The current best practices of the industry-leading brand valuations firms are largely based on financial research and expert consultations. Therefore, in this paper we propose a methodology to complement brand valuations by deriving public opinion about a brand from social media data. To achieve this, we collected text data from 16 brands’ Facebook pages from four industries, manually coded 16,000 comments and post and utilized supervised machine learning to analyze the conversations. To analyze the communication, we built four Domain-Specific text classification models to classify the text data according to a probabilistic method based on fuzzy set theory. The consequent result from these models led to the creation of ‘The Customer Opinion Scorecard’, a group of metrics designed to indicate different aspects of customer perceptions. In the process of analyzing the data and devising the scorecard, several interesting findings was found: 1) Facebook is being used as a feedback channel for brands, and the airline industry are more prone to receive complaints via this channel than the other three industries, 2) high customer satisfaction outweigh negative opinions from temporary PR-crises, 3) not all actions taken on a brands’ Facebook page are communication directed towards the brand, and can therefore not be considered as customer opinions used for brand valuation, and 4) there exists an interdependency between what brands posts on its Facebook page, and what the customers respond. The proposed methodology in this paper is meant to serve as a template for an actual implementation, where all aforementioned findings were adhered to in devising the scorecard and presenting the results. Lastly, the limitations of the proposed methodology and the challenges for implementing it in practice are presented, followed by considerations and recommendations for overcoming these, thus paving way for future work in the area.
|Educations||MSc in Business Administration and E-business, (Graduate Programme) Final Thesis|
|Number of pages||122|
|Supervisors||Raghava Rao Mukkamala|