本文以当前我国平台经济中最为活跃的领域——直播销售作为研究对象，分别从主播对消费者的劝说和销售转化、平台企业对主播的流量分配算法、以及不同平台企业之间的流量竞争三个层面来分析直播销售经济活动。主播能够通过信息传递的方式，将消费者的观看行为转化为实际的购买行为，这种由主播促成的消费者行为的转化构成了直播销售的底层逻辑。为此，本文构建了贝叶斯劝说博弈模型，并结合主播特有的网络效应特征，刻画了主播的经济作用。相应的，平台企业为直播销售提供了基础支撑。平台的算法能力会改变用户流量分配的结果，从而影响主播的流量需求；而平台的服务能力会改变用户的参与情况，从而影响平台的流量规模。结合该问题，本文构建了完全信息动态博弈模型，探究平台关键能力对流量分配收益、直播销售业绩的具体影响，以及对流量竞争的作用效果和影响机制。最后，本文运用相关平台的数据进行了计量检验，得到了扎实的研究结论。论文的主要发现有如下三点：第一，随着主播的网络效应增加，其采用最优信号结构对消费者的贝叶斯劝说效果和销售转化率会提高，由此形成了网络直播销售的贝叶斯劝说机制。第二，在算法中立原则下，平台算法能力的提升能够使得主播的转化率水平产生积极的溢出影响，从而增大各类主播的流量需求，并增加了主播的流量竞争程度。第三，以物流服务为代表的服务能力是不同网络直播平台之间的竞争手段，提高平台服务能力有助于增加平台的用户流量，进而改善平台的直播销售收益。上述发现科学的揭示了网络直播销售的经济逻辑。与现有文献相比，论文的主要创新之处有四点：第一，本文运用贝叶斯劝说理论解释了网络直播销售的经济学逻辑，凸显出主播关键的经济作用。第二，本文将平台的算法能力与流量分配相结合，揭示出算法能力和流量分配原则对主播流量竞争的作用机制。第三，本文基于平台服务能力构建了平台流量竞争模型，揭示出服务水平与平台流量竞争的内在逻辑。第四，本文还运用头部平台直播交易数据对主播劝说机制的作用效果进行了规范的检验，对平台算法能力提升与主播流量竞争的关系进行了差异化分析，为上述理论发现提供扎实的现实基础。

This paper focuses on the most active field in China's current platform economy——live streaming sales, and analyzes the economic activities of live streaming sales from three different levels: the persuasion and sales transformation of live streamer to consumers, the traffic allocation algorithm of platform to live streamers, and the traffic competition among different platforms.Live streamer can transform consumers' watching behavior into purchase behavior through information transmission, and the transformation of consumer behavior facilitated by live streamer constitutes the underlying logic of live broadcast sales. To this end, this paper constructs a Bayesian persuasion model, and describes the economic role of live streamer in combination with their unique network effects. Accordingly, platform provides basic support for live streaming sales. The algorithm capability of the platform will change the result of user traffic allocation, thus affecting the traffic demand of the live streamer. The service capability of the platform will change the participation of users, thus affecting the traffic scale of the platform.Combined with this problem, this paper constructs a dynamic game model with complete information to explore the specific impact of platform key capabilities on traffic distribution revenue and live streaming sales performance, as well as the effect and influence mechanism on traffic competition. Finally, this paper uses the data of relevant platform to conduct econometric tests and obtains solid research conclusions.The main findings of this paper are as follows: first, with the increase of the network effect of the live streamer, the persuasion effect and sales conversion rate of the optimal signal structure on consumers will increase, thus forming the Bayesian persuasion mechanism of live streaming sales. Second, under the principle of algorithm neutrality, the transformation rate of live streamer will have a positive spillover impact on the traffic demand of live streamer when the algorithmic capability of the platform is high. Third, the service capability represented by logistics service is the means of competition among different live streaming platforms. The above findings scientifically reveal the economic logic of live streaming sales.Compared with the existing literature, the main innovations of this paper are as follows: first, this paper uses Bayesian persuasion theory to explain the economic logic of live streaming sales, highlighting the key economic role of live streamers. Second, this paper combines the algorithmic capability of the platform with traffic allocation to reveal the mechanism of algorithmic capability and traffic allocation principle on the traffic competition between live streamers. Thirdly, this paper constructs a platform traffic competition model based on platform service capability, revealing the internal logic of service level and platform traffic competition. Fourthly, this paper also uses the live streaming transaction data of the head platform to conduct a standardized test on the effect of the live streamer’s persuasion mechanism, and makes a differentiated analysis on the relationship between the improvement of platform algorithm capability and the traffic competition between live streamers, providing a solid practical basis for the above theoretical findings.