Determinants of agile team performance: a systematic literature review

Agile methodologies have become one of the most widely adopted approaches for organizing software development projects in dynamic and rapidly changing environments. Agile development emphasizes iterative delivery, continuous customer feedback and close collaboration among team members to respond effectively to evolving requirements and technological change (Beck et al., 2001).

Software development requires coordination among individuals with diverse technical skills. Agile approaches organize work around small cross-functional teams that collaborate intensively throughout short development cycles. As a result, the performance of agile teams has become a critical factor influencing project success, product quality and delivery speed. Empirical research consistently demonstrates that effective teamwork contributes to improved outcomes in software development environments.

Prior studies highlight the importance of teamwork quality in determining team effectiveness. Teamwork quality includes factors such as communication, coordination of expertise, cohesion, trust and mutual support among team members. Evidence from agile development teams shows that these factors positively influence team performance, learning and work satisfaction (Lindsjørn et al., 2016). Similarly, research examining teamwork models identifies leadership behaviors, team orientation and peer feedback as key elements that support effective collaboration and team performance (Jaroliya and Gyanchandani, 2022; Strode et al., 2022). These findings suggest that agile team performance is influenced by a combination of interpersonal processes and structural team characteristics.

Other studies have examined how organizational and process-related factors influence agile team effectiveness. For example, team autonomy and competence have been shown to influence software development agility through communication and collaborative decision-making processes (Rathor et al., 2024). Team agility has also been conceptualized as an emergent property of complex adaptive systems (Werder and Maedche, 2018), complemented by team delivery capability (Xia et al., 2024). Research also indicates that psychological safety, team climate and social interaction patterns play an important role in enabling learning and performance within agile teams (Ahmad et al., 2022).

Despite the growing volume of research on agile teams, the literature remains fragmented across several academic fields including information systems, software engineering, project management and organizational behavior. Individual studies often investigate specific aspects of agile teamwork such as leadership, communication, team structure or psychological safety without integrating these insights into a comprehensive understanding of agile team performance. Consequently, knowledge about the determinants of agile team effectiveness remains dispersed across individual studies. Earlier systematic literature reviews (SLRs) established this approach’s value for the field (Dybå and Dingsøyr, 2008) but did not focus specifically on team-level performance determinants.

Systematic literature reviews provide a rigorous approach for synthesizing existing research and developing a consolidated understanding of a research domain. Such reviews systematically identify, evaluate and synthesize relevant studies to summarize existing knowledge and highlight research gaps (Paré et al., 2015). By integrating findings from multiple studies, systematic reviews enable researchers to identify recurring themes, patterns and relationships within the literature.

Although research on agile development teams has expanded considerably, there remains limited synthesis of the determinants of agile team performance. Existing studies frequently examine individual determinants in isolation, which makes it difficult to understand how different factors collectively influence agile team effectiveness. Furthermore, the coordination mechanisms and contextual conditions that enable agile teams to achieve high performance have not been comprehensively consolidated in the literature.

To address these limitations, this study conducts a systematic literature review of research on agile team performance. The review aims to synthesize existing evidence and provide a structured understanding of the factors that influence teamwork effectiveness in agile development environments.

The study is guided by the following research questions:

This review makes three contributions to the literature. First, rather than cataloguing determinants of agile team performance as an unordered list, it synthesizes the reviewed evidence into an integrative process model in which organizational conditions and team structural characteristics operate as antecedent, enabling factors, while team processes and emergent states, including communication, coordination, psychological safety and reflexivity, function as the proximal mechanism through which those antecedents translate into performance outcomes. This addresses a tendency in prior work to treat team-level determinants as parallel, independent predictors rather than as elements of a coherent causal structure. Second, the review highlights a boundary condition that has received little attention to date: much of the underlying evidence on team processes was generated in co-located team settings, yet agile software development increasingly occurs in remote and hybrid arrangements (Adzgauskaite et al., 2025; Khanna et al., 2024; de Souza Santos and Ralph, 2022). Whether communication, coordination and psychological safety operate with the same strength once teams are geographically distributed remains unclear, and this review makes that gap explicit rather than assuming continuity across contexts. Third, the review draws attention to substantial heterogeneity in how agile team performance has been operationalized across studies, spanning productivity and delivery measures, software quality indicators, project success ratings and team satisfaction, a heterogeneity that complicates comparison of findings and has not been made explicit in prior reviews of this literature. In addition to these three contributions, the review is deliberately scoped to agile teams in software development contexts, offering a more precise, mechanism-level account than broader cross-context reviews of agile team effectiveness (e.g. Steegh et al., 2025) while still drawing on established teamwork theory (e.g. Ilgen et al., 2005; Edmondson, 1999) to ground the software-development-specific findings in established models of team functioning. General models of teamwork similarly treat team processes as the mechanism linking team- and organization-level conditions to outcomes (Salas et al., 2005; Kozlowski and Bell, 2013).

