A study of Unilever’s online consumer-brand engagement in Portugal

• Autor(a) principal: Augusto, Manuel Pereira
• Data de Publicação: 2020
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10071/21807
• topic: Online consumer-brand engagement; Text mining; Sentiment analysis; Árvore de decisão -- Decision tree; Unilever; Análise de sentimentos

The consumption of online social media is growing around the world. In 2020, Facebook registered a record of 2.5 billion monthly active users. Consumers all around the world are increasing their online presence, especially in social networks as Facebook, where they communicate and look for advice from other consumers in ways that firms are finding harder to control or predict. Some brands have taken advantage of the potential of these platforms, trying to understand how to connect with their consumers online, how to measure those efforts and how to improve its consumer-brand engagement. This in-company project is elaborated within the scope of Unilever. Its purpose is to analyze the company’s Brands consumer-brand engagement. In this project, we evaluate and compare Unilever Brands’ engagement performance in Facebook, by understanding which brands and topics are generating more positive or negative engagement and by analyzing which communication patterns lead to more positive sentiments from consumers. To test the hypothesis created, a Text Mining and Sentiment Analysis were performed to over 7000 interactions between the brands and its consumers on Facebook, crosschecking this data with the communication patterns and assessing which ones lead to a higher sentiment score. Results show that Facebook Brand Posts communicating the Brands’ Purpose, a Product Innovation, and a Game/Challenge lead to more positive sentiments from the consumers. A plan to integrate a Sentiment Score Analysis in the Brands’ strategy is included, based on the results gathered from this project.