使用用戶生成內容進行品牌聯想萃取及聯合品牌評估

Abstract

聯合品牌是促進品牌形象或品牌知名度的一個很重要的品牌策略。聯合品牌泛指兩個品牌一起進行營銷活動的一種形式,包含廣告、行銷或產品等,一個好的聯合品牌不僅可以為兩個品牌帶來知名度、創造營收,也可以藉由彼此的合作創造更大的價值。在進行聯合品牌評估的時候,品牌聯想是其中很重要的一個面向,品牌聯想泛指消費者的記憶中對於品牌的認知與偏好,尤其是進行消費選擇時所產生的任何關聯,而認清品牌聯對於企業以及消費者雙方都有極大的好處。過去品牌聯想的相關研究,主要都是透過問卷的方式去萃取各品牌的品牌聯想,這種方法不僅費時,亦無法了解現實中眾多實際品牌的情況,但隨著網路的世代來臨,許多網站成了消費者提供意見的平台,這些累積的使用者生成內容,逐漸成為取得品牌聯想的資料來源。而過去聯合品牌的相關研究也多是以問卷方式為主,這種方法費時之外,可以取得現實中聯合品牌成效的資料也都較少。 本研究將建構一個系統化的方法來進行聯合品牌的評估,透過文字探勘的方法和使用者生成內容,將之應用在品牌聯想的萃取,並將萃取的結果應用於聯合品牌的評估方法中,以建立一個聯合品牌的預測模型。而在我們的實驗結果中也證實利用品牌聯想所產生的變數是可以有效地評估聯合品牌的成效。

Co-branding is an important branding strategy that can bolster brand image and brand awareness. It is defined as “two brands are deliberately paired with one another in a marketing context such as in advertisements, products, product placements, and distribution outlets.” When making a co-branding decision (selecting the right partner for a co-branding), the brand associations of two brands represent an important source of information to examine. Brand associations are customers’ perceptions, preferences, and choices in memory linked to a brand. Traditionally, the process of eliciting brand associations relies on the use of the survey-based approach, which often time consuming and labor intensive. Similarly, co-branding assessment typically is conducted via the survey-based approach; thus, it incurs the same limitations as brand association elicitation. As the Internet embeds in people’s lives, consumers start to share everything online. Such user-generated data (UGC) becomes an alternative data source for extracting brand associations and helping assess the effectiveness of co-branding decisions. In this research, we propose four methods to extract brand associations from UGC (specifically, from online product reviews). Moreover, we develop and construct a predictive model to assess the effectiveness of a co-branding, using the variables derived from the brand associations of the two brands involved in the focal co-branding. Our empirical evaluations suggest the utility and satisfactory effectiveness attained by our proposed brand association extraction methods and the co-branding assessment technique.