Towards Pricing for Fresh Data Acquisition Services

Data generated through interconnected devices facilitates timely decision-making, paving the way for fresh data acquisition services. In particular, a service provider uses limited resources to acquire fresh data for multiple clients, and each client pays for the service according to the price set by the provider. In this paper, we aim to understand two key challenges: 1) how often a client uses the service in the presence of acquisition time and price, and 2) how the provider prices the service based on limited resources, acquisition time, and incomplete information about the clients. We first analyze a single-client Stackelberg game with a finite time horizon. Our analysis shows that the client's optimal decision is to use the service at equal intervals whose length depends on the minimum acquisition time and price. Leveraging the equal-interval insight, we further analyze the provider's pricing from a multi-client Stackelberg game with an infinite time horizon. We devise efficient algorithms for the multi-client problem with complete and incomplete information settings. Moreover, numerical results reveal that the realized profit from the complete information setting can be much lower than the expected profit from the incomplete information setting. We, therefore, develop a method for the provider to make the trade-off between profit and the risk of profit loss. Our numerical results suggest that lower pricing to serve a smaller number of clients reduces the risk of profit loss.