透過您的圖書館登入
IP:3.87.133.69
  • 學位論文

利用非動態資料庫之銀行顧客分群研究

Design of Customer Cluster Analysis Using Non-dynamic Customer Data of Bank

指導教授 : 黃崇興

摘要


大部分有關於顧客關係管理應用在產業資料庫的分析研究上,都是以動態的資料庫為對象,資料不僅豐富而且具有即時性,由顧客的消費交易資料中可以看出顧客的整體消費趨勢,具有一定程度的連續性,如銀行之信用卡刷卡中心資料庫。但對於一般傳統產業之資料庫,如一般商業銀行之顧客帳戶資料庫,由於傳統組織結構的關係與資料庫建置上的歷史限制,使得其對於顧客之動態資料難以取得,顧客交易資料相當有限,因此在進行相關顧客關係管理的研究上就很難產生重大的成果。 故本研究是以銀行業的非動態顧客交易資料為對象,嘗試以最典型的RFM模型中的三個指標為基礎,來重新定義出銀行中傳統交易型顧客的交易行為變數,分別為R(最近一個月之交易次數)、F(六個月之交易總次數)、M(平均餘額),而以此三個交易行為變數來做為顧客分群的基礎,進行統計集群分析。對資料庫中之顧客加以分群,並透過差異性檢定,以確定這樣的分群是否具有意義。在確定分群具有意義後,以各集群顧客在人口統計變項與交易行為變項上的表現,對各個顧客集群特性進一步加以命名與描述,並將分群結果提供給銀行,以使其得知哪些是對銀行有較高價值的顧客,哪些顧客是需要注意並加強服務的顧客,哪些顧客對銀行的價值是低到可以放棄的,哪些是已流失的顧客,並加以描述每一群顧客的交易行為,以提供給理財專員進行適當的行銷策略,且能提供適當的金融服務資訊給適當的顧客。 本研究對L銀行之非動態顧客資料進行分群,結果可將顧客分為七群,分別是:最有價值之顧客(最重要之顧客)、次重要之顧客、中低價值之顧客、一般顧客、即將流失之顧客、已流失之顧客與無價值之顧客,研究的具體成果為對各個集群之顧客提出行銷上的建議,以做為銀行理財專員在行銷上的參考,並提供給各群顧客適當的產品與服務。

並列摘要


Most research and analyses of the industry database conducted by the application of Customer Relationship Management are based on dynamic database, which is featured with affluent and instant materials. The aggregate consumption trend can be investigated by analyzing the customer’s transaction records, for example, the credit card transaction records. As regards the database of the traditional industry, such as commercial banks, it is difficult to acquire customers’ dynamic data due to the characteristics of the traditional organization structure and the restriction on the database construction. As a consequence, there are limited customer transaction records in this field and it is difficult to get great results in research related to CRM. This research regards non-dynamic customer transaction database of the banking industry as the investigation subject, trying to define the transactional behavior variables of traditional customers based on three indexes used in RFM model, i.e. Recency (transaction orders in the latest month), Frequency (the total transaction orders in six months) , Monetary (average balance). This research takes these three variables as the clustering factors to conduct cluster analysis in order to divide customers in the database into clusters. After the assessment of differential testing and being assured of the validity, we can define and describe each cluster according to its demographic variance and transaction behavior variance. The results would provide full information to bank practitioners. According to the result of this research, the customers in the Bank L’s non-dynamic customer database can be divided into seven clusters: the most valuable customers (the most important customers), second valuable customers, medium important customers, general customers, draining customers, customers already drained, and non-value customers. This article provides marketing implication and suggestions about each cluster to practitioners in order to help banking specialist to offer proper products and services to every cluster of customers.

並列關鍵字

cluster analysis CRM RFM model

參考文獻


16. Kalakota, Ravi and Marica Robinson, 1999. e-Business: Roadmap for Success, Addison Wesley Publishing Company.
13. Hughes, Arthur Middleton, 1996. Boosting Response With RFM, American Demographics:4-9.
14. Hughes, Arthur Middleton, 1994. Strategic Database Marketing, Chicago, IL: Probus Publishing Company.
5. Chen, L. D., T. Sakaguchi and M. N. Frolick, 2000. Data Mining Methods, Applications, and Tools, Information Systems Management, 17(1):65-70.
6. Davids, Meryl, 1999. How to Avoid the 10 Biggest Mistake in CRM, The Journal of Business Strategy, 20(6):22-26.

被引用紀錄


黃于真(2013)。運用統計與資料探勘方法進行顧客購買行為分析〔碩士論文,長榮大學〕。華藝線上圖書館。https://doi.org/10.6833%2fCJCU.2013.00046
王偉丞(2015)。從信用卡交易紀錄探勘消費者衝動性購買行為〔碩士論文,國立臺灣大學〕。華藝線上圖書館。https://doi.org/10.6342%2fNTU.2015.00060
劉向桓(2008)。市場分群模式與顧客價值模式研討─以電影市場為例〔碩士論文,國立臺灣大學〕。華藝線上圖書館。https://doi.org/10.6342%2fNTU.2008.10679

延伸閱讀