Road freight pricing via parametric models

Abstract

In reference to product cost structure, transportation can represent up to 60% of its total value. In Brazil, road transportation mode is the most used kind, making up to 61% of the freight operated as a whole. However, road transportation pricing is not a trivial process because it involves variables that go beyond the basic costs of the activity. The objective of this work is the development of a freight pricing model with a higher degree of precision and reduced response time, using multiple linear regression techniques. The study is applied to a logistics operation company, which business model is the connection between the carrier and the freight demander. The pricing model developed obtained a coefficient of determination of R ^ 2 = 0.9966. In other words, a highly explanatory model, once it explains 99.66% of the variability of observed freight values. The predictive variables that make up the model are distance, driver cost, type of customer, invoice amount, origin and destination. With the model developed, the variables had their weight statistically calculated, giving automation to the process and making the routine of the company's analysts more agile.