Knowledge that Transforms

Open innovation and crowdsourcing have become standard tools for governments, foundations, and firms seeking solutions to hard problems. The implicit assumption is that more participation and more collaboration produce better outcomes. This research complicates that picture. Using agent-based simulation calibrated against one of the world’s largest crowdsourcing platforms, we show that spontaneous team formation during contests trades off parallel exploration for collaborative depth and that this tradeoff is not always favorable. For simple, well-defined problems, enabling teaming at the right moment improves the odds of finding the best solution. For complex, interdependent problems, precisely the kind that public innovation challenges tend to target, poorly timed or signal-driven teaming can suppress the diversity of search and reduce the likelihood of breakthrough outcomes. Policymakers designing grand challenge competitions, innovation prizes, or open problem-solving initiatives should attend carefully to the structural features of their platforms: what information is made visible, when teaming is permitted, and how incentives interact with collaboration norms. The design of the information environment is not neutral; it shapes who collaborates with whom, and ultimately whether the crowd delivers on its promise.

Many firms buy cloud services from cloud vendors, such as Amazon Web Services to serve end users. One of the key factors that affect the quality of cloud services is congestion. Congestion leads to a potential loss of end users, resulting in lower demand for cloud services. Although discount can stimulate demand, its effect under congestion is ambiguous; a higher discount leads to higher demand, but it can further lead to higher congestion, thereby lowering demand. We explore how congestion moderates both cloud vendor pricing and the buyer’s fulfillment decisions. We seek to answer how the congestion sensitivity of the end users and the cost of technology impact buyer profitability and the cloud vendor’s choice of discount. We also examine how the cost of technology determines the buyer’s willingness to pass on savings to end users. Our results show that the buyer is not necessarily worse off even when the end users are more intolerant to congestion. In fact, when end users are more congestion sensitive, the demand for cloud services can sometimes increase, and the discount offered by the vendor can decrease. We also observe that a lower cost of technology can sometimes hurt the buyer, and the buyer can pass on lower benefits to end users.