Effectively collecting real-time data is a fundamental problem in IoT (Internet of Things) searches. In the IoT, most data are linked with the owner's private information and cannot be publicly released on the Internet. This invalidates the use of crawlers to collect data in IoT searches. As a result, effectively motivating potential data providers (PDPs) to provide real-time on demand data becomes a key requirement for the development of an IoT search service. To address this problem, we acknowledge the realistic assumption of incomplete information, and propose a buyout-auction framework, with the constraint of QoD (Quality of Data), to collect real-time data and maximize bidders' payoff. Simulation results demonstrate that our approach can drive PDPs to participate in bidding in a timely manner and provide data under the constraints of QoD to IoT search service providers.