Investigating Metric Management in Large-Scale Agile Software Development

Abstract

The world of software engineering has transformed rapidly in recent years. Two decades ago, the waterfall method with its plan-oriented approach was predominant. However, with the evolution of the Agile Manifesto, more adaptive methods have almost completely replaced the prevailing methodologies. To address the increasing complexity of software, scaling these methods has been a significant research focus in recent years.
With the increasing number and complexity of frameworks and best practices, the implementation of large-scale Agile software development faces some challenges, such as using scaling frameworks as cooking recipes.
Improving the maturity of an Agile program seems to be a solution to this difficulty. Researchers have already developed maturity models for Agile and large-scale Agile programs. Nevertheless, the focus in the field of large-scale Agile programs was mainly on benchmarking. According to maturity model research, a worthwhile goal is also to provide recommendations for improvement.

This thesis proposes a novel approach combining checklist-based documents with a maturity model applicable to all large-scale Agile programs regardless of the framework. Moreover, it enables benchmarking of a large-scale Agile program and provides recommendations for improvement.

To achieve this, the thesis examines the current state of research on maturity models in Agile environments. 
Then, a maturity model is designed using the Design Science Research approach and the latest research results on maturity model development. As part of the iterative approach, researchers and practitioners from six different organizations contribute to the designing process. The alternative assessment method combining checklists and maturity models is also presented. Furthermore, an Excel Spreadsheet and a web-based tool for conducting the assessment are presented. Accordingly, the artifacts are evaluated. Finally, the key findings are summarized, and an outlook on possible further research is given.

Research Questions

Research question 1 (RQ1): What is the state of the art in research related to maturity models and maturity models in Agile and large-scale Agile environments in particular?

Research question 2 (RQ2): How can a maturity model be designed for large-scale Agile software development?

Research question 3 (RQ3): How can a prototypical implementation support practitioners in the usage of the maturity model?