Nonlinear Pricing of Information Goods信息商品的非线性定价.pptxVIP

Nonlinear Pricing of Information Goods Arun Sundararajan New York University, Leonard N. Stern School of Business Associate Professor of Information, Operations and Management Sciences Management Science, Vol. 50, No. 12 (Dec., 2004), pp. 1660-1673 1 描述故事，從provider角度給出合理的切入點說明provider的strategy應為何(formulation, figure) 有了transaction fee之後故事要重跑 2 Ground Truth Customer type: θ Customer distribution: F(θ) Customer Utility function: U(q, θ) Provider Pricing Scheme: - Usage-based menu: (q(θ), τ(θ)) - Fixed-fee: T Question: How does the provider set his parameters to maximize his revenue (or profit) 3 Assume that U(q, θ) and F(θ) are known to the provider Assume that the customers are rational and would choose the best selection Provider use some techniques to set the menu (q(θ), τ(θ)) and the fixed-fee T to maximize his profit 4 Monopoly firm Selling information goods Either a homogeneous good (ex. bandwidth) or a bundle of related heterogeneous quantity units (ex. a library of MP3 songs) Variable costs of production are zero That is, the provider could provide unlimited goods to customer (fixed-fee pricing scheme) While the spectrum is limited, the fixed-fee pricing scheme is inappropriate, and the usage-based pricing scheme should be under some constraints Better suits in the cloud service case Discuss more later 5 Customers are heterogeneous Indexed by θ in this paper Indexed by two parameters μ λ in previous work We can use this feature to simplify our work Note that we should examine the properties of the utility functions though 6 Customer type distribution F(θ) and their utility function are known Standard assumption in many works In our previous work, we didn’t mention the distribution F(θ) But by revelation principle, we can assume that each MVNO would truly report his type (want to adjust μ λ to θ) 7 Also, we use acceptance function A(b, p) and W2P and tW2P from the demand function rather than using the utility function 從[*]懷疑 U(q, θ) = ∫p(q, θ), where