- Data Analytics
The truckload transportation industry is an established industry, in the US, with annual revenue for for-hire truckload reaching greater than 300 billion dollars in 2019. A major problem encountered by for-hire truckload carriers is a sudden, unexpected and sustained reduction in shipment volume over lanes referred to as ‘churn’. Churn leads to a significant problem in the balance in the carrier’s network which, in turn, drives up costs, reduces revenue and decreases driver satisfaction. In this capstone, which is a first-of-its-kind study within the truckload industry, we leverage data from our sponsor - a large national trucking firm - to formally define churn using three parameters: base, drop and duration. Based on this definition we identify churn by origin within the carrier’s network and then establish correlations between the characteristics of an origin and the likelihood of churn at that origin. This framework allows carriers to quickly detect churn before it materializes and take proactive steps to mitigate its negative impact. Our research on churn opens up avenues for further study in this area, within the TL industry, including studying churn at a larger scale to develop more widely applicable ways of defining, identifying and detecting churn.
Access the full capstone paper on DSpace