Entrance tickets for cultural venues are often sold through online retailers. Internally, these retailers have access to demand insights of a large amount of globally distributed venues. Using Tiqets as our industry partner, we were able to access data from one of the largest retailers with global reach. This thesis explores the possibility of leveraging prices to adjust the demand for a venue. In finding the relation between price and demand from historical sales records we encounter several obstacles. This thesis explores three different ways of minimizing the negative effects of using observational data. By introducing a decomposable time-series regression model to predict demand we can isolate the effects of trend, seasonality and recurring events without having to introduce new data. Introducing price as a module enables us to control for the desired future demand. Secondly, we explore exogeneity in the variation observed from historical price changes and introduce ways of ensuring exogeneity on future data. Finally, we look for an automated way of finding similarity across venues using the associated historical sales data and use this to estimate the variation in error when grouping venues in a controlled experiment.