In the consumer-conscious time, marketing administrators do not face a homogeneous mass market, but rather a heterogeneous market. Relationship marketing which accentuates long run and stable relationship with customers and one-to-one marketing which considers heterogeneities between customers have become the mainstream. The goal of this thesis is to apply hierarchical Bayesian theory to the parameters estimation in a stochastic customer value model that can be utilized to estimate customers’ individual values. Our proposed model consists of the following steps: 1. Construct a stochastic model of customers’ lifetime values by combining the RFM and Markov Chain theories. 2. Apply hierarchical Bayesian theory to the derivation of parameters’ distribution in the above stochastic model. 3. Estimate model’s parameters by using the Markov Chain Monte-Carlo procedure. 4. Calculate customers’ individual lifetime value. The main contributions of this thesis are to apply hierarchical Bayesian theory to the derivation of parameters’ distribution and to estimate model’s parameters by using the Markov Chain Monte-Carlo procedure. In the empirical validation, the proposed method has been validated using an industrial computer company’s sales data. Results of our study show that our model can catch heterogeneity between customers and predict each customer’s purchase behavior very well.