Analyze the US freight market to benchmark freight rates so that Shippers and carriers can optimize their pricing using the benchmark
The obtained dataset was first cleansed and prepared to obtain a standardized format. After various validation checks were performed on this data to identify outliers, the outliers were removed from the model. The cleansed data was then moved to SPSS where the regression was performed at a 95% confidence interval
A linear regression model was used to estimate the freight rates that were then compared against the actual rates to give a benchmark (market study). A LP model was developed for lane matching which identified the dead miles during a transaction and realized potential where collaborating with non-competitive shippers could reduce it.
For the TL model it was identified that distance was the most influential factor driving the overall costs, followed by fuel and geography.
Lane by lane cost prediction was obtained which enabled the client to gauge the shipper’s performance for each transaction, and use it as a benchmark for further contract negotiation on that lane.
Lane matching helped client suggest shipper collaboration to reduce the dead miles and improve the overall cost savings.