Application Development with Open Data in Hackathon

Abstract

Organizations seek innovation not only by themselves but also through open innovation activities. This exploration is manifested in various activities such as innovation contests, design competitions and innovation communities. In this context, one of the key activities that promote open innovation is the “hackathon.” The hackathon is an event in which voluntary members, such as software engineers and designers, are gathered and develop useful applications in a short time. This new type of collaboration has received much attention as a means of promoting open innovation. Previous studies have addressed the motivation of participants, the demographics of participants and prize structures for open innovation activities. However, they cannot explain how we should design the process of activities. Organizers of hackathons face high uncertainty regarding the outcome of these events. It is necessary for managers and policy makers to identify the key factors that determine the performance of ad-hoc teams in hackathons. The purpose of this study was to derive specific design principles that can improve team performance in hackathons. Here, we demonstrate that data-driven application development in hackathons is a more effective process design than the needs-driven application development that has typically been implemented in recent hackathons. To test our hypotheses, we held two hackathons with open data resources, conducted surveys and statistical methods (regression analysis and t-test). We find that data-driven application development can create products by combining existing services, technologies and data. In addition, the output of data-driven application development can be evaluated as a highly developed application. This result indicates that data are a key factor in promoting innovation via new combinations of existing resources, knowledge and methods. Although the process design of hackathons was analyzed in this study, the discussion presented can be understood as engaging the traditional problem of collaboration design and should thus aid in further research.