Managing Dependencies in Large-Scale Agile

Abstract

Delivering results iteratively and frequently in large-scale agile requires efficient management of dependencies. We conducted semistructured interviews and virtual observations in a large-scale project during the Covid-19 pandemic to better understand large-scale dependency management. All employees in the case were working from home. During our data collection and analysis, we identified 22 coordination mechanisms. These mechanisms could be categorized as synchronization activities, boundary-spanning activities and artifacts, and coordinator roles. By using a dependency taxonomy, we analyzed how the mechanisms managed five different types of dependencies. We discuss three essential mechanisms for coordination in our case. First, setting Objectives and Key Results (OKRs) in regular workshops increased transparency and predictability across teams. Second, ad-hoc communication, mainly happening on Slack because of the distributed setting, was essential in managing dependencies. Third, the Product Owner was a coordinator role that managed both inter-team and intra-team dependencies.