“When in Rome, Do as the Romans Do”: Cultural Barriers to Being Agile in Distributed Teams

Agile ways of working stem from a group of methods united by a common philosophy, values, and principles. It emphasizes teamwork and heavily relies on the ability of a software team to self-manage [8, 19]. The principles of self-management and autonomy, central to agile ways of working, put certain demand on the organizational culture, team composition and behavioural norms [8, 21, 28]. Morgan [23] emphasises the importance of teams’ ability to engage in self-learning and drive continuous improvement, and ability to act upon minimum critical specification. van Solingen et al. [26] argue that the prerequisites for improvement and learning are openness and the ability to discuss the underlying problems. Based on two large surveys of agile teams, Williams captures practices essential for teams to be considered agile being related to their ability to satisfy the customer through early, continuous and frequent delivery of valuable, working software; the prerequisite for which is, among others, staffing projects with motivated individuals who are given the needed resources and authority to get their job done [31].

A number of studies investigated the relationship between organizational culture and the use of agile methods [8, 12, 13, 29]. Based on a multi-case study of nine projects Strode et al. [29] found that specific organizational culture factors correlate with effective use of an agile method. Their findings suggest that an organization is more likely to be successful if the organization values feedback and learning; social interaction in the organization is trustful, collaborative, and competent; the project manager acts as a facilitator; the management style is that of leadership and collaboration; the organization values teamwork is flexible and participative and encourages social interaction; the organization enables empowerment of people; the organization is results oriented; leadership in the organization is entrepreneurial, innovative, and risk taking; and the organization is based on loyalty and mutual trust and commitment [29]. Similar findings emerged from studying 58 agile practitioners from 23 organizations in New Zealand and India [8]. Hoda et al. found that the prerequisite for self-organizing agile teams to establish and flourish is senior management support, in terms of providing freedom and establishing an organizational culture of trust. They also suggest that an organization with a strict hierarchical structure is not conducive to self-organizing agile teams, because the hierarchy enforces a lack of openness marked by restricted and indirect lines of communication and feedback, which in turn leads to an environment of fear [8]. Based on a multi-case study, Kautz et al. [13] found that agile development thrives in different organizational cultures, even in hierarchical ones, as long as the 4 core values are present to a significant extent. Furthermore, they argue that while organizational culture has an impact on the way agile development is enacted, in practice it is often the method which is adjusted to the organization. Similarly, Iivari et al. argue that the relationship between an organizational culture and agile ways of working is dynamic and therefore will continuously evolve [12]. This means that time perspective matters and studies on the compatibility between the culture and agile ways of working shall take the dynamic nature of this relationship into account.

Another reason to look at the organizational culture from a time perspective is the staff changes. When adding new people to an already established agile team, it is essential to support the new team members in adapting to the existing teams culture and ways of working, which is especially difficult in virtual setups. In their study of onboarding Portuguese developers into existing Norwegian agile teams [22], Moe et al. conclude that the most important success factor is finding people that matched the culture of the existing teams. Therefore, during an onboarding process, all interviews and visits need to focus on communicating the values and culture and on giving insight into the existing organization’s norms.

Since, national culture is said to have significant influence on the organizational culture [9] and organizational culture may impact the use and success of agile ways of working [13, 29], there is interest in understanding the use of agile methods and practices in companies located outside the locations of early adopters of agile methods. In particular, researchers and practitioners have wondered about the abilities of the companies and engineers from the Asian region, the primary recipients of offshoring contracts, to adopt the agile ways of working, which are so distinct to their national culture.

To address these questions, a number of studies sought evidence of successful use of agile methods in offshored projects [10, 25]. Some researchers infer the successful adoption from the large number of practices reported as being followed [2, 32]. However, the validity of these studies as well as the research approach are questionable, because high level of commitment to the use of agile practices can be explained by the readiness to accept the established rules in hierarchical (i.e., high power distance) cultures, as found, for example, in a study of agile adoption in Malaysia [1]. Other research studies tried to improve the understanding of what specifically impedes the adoption of agile ways of working in Asian cultures [1, 5, 8, 14, 15, 30] and how to succeed [8, 25]. In Table 1, we summarize a list of impeding behaviors reported on the managerial and engineering levels in related studies.

The cited studies cover different countries within the Asian region, including India [5, 8, 27, 30], Malaysia and Singapore [1] and Asia Pacific in general [15], and are either based on interviews or own experiences.

