DETERMINING CUSTOMER SEGMENTATION BASED ON RFM & DEMOGRAPHIC VARIABLE USING CLUSTERING TECHNIQUE ANALYSIS TO INCREASE PT X'S CUSTOMER RETENTION

• Main Author: Rafiq Azka, Dhiya'
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/79604
• Institution: Institut Teknologi Bandung

PT X is an MSME that’s focused on fashion products. Initially, sales of PT X's products were only done online, through the website, social media, and e-commerce. As the business grew, PT X decided to sell the products offline by opening a shop. However, it was found that the customer retention which can be seen from the repeat purchases data or the repurchase rate of PT X's customers is very volatile. Fluctuating customer retention can be caused by a rigid (not personalized) customer retention strategy. For now, the customer retention strategy is only based on customer segmentation which is done intuitively and not based on historical data. Proper and data-driven customer segmentation is the key in designing a more effective customer retention strategy. This study aims to determine the customer segmentation of PT X's customers using clustering analysis/technique. Due to PT X’s limited data, the research has to be carried out using data that’s obtained from PT X’s digital customer survey. This research methodology follows the stages of the Cross-Industry Standard Process for Data Mining (CRISP-DM). The customer segmentation is based on the recency, frequency, and monetary model (RFM model) which is able to identify customer buying behavior and demographic variables to determine customer characteristics. The customer model segmentation is built with a combination algorithm of hierarchical clustering and k-means clustering. This research produces a customer segmentation and a customer retention strategy recommendation based on the produced customer segment. The produced customer segment is more precise and data-driven. The customer retention strategy that’s made based on customer segmentation can reduce the fluctuation of PT X’s customer retention because it can help PT X in the process of designing a more effective customer retention strategy.