This study adopts an SLR approach to synthesize existing research on agile team performance. Systematic literature reviews provide a structured and transparent method for identifying, evaluating and integrating findings from a body of research (Kitchenham and Charters, 2007; Tranfield et al., 2003). This follows established guidance on literature review methodology (Snyder, 2019; Brereton et al., 2007).

Research on agile software development has expanded substantially during the past decade, with scholars examining topics such as teamwork quality, team autonomy, leadership distribution, coordination mechanisms and psychological safety within agile teams (Lindsjørn et al., 2016; Hennel and Rosenkranz, 2021; Malik et al., 2021; Strode et al., 2022). However, these studies are distributed across several research domains, including software engineering, information systems, project management and organizational behavior. Consequently, empirical insights into agile team performance remain fragmented across disciplines. A systematic literature review is therefore appropriate for consolidating these dispersed contributions and identifying recurring themes in the literature. The review process followed established guidelines for systematic reviews in management and software engineering research (Kitchenham and Charters, 2007; Tranfield et al., 2003). Reporting of the study selection procedure was aligned with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 framework (Page et al., 2021).

A comprehensive search strategy was developed to identify studies examining team performance and related constructs in agile software development environments. Multiple academic databases were selected to ensure broad disciplinary coverage and to minimize the risk of omitting relevant publications.

The primary databases included Scopus and Web of Science Core Collection due to their extensive indexing of peer-reviewed journals and citation tracking capabilities. To capture research within the software engineering and computer science communities, IEEE Xplore and the ACM Digital Library were also searched. Additional databases such as ScienceDirect, Wiley Online Library and Emerald Insight were included because they index journals in information systems, project management and organizational research.

Google Scholar was used as a supplementary tool for backward and forward citation chaining rather than as a primary search source (Haddaway et al., 2015).

The search covered studies published between 2014 and 2025. This time period reflects a phase during which research on agile methodologies increasingly focused on team-level dynamics, collaboration mechanisms and performance outcomes rather than on early adoption debates.

Search expressions were developed iteratively to capture variations in terminology across the literature. Keywords associated with agile development approaches, including agile software development, Scrum and extreme programming, were combined with constructs related to teamwork and performance. These constructs included team performance, team effectiveness, teamwork quality, team dynamics, autonomy, leadership, communication and psychological safety. Boolean operators were used to combine search terms and refine results.

Table 1 summarizes the search strategy and total records identified.

Following the database search, backward and forward citation tracking was conducted to identify additional relevant studies. Citation chaining is widely recommended in systematic literature review methodology because it helps identify influential publications that may not appear in keyword searches (Wohlin, 2014).

Inclusion and exclusion criteria were defined before screening began to ensure transparency and reduce bias (Kitchenham and Charters, 2007; Tranfield et al., 2003).

Studies were included if they examined agile software development teams and addressed team-level constructs related to performance or effectiveness. Relevant constructs included teamwork quality, coordination, communication, leadership, autonomy, psychological safety, self-organization and team agility. Both empirical studies and theory-based research investigating team dynamics within agile environments were considered.

Only peer-reviewed journal articles and conference proceedings were included, reflecting publication norms in software engineering, where high-quality empirical work often appears at leading conferences (Kitchenham et al., 2009).

Publications were excluded if they focused exclusively on technical aspects of software development, such as programming techniques or software architecture, without addressing team-level mechanisms. Studies examining organizational transformation without analyzing team processes were also excluded. Grey literature, including technical reports, dissertations, editorials and white papers, was not included in the review. The analysis was limited to publications written in English.

The study selection process followed four stages: identification, screening, eligibility assessment and final inclusion. Records retrieved from database searches were first exported into reference management software, where duplicate entries were removed. A total of 612 records were identified through database searching, of which 112 were removed as duplicates, leaving 500 records for title/abstract screening. Duplicate removal is an essential step in systematic reviews because repeated records can distort the screening process (Kitchenham et al., 2009).

In the screening stage, titles and abstracts of the remaining records were reviewed to determine their relevance to agile software development teams and team performance constructs. Publications that focused exclusively on technical development practices or software tools without examining team dynamics were excluded during this stage. At this stage, 420 records were excluded and 80 records were retained for full-text assessment.

