Web Based Brand Equity. Web mining as an alternative approach to Customer Based Brand Equity measurement

Abstract

Brand equity denotes the value that is transferred from a brand to its products, and which translates into higher sales volumes and higher profit margins. Consumer based brand equity (CBBE), a perspective of brand equity as a set of associations held in consumers’ minds, was conceptualized in the early 1990’s. CBBE measurement has always been conducted by the use of surveys, administered as personal interviews, telephone calls and the like. Surveys are expensive and time-consuming, which means that alternative approaches for CBBE measurement should be investigated. The internet offers multiple types of information which may relate to CBBE in various ways. This research investigated whether publicly available online resources can be collected and analyzed with regard to CBBE. In order to do so, the author re-conceptualized CBBE as applicable for measurement using online resources. The online conceptualization was then operationalized in terms of relevant information available online, resulting in an online CBBE scale consisting of 11 metrics. The metrics were gathered from four domains: Google search results, Google search statistics, social media, and online reviews. Web mining is the process of automatic web information retrieval. In order to automate web retrieval procedures within the four domains, the author designed a computer application, intended to work across product categories. The web mining approach was performed on the product category of headphones, featuring 62 headphone models from 18 different brands. Product information was supplied by leading Norwegian electronics retail chain Spaceworld Soundgarden, and included all headphone models sold in-store or online in 2016. The study revealed that several online measures of CBBE are correlated with sales. One of the key findings of the multiple regression analysis was a general shift of explanatory power, when comparing expensive models with budget models. In the high-end segment, product-specific measures, such as professional review scores, best explained differences in sales, while several brand-specific measures were negatively correlated. In the low-end segment, the result was the opposite. The web mining approach is shown to be able to provide insights on CBBE, but generalizability across product categories is limited, due to reliance on social media and reviews. Recommendations for future research are proposed, such as utilizing nationwide sales figures, and analyzing products that are in direct competition. Expanding the set of algorithms, such as including sentiment analysis, may also increase the performance of the web mining approach to CBBE measurement. The research concludes with the argument that the web based CBBE may, with refinement, become a valid alternative to survey research.