Within the financial sector, insurance companies generate significant amounts of data on a daily basis, from policy transactions to customer interactions and risk evaluation. This accumulation of data presents an important opportunity for all of these companies to use it for their own strategic advantage. Therefore, through a detailed analysis, this paper presents a case study on how several supervised classification machine learning algorithms (Logistic Regression, Random Forest, Support Vector Machine, etc.) can help in identifying and classifying new clients whether they are potential or non-potential clients for a company in the insurance sector. Moreover, this paper also considers a hybrid ensemble mechanism with the best-performing models among all evaluated models, alongside a decision threshold variable to balance accuracy and minority class recall. The results demonstrate the ability to choose different models to enhance the company’s decision-making in different market scenarios.