Identifikasi Faktor-Faktor yang Memengaruhi Waktu Pembelian Kendaraan Kedua Menggunakan Metode CART
Keywords:CART, regression tree, vehicle purchase distance
Increased competition between personal vehicle dealers make them need strategies to hold their customers and increase their sales. One of the strategies they could apply is prospecting their customers at the right time. We could predict the right time by identifying the relationship between the length of their purchase time and its factors based on the transaction data of Z Company from year 2002 to 2015 using Classification and Regression Trees (CART). Data analysis is separated between groups of customers who made the second purchase maximum of 10 years after the first purchase (group A) and more than 10 years after the first purchase (group B). Group Aâ€™s regression tree produces 8 terminal nodes with MAD value 1.84 years. The independent variables that plays a role are tenor, job, age, and brand. Group Bâ€™s regression tree produces 4 terminal nodes. Authorized service and job come out as independent variables which affect the splitting process. MAD value for Group Bâ€™s regression tree is 0.56 years.