Analysis of commercial documents and reports of all branches to identify possible abnormalities between the separate branches (e.g. turnover, clients, profits, and personnel). 2. Process and meeting (between site personnel and branch managers) observation in branches to assess the flow of information between construction site personnel and branch managers. 3. Expert interviews at manager level. The interviews will be semi-structured. This ensures a set of important questions are combined with the necessary flexibility to vary the order in which questions are asked and to add new questions in the context of the research situation (Bryman & Bell, 2011; Saunders, Lewis & Thornhill, 2012).

To achieve the objectives of this study, research depth is crucial. That means the researcher needs to know what is critical for business success of decentralised construction companies. Therefore, detailed insights into the studied organisation, internal processes and contexts are significant. As the management accounting system to be developed will be based on these findings, an inductive research approach, implying moving from specific observations to broader generalisation (Bryman & Bell, 2011), will be applied. To ensure in-depth insights, a qualitative research approach by using the case study method will be adopted (Bryman & Bell, 2011; Saunders, Lewis & Thornhill, 2012). This choice is also comparable to previous research in this field (e.g. Rockart, 1979; Boynton & Zmud, 1984; Jones, 1995). In this study, the role of the researcher is crucial. To create non-biased findings, it is necessary that the researcher obtains insights into the everyday workings of the studied organisation (Yin, 2014). To achieve that, the researcher works as a member of the organisation, but maintains detailed records of his experience (Ryan, Scapens & Theobald, 2002). The researcher depends on the organisation’s managers’ knowledge and experience, and thus he has to ensure adequate collaboration between the managers and him. To get into the role of a participant (Ryan, Scapens & Theobald, 2002), the researcher has chosen the company where he is employed (a German decentralised construction company consisting 20 branches and 900 employees) to develop a single case study design. The application of a multiple case study is not reasonable, as it would not provide the advantages connected with the researcher’s role as participant. There are also examples of previous research in this field using the single case study method (e.g. Bullen & Rockart, 1982; Boynton & Zmud, 1984).