基于数据挖掘的电信企业客户行为分析研究

【作者】 全昌文；

【导师】 邵培基；

【作者基本信息】 电子科技大学 ， 管理科学与工程， 2005， 硕士

【摘要】 我国电信企业改革和重组后,电信运营业面临新的竞争格局,由网络资源的竞争转向服务竞争,市场变得完全竞争,客户资源成为企业竞争的焦点。客户关系管理(CRM)成为电信运营的必然选择。根据调查,目前电信企业的CRM 还很不完善,对于客户的细分主要是依据统计分析的结果而进行,客户细分的依据的标准是消费额或者消费额、年龄等的组合。这种方法难以真正反映客户的需求,忽略了客户消费的偏好和规律。目前,对CRM 客户的研究也主要是针对客户价值进行研究,以及以客户价值为基础客户细分、客户保持等方面的研究,以客户价值为基础的研究能够识别谁是有价值的客户,但不能反映客户的需求动机、消费偏好、需求变化等。为了准确的把握客户及有效展开营销活动,有必要从理论和实践上对电信企业的客户行为进行研究,但传统的统计方法不利于客户知识发现。本文对目前我国电信企业的竞争格局做了初步分析,针对电信企业CRM 现状及客户细分存在的问题,对CRM 的理论框架、数据挖掘技术在客户行为分析中的优势及作用、客户行为等做了系统性的分析。作为本研究的实证分析,本文选取M 移动公司作为研究对象,在对M 公司访谈、客户问卷调查的基础上,对该公司的全体“全球通”客户某三个月的消费数据进行聚类分析和关联分析,本文提出该公司的客户可分成五个群类,电信业务中新业务具有很大的发展潜力。通过研究,本文还发现了该公司客户行为的一些规律。最后,本文对M 移动公司提出了技术上和营销上的建议,这些建议对于其他电信企业也具有借鉴意义。更多还原

【Abstract】 After reform and Reorganization of Chinese communication companies, mobile companies are faced with new competition situation. The competition situation is from the competition of network resource turning to service, perfect competition market and customers as the competition focus of enterprises. So it is necessary for mobile operators to choose Customer Relationship Management (CRM). At present, the Customer Relationship Management (CRM) of mobile companies exists some problems. The base of customer segment is the result of statistic analysis and the criterion is expenditure or the combination of expenditure, age and so on. This method can ‘t well reflect the demand of customers and neglects their preference. The research of CRM mainly aims at studying customer value and the customer segment and holding based on its value. The study based on customer value can identify valuable customers, but not reflect customers’requirement motivation and change, consumption preference and so on. So it is necessary to study customer’s behavior. But the traditional statistic method has limit to the new customer knowledge finding. This dissertation illustrates the competitive situation of telecom companies in China at present. And the theoretical frame of CRM and customer behavior, the advantage and influence of data mining technology used in customer behavior analysis, are analyzed systematically. This paper select M mobile company as research object to conduct empirical study. Based on interview and questionnaire investigation, cluster analysis and correlation analysis are conducted with three month consuming data of whole “global communication”customers of this company. This paper proposes that customers of this company can be divided into five clusters and there are potential developments in new ones of telecom businesses. Besides, through studying, some rules of customer behavior are found in this company. At last, several suggests are proposed in technology and marketing for this company, which also have some reference values to other telecom companies.更多还原