| 摘要: | 隨著B2C電子商務的普及,網路購物已經成為業者的兵家必爭之地。線上賣家必須透過商品和服務的品質以留住舊顧客並開發新顧客,線上買家只能透過賣家和平台業者所公開的資訊作為選擇交易對象以及購買決策的依據。一般而言,獲取一位新顧客是留住一位舊顧客成本的五倍,因此避免顧客流失為現今網路賣家的一個重要課題。本研究的目的在藉由購物網站之實際交易資料分析,找出顧客流失的影響因子,進而評估顧客流失的可能性。
本研究使用網頁內容資料探勘的方式,蒐集淘寶網店舖街中的女裝店舖賣家在2013年1月1日至4月30日期間每日的賣家相關公開資訊。我們以線上顧客消費行為和線上賣家之特徵與經營績效兩類的變項,建立一個B2C電子商務顧客流失的預測模型。線上顧客消費行為採用RFM的模型,包括最近交易間隔天數、累積交易次數、以及平均交易金額等三個變項。線上賣家的變項包括評價分數、收藏人氣、商品販售種類、主營占比、服務評價、與回購率等共六個變項。運用模型適配度檢定和羅吉斯迴歸分析,研究發現除了RFM三個變項外,亦有其他的變項和顧客的流失率顯著相關,是網路賣家不容忽視的。因本研究建立之模型可以有效地預測線上消費者的流失行為,對於網路購物之平台業者或賣家,可以提供有效的經營策略之指引方針,藉此設法留住顧客,達到更大的獲利。
;With the popularity of B2C e-commerce, online shopping has become an important issue. Online sellers of goods and services must be through the quality to retain old customers and develop new customers, online buyers only as a selection of trading partners and purchasing decisions through information which platform provider seller disclosed. Generally, the cost of obtaining a new customer is five times as large as the cost of retaining an old customer, thus avoiding the churn of customers is an important issue nowadays for Internet sellers. The purpose of this study is to identify influencing factors of the churn of customers by the actual transaction data analysis for shopping website, and then assess the likelihood of churn of customers.
Using web content mining technique, we collected real data of seller′s information from dress shop in Taobao shop street during the January 1, 2013 to April 30, 2013.We consider that online consumer behavior and online seller’s operating performance are two variables which we used to build a predictive model for B2C e-commerce customer’s churn. Online consumer behavior using the RFM variables, including recent trading interval, the cumulative number of transactions and average transaction amount. The variables of online seller, including the rating score, collectibles popularity, assortment size…etc. Using logistic regression analysis, the study found that in addition to the three RFM variables, and there are other variables significantly associated with customer churn rate, the network seller cannot be ignored. Because the model established in this study can effectively predict the churn of the online consumers. On the other hand, this study can provide effective business strategies for Internet shopping platform provider to retain customers and to achieve greater profitability . |
