近年來行銷思潮逐漸轉為一對一行銷,而其中最容易因廣告促銷而購買的客戶往往是俱有衝動性購買傾向高的顧客。因此本研究嘗試以某銀行信用卡資料庫為例,結合資料庫行銷與統計方法,以估計顧客之衝動性購買侵向以及期衝動性購買狀態之遷徙行為。首先,結合馬可夫鏈模型及層級貝氏模型以估計顧客之移轉機率矩陣,以預測顧客未來狀態,達到平均66.5%的擊中率。並設定6條路徑,以逐步回歸分析其遷徙路徑與顧客特質之間的關聯性。本研究之結果將可幫助企業了解其顧客之衝動性購買遷徙情形,並提供模型估計顧客未來之衝動性購買狀態,以利銀行規劃未來行銷策略及對顧客的價值做衡量。
This main idea of this study is trying to estimate the impulse buying attitude and their migration behavior with database marketing and statistical method, which is based on a domestic bank’s database of credit card customer. In the first, we established individual user’s Transition Probability Matrix with Markov Chain and Hierarchical Bayesian Model to predict the future states of them with an average hit rate 66.5%. Second, we set 6 meaningful migration paths and analyze the relation between the transition paths and demographic variable. Though this analysis, we could realize the impulse buying behavior and help the bank plan their marketing strategy and evaluate the value of their customer.