資料探勘技術於客戶價值分析與行銷策略之探討－以台灣生技業銷售為例

Customer Value Assessment and Marketing Strategies through Data Mining Techniques - A Case Study of Taiwanese Biotechnology Industry

劉業萍(Yen-Ping Liu)

指導教授 ： 蕭瑞祥

淡江大學/商管學院/資訊管理學系碩士在職專班/碩士(2016年)

https://doi.org/10.6846/TKU.2016.00063

摘要

「生物經濟」將接替資通訊科技產業成為新的未來台灣經濟產業命脈。台灣生技產業面對產業目標5000億元的挑戰，如何利用電子資訊和生物經濟雙引擎動能，應用於實際銷售狀況和有效率地把手邊的資料轉化成有用的資訊，找尋出最佳行銷活動依據，合理化的進行資源分配，深耕客戶關係，以建構台灣發展生技產業的競爭優勢，有助於台灣生技產業經濟的永續成長，為目前生技業界重要的議題。
本研究應用資料探勘技術於實際交易資料庫，並參考CRISP-DM為基礎流程，規劃建立適合生技業顧客價值的分析與預測的一套標準作業程序Standard Operation Procedure(SOP)。過程中彙整客戶交易資料後，運用RFM(Recency Frequency Monetary)模型三項指標，作為客戶價值分類的基準，將醫療院所分成四種類型客戶。應用資料探勘工具的決策樹分析，挖掘出不同客戶群的銷售數據規則。最後，將資料探勘的結論交由專家訪談評估適合的解決方案，歸納不同價值的客戶群對應的行銷支援與策略規劃。研究結果，除提供生技業瞭解醫療院所的交易特性，有助於產品行銷外，亦可提供其他Business to Business (B2B)的業者，規劃銷售管理及維繫客戶關係管理的參考。

關鍵字：資料探勘、客戶價值、決策樹、生物科技業

關鍵字

資料探勘 ； 客戶價值 ； 決策樹 ； 生物科技業

並列摘要

“Bio-economy” will replace the information communication technology (ICT) industry as a core sector of Taiwan’s economy in the future. In the face of the NT$500-billion target of Taiwan’s biotechnology industry, using the kinetic energy of the double engines of electronic information and bio-economy in actual sales to seek the best marketing activities, reasonable resource distribution, and deeper customer relationships to construct national competitiveness for biotechnology industry development can benefit the industry’s sustainable growth. This is one of the most important issues in the industry today.
This study aims at the application of data mining technology to the actual transaction database. With process planning based on CRISP-DM, it is able to build a set of standard operating procedures (SOP) for the analysis and prediction of customer value in the biotech industry. After collecting customer transaction data, first, it takes the RFM (Recency. Frequency, and Monetary) classification model as three indicators for the benchmarks of customer value, dividing customers of medical institutes into four types. Next, through decision tree analysis of data mining tools, it can dig out different customer sales data rules. Finally, expert interviews based on the data mining results can evaluate suitable solutions, concluding marketing support and strategic planning according to different values of customer groups. The study results allow the biotech industry to understand the trading characteristics of medical institutes, and this also contributes to product marketing. As for other Business to Business (B2B) operators, the results serve as effective references for the planning of sales management as well as maintenance of customer relationship.

並列關鍵字

Data mining ； Customer value ； Decision tree ； Biotechnology industry

參考文獻

[7] 辛昱辰、侯藹玲、江志民 (2013)。『應用 Crisp-dm 建立農產價量預測模型-以甘藍為例』。Journal of Data Analysis，8(1)，109-137。

[18] Altınçay, H. (2007). “Decision trees using model ensemble-based nodes,” Pattern Recognition, 40(12), pp. 3540-3551.

[19] Altuntas, S., Dereli, T., & Kusiak, A. (2015). “Analysis of patent documents with weighted association rules,” Technological Forecasting and Social Change, 92, pp.249-262.

[20] Anderson, E., & Schmittlein, D. C. (1984). “Integration of the sales force: An empirical examination,” Rand Journal of Economics, 15(3), pp.385-395.

[21] Belcourt, M. (2006). “Outsourcing: The benefits and risks,” Human Resource Management Review, 16(2), pp.269-279.