摘要
选择合适的客户并对其推荐合适的产品成为企业提高交叉销售业绩、获得竞争优势的重要方面。本文以某商业银行的18515位个人客户的历史交易数据为研究对象,构建序列模型,利用逻辑回归分析的方法在仅利用银行现有数据的基础上来预测客户下次购买不同产品的概率。实证结果显示,本模型对每种产品的预测准确率达到72%以上,这样不仅能帮助银行提高客户经理在产品推荐时的准确性,而且可以使银行的营销决策更有针对性,从而给银行带来更大的收益。
"Recommended right product to right customer" is an important aspect to improve the cross-selling and get competitive advantage.This research selects historical transaction data of 18515 customers from a commercial bank as samples to build sequence model and to use Logistic regression to forecast the probability of customer next product to buy only based on existing data.The results show that the accuracy of each product has reached more than 72%,so there were more targeted to different customer use different cross-selling strategy,thereby increasing the accuracy of products which recommended by manager,bring greater benefits to the enterprise.
出处
《预测》
CSSCI
北大核心
2011年第6期13-18,共6页
Forecasting
基金
国家自然科学基金资助项目(70872087
71002102
71172133)
陕西省教育厅基金资助项目(08JK082
2010JK141)
陕西省普通高等学校哲学社会科学特色学科建设资助项目
西安工业大学科研创新团队建设计划资助项目
西安工业大学校长基金资助项目(XJJ200724)