Many IT organizations turn to agile software delivery approaches such as DevOps in order to reduce the number of IT projects that are running behind schedule and above budget. However, the DevOps paradigm calls for an increased set of capabilities that need to be built and aligned with their context in order to ensure superior team performance. This research aims to develop a context-capability coalignment profile for DevOps teams in public organizations. This profile and the corresponding design approach may serve as a model for other software production teams seeking to enhance their performance through improved coalignment. The resulting set of design principles places the traditional information systems theories of dynamic capabilities and contingency theory in a modern context. We adopt a longitudinal action design research approach centered around a DevOps team working in the IT department of a Dutch public organization. A mixed method design including scientific questionnaires, workshops, expert opinions and semi-structured interviews is employed to build and evaluate the profile. The resulting profile is characterized by technological complexity, a highly regulated environment, departmental interdependencies and high system relevance. The evaluation phase supports the validity of the artifact and suggests moderately improved coalignment of context and team capabilities after the research period, as well as a positive influence of coalignment on team performance. It is contended that software teams in public organizations can benefit from improved coalignment between context and DevOps capabilities by following the presented approach. We argue that it is important to create a profile which is internally consistent and views coalignment as a continuous process in order to maximize the positive effect on team performance.

Abstract This paper reports on the design and validation of a capability measurement instrument for software delivery teams that make use of the DevOps approach. The instrument is based on the results of a systematic literature review and was developed and validated by involving a total of five domain experts and conducting a field study among six DevOps team members. To this end, we used qualitative and survey-based data collection methods from participatory action research as well as design science. The resulting instrument encompasses five dimensions, covering seventeen capabilities and thirty-eight associated practices. The practices are evaluated on five capability levels. The results of the validation process indicate clear agreement of the domain experts and team members with all aspects of the instrument. As a contribution to practice, this research offers a pragmatic tool for IS practitioners which provides insight into the status of their DevOps transformation and offers directions for improving DevOps team performance. Furthermore, this research contributes to the ongoing research stream on DevOps by providing novel insights into the nature of DevOps capabilities and their potential configurations.