以產品屬性定義消費者購買訴求及顧客分類之研究

Abstract

過去顧客關係管理在顧客分類的應用分析上，最常使用的分類原則是人口統計變數及交易行為變數，不但做法客觀，資料取得也相當容易。然而這種分析數據對於顧客購買行為的解釋能力卻相當有限，例如同一顧客所購買的產品並不只侷限於單一產品類別，因此無法從顧客購買某一類別產品的時間、次數及金額進行明確的市場區隔；而有相同環境背景的顧客亦不代表具有相同的購買訴求，同一性別、年齡或職業的顧客可能會產生截然不同的消費行為。 故本研究以國內電視購物頻道E公司的顧客資料庫作為研究對象，打破過去以功能別將產品歸類的方式，利用消費者行為的心理訴求重新賦予產品屬性，並以新的產品屬性分類架構作為顧客分群的基礎，從顧客的購買紀錄來推測顧客的購買訴求，以進行統計集群分析。在對分群的結果加以檢定後，可確定此種分群方法是有意義的，且群與群之間具有顯著的差異性，並透過成偶檢定的正負面描述，進一步對各集群顧客的消費特性加以描述與命名，希望透過分群的結果，讓企業瞭解不同顧客的購買偏好與需求，使行銷人員可以針對不同集群的顧客訂定不同的行銷策略，且提供適當的產品資訊給適當的顧客。 本研究首先利用內容分析法將跨功能別的產品屬性抽離出來，歸納成五大產品屬性類別，包括高貴時尚類、科技功效類、平實親和類、健康專業類及怡情消遣類，並以此五大產品屬性類別作為顧客的分群因子，對E公司2004年之常購顧客資料庫進行分析，結果可將顧客分為四群，分別為：實在精算型、知性休閒型、獨善功利型、低價取向型，研究的具體成果為對各集群顧客提出行銷上的建議，協助企業針對不同集群顧客的購買訴求推出適當的產品，發揮行銷策略最大的效益。

In the past, scholars usually used demographic variables and buying behavior variables as the rule of the customer classification in research conducted by the application of Customer Relationship Management. The method helped researchers collect data objectively and easily, but was limited to explain the buying behavior. For example, one customer could buy many categories of products. It’s difficult to segment the market according to the time, frequency, and amount of buying only one category. Customers who have the same backgrounds might have different buying behaviors with different buying appeals as well. This research regards the customer transaction database of TV- Commerce Company E, as the investigation subject, trying to replace product categories with product attributes from psychological appeals of consumer behavior. As the basis of the customer classification, this new product classification can speculate customers’ buying appeals from their transaction records and cluster them into different groups. After the assessment of differential testing and validity analysis, we can describe and define each cluster according to the characteristics of their buying behaviors. The results would help the marketing department to provide proper product messages to proper customers by different marketing strategies. This research extracts the product attributes from different product categories by Content Analysis first, and then generalizes five categories as the clustering factors in the customer classification. The loyal customers in Company E transaction database in 2004 can be divided into four clusters. This article provides suggestions about each cluster in order to help the marketing department to reach the maximum of marketing efficiency.