolo, 2018) in both everyday activity and organizational development efforts.The management promoting relational engagement among teams presents a counterforce to such challenges.Management is thus considered to have an important role in enabling cohesion among teams.According to the interview material, this means avoiding situations where teams end up competing against one another, providing support in solving conflicting messages, encouraging the asking of questions, and ensuring no one is left alone.Moreover, as noted while reviewing the findings, designing the performance metrics of teams in a way that promotes inter-team collaboration instead of competition helps contribute to the creation of an efficient software R&D pipeline.
Further, nurturing trust and information sharing are valued.Based on the interviews, achieving mutual trust requires openness, transparency, accountability, and bearing responsibility.The importance of information and knowledge sharing becomes emphasized in global collaboration influencing not only the daily collaboration but also common learning: [I]t needs to be emphasized, the importance of [building trust], because if we don't trust each other, we won't necessarily share all the information with each other nor necessarily talk about some slip-ups we have made, so that others would learn -.
Engaging in mutual conversation with adequate respect and humbleness means tactfulness in giving feedback, encouraging the open and timely communication of issues, and ensuring an approach of respect toward one another despite potential decision-making power.It materializes as "try-[ing] to -find a common note -." Similarly, finding solutions together and engaging in mutual conversation by asking, listening, and responding are appreciated.More specifically, striving for understanding the facts and the needs of others in a mutual way facilitates effective collaboration, conversation, and action: "always mutual, to understand the need of the others."Mutual understanding helps create a sustainable basis for collaboration and is highlighted especially in collaborations without a manager-employee relationship.
Finally, the approach of operating as a team gained emphasis.The approach includes acknowledging needing and supporting one another; being willing to succeed, fail, and find a way forward as one team; sharing mistakes for learning as a team; aiming for a common goal; and looking at things as one team of teams across borders.As an informant crystallizes it, "the will for common success -. [T]hat it's not only this our team that could make [it], but perhaps a little like a band -." Overall, these aspects resemble the relational engagement and productive dialogue called for by Salvato and Vassolo (2018).Nurturing openness and engagement in a global work community contributes to flowing collaboration in the software R&D pipeline with a clear goal setting and effective issue resolution across cultural and organizational boundaries.Figure 4 summarizes the findings from the practice of global open engagement as nurturing openness and engagement in global communication and collaboration.While the management set the stage for engagement among teams, following Feldman and Orlikowski (2011), we suggest that only the everyday actions of all employees create the practice.The figure is organized in a similar manner as in the previous sections.

