資訊管理學報
張心馨;蔡獻富;
頁: 161-203
日期: 2004/10
摘要: 資料庫與網路科技興起，加速資料採礦技術於顧客關係管理的應用，企業評估績效的指標由市場佔有率變成顧客佔有率。然企業資源有限，且開發新客戶的成本是舊顧客五倍以上，因而必須辨識出顧客價值與企業間的關係，以保留高價值的顧客。基於大型量販店市場快速興起，並累積大量的顧客與歷史交易資料，但目前仍未被廣泛地應用。本研究擬完成：(1)以大型零售量販店為研究對象，修正傳統RFM顧客分群的方式，採用SOM與K-Mean兩階段分群方法，運用資料採礦技術有效萃取不同顧客群體的區隔。(2)分群變數之策略性區隔的LRFM資料模型，並以顧客價值矩陣和關係類型矩陣所構成的多維度顧客分群規則來解釋結果。(3)結合實際個案的資料，剖析資料採礦技術和顧客獲利性的影響因子，並深入了解不同顧客群體活動績效表現。經由個案研究及資料分析結果驗證，當活躍顧客與企業間的交易關係長度維持越長久，顧客獲利性越高；活躍顧客群中忠誠顧客和顧客佔有率越高，對企業的利潤貢獻較高。以沉寂長度做為衡量的流失顧客的指標，流失顧客對企業的利潤貢獻度不一定較低。具較高顧客獲利性的活躍顧客，對於活動的績效表現越佳；但在連續性的活動中，活躍顧客的顧客獲利性與活動績效表現間相關性並不一致。
關鍵字: 顧客關係管理;顧客區隔;資料採礦;自我組織;顧客獲利性;
Abstract: Database, Data Mining, and Internet have been hailed as a significant conceptual advance in the evolution of marketing, which nevertheless remains based on the Analysis of Consumer Behavior. Accompanying this customer-centric approach, there has been a shift from market share to profitability as the preferred business performance index. The case study and data analysis reported here was motivated by: (1) the availability of large customer activity datasets derived from shopping center transaction records. (2) The requirement to utilize this data to help optimize the use of limited business resources. (3) The availability of Database and Internet based technologies, which have been applied to CRM and Data Mining, but rarely in the context of large shopping centers. This study employed traditional RFM segmentation to extract shopping center customer clustering patterns, and analyzed case study data for correlations between various factors and profitability. The sample used in this study is taken from a single shopping center in the corporate retail sector. The research design used two stages clustering to derive Active Customer Segmentation based on an LRFM data model itself derived from the Strategic Segmentation technique of Sung and Sang. The results were incorporated into a CRM Model using multi-dimensional customer clustering, combined Customer Value Matrix, and a customer relationship type matrix. The results indicate that longer-term active business relationships between customer and corporation are associated with higher profitability.
Keywords: Customer Relationship Management CRM;Customer Segmentation;Data Mining;Self of Organization Map SOM;Profitability;
瀏覽次數: 15528 下載次數: 110
引用 導入Endnote
頁: 161-203
日期: 2004/10
摘要: 資料庫與網路科技興起，加速資料採礦技術於顧客關係管理的應用，企業評估績效的指標由市場佔有率變成顧客佔有率。然企業資源有限，且開發新客戶的成本是舊顧客五倍以上，因而必須辨識出顧客價值與企業間的關係，以保留高價值的顧客。基於大型量販店市場快速興起，並累積大量的顧客與歷史交易資料，但目前仍未被廣泛地應用。本研究擬完成：(1)以大型零售量販店為研究對象，修正傳統RFM顧客分群的方式，採用SOM與K-Mean兩階段分群方法，運用資料採礦技術有效萃取不同顧客群體的區隔。(2)分群變數之策略性區隔的LRFM資料模型，並以顧客價值矩陣和關係類型矩陣所構成的多維度顧客分群規則來解釋結果。(3)結合實際個案的資料，剖析資料採礦技術和顧客獲利性的影響因子，並深入了解不同顧客群體活動績效表現。經由個案研究及資料分析結果驗證，當活躍顧客與企業間的交易關係長度維持越長久，顧客獲利性越高；活躍顧客群中忠誠顧客和顧客佔有率越高，對企業的利潤貢獻較高。以沉寂長度做為衡量的流失顧客的指標，流失顧客對企業的利潤貢獻度不一定較低。具較高顧客獲利性的活躍顧客，對於活動的績效表現越佳；但在連續性的活動中，活躍顧客的顧客獲利性與活動績效表現間相關性並不一致。
關鍵字: 顧客關係管理;顧客區隔;資料採礦;自我組織;顧客獲利性;
Integrating of SOM and K-Mean in Data Mining Clustering: An Empirical Study of CRM and Profitability Evaluation
Abstract: Database, Data Mining, and Internet have been hailed as a significant conceptual advance in the evolution of marketing, which nevertheless remains based on the Analysis of Consumer Behavior. Accompanying this customer-centric approach, there has been a shift from market share to profitability as the preferred business performance index. The case study and data analysis reported here was motivated by: (1) the availability of large customer activity datasets derived from shopping center transaction records. (2) The requirement to utilize this data to help optimize the use of limited business resources. (3) The availability of Database and Internet based technologies, which have been applied to CRM and Data Mining, but rarely in the context of large shopping centers. This study employed traditional RFM segmentation to extract shopping center customer clustering patterns, and analyzed case study data for correlations between various factors and profitability. The sample used in this study is taken from a single shopping center in the corporate retail sector. The research design used two stages clustering to derive Active Customer Segmentation based on an LRFM data model itself derived from the Strategic Segmentation technique of Sung and Sang. The results were incorporated into a CRM Model using multi-dimensional customer clustering, combined Customer Value Matrix, and a customer relationship type matrix. The results indicate that longer-term active business relationships between customer and corporation are associated with higher profitability.
Keywords: Customer Relationship Management CRM;Customer Segmentation;Data Mining;Self of Organization Map SOM;Profitability;
瀏覽次數: 15528 下載次數: 110
引用 導入Endnote