Socio-intentional framework for agile methods tailoring

Agile is nowadays one of the most-used software development approaches. To gain the full benefit from adopting agile methods, software development teams can choose to adopt agile methods on a custom-basis depending on the context and defined criteria. Over the years, a vast amount of empirical studies aiming to share adoption experiences and introduce the possibility to customize agile methods has been published. Two key aspects in these studies are the goal- orientation and the social aspect. Intuitively, a team adopts agile methods because there are things it wants to achieve. Once the team identifies the right practices to achieve its goals, the adoption result highly depends on the social factors, including the individuals, their interactions and collaborations. In this thesis, we propose a socio-intentional framework for tailoring agile methods that allows practitioners to analyze agile practices and to define the right strategies for adopting them. The objective of our framework is to help practitioners decide which agile practices to adopt based on what they want to achieve and identify beforehand how team members should work together to successfully adopt them. To achieve this objective, we proceeded through four main contributions. In the first contribution, we conducted a Systematic Literature Review (SLR) to extract the motivation behind each agile method’s adoption and compared it with the precepts given in the Agile Manifesto. The results show that the Agile Manifesto is highly relevant to these motivations. We conclude that the Agile Manifesto can be used as a criterion for practice selection. In the second contribution, we made knowledge about agile practices adoption reusable in a systematic manner by building an ontology. We started by conducting another SLR to gather knowledge. Based on this, we have built and theoretically validated our ontology. Finally, we conducted a survey with agile experts for empirical validation. In the third contribution, we have built a user-friendly tool that allows practitioners to get the information from the ontology easily. In the last contribution, we proposed a methodology to analyze agile practices on two different levels. The tactical level allows practitioners to decide what agile practice they should adopt based on what they want to achieve and to check the suitability of the selected practice(s) to the team’s situations. The operational level allows practitioners to identify beforehand the vulnerabilities in a practice adoption and define how to avoid or solve them. For each level, we showed how to get the knowledge from our tool and to use modeling techniques to ease the analyzing process. Finally, we showed how our framework can be applied within a real software development team.