Facilitated shared learning
Organizational learning is deeply entangled with the DCs of an organization (e.g., Ghosh & Srivastava, 2022).In this context, it can be identified to have at least two distinct goals, organizational learning for effective day-to-day operation and innovative solutions.In the latter, the concept of modifying organizational routines by DCs (cf.Zollo & Winter, 2002) is here extended to emphasize the even more explorative (cf.e.g., Levinthal & March, 1993) aspects of not only modifying but also looking beyond the existing routines for new, creative solutions.
A continuous challenge in a complex, fast-changing operation appears to be the balance of simultaneously gaining and maintaining a broad and deep understanding of the environment and problem space.First, this relates to the ability to create innovative solutions.Giving exposure to a wider space is perceived to facilitate creativeness.While wide understanding of the problem space enables new solutions with a broader perspective, increasing the visibility and understanding of the current environment is required to make future visions executable.In other words, "you need to know -your problem space,you need to find the necessity, -you need to know as wide problems as possible." Second, in the daily software R&D work, the challenge appears to relate to the axis of stability and the dynamism of tasks and responsibilities.On the one hand, while stability enables effective, streamlined processes with clear responsibility and accumulated expertise, it may introduce the risk of becoming overly stable, particularly in terms of continuous organizational learning.On the other hand, if the task and domain repertoire in which the experts are engaged is very dynamic with continuously changing responsibilities and a highly distributed task allocation, it may help in achieving a broad range of competences.At the same time, a risk of fragmentation of competence and collaboration seems real.Therefore, supporting a balance between stable and dynamic scope of responsibilities, and in building broad and deep expertise seems to help organizational learning in this context."[P]erhaps, to an extent, here as well, the middle ground can be a good approach -," one informant said.
Similarly, to enable learning, there needs to be a sufficient capacity for learning and moving in the right direction within teams.In the end, it is the management's responsibility to ensure there are enough people at the core with good learning skills and the capability to understand the entity.Moreover, it is important to identify and enable such people: "-to observe and understand that these kinds of persons exist, and -let them, and perhaps pursue them a little, to develop and take -different kinds of roles." Whichever mode of learning is in question, it appears that, again, the role of communication and engagement should be embraced in global work and idea cultivation.As summarized, "[t]here should be freedom of space for others to engage, to communicate, to share, because the more you share the -better your idea gets in the evolution."These issues become tied to both innovativeness and sharing of learnings.
When striving for understanding and learning through sharing of learnings, mutual, frequent knowledge and learning sharing among teams is appreciated.Similarly, as in the findings on organizational culture, the open sharing of mistakes is emphasized and encouraged for learning and improvement, as is highlighted below.
[T]o learn from mistakes and share mistakes.-That when you make a mistake, it is a very good practice to also learn, and for that we have also encouraged to share those mistakes also with other teams, so we could learn from them -.
[E]very time we release a product, we should have a lessons-learnt to the next release that what [we] did, what did go well, and what did not go well from the previous release, so that we do not repeat the same mistake again.
Finally, silos in knowledge and learning sharing could hinder gaining a broad understanding, for example, if one's focus remains solely on one's own product.Therefore, systematic utilization of mechanisms and processes of knowledge and learning sharing is identified as supportive in a complex and large organization.
Many of the aspects discussed in this section relate to the preceding sections of organizational culture and openness of communication.However, highlighting the perspective of organizational learning as its own entity illustrates that while organizational culture can be seen as an enabler, or even as a prerequisite, organizational learning, then, could be seen as one result of such a culture.Further, in reflection of the literature, we see qualities of both "tolerance for failure" and the practice of "failure analysis," where mistakes are not only regarded as opportunities for learning but are also openly analyzed and examined for lessons learned (cf.Danneels & Vestal, 2020).Figure 5 summarizes the findings from the practice of facilitated shared learning as organizational learning and idea cultivation through the management's facilitation and mutual sharing of gained experiences.

