Research On Behavioral Interaction And Strategy Optimization Of Social Commerce Participants

As a new type of e-commerce model that embodies social value,social commerce(SC)has entered its heyday.It has been reflected in the unprecedented market scale,surging number of employees,continuous favorable policies,and SC layout of leading e-commerce platforms and enterprises.However,some issues have also emerged in the development of the SC market.For instance,the sharply increasing marketing cost of SC platforms and enterprises results in their compressed living space,notably for small and medium enterprises.It is difficult for them to acquire new users and maintain active users.In addition,there exist unhealthy phenomena such as false publicity,inferior product and misleading price,which not only damage the legitimate rights and interests of consumers,but also restrict the high-quality development of SC industry.These issues are closely related to the operational decision-making and governance of SC.As important parts of the digital ecosystem,consumers,enterprises,platforms and the government in SC are no longer isolated individuals,and the behaviors and decisions of each participant are interrelated and influenced.Based on the Identity Theory and following shifts of perspective of “consumer → enterprise → platform → government”,this paper explores the behaviors and decisions of each participant in SC in different complex environments,to provide management implications for solving practical issues in the development process of SC.The main research contents are shown as follows: Firstly,from the consumer perspective and combined with the social commerce value network(SCVN),this paper introduces the Graphical Evaluation and Review Technique(GERT)to construct the SCV-GERT model and investigates the impacts of various consumers and their interactions on the SC value flow.Secondly,from the enterprise perspective,this paper constructs decision-making models of e-commerce enterprises under the traditional e-commerce platform(TECP)and three kinds of SC marketing channels involving “e-commerce platform + social”(ECPS),“online community + business”(OCB)and “social media + business”(SMB).Then,this paper analyzes the selection strategies of e-commerce enterprises under multiple marketing channels.Thirdly,this paper focuses on the enterprise perspective and takes “forwarding and bargaining context” as an example,assuming a Stackelberg game model between an SC platform and an e-commerce enterprise on the basis of the traffic benefit and social difficulty,to investigate their strategy changes and profits in four scenarios.Finally,from the government perspective,this paper investigates the behaviors and interactions of multiple participants in the governance context of live streaming commerce(LSC),constructs interactive models under four different governance patterns,and uses Deep Reinforcement Learning(DRL)to explore optimal strategy changes of each participant.The main conclusions of this study are shown as follows:(1)Forwarding consumers with high social trust input contribute the most to the SCVN.(2)Consumer base size and product added-value significantly affect the marketing channel selection of e-commerce enterprises.SMB is the preferred marketing channel for most e-commerce enterprises,but SC marketing channels(ECPS,OCB and SMB)are not necessarily better than TECP.When e-commerce enterprises have low consumer base size and low product added-value,choosing TECP can bring maximum profits for enterprises.(3)Platforms in the early-mid stage tend to adopt a subsidy strategy to obtain massive user and traffic benefits.For enterprises operating highquality but low-added-value products,it is unnecessary to join a developed social commerce platform.(4)Platforms as direct beneficiaries can always gain more profits than can enterprises from increased traffic benefits,and platforms desire to reduce the forwarding cost,whereas enterprises favor maintaining a higher one.(5)The proposed method – Stack-DDP – is verified to be highly stable and accurate in fitting the optimal strategy of multi-agent Stackelberg games.Joint governance mechanisms of multiple participants,i.e.,adding a consumer reporting mechanism,platform supervision mechanism and government regulation mechanism to the basic governance model,can effectively curb the false propaganda and price-misleading behavior of broadcasters.The main innovations and theoretical contributions of this study are:(1)Considering the behavior of digital influencers under different SC business models,this paper proposes a nuanced division and definition of them – forwarding consumers,agent consumers and commenting consumers – and explores their differentiated contributions in the SCVN.(2)This study introduces GERT to the SC scenario and constructs the SCV-GERT model,which is based on the identity and behavior of consumers participating in SC and their interactions with their peers.With the SCV-GERT model,this paper explores the value contribution of various types of consumers with social resources as the value flow to the whole network,which enriches the empirical method of SCVN.(3)A more realistic Stackelberg game model of SC platforms and e-commerce enterprises in the context of forwarding and bargaining is constructed to study the effects of consumer forwarding cost,platform traffic benefits,and consumer base size on platform profit-sharing and subsidy strategies as well as enterprise pricing strategies.(4)In the governance context of LSC,this paper constructs a multi-agent interaction model and proposes the Stack-DDP algorithm to obtain equilibrium solutions,which is the first attempt to introduce reinforcement learning to this field.Stack-DDP demonstrated to be highly stable and accurate by evaluations,enriches the solving methods of the equilibrium strategy under the complex game of multiple participants in LSC.