Data-driven customer segmentation in the medical beauty industry: clustering models and natural language processing techniques

• Autor(a) principal: Indorf, Fenja
• Data de Publicação: 2022
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10362/161317
• topic: Business analytics; Customer segmentation; Clustering; Natural language processing; Domínio/Área Científica::Ciências Sociais::Economia e Gestão

This research study performs a customer segmentation on a medical beauty case to inform data driven decision making and refine marketing efforts. RFM analysis with a K-Means clustering algorithm, and an HDBSCAN after UMAP dimensionality reduction have been applied. Natural language processing, Doc2Vec, served as the means for sequence embedding. A decision tree algorithm is used to interpret the clustering results. HDBSCAN achieved the best clustering results, and customers were grouped into five clusters based on their purchasing behavior and demographics. The recommended model is the RFM model, and the resulting three customer groups are used to derive marketing strategy recommendations.