Discussion
In this paper, we set out to better understand how collaborative practices, as enacted by management and employees, enable interpersonal DC in the context of global software R&D work.As findings, we identified four collaborative emerging practices, actors and activities influencing the practices, and the goals contributed to by the practices.The findings denote the three approaches, the what, the how, and the why, of practice theory (Feldman & Orlikowski, 2011).
From the DC perspective, it can first be suggested that these empirical findings support the multi-level framework by Salvato and Vassolo (2018) in terms of the "interpersonal (meso) level" mechanisms of DC creation in a global software R&D organization.Second, the findings underline the importance of the inclusion of all employees in the enactment and creation of DCs in such a work environment.We believe this strengthens the view of how DCs should be considered as social processes crossing organizational layers (cf.Salvato & Vassolo, 2018) rather than primarily managerial capacities.
However, the scope of this argument depends on the context (McLaren & Durepos, 2021) and the lenses.If we adopt the view of DCs as strictly strategic capabilities where the management sets goals, directions, and structures for employees to follow, then yes, DCs appear primarily as a managerial capability.If we view DCs with a broader, more practice theory-oriented lenses (Feldman & Orlikowski, 2011;Nicolini, 2012), seeing the world as connected, emergent webs of actions, we almost inherently need to put more focus on the expertise across the organization.In today's complex, dynamic, and in many ways connected world, we might go even as far as to suggest that limiting the DCs to the management ends up ignoring the full potential of an organization, not only in becoming more ready for change but also in driving it (cf.Wenzel et al., 2021).
Even though the practices per se can quite easily be connected back to the various contexts of GDW, such as the importance of effective communication and collaboration (Kotlarsky & Oshri, 2005), or tensions relating to differing cultures (Cramton & Hinds, 2014), the most interesting part to us here lies elsewhere.It is in how the aspects of dialogue, participation, inclusion, and striving for understanding reappear in each of the practices, tying them together into social and collaborative accomplishments (cf.Salvato & Vassolo, 2018).
When looking at the goals in Figures 2-5, it seems that they have the potential to support the dynamic capacity of the organization in a number of ways.On the one hand, common organizational practices, processes, tools, and policies (Figure 2) create a stable ground.On the other hand, allowing enough room for variation according to the needs of the diverse sub-organizations supports dynamism, nimbleness, and routines' capability to adjust as needed (cf.Tallon et al., 2019;Wenzel et al., 2021).This could help prevent overly rigid organizational structures.The goals of a common supportive organizational culture (Figure 3) and smooth collaboration across the software R&D pipeline (Figure 4) promote organizational cohesion and inclusion, as well as the high quality of interpersonal relationships as part of the interpersonal level of DCs.Further, where the goals in Figure 2 addressed the balance of simultaneous dynamism and stability through a dialogic approach to organizational development, the goals of organizational learning in Figure 5 pose a dual, ambidextrous ambition of learning for both day-to-day operations and for new innovative solutions (cf.Marabelli & Galliers, 2017).In that sense, these goals address tensional or paradoxical phenomena (cf.Brooks et al., 2020;Wenzel et al., 2021).Finally, in some of the identified practices, actions appear more prominent, while in some others, structures dominate.For example, in facilitated shared learning, management providing the structures for collaborative learning seems paramount both from the perspectives of work allocation and facilitating learning forums.In constructive working culture, then, the enactment and perceptions of the culture by employees appear in the foreground.
These findings make the role of employees in generating DCs in a global software R&D organization visible and hopefully more concrete by explicating the activities in which the employees are engaged in terms of the identified practices.The findings also shed light on the enabling practices in an organization, when exploring the multi-level DC framework (Salvato & Vassolo, 2018) from an interpersonal, meso-level perspective.Thus, the findings tie the collaborative practices to organizational development in a global software work environment with the DC link.These notions apply from the practical perspective as implications to management, and on how the DC theory could be expanded to concern the role of all organizational participants (cf.Feldman & Orlikowski, 2011) more comprehensively.It could be seen that while the management facilitates the structures, all employees realize and generate the practices enabling dynamic interpersonal capability in an organization (cf.Giddens's structuration theory in Feldman & Orlikowski, 2011).Illustrative of this is the practitioner note that the best ideas and the culture of continuously improving starts from motivated teams who feel they can make a difference.
We further suggest that the identified practices can be evaluated in an organization in different dimensions.Figure 6 illustrates the four practices across two dimensions, the reach and type of organizational activity.The dimension of reach encompasses considerations, such as local vs.global, team vs. organizational level, and deep vs. broad scope.The dimension of type differentiates between exploitative (utilizing existing capacities) and explorative (developing new capacities) activity (Levinthal & March, 1993).The activities can be, for example, new competence development, process enhancements, or the sharing of learnings.The arrows and dashed circles in the figure exemplify how the emphasis of the practices may shift dynamically between the dimensions influenced by time, emerging situations, and organizational goals.For example, deep explorative learning may shift or be driven toward broad explorative learning, or local efforts of knowledge sharing culture expanded to cross team boundaries.The model could be utilized to reflect on and evaluate the different conditions and their sufficiency at different times in an organization, thus encompassing the aspects of temporality and dynamism of organizational development according to a process theoretical approach (Markus & Robey, 1988).
The study has limitations and contextual considerations that need to be discussed to position the findings in the body of DC and GDW knowledge (cf.Davison & Martinsons, 2016;McLaren & Durepos, 2021).First, the study could be judged to represent "the weak programme" of practice theory, where theory is utilized as a tool to identify the correct level of analysis rather than applied as "the strong programme" diving deep into the daily practices by methods such, as observation "of practice as it happens" (Nicolini, 2012, pp. 12-14).We identify this as a methodological issue, as we utilized interviews as the primary method of data collection, which Nicolini (2012) criticizes.However, we follow our understanding of Peppard et al. (2014) and Feldman and Orlikowski's (2011) perhaps more philosophically oriented approaches and utilize practice theory primarily as a lens.Additionally, the study could have benefited from a stronger organizational learning lens (Järvinen & Poikela, 2001;Levinthal & March, 1993).Nevertheless, we found the practice perspective most fitting considering the "complex, dynamic, distributed" environment heavily loaded with social processes (Feldman & Orlikowski, 2011, p. 1240).
Second, the number of interviewees is limited to eight in our study, which is close to the lower end of six informants of published qualitative IS research (Sarker, 2021).However, here we follow Sarker's (2021) notion of emphasizing the rich quality of the interviews and the selection criteria of the interviewees.To address this limitation, the findings were also validated by review and comments from the case organization representatives providing additional voices beyond the key informants.
Third, the study focuses on a northern European site of a large GDW company which presents a limitation in terms of generalizability of the findings.However, given the common organizational software processes, high experience of the organization in the GDW model, and the versatile work roles of the informants close to the GDW phenomena, we argue the findings provide insight into the enhancement of DC theory, especially in similar contexts of GDW.Examples of such contexts include the senior professionals' perspective, distributed organizations with common software processes, distributed organizations with high expertise across multiple locations, and culturally similar country contexts as in the present study.At the same time, we suggest the findings also provide value as input in the exploration of differing contexts of GDW, for example in comparing and contrasting the different findings.In sum, instead of aiming to generalize to a population or across different contexts (Tsang & Williams, 2012), we believe the paper makes a meaningful "contribution of rich insight" as an interpretive IS case study (Walsham, 1995, p. 79) in understanding how interpersonal-level DCs can operate beneficially in global software R&D work.

