Open Innovation – a concept adopted from the private sector – was originally designed to open the organization’s boundaries to invite knowledge and ideas from outsiders into the R&D process to generate innovative products and services. In the public sector, open innovation as part of the Open Government Initiative represents a new policy instrument to conduct prizes and contests under the America Competes Act. Government agencies are using a central platform called Challenge.gov inviting problem solvers to submit solutions to problems government cannot solve on its own. This crowdsourcing process has been adopted by 55 agencies in the U.S. federal government who ran 204 online contests and prizes on the platform during the last three years on topics ranging from technology and defense solutions, to health or international affairs. Problems and submissions are publicly observable on the platform, however it is less obvious how the awarded crowdsourced innovations are used and implemented in government. Does the use of a new policy instrument and a digital platform increase the quality of innovations and are agencies willing to integrate the solutions into their own standard operating procedures? To answer these questions, qualitative interviews with program managers at the participating agencies were conducted. The semi-structured interview are part of a larger study looking at the strategic decision-making, managerial processes, day-to-day practices, as well as performance measures of linking prizes to outcomes in form of submitted innovations. The interviews were used in conjunction with online observations and coding of digital interactions on Challenge.gov. Each interview was transcribed and coded using the qualitative data analysis software NVivo. Initial themes were derived from the existing literature on open innovation, crowdsourcing, public sector innovations, performance measurement and additional themes emerged from the data. The conclusions discuss implications for both theory and practice.