A closer look at behaviors of engineers in India and neighboring countries reveals that most if not all impeding behaviors are likely to be caused by the hierarchical culture of the organizations and related management behavior, as suggested in related research [8, 29]. For example, Ayed et al. [1] report that Malaysian and Singapore engineers lacked the freedom to decide about their ways of working and therefore did not see the point in self-learning. But what if the hierarchical culture of command-and-control highlighted in numerous studies as poisonous to the agile ways of working [1, 5, 10, 14, 15, 30] would be replaced with the more empowering onshore management; would the offshore engineers working in mixed onshore-offshore teams be able to adopt the agile ways of working? Or would the less hierarchical Western companies fail to ignite the agile culture in their offshore collaborations? The answers to these questions are of high importance for shaping the understanding of the compatibility of agile ways of working with the use of offshoring.

The context of our study is a collaboration between an outsourcer/customer company from Sweden working in the telecommunication industry and an outsourcing vendor/a consultancy company from India. For confidentiality reasons we are prohibited to disclose the names of either of the companies. Our investigation targeted five distributed agile teams composed of team members from both the Swedish company and the offshore vendor from India. The selection of the teams was done with the help of the companies and represented all important business areas.

The studied teams were set up as DevOps teams consisting of a few smaller mixed Dev and Ops teams working accordingly in the development stream or the operation stream with supporting roles around. Each smaller team was cross-functional and involved developers, testers, a System lead, a Team lead and a Product owner (see the structure in Fig. 1 and profiles of the studied DevOps in Table 2).

DevOps teams followed agile principles and ways of working with iterative development (Scrum or Kanban, decided by each team individually), daily stand ups, and team retrospectives as the primary rituals. The offshore members of the teams were expected to follow the same agile principles and philosophy as the contracting organization. This is why cultural incompatibility across locations was seen as a threat.

A mixed approach was employed to study the impact of cultural differences on the collaboration between the Swedish customer and the Indian outsourcing vendor. Data collection was done in several steps including a quantitative data collection approach when exploring the personally experienced misunderstandings and relevant behaviors impeding collaboration, and a more qualitative approach when checking the occurrence of impeding behaviors in five distributed DevOps teams, as visualized in Fig. 2 and described in the following sub-sections.

The generated items and the recorded notes were analyzed in iterations. First, we aggregated elicited items in one list of misunderstandings grouped by similarity. This list contained seven larger categories related to cultural differences and their impacts. Then we revisited each category one by one, read through the session notes and formulated items in a particular form: As a <role and/or site representative> it is confusing for me when <role and/or site representative><behavior> (when/in <situation>).

In this subsection, we discuss the limitations and issues that might threaten the validity of our results. Given the qualitative nature of our study, in the following we discuss the validity, reliability and generalizability threats following the guidelines by Leung [17].

Reliability in qualitative research refers to the replicability of the results. A margin of variability in results is accepted when dealing with qualitative research [17] or mixed methods, since the subjectivity of the researcher is embedded in the roots of the analysis. The main threat is then related to consistency. To mitigate this threat, we let the participants report their responses in a data collection tool, by being systematic when taking notes and documenting the discussions during the sessions, and by keeping the quotations from the participants as exact as possible. Furthermore, we conducted feedback sessions to validate our interpretation of the impeding behaviors and situations.

Possible threats to the validity and reliability of our results are related to the reluctance of offshore participants to express their opinion in front of managers or onshore peers, as well as reluctance to talk about compromising issues. To alleviate these threats, we separated the Indian and Swedish participants in the initial group interviews and feedback sessions to be able to talk more openly. In the mixed workshop sessions, we explained the importance of truthful responses and asked onshore team leads to encourage openness. We also used a survey form that allowed participants to provide their responses anonymously, which remained untraceable to individuals even during the discussions. The reported problematic behavior in the survey, confirming comments received from the onshore participants and reflections voiced by more experienced offshore participants (in other words triangulation of the data sources) make us believe that we have elicited as honest and open responses as possible.

Data triangulation is the core principle of case study research [24]. To enhance the validity and reliability of our results, the individual responses elicited during the group interviews were first discussed in the respective groups and compared across the onshore and offshore groups. We then elicited quantitative data from larger groups of participants (survey responses) to minimize the bias towards selected individuals. The quantitative data elicited through the survey was further triangulated with the qualitative data (notes) from the discussions held during the workshops.

Generalizability of the conclusions drawn from our results are of course limited to the studied context. However, our results disprove an existing view that cultural barriers are likely to remain since the major differences in norms and values cannot be harmonized [14]. We believe that it is fair to assume that the gradual changes in behavior that we observed as a result of the gained experience with working in a mixed environment may also happen in similar contexts in other organizations.

