Data science may feel like a new frontier for truckload transportation, but concepts like “big data” have always been a part of the industry. In fact, the problem has never been a lack of big data, but rather a lack of technology able to manage the vast quantities of data that are constantly being produced and collected.
That’s changed, thanks to machine-learning and data science.
At CSCMP EDGE 2021, Dr. Chris Caplice spoke about how data science is transforming the transportation industry. Here’s a summary of how shippers can use data-driven insights to answer four key questions and make proactive, informed decisions.
What is data science?
It can be difficult to appreciate just how big the transportation market is. In 2020 alone, it was worth $700 billion (3 percent of the GDP). The market includes millions of shippers, hundreds of thousands of carriers, and tens of thousands of intermediaries, all of which are producing millions of data points daily.
The importance of making sense of those data points can’t be underestimated. After all, transportation networks are incredibly complex to begin with. Further, with the amount of “big data” that is produced by every single shipment, getting the necessary insights can be a challenge. That’s where data science comes in.
By definition, data science is a collection of methodologies and techniques that help analyze data and extract meaningful insights in order to make more strategic decisions.
Data science is not independent of domain expertise or human judgment. It’s a tool that empowers people to use their judgment more effectively. Whereas previously we would’ve used statistical averages to inform decision-making, now we’re able to get deeper answers to essential questions. This gives us a nuanced and intricate look at the marketplace that otherwise wasn’t possible.
4 key questions answered by data science
Before data science, it was difficult to analyze more granular pieces of information that are key to transportation in order to get the full story for a lane, region or contract rate.
With data science, we no longer need to rely on statistical averages — we can get the full picture. In order to build this comprehensive picture, data science answers the following four questions:
Descriptive: What happened in the past?
To move forward, we first need to understand what happened in the past. That’s true for any industry, but it’s particularly important when it comes to truckload transportation where the massive amount of variability in play at any given time can make it hard for shippers to predict the future.
Data science goes beyond averages to explain what happened in the past and how you fit into that data. For example, you could use data science to break down lane rates into contract and spot over a set period of time. You could then break down the market contract rate further to see what the rate was for certain types of loads, trucks and more, so you can narrow your insights to focus only on the relevant data. You can even perform an analysis of your company’s specific data so you know exactly what happened within your business.
Diagnostic: Why did this happen?
Once you know what happened, it’s time to figure out why. Did something cause past behavior? Can you change it for the better? Understanding the underlying issues from the past helps you address them proactively for the future.
For example, data science can be instrumental in helping you understand your past costs. This is another area where averages barely scratch the surface. Lane rates can vary wildly based on a number of factors including distance, equipment type, lane characteristics, macro-economic conditions at the origin and destination, and more — not to mention the difference between forehaul and backhaul rates that can be lost in averages.
With data science, you can go beyond anecdotes to actually understand how lanes work. You can estimate a contract rate, monetize the impact of different factors, and gain new insights to adjust your approaches in the future.
Predictive: What is most likely to happen in the future?
For many shippers, this may be the most valuable question that data science can answer. Machine-learning models are making incredible strides and revolutionizing forecasting by including other potential explanatory variables that can affect contract or spot rates. This improves accuracy and lets you shift from lane-level forecasts to shipment-specific forecasts, giving you more control than ever before.
One of the most exciting areas of data science is predicting and counteracting churn. Combined with human judgment, data science can help shippers identify customers and carriers at risk of churn so they can then take proactive steps to keep them engaged.
Prescriptive: What actions should be taken?
Once the data has been analyzed and data science has offered predictions for the future, it can offer some possible actions to maximize positive outcomes. Of course, it’s up to human judgment to determine which actions to actually take.
The example about churn highlights the way that data science is not a robot-like replacement for human judgment, but rather an asset to it. While data science may identify customers and carriers at greatest risk of churn and even offer some possible actions to take, it’s up to the shipper to decide, first, if they’d rather have false positives or false negatives mixed in with that data and, second, what specific intervention to take with each customer or carrier.
The beauty of data science is that it eliminates many of the mundane tasks once associated with keeping a shipping business afloat and instead leaves human workers free to put more of their time and energy into making critical decisions.
Empowering people with data science
Data science offers a lot of potential for the shippers who embrace it, but it’s important to remember that data science is NOT a replacement for human decision-making. It’s a collection of techniques and methodologies that helps extract insights from data and leaves humans free to spend more time making the important decisions that drive your business.
Today, data science is a required tool in your transportation toolbelt — and that alone is changing the industry. By helping shippers understand their past, present, and future, data science is transforming truckload transportation for the better.
To empower your team with data science, reach out to DAT to speak with your Transportation Analytics Expert.