CFE 2019: Start Registration
View Submission - CMStatistics
B1893
Title: A customer segmentation method for UK open banking-type data Authors:  Andrej Svetlosak - University of Edinburgh (United Kingdom) [presenting]
Raffaella Calabrese - University of Edinburgh (United Kingdom)
Miguel de Carvalho - FCiencias.ID - Associacao para a Investigacao e Desenvolvimento de Ciencias (Portugal)
Abstract: An unsupervised method is developed to segment individuals based on demographics, financial status and financial behaviour tailored for transactional open banking-type data. We analyse transactional records of 9,875 customers in 2017 and their registered accounts across 49 UK financial institutions. We propose a segmentation method that takes advantage of the information hierarchy between demographics, financial status and financial behaviour, creating a three-level segmentation that is easy to interpret. After obtaining the segments we assess their quality and highlight how the groups differ in demographics, financial status and financial behaviour.