Owing to the requisite costs and labor for data collection and analysis, the purchase and exchange of thirdparty data in the market has become a viable option for many institutions. However, the debate on the dynamic interactions among various types of data, regulations, and market participants remains unresolved. Thus, this study examined the impacts of buyer behaviors and market size on data distribution and pricing dynamics in data marketplaces. This study developed a model on data provided by marketplaces through a random walk on the co-occurrence networks of variables to reflect the characteristics of real-world data networks. Buyer agents with three purchasing strategies (Random, Related, and Ranking) were simulated across six scenarios by varying the number of datasets and buyers. The experimental results demonstrated that the proposed model generated data networks that reflected the principal characteristics of actual data marketplaces. Markets, wherein buyers acquired data related to their existing data, yielded increasingly prevalent datasets with more variables, resulting in marketplace imbalances. The Random strategy resulted in the highest utility for buyers across all scenarios, whereas the Related strategy led to the highest prices for popular datasets. As the number of buyers increased relative to the available datasets, disparities in the distribution based on the strategy became more pronounced. The strategic adjustment of buyer approaches in accordance with the market size could mitigate the emergence of disproportionately popular data and promote price stability. Thus, this study provides insights for data marketplace operators and institutional design through the modeling of realistic marketplace dynamics and an analysis of the outcomes across different market conditions.