With the vigorous development of platform economy, User-Generated-Content (UGC) video platforms have attracted a large number of users with their unique competitive advantages. The core logic of the business model of UGC video platform is Site-traffic-to-cash Conversion. Different from traditional video platforms, UGC video platforms do not directly purchase or participate in the production of videos. Instead, they encourage each user to become a video producer, and the platform conducts traffic-driving of the produced videos so as to obtain high video traffic. Based on this traffic, the platform can carry out commercial promotion to obtain advertising revenue, and the content creator can obtain a certain percentage of revenue based on the video views. This revenue becomes an important motivation for producers to create videos, which prompts the content quality, forming a virtuous cycle of traffic monetizing.In practice, the main traffic-driving modes of the platform can be summarized into two types: one is the Content-Traffic-Driving (CTD) mode that directly acts on the video content, and the other is the Fan-Traffic-Driving (FTD) mode that acts on the producer. Different from the CTD that uses technical means to recommend video as much as possible, the FTD refers to a mode in which the platform uses its own resource advantages to help producers expand the audience group and make use of the high willingness of fans to achieve the purpose of high content playback flow.The choice between different modes reflects the different ways in which the platform allocates the flow resources, and also determines the advantages and disadvantages of the flow realization effect. Therefore, this paper studies the issue of how to choose the flow realization strategy from the media platform from the following three aspects: First, how can the platform rationally allocate its own public domain flow resources to maximize revenue? Secondly, how do the two different drainage modes affect the profit of the platform and the content creator and the degree of participation of both parties? Finally, how to effectively coordinate the conflicts in the drainage process? In the first part of this paper, the above research background and the main research issues are introduced in detail, and the related literature is reviewed.In the second part of this paper, we construct a differential game model to describe the platform-dominated drainage behavior. The benchmark model includes two parts: Content-Traffic-Driving (CTD) mode and the Fan-Traffic-Driving (FTD) mode. The essential difference between these two modes lies in the different allocation of platform traffic resources. In the CTD, the platform puts resources into the promotion of content, and directly converts part of the Public-domain traffic into the content viewership; In the FTD, the platform invests resources to increase the number of fans of creators, and some Public-domain traffic is converted into Private-domain traffic of creators. In the game, the platform needs to select an appropriate traffic resource allocation model and decide the amount of public domain traffic resources invested; The content creator is responsible for producing the specific content and providing the carrier for realizing the traffic. Finally, this section obtains the equilibrium effort between the platform and the content creators.In the third part of this paper, we compare the profit and equilibrium of the two modes. It is found that when the traffic realization ability is high (low), participants from both sides simultaneously choose the FTD (CTD) mode; Otherwise, there will be a “Conflict Region” in which both parties cannot reach an agreement on the mode selection. There are two main reasons for this conflict: the lack of incentives in revenue sharing contracts and the asymmetry of decision-making between participants. In addition, the content creators′ decision preference of flow realization model is not related to their efforts. That is, although the content creators will get higher returns from their dominant strategies, they have no motivation to improve their efforts. The analysis shows that the improvement of profit comes from the improvement of platform effort, and the creators tend to “hitchhike”.In order to solve the above two problems, a subsidy contract model based on the number of fans is proposed in the fourth part of this paper. The results show that this Fan-number Contract effectively improves the creator′s efforts and brings Pareto improvement to both participants. In addition, the introduction of this contract lowers the threshold of FTD, improves the applicability of CTD, and reduces the possibility of conflict between participants to some extent.In the last part of this paper, we summarized the conclusions and gave some interesting management opinions: First, UGC platform needs to pay more attention to the traffic realization ability of content creators, and provide appropriate traffic realization arrangements for them according to their traffic realization ability. Secondly, our conclusion implies that the platform needs to give weight to those “promising creators” who are not top-notch in ability but have high potential, and cooperation with them can achieve higher traffic-driving efficiency. Finally, the Fan-number contract can be used as a lever to motivate creators and ease strategic conflicts. However, the platform needs to set a moderate subsidy rate to avoid the problems of excessive and insufficient incentives.