消費者再購商品之研究－以某電腦補習班為例

The Research of Consumer Repurchase －A Case of Computer Educational Institution

林志堅(Chih-Chien Lin)

指導教授 ： 邱光輝

國立臺北大學/商學院/企業管理學系/碩士(2011年)

摘要

近年來無論是在升學或是求職上，電腦專業證照已成為個人專業能力的證明之一。因此民間成立的電腦教育機構遂成了學生和求職者主要的學習管道，但在市佔率、知名度與口碑較高的電腦教育公司較易取得國際軟體公司之授權及指定考場等優勢。

本研究係以某電腦補教業為資料來源，蒐集2008至2010年三年間此補教公司之顧客資料，進行可能的產品組合分析之，期望提供符合客戶的服務需求，整合滿足客戶一站購足 (one stop shopping) 的方便性。

本研究以資料挖掘技術中的序列型樣分析為工具，試圖在一共3,018筆依續購買二項課程之顧客進行關聯規則分析。經由計算個別類別或各項課程之比率，並建立類別或課程的條件比率、計算出信心指數最後再依信心指數計算出改善指數 (improvement)，進而推薦出適合的產品組合。

本研究經實證分析，獲得之研究結論如下：
1. 顧客第一次購買最多為就業訓練類 (第六類)，高達46.1%。第二次購買最多為動漫影視設計產業類 (第二類)，比率為20.40％。
2. 顧客第一次購買最多的項目為第六類的第一項 (數位動畫)，比率為19.68%。而第二次購買最多的產品項目為第三類第一項 (AutoCAD)，比率為19.12%。
3. 若以類別區分則改善指數最高者為先購買類別1 (互動商業廣告媒體設計產業) 再購買類別2 (動漫影視設計產業)，指數為2.83，因此在推銷類別2時，可優先向已購買過類別1者推銷之，成功比率為單獨向未曾購買之顧客推銷之2.83倍。
4. 但若以產品項目區分則改善指數最高者為先購買過項目2 (手機多媒體設計) 再購買項目3 (網頁設計)，指數為402.4，因此在促銷項目3時須優先考慮已購買過項目2之顧客以提高成功比率。

關鍵字

再購 ； 產品組合 ； 資料探勘 ； 序列型樣分析 ； 顧客忠誠度

並列摘要

Both in the academic field and job-seekers field in recent years, computer professional certificate has become one of the individual verification of professional competence. Private computer educational institutions are established for students mainly learning gateway, but rate, popularity and high reputation of computer education companies are authorized easier in international software, designated to the examination room and have other advantages.
I studied one computer educational institution as a data source and collected 2008-2010 three years of customers data of this computer educational company which was processing for possible product bundle analysis, providing a diversity of service needs and integrating customers one-stop shopping. Further improvement is for overall profit of this company.
This research used sequential pattern analysis of the data exploration techniques for tools and tried to do association rule analysis in 3,018 continuous purchases data of total customers who purchases two courses. By calculating the individual category or probability of the various courses, it established the categories or the course of conditional probability, calculated confidence index and then used confidence index to calculate improvement index. Moreover it recommended suitable products combined. In addition it can consolidate customers, and increase the profits of enterprises both.
Empirical analyses of the present study, access to research findings are as follows:
1. The NO.6 class－employment training customers is purchased most at the first time, up to 46.1%; The NO.2 designing for animation, film and television industry are purchased most at the second time, chance of 20.4%.
2. The Course 6-1 is purchased most at the first time, up to chance of 19.68%.The course 3-1 is purchased most at the second time, up to chance of 19.12%.
3. In categories, the highest improvement index is purchased category 1first and then to purchase category 2. Index is to 2.83. Therefore we can promote category 2 to the customers who have purchased category1. The chance of success will make 2.83 times to those customers who have never purchased.
4. In product items, the highest improvement index is for purchase product 2 first to purchase product 3.The index is 402.4. Therefore, for raising the successful ratio, as we promote product 3, we must primordial considerate those customers who have purchase product 2.

並列關鍵字

repurchase ； product bundles ； data mining ； sequential pattern analysis ； customer loyalty

參考文獻

5. 陳淑芳(2002)。顧客關係、服務品質、品牌形象、顧客滿意度與再購意願間關係之研究—以銀行業為例，大同大學事業經營研究所未出版之論文。

6. 陳澤義(2005)。服務管理。台北：華泰

12. 顏博文(2003)。應用資料探勘技術分析學生選課特性與學業表現，中原大學資訊管理學系未出版之碩士論文。

3. 林素蘭(2005)。顧客滿意度、品牌忠誠度對再購意願之影響－以汽車銷售業為例，中原大學企業管理學系未出版之碩士論文。

4. 周碩雄(2010)。以社會網絡分析方法解釋產品組合之研究－以符號學詮釋為觀點，台北大學企業管理學系未出版之博士論文。