USULAN KONTEN REKOMENDASI PRODUK BERDASARKAN SEGMENTASI KONSUMEN SMARTPHONE PADA SITUS M-COMMERCE PT X DENGAN MENGGUNAKAN DATA COOKIES

• Main Author: PUTRA HARJOPRAKOSO NIM 14414049 , ADHINUGRAHA
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/24981
• Institution: Institut Teknologi Bandung

PT X, an e-commerce company in Indonesia, wanted to increase conversion rate for its m-commerce because of their low conversion rate. A suitable approach is to increase its purchase probabilities by improving their product recommendation system. Purchase probabilities could be increased by understanding which product attributes drives consumer purchase decision making. Therefore, PT X could give the right recommendation content to simplify decision making process. <br /> <br /> This study uses cookies data consists of view product data, product catalog database, and transaction data during 2 November 2017 to 15 November 2017. The cookies data are processed using Python programming language to extract the needed variables. Variables used in this study consists of smartphone product attributes which established from previous study on theory of planned behavior (TPB), from which 15 hypotheses of this study are derived. The amount of data analyzed is 2198 observation or 1397 consumer. <br /> <br /> Latent class logistic regression is utilized to segment PT X m-commerce smartphone market into four classes based on its purchase decision making behavior. The four segments consist of impulsive camera-oriented buyer, promotional utilitarian buyer, impulsive price-oriented buyer, and utilitarian buyer. The findings of this study are used to generate recommendation system content based on attribute products that drive consumer purchase decision making for each class.