Stochastic Modeling of Customer Value Analysis

隨著資訊科技的快速進步以及相關演算法的發 展，現今顧客終生價值已經可以得到比以往更好的估計 結果。本研究的主要目的是提出一個能夠估計顧客終生 價值的系統化模型而且進行實際模型的驗證。本研究系統化模型包含下列步驟：1. 建構顧客購買行為之隨機模型以及提出顧客購買頻率及金額之機率分配假設。2. 利用馬可夫鏈描述顧客購買行為的改變，並且根據RFM(recency-frequency -monetary)模型定義顧客之購買狀態。3. 根據貝氏定理推導顧客購買狀態之移轉機率。4. 建立顧客之移轉矩陣及利潤矩陣，並且計算顧客之終生價值。本研究亦利用某工業用電腦廠商之實際銷售資料進行本模型之驗證，得到以下結論：1. 本模型之估計結果的確優於其他業界常用之顧客 價值分析方法，並能得到一個非常準確的顧客價值之估計。2. 我們所提出的模型不但可以捕捉顧客的異質性， 並且可以有效預測每個顧客的購買行為。With the rapid progress of information technology and development of related algorithms, customers’ lifetime value can now be better estimated. The goal of this research is to propose a systematic method to estimate customers’ lifetime value and conduct empirical validation. Our proposed model consists of the following steps: 1. Construct a stochastic model of customers’ purchase behavior and propose the underlying probability distributions of customers’ purchase frequency and monetary. 2. Model the customer’ purchase behavior as a Markov chain and the associated transition states based on RFM (recency frequency monetary) model. 3. Use Bayesian theorem to derive transition state probability. 4. Construct customers’ transition and reward matrices and calculate customers’ lifetime value. The proposed method has been validated using an industrial computer company’s sales data. Results show that: 1. Our model outperforms the current empirical estimated practices and achieves a very accurate estimation of customers’ value. 2. The proposed method is able to model customers’ heterogeneity and predict each customer’s purchase behavior