Using User-Generated Content for Brand Association Extraction and Co-Branding Assessment

• Main Authors: Wen-Ling Chu, 朱雯伶
• Other Authors: Chih-Ping Wei
• Format: Others
• Language: en_US
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/k74x7g

碩士 === 國立臺灣大學 === 資訊管理學研究所 === 105 === 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.