The remaining studies were subjected to full-text assessment to evaluate their eligibility. Each article was examined against the predefined inclusion and exclusion criteria to determine its conceptual relevance and methodological clarity. Studies that mentioned agile teams only marginally or did not analyze team processes or performance outcomes were excluded. At this stage, 45 records were excluded, consistent with the counts reported in Figure 1.

After completing this procedure, 35 studies met all selection criteria and were retained for analysis.

Records excluded at full-text review were removed according to five criteria:

1. technical focus without team-level analysis;

2. organizational-transformation focus without analysis of team processes;

3. marginal or nonsubstantive treatment of agile teams;

4. grey literature (technical reports, dissertations, editorials and white papers); and

5. publications not written in English.

The distribution of exclusions across these five criteria individually is not reported; only the aggregate counts at each stage, shown in Figure 1, are presented.

Data extraction was conducted using a structured coding framework designed to capture key characteristics of the included studies. Coding and synthesis followed established procedures for thematic synthesis of qualitative and mixed-methods research in systematic reviews (Thomas and Harden, 2008). Information extracted from each article included publication year, research method, study context, theoretical foundations, constructs examined and reported performance outcomes.

The coding process followed an iterative procedure. Initial coding identified recurring constructs across the literature, including autonomy, communication quality, shared mental models, leadership structure, teamwork quality, psychological safety and coordination mechanisms. These constructs were then grouped into broader categories representing major dimensions of agile team functioning.

The systematic review resulted in a final sample of 35 studies published between 2014 and 2025. The increasing number of publications during this period reflects growing scholarly interest in understanding the mechanisms that influence team performance in agile software development environments. The emergence of agile methodologies has shifted research attention toward collaborative development practices, team coordination mechanisms and the organizational conditions that support agile work.

Quantitative survey research was the most frequently used method. For example, Lindsjørn et al. (2016) demonstrated that teamwork quality significantly influences team productivity and project success, and Malik et al. (2021) showed that psychological empowerment predicts stronger collaboration and effectiveness. Among the subset of included studies with verified methodological detail reported in Table 2, quantitative survey designs predominate, consistent with the pattern described above.

Qualitative case studies also featured prominently. Moe et al. (2010) examined coordination practices, highlighting self-organization and collective decision-making, while other qualitative work has investigated large-scale environments where multiple teams collaborate (Dingsøyr and Moe, 2014).

Several studies employed mixed-method designs, using interviews or observations to identify teamwork practices before testing relationships with survey data, capturing both social dynamics and measurable outcomes.

The reviewed studies were conducted in software firms, technology companies and large-scale agile environments. This diversity of contexts indicates that agile team practices are widely adopted across industries and that research on agile teamwork is relevant to multiple organizational settings. Despite differences in research methods and contexts, many studies examined similar constructs related to team processes, team structure and organizational support.

A further observation from the descriptive analysis concerns how agile team performance itself was measured across the reviewed studies. Rather than relying on a shared outcome construct, the literature operationalizes performance through a range of distinct measures, including productivity and delivery-related indicators, software quality, project success ratings and team satisfaction. Table 2 illustrates this heterogeneity using studies discussed elsewhere in this review. This inconsistency in outcome measurement complicates direct comparison of effect patterns across studies and has not been made explicit in prior syntheses of this literature; it is treated here as a distinct methodological gap that future research, including future systematic reviews and meta-analyses, should address through a shared outcome taxonomy.

The thematic synthesis of the literature identified three major categories of factors influencing agile team performance: team processes, team structure and composition and organizational conditions. These themes emerged consistently across the reviewed studies and reflect the multilevel nature of teamwork in agile software development environments.

RQ1 examines the factors influencing agile team performance. A consistent finding across the reviewed studies is the importance of team processes that facilitate collaboration and coordination among team members. Agile development practices emphasize continuous interaction through mechanisms such as daily standup meetings, sprint reviews and retrospectives. These practices create opportunities for teams to exchange information, coordinate tasks and adjust their strategies during development cycles.

Communication quality emerges as one of the most frequently cited drivers of team effectiveness in agile environments. Empirical research shows that frequent and transparent communication improves coordination and reduces misunderstandings among team members. Lindsjørn et al. (2016) demonstrated that communication and coordination are central elements of teamwork quality and are strongly associated with project success in agile teams. Similarly, Schmidt et al. (2014) showed that agile practices enhance team performance by supporting shared mental models that allow team members to anticipate each other’s actions and coordinate complex tasks more effectively. This theoretical mechanism has also been examined directly in agile contexts using shared mental models theory (Yu and Petter, 2014).

