Data-driven decision-support system for improving store expansion decisions in retail

In this thesis, we propose, develop, and evaluate a data-driven decision support system, named Dexter, that provides insights and recommendations for retail store expansions. We propose an ensemble algorithm that incorporates these key factors to generate retail store expansion recommendations. We have co-designed Dexter’s decision-support dashboard jointly with experts in the retail industry by following design science research methodology to ensure its purpose and fit. Dexter’s ensemble algorithm is evaluated using real world operational data obtained from an industry partner (a leading retail business) and its usability was assessed through a post-study system usability test with industry experts. The results show that a systematic data-driven approach makes the decision-making process more efficient and has the potential to assist store owners in making informed decisions for effective retail store expansions.