How Data Analytics Helps Manage Logistics Costs
Cutting transportation costs is one of the top concerns in the logistics industry, but a challenging problem to solve. Every item moved is needed by someone, somewhere, at a certain time, and in a certain condition. Those needs often allow for low cost solutions like transport by train or boat. In other cases, the only acceptable solution is an immediate flight and then driver to the destination. In the worst cases, with a little more planning the low-cost solution would have worked, but poor planning made a much more expensive solution necessary.
In this post we will focus on two ways data analytics can help reduce transportation costs. Firstly, we will discuss how the right tools or visualizations can help identify anomalies in transportation when they occur. Then we will show how using comparisons between data sets can highlight where the status quo itself has problems.
Automated anomaly detection is an active area of work in machine learning and data analysis. Configuring one of these tools to highlight changes in your data is getting easier and resulting in more valuable insights every day. For example, using anomaly detection you could identify a spike in upgrades or a drop in on-time deliveries. In either of these cases, getting to the bottom of the issue can help to either save or make money before the anomaly becomes the new normal.
But even without those tools, the right data visualizations can highlight issues. As we have mentioned elsewhere on our blog, humans are remarkably good at pattern recognition. In a recent project at Mashey, we plotted shipping revenues vs margin by category. As a user walked through time in the BI tool, we could identify outliers quickly that revealed underlying anomalies in the data.
While anomalies can certainly hurt company performance, it is typically the status quo that drives the overall success of a company. Reducing transportation costs often starts with selecting the right shipping option, given the needs of the current shipment. For example, a blanket policy of opting for 2-day shipping may make sense at first. But local orders by most shippers get to their destinations in the same amount of time whether you pay extra for a guaranteed delivery date or not.
Fortunately, data analytics can help solve these problems as well. By showing the right comparisons between different data sets, we can use “natural experiments” to find opportunities. For the 2-day shipping problem above, showing average days to deliver by region and shipping method, an analyst could quickly notice that even for products shipped without a guaranteed delivery window, local shipments arrived the next day.
Here at Mashey we are passionate about using data analytics to drive success for your business. Contact us to learn more about what we can do for you.