Based On The Principal-agent Crowdsourcing Innovation Incentive Mechanism Research

The thesis studies on incentive mechanism, operating mechanism and operating efficiency of Crowdsourcing innovation based on theories of innovation, economics and management.Crowdsourcing innovation is an emerging multi-agents innovation method which is promoted rapidly around the world. It refers to a company subcontracting tasks in a voluntary form to non-directed internet public(usually a huge number of) which used to be done by employees inside the company. These tasks can be to develop a new technology, to complete a new design, or to improve an algorithm for massive data analysis. Tasks of Crowdsourcing are usually done by individuals, but if need to collaborate to complete, the task may also rely on open-source form of production. Companies pick the most desired innovative work or solutions proposed by network public and offer this people money prize.With the development of the Internet, Internet-based innovation model, innovation processes are becoming more and more popular. Crowdsourcing Innovation is a kind of innovative ideas and innovative models based on the interaction functioned by a new web2.0, which present democracy, equality, participatory culture and collective intelligence, and it has the characteristics of innovation of network organization. Thus Crowdsourcing Innovation is accelerating the generation of ideas to change innovations and designs of traditional R&D enterprises, making this innovative model very attractive to many Fortune500companies and emerging start-ups. Crowdsourcing is making innovation to be a social behavior. Open innovation based on Crowdsourcing will likely become a major research and development business models in the future.The main techniques used in this article is principal-agent theory and related mathematical analysis tools, simulation modeling and social network theory and other methods as quantitative analysis and case study as qualitative analysis. This paper presents the operation mechanism of Crowdsourcing, and the efficiency of innovation operation impact by different incentives mechanism. The main contents and conclusions of this paper are as follows:Increasing competition among enterprises, rising research and development costs and diverse consumer demand force companies to promote demand-oriented open innovation model. On the other hand, the continuous development of network technology and the growing group of enthusiasts formed a great wisdom supply. Crowdsourcing innovation came into being and shows great charm, getting a number of firms. Crowdsourcing Innovation’s core competence is the collective intelligence:a huge number of "idea exchange", participants with a variety of ways of thinking with the function of the weak links are often able to solve problems business problem inventively. And with the Internet as a catalyst, this power of collective intelligence has been maximized and Crowdsourcing innovation has become a "borrow brains globally" model of innovation.The same Crowdsourcing innovation concept can produce many different Crowdsourcing innovation models. Enterprises should select the appropriate Crowdsourcing model based on the specific characteristics of problem, with Simple problem, we can use a single-pick Crowdsourcing. Such problems as large-scale programming which can be modular, we can use modular multi-pick Crowdsourcing. With problems such as participants adding Wikipedia tips or developing Android system, we can take open source form. With business activities based on pictures, video, music and other media library, we can take advantage of Crowdsourcing to create media libraries.Crowdsourcing organizers are rational, profit-oriented and attract more participants is desirable. Typical domestic and abroad influential Crowdsourcing sites are basically attract participants by money. So it is an indisputable fact that money incentive is very important. So a reasonable and efficient prize incentive structure is the basic to improving overall Crowdsourcing innovation operating efficiency. Study the prize structure is very important.By deduction of interaction index based on principal-agent model and calculation of incentive efficiency index, participation efficiency index and time efficiency based on simulation model, we finally get the consistent conclusion:In terms of Crowdsourcing incentive efficiency and operating efficiency, linear incentive is better than fixed bonus incentives and incentive with participant-interaction is better than that with no participant-interaction. Of course, not all tasks are suitable for participants to interact. the interaction mechanism should be established with the primary consideration of characteristics of the task itself.On the specific design of linear incentives, we provide inspiring ideas:Higher incentive index should be set if overall participants’ skill level on Crowdsourcing platforms is higher. And lower incentive index should be set if overall participants’ effort cost level on Crowdsourcing platforms is higher. So the organizers of Crowdsourcing innovation should spare no effort to collect data to understand participants on the Crowdsourcing platform and set up reasonable and efficient incentives on the basis of that.