Psychological safety is another process repeatedly identified in the literature. Psychological safety refers to a shared belief that team members can express ideas and concerns without fear of negative consequences (Edmondson, 1999). Studies examining agile development teams show that psychologically safe environments encourage open discussion of problems and learning behaviors, which contribute to improved software quality and project outcomes (Alami et al., 2023; Hennel and Rosenkranz, 2021).

Team reflexivity also contributes to agile team effectiveness. Reflexivity refers to the collective process through which teams evaluate their goals, strategies and work practices to improve performance. Agile retrospectives are designed specifically to support such reflection and learning. Research suggests that teams engaging in regular reflection are better able to adapt their work practices and respond to changing project requirements (Buvik and Tkalich, 2021). Consistent with this, reflexive software development teams have been found to outperform less reflexive teams on delivery outcomes (Kakar, 2017).

RQ2 examines the structural mechanisms that shape agile team performance and how these relate to team processes. Agile methodologies rely on cross-functional teams composed of members with complementary technical skills and expertise. Structural characteristics of teams therefore influence how tasks are organized and how team members collaborate during development activities.

Team autonomy is widely recognized as a defining feature of agile team design. Autonomous teams are responsible for organizing their own work processes and making decisions regarding task allocation and development practices. Empirical studies show that autonomy enhances team motivation and adaptability in complex project environments (Tripp et al., 2016). Research also indicates that autonomous teams are better able to respond to evolving customer requirements and technological challenges (Rathor et al., 2024). Autonomous teams must also actively negotiate the boundaries of that autonomy with the surrounding organization (Mikalsen et al., 2019).

Leadership arrangements also influence team performance in agile environments. Rather than relying on hierarchical leadership structures, agile teams frequently adopt shared leadership models in which leadership responsibilities are distributed among team members. Strode et al. (2022) argued that shared leadership improves coordination and collective problem-solving by enabling team members to contribute their expertise during decision-making processes. This pattern extends to multinational agile teams, where leadership style shapes collaborative attitudes and outcomes (Quill et al., 2024).

Team size and skill diversity further affect team effectiveness. Smaller teams facilitate communication and reduce coordination complexity, while teams composed of members with diverse expertise are better able to address complex development tasks. Research on agile development teams indicates that balanced skill distribution supports collaboration and improves project outcomes (Dingsøyr et al., 2024). Individual-level traits also matter: personality has been linked to Scrum team effectiveness, alongside skill diversity (Truong et al., 2025).

RQ3 concerns the role of the organizational environment in shaping agile team performance and how organizational conditions relate to team processes. The literature indicates that agile teams do not operate in isolation but are embedded within broader organizational structures that influence how they function.

Organizational culture plays an important role in enabling team effectiveness. Prior research indicates that organizational culture and employee engagement significantly influence employee performance and work outcomes (Singh and Randhawa, 2020; Hennel and Rosenkranz, 2021). Agile teams operate within broader organizational systems, including digital workplace practices affecting engagement (Rani and Suneja, 2025); consistent with the proposed model, agile working relates to performance partly through team engagement (Peeters et al., 2022). Management support also emerges as a key organizational factor influencing agile team performance. Leadership commitment to agile practices facilitates the allocation of resources, training programs and organizational alignment required for agile teams to operate effectively (Malik et al., 2021). Comparable evidence from other sectors confirms this pattern (Foli and Caesar, 2020).

Coordination mechanisms at the organizational level become particularly important in large-scale agile environments where multiple teams collaborate on complex projects. Studies examining large-scale agile development highlight the importance of governance structures and communication mechanisms that enable coordination across teams (Dingsøyr and Moe, 2014). Case studies identify specific enabling mechanisms, including synchronization meetings and boundary-spanning roles (Bjørnson et al., 2018), while others highlight the difficulty of tailoring standard practices to large, multiteam programs (Rolland et al., 2016). The findings indicate that agile team performance emerges from the interaction of team processes, structural team characteristics and organizational conditions. These factors operate at different levels but collectively shape the effectiveness of agile software development teams.

The findings reported under RQ1–RQ3 are not independent of one another. Read together, they indicate that organizational conditions and team structural characteristics do not influence agile team performance directly, but operate as antecedent, enabling factors whose effects are carried by team processes and emergent states. This section makes that mediating logic explicit, drawing on relationships already implied by individual studies within the reviewed literature but not previously consolidated into a single model.

