The Integrated Analysis of Customer Value and Long-tail Theory-Take Amazon.com as an Example

自2004年長尾理論一書問世以來，許許多多的企業開始對於暢銷商品與冷門商品有了不一樣的看法，特別是電子商務相關企業。由於網路所具備的成本優勢與多元資訊，使企業能夠且願意提供多樣化的產品與服務，來滿足消費者過去未被少數大眾暢銷商品滿足的獨特需求。這些過去在80/20原則下不重要的冷門商品(長尾商品)，開始因為科技的進步，使人無法忽視其總合所產生的力量。然，對企業而言，最理想狀態是能實行一對一行銷，針對每位顧客的異質性提供不同的產品。但現實狀況往往是企業囿於有限的資源，必須辨認並區隔出對企業獲利貢獻較高的顧客，優先將資源運用在維繫這群高價值顧客上，期望能提高顧客之終身價值，使企業利益最大化。因此，唯有了解顧客之購買偏好與購買行為模式，才能有效以正確的商品與方式與之建立關係、滿足需求。研究主要以亞馬遜網路商店2004年之資料庫作為分析對象，並將產品鎖定在書籍類別，檢視不同價值顧客群是否存在著不同的購買偏好？是否受書籍類別影響？並藉由集中度比值的分析觀察不同價值顧客群在網路搜尋行為上的差異，從中推測其行為背後的原因，再配合其購買偏好的結果，將可幫助企業更貼近顧客以及提供更適合的商品。結果顯示，亞馬遜網路商店書籍類別之高價值顧客顯示出對長尾商品的偏好，相對地，低價值顧客則對暢銷品較偏好；深入至各類別書籍觀察，則的確會因類別不同影響分析結果。此外，在網路搜尋行為上，顧客也因購買商品與分屬不同價值顧客群之差異，在網站造訪次數與平均瀏覽頁數的集中程度上也互異。對於如亞馬遜網站一般的電子商務企業來說，若能掌握上述的顧客資訊，將可使企業在資源應用上更有效率，並幫助行銷人員提出更有效的策略建議。The concept of segmentation is related to the CRM, relationship marketing, and customer value. Company has to be able to recognize which customer belongs to the “most valuable customer” group because of the limited available resource. Companies should find those high-valued customers’ preferences and try to satisfy their needs. Thus they can improve profit and get sustainable growth by maximizing customer lifetime value. nd with the emerging of “long-tail theory”, many companies start to supply as many kinds of products as they can to meet different needs of customers. Although the Pareto Principle is still true, companies are more willing to serve products even if they are not best seller because of the low cost and information accessibility of internet. Therefore, heterogeneity of customers becomes an important issue. For companies, only if they can find out the difference between different groups of customer, and understand their purchase interest and search behavior separately, they can make the right marketing decision. he thesis takes Amazon.com as an example to analyze with a focus on books category. We analyze the purchasing records in the database to see whether there is the difference between different customer groups, and whether the result will be affected by the type of books. We also know the search propensity of different groups of customer by analyzing their concentration ratio. So the analysis contains two dimensions at the same time, customer and product. From the empirical result, we find that high-valued customers have preference on so-called ”long-tail products”. Of course, low-valued customers prefer those “best sellers”. Besides, this situation will be sometimes different if we just focus on a certain type of book. The concentration degree of search behavior is also different between groups. For e-commerce companies like Amazon, the cross analysis above can help them execute relationship marketing better. According to the information, they also can come up with more effective marketing strategies and be more efficient on marketing resource allocation.口試委員會審定書………………………………………………… I詞………………………………………………………………… II文摘要…………………………………………………………… IV文摘要…………………………………………………………… V錄………………………………………………………………… VII次………………………………………………………………… IX次………………………………………………………………… X一章 緒論…………………………………………………… 1第一節 研究背景與動機……………………………… 1第二節 研究目的……………………………………… 3第三節 研究範圍……………………………………… 3第四節 研究流程……………………………………… 4二章 文獻探討……………………………………………… 5第一節 關係行銷……………………………………… 5第二節 顧客關係管理………………………………… 10第三節 顧客終身價值………………………………… 16第四節 集中度………………………………………… 20三章究方法…………………………………………………………… 37第一節 研究對象與範圍…………………………… 37第二節 研究架構……………………………………… 38第三節 RFM模型……………………………………… 39第四節 書籍分類……………………………………… 39第五節 羅吉斯迴歸模型……………………………… 40第六節 顧客集中度…………………………………… 43四章 實證分析……………………………………………… 45第一節 顧客價值與長尾理論之整合分析…………… 45第二節 各類別書籍之整合情況分析………………… 51第三節 消費者特性分析……………………………… 60第四節 集中度分析…………………………………… 73第五節 網路搜尋行為分析…………………………… 80五章 結論與建議…………………………………………… 83第一節 結論…………………………………………… 83第二節 行銷意涵與策略建議………………………… 88第三節 研究限制與後續研究方向…………………… 95考文獻 ………………………………………………………… 9