As a result of mobile internet development, online social networks have become the main carrier of product diffusion, but it is difficult for enterprises to predict the impact of network dynamics on user decision-making. In reality, most networks are dynamic, which leads to fluctuations in consumer group decision-making and increases the difficulty of enterprises making the right product diffusion decisions. Moreover, in a rapidly changing network environment, consumers can exhibit bounded rational behaviors, such as impulse buying or exhibiting psychological resistance in decision making, which bring great uncertainty to the product diffusion process. Therefore, in a rapidly changing market environment, it is highly important that enterprises explore the influence of dynamic network structure evolution and users′ bounded behavior on product diffusion for long-term development and sustained competitive growth. Based on this, the overreaction theory is introduced to describe the bounded rational behavior of consumers, and the evolution mechanism of the network topology structure is constructed through a combination of the following constructs: key opinion leader (KOL) and the key opinion consumer (KOC). By establishing a multiagent simulation model and analyzing the experimental results, the influence of network dynamics and bounded user rationality on product diffusion is obtained.The first part of the paper presents the main model, which is a product diffusion model based on overreaction in a dynamic network. First, relying on complex networks and combined with realistic marketing promotion methods, three network topology evolution mechanisms are proposed, namely, the random network evolution mechanism, degree priority network evolution mechanism and community network evolution mechanism, which include corresponding node entry, connection, disconnection and exit submechanisms, respectively, for accurately describing the dynamic evolution of network topology structures in reality. Next, as users are prone to producing bounded rational behaviors in highly dynamic online social networks, the overreaction theory is introduced on the basis of evolutionary game theory to construct a bounded rational threshold decision model for users and to describe bounded rational user behaviors, such as impulsive consumption and resistance to consumption.In the second part, the agent-based simulation model is designed on the basis of the above model, and simulation software is used to model the multiagent simulation. After the simulation model is established, the next step involves setting the parameters and verifying the model. First, the setting range of the game parameters is explored through a multigroup parameter experiment to determine the game parameters. The effectiveness of the model is subsequently verified through three methods: internal validity, cross-model validation and real case data validation. The model proposed in this study can be used for the evolutionary game analysis of product diffusion in dynamic networks.In the third part, the influence of user bounded rationality and network evolution mechanisms on product diffusion under static and dynamic networks are explored and a series of conclusions are drawn. First, overreaction has a negative effect on user aggregation and a positive effect on the reduction of the average distance between users in the network. Second, an increase in overreaction can promote product diffusion while simultaneously reducing the stability of product diffusion. Third, node entry speed has a positive impact on the number of product adopters, but if the node entry speed is too high, market volatility increases. Fourth, the number of adopters shows a marginally decreasing growth trend with the acceleration of connection speed, but under the degree-based network evolution mechanism, the number of adopters shows negative growth. Fifth, disconnection speed has a negative effect on product diffusion, and this negative effect is more significant under the social network evolution mechanism.The last part of the paper is the conclusion, in which the model and the results analysis are summarized and corresponding managerial insights are offered. Companies can guide users to engage in moderate irrational shopping behaviors through live streaming and other means, but product quality must be ensured. Companies should avoid excessive product promotion, particularly KOL promotion, and should target different user groups through precise marketing. Companies can keep users active by posting regular discussion topics and engaging in interactive activities to prevent product sales from decreasing, especially under KOC promotion.In summary, the modeling ideas and simulation system proposed in this paper consider the bounded behavior of users and rely on an analysis of the influence of various promotion modes on product diffusion in a dynamic network. The effectiveness of the model is verified by classic cases, which can guide enterprises in formulating product promotion strategies.