Assessing customer involvement in continuous software engineering: Development and validation of a reference model-based approach

Software development is a fast-growing tech sector. Due to continuous changes faced by organizations such as, changing user needs, new technologies, globalization and regulations, new software development methods have evolved into continuous
software development.

Continuous software development extends Agile software development by shifting from cyclic to continuous value delivery. To enable continuous software development, the concept of DevOps was introduced. DevOps is operationalized by Continuous Integration, Continuous Deployment and Continuous Delivery processes (CI/CD-processes) based on small increments providing continuous value to the customer.

Since CI/CD-processes provide value to the customer on a continuous basis they need constant customer feedback. Thus, the active involvement of the customer is important. However, prior research reports challenges regarding customer involvement. Examples of challenges are: establishing and applying the right feedback methods and utilizing obtained data, lack of organizational support in order to improve CI/CD-processes, and difficulties in accessing users of the software product. To solve these challenges, measures are suggested such as, apply approaches to convince the customer through blogs, workshops and demonstrations, and establish a culture of open communication among stakeholders.

Nevertheless, literature offers restricted information on how customer involvement can be actively stimulated. Furthermore, literature does not clarify how the extent customer involvement is stimulated can be assessed. To address these problems the
main research objective of this dissertation is the development and validation of a Reference Model of Measures to Actively Stimulate Customer Involvement (RM-MASCI) and a Reference Model-based approach (RM-approach) to assess the extent
customer involvement is stimulated in CI/CD-processes in practice.

This dissertation follows a research strategy based on design science. The dissertation presents a four-step research process.

In the first step, we develop and validate a well-founded and structured model of Critical Success Factors (CSFs) of collaboration meaningful in a DevOps context. While the results obtained are found to be useful, they are also considered to be too abstract to identify the necessary measures and develop RM-MASCI. Therefore, we shift our focus to the identification and validation of CSFs of CI/CD-processes in the next step, since the concept of DevOps is operationalized by CI/CD-processes.

In the second step, we develop and validate a well-founded and structured model of CSFs of CI/CD-processes meaningful for the CI/CD-profession. This model is developed by conducting a systematic literature review and subsequently a classification. The resulting model of CSFs is validated through a multiple case study. We used the real-life experiences from experts in practice to obtain the explanations corresponding to the CSFs.

Based on this result, RM-MASCI is developed in the third step. The model is developed by conducting a systematic literature review and subsequently a classification. The resulting model of measures is validated through a multiple case study. We used the
real-life experiences from experts in practice to obtain explanations and used them to develop RM-MASCI.

In the fourth step, we refine and validate RM-MASCI and develop and validate an RM-approach. The RM-approach encompass the following three concrete concepts: the assessment process, a refined RM-MASCI and the addition of a moderator. We applied
and validated the RM-approach through a multiple case study aiming to validate the refined RM-MASCI.

The results of this dissertation contribute to literature by presenting measures to actively stimulate customer involvement in CI/CD-processes. Furthermore, the applicability of RM-MASCI is confirmed through a validated reference model-based
approach to assess the extent customer involvement is actively stimulated in CI/CD-processes.

The RM-approach, including RM-MASCI, can be used as a tool or instrument to periodically assess the extent customers are involved in the CI/CD-processes. Identified areas of improvement provide an opportunity to improve customer involvement.