Data-Driven Strategies for Fuel Distribution in Indonesia: A Case Study of PT Pertamina Patra Niaga

Abstract

Fuel oil or what is often referred to as BBM is one of the basic needs to drive all community activities. So the government appointed PT Pertamina as a single company which is a state-owned company to facilitate fuel needs for all levels of society. However, with increasing demand, the government formed a new policy to allow private companies to come in to meet all fuel demand. With this, PT Pertamina is no longer the only fuel supplier in Indonesia and must continue to develop mature strategies so that profits do not fade. One way is by examining sales data and predicting customer loyalty. The RFM method followed by the decision tree algorithm and k-means clustering is applied in this research, with the output being able to determine the level of customer loyalty, the level of salesman performance, as well as predicting the potential for customers to churn and its correlation with the salesman's skills. The data used as a reference for the research is sales transaction data obtained from PT Pertamina Patra Niaga Regional Jatimbalinus. And from the research, results showed that the majority of PT Pertamina Patra Niaga Regional Jatimbalinus customers are loyal customers. With a salesman, performance is divided into good performance and less good performance. This grouping is obtained based on the salesman's overall performance track record. As for customer churn predictions, it was found that there was 1 group of customers who were predicted to churn heavily, but this was not influenced by salesman performance, as evidenced by transaction track records in existing data