Agile scaling: Dealing with the unexpected

Agile working methods are regarded as a promising approach for industrial development processes. For example, small teams independently handle sub-projects, enabling greater flexibility and efficiency. However, this can lead to unforeseen conflicts and complications that disrupt or delay the overall project. A joint study by the University of Stuttgart and the RPTU University Kaiserslautern-Landau shows that work processes should not be planned solely from the management level—instead, employees should be empowered to actively help shape them.

Small teams that work on sub-projects of a joint large-scale project in parallel and autonomously of each other within a company are called agile teams. The agile working methods they apply are seen as a beacon of hope for companies: instead of rigidly adhering to a fixed plan, projects are broken down into small, manageable steps that are regularly reviewed and adjusted. “This enables teams to react better and faster to new customer needs and market changes,” says Dr. Christian Mahringer, postdoctoral researcher at the University of Stuttgart School of Management.

However, although many organizations make great efforts to implement and scale appropriate methods, they often reach their limits in practice, for example, because there are complications in the coordination processes between teams. A new study by Dr. Christian Mahringer and Associate Professor Dr. Anja Danner-Schröder, which appears in the Academy of Management Journal—one of the most renowned journals in business administration—gets to the bottom of this problem and provides surprising findings.

Drawing on a dataset obtained through observation of software development teams, Mahringer and Danner-Schröder analyzed how the teams coordinated with each other. They found that when companies try to force agile teams into a fixed design, unforeseen complications and conflicts can arise. A practical example illustrates this issue. Two teams planned to share a key team member to facilitate more effective knowledge transfer. What initially appeared to be an efficient solution unexpectedly resulted in significant coordination challenges. This is because the team member was unable to fulfill his role in both teams due to scheduling conflicts. Unforeseen dependencies and delays emerged that could not have been anticipated in advance.

Mahringer and Danner-Schröder concluded: “The design of a development process cannot be dictated from above. Rather, agile methods thrive on active design. Instead of enforcing rigid processes, managers should create environments that enable team members to respond flexibly, make informed decisions, and organize themselves,” says Mahringer. The teams should therefore be granted a high degree of autonomy and personal responsibility. Employees at lower hierarchical levels should also be empowered to contribute to ongoing design processes. “It's not enough to implement a framework and assume that it will work automatically,” emphasizes Danner-Schröder, who teaches Management Studies in the Department of Economics at RPTU. She continues: “The real challenge of an ongoing design process is to find your way through a constantly changing web of dependencies.”

Mahringer and Danner-Schröder identified four key competencies that influence the success of agile structures within organizations. One of them involves teams first analyzing dependencies to clearly understand how activities and resources are interconnected. Resources should be reconfigured as a further competence: If something is shared, it must work for both teams. Disruptions should be cushioned, for example by shifting priorities or providing mutual assistance. Synchronizing schedules is also essential. "Sprints, releases, and meetings should be carefully coordinated," says Danner-Schröder. While software development already offers organizational structures—such as frameworks designed to align multiple teams—these often reach their limits when faced with unforeseen scenarios that arise during ongoing work. "Companies invest heavily in implementing agile frameworks like SAFe or Scrum@Scale, hoping they can be used to manage complex development processes," adds Mahringer. “But our research shows: These approaches often fail to deliver the promised agility, because they start from the wrong basic assumptions.” Danner-Schröder adds: “The success of agility on a large scale does not depend on how well teams follow a plan. It's about how well they deal with things that can't be planned.”

The study thus provides valuable impetus for companies that want to successfully anchor agile working methods not only at team level, but throughout the organization.

Publication
Christian A. Mahringer and Anja Danner-Schröder, 0: Autonomous, Yet Interdependent: Designing Interfaces across Routine Clusters. AMJ, 0, https://doi.org/10.5465/amj.2022.0853

In our “Publications compact” series, you will find further reports that make research findings from Stuttgart publications tangible - clear, understandable and accessible to all.