Abstract: At present, user knowledge contribution of social Q&A community is gradually showing a “90-9-1” pyramid structure, and the centrality of nodes will increase It has a direct impact on the traffic support and exposure of long-tail users; how to improve the knowledge contribution of long-tail users, incubate and cultivate core users, and encourage them to spread high-quality paid knowledge is very important for community development. This paper analyzes the feedback loop between the knowledge contribution, reputation reward and group conversion coefficient within the segmented group through system dynamics, draw the causality diagram and the flow stock diagram, combine the expert scores to estimate the model parameters, explore various motivational factors and the impact on the conversion of sub-groups and the contribution of knowledge. The results show that the sub-groups all show a rapid growth trend. Compared with the rapid growth of core users, long-tail users show a more stable marginal increasing effect. After the waist users experience the initial cognitive lock-in of the free model, their paid knowledge contribution will show a steady growth trend.