Conclusions
The paper presented a qualitative, interpretive single-case study exploring the research question of how collaborative practices, as enacted by management and employees, enable interpersonal DC in the context of global software R&D work.Based on the inductive and abductive analyses, we identified four collaborative, emerging practices: dialogical organizational development, constructive working culture, global open engagement, and facilitated shared learning.The paper discussed the findings against recent DCs and GDW research by utilizing the practice theory as an analytical tool.The paper also suggested a dynamic model for assessing the identified practices in the dimensions of type and reach of organizational activity.A common thread through the findings was the emphasis on social aspects, such as common understanding, sharing, and openness.
As implications for practice, the paper contributes to a new understanding for management of the practices supportive of organizational dynamic capability in the context of GDW.The importance of the different work roles of employees and the management was highlighted in terms of the facilitation and enactment of the practices.The paper links these collaborative practices to organizational development and provides pointers for their reflection and evaluation.The study encourages a stronger inclusion of employees in DC considerations in organizations in order to fully avail of an organization's dynamic potential.We anticipate that practices such as those discussed here will become even more relevant in the future (cf.McLaren & Durepos, 2021), as digitalization takes new shapes and forms, adding complexity to our worlds.
As theoretical contributions, the findings provide a further empirical grounding to the multi-level and participatory views of DCs in organizations (e.g., Salvato & Vassolo, 2018;Wohlgemuth et al., 2019), enhancing the previous theorizing and research with new insights on the emerging practices.Additionally, the paper provides a case of how the practice perspective can be utilized as an analytical lens in qualitative, empirical DC research.
Even with the limitations discussed in the preceding section, we believe these findings are useful in cases of global and complex operation, such as software R&D and other IT intensive activities.As further research, we suggest a move toward the strong practice research programme (Nicolini, 2012) in studying collaborative practices in the context of interpersonal DCs in globally distributed software R&D work.While the findings were accomplished by analyzing interview material, we could increase the understanding and further put it to the test by applying the tool kit suggested by Nicolini (2012).This would include applying observational techniques to a chosen set of practices in real time in an organization.It would also be beneficial to include several locations or units from a case organization to gain a more complete picture, for example, to help identify similarities, differences, and potential conflicts among the practices.Overall, these findings encourage including employees more closely into DC research in the future.
To conclude, by applying the practice lens, we studied practices enabling interpersonal-level DC in a global software R&D organization.With this study, we wish to contribute to the understanding of the collaborative nature of DCs crossing organizational layers, both in theory and management practice, and how the practice theory can be used as an analytical lens in organizational IS research.While the study has its limitations, we believe it contributes to the scholarly and managerial discussion on the development of dynamic capabilities through collaborative activities engaged in by different actors in global software R&D work.

Figure 1 .
Figure 1.The analytical lens utilizing the DC theory and practice perspective (PP).

Figure 2 .
Figure 2. A summary of the findings in the dialogical organizational development.

Figure 3 .
Figure 3.A summary of the findings in constructive working culture.

Figure 4 .
Figure 4.A summary of the findings in global open engagement.

Figure 5 .
Figure 5.A summary of the findings in facilitated shared learning.

Figure 6 .
Figure 6.A dynamic model exemplifying the four practices in two dimensions.