This paper examines dynamic interactions among online travel agencies (OTAs) and hotel chains. The relationship between OTAs and hotel chains is multifaceted: cooperators and competitors. More importantly, the multifaceted relationship may change over time. In this study, we propose an approach which can incorporate the dynamic interactions between OTAs and hotel chains using a dynamic mutually correcting marked point process model. We also account for consumer heterogeneity by incorporating individual random effects. We apply our approach to comScore click stream data which contains individual-level website visit and online hotel booking information. We find strong empirical evidence of dynamic interactions among OTAs and hotel chains. We also find that key managerial metrics such as demand predictions and transition probabilities can be distorted, if we disregard the dynamic aspect of the relationship between OTAs and hotel chains. Finally, we provide the potential of this approach for better understanding the competition structure of hospitality industry and therefore making more reliable polices such as the optimal amount of commission fees to OTAs and the approval for M&A between large OTAs.