Understanding passenger reaction to dynamic prices in ride-on-demand service

In recent years emerging ride-on-demand services (eg., Uber or Didi) are penetrating into the market of traditional taxi service. In these new services mobile devices are a key enabler: they serve as the intermediary between passengers/drivers and the service provider, tracking the locations and behavior of both passengers and drivers. On the other hand, the use of mobile devices also help us to capture huge amount of data for analysis. Through collaboration with a leading service provider in China, we collect vast amount of accurate data and analyze, in this paper, passenger reaction to dynamic prices in such a service. We consider the analysis as an important step towards making the service more efficient and more attractive to the passengers. We present the patterns of passengers' reaction, and discuss if it is useful to estimate the trip fare for multiple times in order to get a lower price. Our findings pave the way for future study on system optimization and policy considerations.