Due to their complex nature, data warehouse (DWH) development projects often fail or significantly exceed budgets. Prior literature has revealed that achieving common understanding in the requirements elicitation phase is a key factor for the successful knowledge transfer and overall success of information systems (IS) development. To closer observe these phenomena, we conducted in-depth interviews with participants of several DWH projects. We propose the concept of 'brokering situations' to examine social interactions between project stakeholders and we analyze multiple brokering situations in reported DWH projects. Brokering situations help to understand the time-dependent development of brokers, the boundary objects, and their relations with regard to knowledge transfer. We augment Carlile’s (2004) classification of the boundary objects according to boundary complexity and development time. We discovered that in case of a low knowledge fit between project participants, involvement of a 'knowledgeable' broker seems important.
Available at: http://works.bepress.com/cristoph_rosenkranz/20/