Electronic commerce involves data-rich business models that offer many application areas for machine learning. In electronic commerce, retailers aim to optimize the product listing on their websites to increase revenues. The product listing on websites often uses a sales forecast for placement optimization that requires historical data, which are sometimes unavailable at small and medium retailers. Thus, we apply a design science research approach to implement three methods for systematically generating historical sales data that improve elasticity information. The non-availability of historical data is exemplified in a case study of a medium-sized German retailer. Simulating these methods for a more extensive data set revealed that the clustered data generation method generates elasticity information so that the placement optimization performs best in terms of expected profits and computational time. Additionally, we propose an enhanced process for the introduction of data science in electronic commerce.
Hütsch, Marek and Wulfert, Tobias, "Implementation of Placement Optimization in Electronic Commerce: Towards a Data Generation Method to Increase Forecast Elasticity" (2021). PACIS 2021 Proceedings. 73.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.