Bayesian pricing strategy for information goods

This paper studies Bayesian pricing strategy for subscription-based information goods with uncertain demand. Considering the changing environmental issues, financial risks, geopolitical instability and other uncertainties, it is almost impossible to perfectly estimate the demand for new products, especially for information goods with a short life cycle. By exploring the demand pattern of most information goods over their whole life, we assume the phase-out time of information goods is Weibull distributed with the parameter estimated in a prior distribution. The solution to the problem involves a dynamic programming formulation. We present a Monte Carlo-based approach to solving the problem, where the posterior prediction of the parameters in Weibull distribution can be derived from the observed demand in the previous periods. By numerical experiments, we find the positive association between the total expected revenue and the renewal rate, while a negative association between the total expected revenue and the price sensitivity parameter.