Based on the interviews, we identified 19 sources of misunderstandings reported by the Swedish participants and 14 sources of misunderstandings reported by the Indian participants. Notably, both sides reported what was confusing in the behavior of their counterparts as well as the own behavior that have led to misunderstandings or confusion. Our further analysis of the situations in which differences in behavior were seen as barriers for collaboration led to an aggregated list of 26 behaviors, which was triangulated with related literature. Of these, 12 behaviors were prioritized as frequently occurring and important to discuss as determined in the feedback sessions. In this paper, we provide an analysis of six of these behaviors that can be classified as impediments to agile ways of working (see Table 3).

The six reported behaviors impeding agile ways of working surface in the daily meetings, task allocation and content discussions, and team retrospectives; and appeared all but one as common sources of misunderstandings between the offshore and onshore members (see the column Happens in Table 3). Evidently, the most typical impediment is willingness to say yes to most requests in deference to superiors and reluctance to warn about non-feasible deadlines. Other impeding behaviors had varying frequency of occurrence. Some respondents indicated that although the impeding behaviors occurred in the past (“Used to happen” in Table 3), social integration of the Indian members led to the assimilation of the established ways of working and put an end to behavioral differences. This is why, in the next step we sought to further understand how common the impeding behaviors are in different teams, and what stimulates cultural integration.

Table 4 summarizes our results from surveying offshore members from five distributed DevOps teams regarding their likely behavior in six situations (impeding behaviors are emphasized in red color). Our results suggest that accepting unfeasible tasks in deference to superiors (#1), seeking immediate manager’s approval for tasks (#2) and reluctance to confess about lagging behind schedule (#4) are behaviors experienced in all five distributed teams, while confusing behavior in situations #5 and #6 are not that common. It is also evidently from the results that DevOps 1 and DevOps 5 appear quite successful with the cultural integration of their Indian members, while DevOps 2, 3 and 4 are challenged. This is confirmed by a more detailed analysis, which reveals that the most common respondents of impeding behavior are the recently onboarded members. In the following, we further detail our findings from analyzing survey responses and notes capturing the discussions the team members held during the sessions.

In response to the first research question, we have identified several behaviors rooted in the hierarchical forms of organization as impediments for agile ways of working, which confirm previous research that highlight cultural barriers common to offshore organizations. In particular, we found the reinforcement of deference to superiors [5, 8, 14] to be a common barrier for the studied teams. This in our case led to a willingness to say yes to even unrealistic requests (similarly to [1, 5, 30]), reluctance to expose problems (similarly to [1, 5, 8, 30]) and reluctance to reveal the lack of understanding and ask questions to a superior. At the same time, our findings suggest that the empowering culture and democratic leadership from Sweden encourage the trust and transparency-based behavior among the offshore members. However, the behavioral transformations take time and cannot be taken for granted. In the presence of hierarchical structures and command-and-control leadership locally, cultural integration might take longer, or result in that offshore team members are forced to assimilate different behaviors in parallel. In fact, during one of the workshops an offshore team lead told us that offshore engineers that move between customers might not be rewarded for what is a typical behavior in an agile environment and need to revert to the behaviors typical to hierarchical cultures.

When contrasting our findings with the organizational culture factors that correlate with effective use of an agile method [8, 29], we can say that the culture of the outsourcing vendor did not confirm with the highlighted values and management style. The implication of this is that agile companies might be more likely to succeed with offshoring through establishing their own sites; working with vendors that are already used to agile values and ways of working or recruiting people that matched the culture of the existing teams [22]. In particular, the companies may want to assess the vendors management style, i.e., facilitation-leadership, collaboration-oriented management style and focus on empowerment as crucial factors for adopting agility [29].

Unlike researchers who suggest that cultural barriers are likely to remain since the major differences in norms and values cannot be harmonized, as they derive from deep-seated differences in cultural background, education, and working life [14], we found that behavior seem to change when engineers are exposed to the culturally distinct environment. Our findings demonstrate that Swedish members succeeded in integrating members of a hierarchical culture and stimulating the changes in their behavior. Our findings are in accordance with Iivari et al. that found that the relationship between an organizational culture and agile ways of working is dynamic [12]. Our experience, however, shows that distributed teams are often left to experiment and adjust their ways of working, since cultural awareness is often gained with experience. This has been noted by Casey [3], who found that the importance of and requirement for cultural training is often not recognized before a lot of time, effort and resources are wasted. We therefore recommend companies and distributed teams to run cross-cultural communication courses to discuss the values that should govern behavior as also suggested in [16]. For team leads and onshore managers struggling to overcome the same cultural barriers as reported in this article, we can recommend the following: