Customer Clustering and Classification Based on Location and Online Store Transaction: a Case Study

XYZ Store is an online retailer that offers Rhea Health Tone (RHT), a supplement that helps build immune system strength. Until now, the shop has never contemplated grouping customers and distinguishing successful transactions from failed ones. The research aims to categorize consumers to determine which are purchasing RHT from local retail outlets. It is possible to create a distinction between successful and failed transactions. We have got two datasets to tackle. Tokopedia, an Indonesian marketplace, shared transaction data from its platform spanning three months and 26 days. The second dataset provides additional details, such as the outcomes of the transactions and their dates between September 2020 and June 2021. The purchase data is segmented into five data sets depending on the different locations where data is collected, and the characteristics of the customer base are determined through cluster analysis. This study will help to make market tactics more adaptable via the use of its findings. Based on the results classifications, we figured out that our data had been prepared poorly, which caused an imbalance in data. A better result, more accuracy, and a higher AUC might have been achieved had we prepared adequately.