網購退貨：顧客服務構面、產品構面、安全構面、退貨機率

Reasons of Returning Goods for Online Shopping

現今網購業者面臨的困境是，顧客退貨出現快速成長，這些退貨的相關損失跟事後處理機制導致店家產生額外的龐大成本，可能將辛苦賺來的毛利賠掉，但國內外文獻以機率與量化模型來探討網購退貨原因幾乎沒有。此外，業者常迷者的滿意度，但之後經歷服務失誤後的退貨率就真的會降低嗎？本研究透過三種管道收集資料：國內最大電子佈告欄（BBS）之網購版、以及分別透過電子郵件、紙本型式對管理學院學生發放問卷，目的希望透過不同的發放管道瞭解受測者在填答時反應是否相異，以及本研究模型是否對不同族群及溝通管道會有所差異。問卷共有效回收544份，層級羅吉斯迴歸分析結果顯示：三種不同資料蒐集管道，「顧客服務構面」、「產品構面」皆顯著影響經歷服務失誤後消費者退貨與否，而「安全性構面」則無顯著影響，此外，提升消費者的滿意度，退貨率並不會顯著降低。研究並發現對「顧客服務構面」重視的顧客，會選擇退貨的機率是對「顧客服務構面」相對不重視者的1.948倍；對「產品構面」重視的消費者，其會選擇退貨的機率是對「產品構面」不重視的消費者的1.509倍。研究結果也指出網路購物經驗為半年的顧客，會選擇退貨的機率只有兩年以上者的0.204倍，但當消費者的網購經驗為兩年以上且有過網購退貨經驗，其退貨機率則顯著較其他人高出許多。本研究建立的模型預測準確率在三種不同的資料蒐集管道，皆達75%以上，而根據本研究模型的預測，經歷服務失誤後，顧客的網購退貨機率幾近70%，研究最後並提供管理意涵。

The current dilemma facing the online shop is the dramatically increasing cost of returning goods. The loss and required mechanism of handling the returning goods would significantly cut the profit. Nevertheless, the research literature using the quantitative and probability models to deal with the issue of returning goods is scarce. This study proposes the quantitative and probability models and collects data through three ways: the online shopping discussion board of the largest BBS in Taiwan, and sending email and paper survey to the students of the college of management, reporting data collected from 544 respondents. The results of the analysis of hierarchical logistic regression show that the customer service dimension and product dimension significantly influence whether the customers return goods after experiencing service failure, while the security dimension does not. The probability of returning goods for the customers thinking highly of customer service is 1.948 times than that of the customers who do not. The probability of returning goods for the customers thinking highly of product dimension is 1.509 times than that of the customers who do not. The results also indicate that online shopping experience significantly affects whether the customers return goods or not over the Internet. The probability of returning goods for the customers with half a year of online shopping experience is 0.204 times than the ones with more than two years of online shopping experience. Additionally, the probability of returning goods for the cutomers with more than two years of online shopping experience and experiencing returning goods is significantly higher than the ones that do not. The precision rate of predicting whether the customers return goods, using our models, is higher than 75%. According to our models, the probability of returning goods for the customers who experience the service failure is 70%. The managerial implications will be provided.