Three pieces of evidence from the reviewed studies support this structure. First, team autonomy and competence have been shown to influence software development agility specifically through communication and collaborative decision-making, rather than directly (Rathor et al., 2024), indicating that a structural antecedent (autonomy) operates through a team process (communication and decision-making). Second, psychological empowerment, a structural and motivational condition, has been linked to team performance via its effect on collaboration, again consistent with a process-mediated rather than a direct pathway (Malik et al., 2021). Third, supportive organizational cultures have been shown to foster the psychological safety that in turn strengthens teamwork effectiveness and software quality, rather than acting on performance outcomes directly (Hennel and Rosenkranz, 2021). In each case, a structural or organizational condition is associated with performance only in conjunction with or through a team process.

Accordingly, this review proposes that: (P1) organizational conditions, particularly management support and organizational culture, shape the structural design choices available to agile teams, including autonomy and shared leadership arrangements; (P2) structural characteristics of teams influence performance indirectly, through their effect on team processes and emergent states such as communication, coordination, psychological safety and reflexivity; and (P3) team processes and emergent states function as the proximal mechanism through which organizational and structural antecedents are translated into agile team performance outcomes, including productivity, software quality and team satisfaction. Agile team performance, in this view, cannot be explained by any single factor; it emerges from the alignment of organizational conditions, team structure and team processes, with team processes carrying the effects of the more distal antecedents.

This review’s primary theoretical contribution is the antecedent–process–outcome model developed in Section 3.3, which reframes team-level determinants as operating at different causal distances rather than as a flat list of correlates. Studies that treat autonomy, shared leadership or organizational culture as directly predicting performance, without accounting for the team processes through which they operate, risk misspecifying the underlying causal structure.

Beyond the model, the co-located origin of most team-process evidence suggests existing teamwork-quality and psychological-safety models may need re-examination in remote and hybrid contexts. Similarly, the measurement heterogeneity in Table 2 shows the field lacks a shared operationalization of performance, a first-order concern for theory-building, not a footnote.

The practical implications follow directly from the mediation model. Organizations should not treat autonomy, shared leadership or culture as ends in themselves; these conditions improve performance only when translated into stronger communication, coordination, psychological safety and reflexivity. Structural redesign alone, such as granting more autonomy, without attention to whether it improves team processes, is unlikely to produce the intended gains.

Second, organizations operating remote or hybrid agile teams should not assume co-located practices transfer directly; managers should monitor whether communication and psychological safety are sustained under distributed conditions. Third, organizations benchmarking performance across teams should be explicit about which outcome, productivity, quality, delivery speed or satisfaction, is being compared, given the field’s measurement inconsistency.

Despite the contributions of this review, several limitations should be acknowledged. First, the review focused on studies published in peer-reviewed journals and conference proceedings written in English. As a result, relevant studies published in other languages or grey literature may not have been included. Second, although multiple databases were searched, it is possible that some relevant studies were not captured due to variations in terminology used to describe agile teams and team performance.

Future research can extend this work in several directions. First, empirical studies are needed to test the mediating relationships proposed in this review, specifically whether organizational conditions and team structural characteristics influence performance indirectly through team processes and emergent states, rather than directly. Second, given that much of the underlying evidence on team processes originates from co-located settings, future research should examine whether communication, coordination and psychological safety operate with the same strength and function in remote and hybrid agile teams, building on emerging work in this area (Adzgauskaite et al., 2025; Khanna et al., 2024; de Souza Santos and Ralph, 2022). Third, future systematic reviews and meta-analyses should work toward a shared outcome taxonomy for agile team performance to address the measurement heterogeneity identified in Table 2, enabling more direct comparison of findings across studies. Finally, longitudinal studies may provide deeper insights into how agile team processes evolve over time and influence project outcomes.

This review synthesized existing research on the determinants of agile team performance. Based on 35 studies, it finds that team processes and emergent states function as the proximal mechanism through which organizational conditions and team structure translate into performance outcomes, rather than as a parallel category of determinants.

This model is the review’s central theoretical contribution. It also surfaces two further gaps: the co-located origin of the team-process evidence base and the inconsistent operationalization of performance across studies. Addressing these, alongside testing the proposed mediating relationships, is a promising agenda for future research.

For practice, the findings suggest that organizations seeking to improve agile team performance should not treat team autonomy, shared leadership and management support as ends in themselves but as conditions that must be translated into effective communication, coordination and psychologically safe team climates to shape performance outcomes.