TMS Machine Learning Generates Accurate Comps After Recent Disruption
The chaos of 2020 wreaked unprecedented havoc on the freight transport industry. A true Black Swan event, the COVID-19 pandemic derailed every phase of the supply chain and created backlogs some shippers are still working through today. As the ripples continue to affect global trade, a new problem is emerging: the challenge of pulling accurate comps for future data forecasting. The numbers from 2020 and 2021 are heavily skewed — which makes it difficult to look forward with any real clarity.
This presents a real problem for shippers. Without good comps, it’s difficult to facilitate freight ops. How do you budget accordingly? Or benchmark rate averages? How do you negotiate rate contracts in good faith? It’s difficult — but not impossible.
2020 skewed comps and much more
Freight comps were all over the place in 2020. If you pull numbers from the early part of the year, there’s likely some consistency from past years, but pull from the second quarter, and you’ll find major lows. The third quarter saw rates spike — and they’ve only continued to rise since. Clearly, there’s no good way to pull prior-year comps.
To make matters worse, shippers can’t even rely on sales data from 2020 to forecast freight demand. Guidance regarding sales projections is even worse, which only adds to the confusion as shippers and carriers scramble to plan for the end of 2021 and the start of 2022.
Realigning comps amidst disruption
So, how can shippers and carriers come to terms on rates? Without the ability to extrapolate from prior-year data in conjunction with forecasts, shippers are turning to more creative modeling methods. The best methods for realigning comps to approach accurate benchmarking include:
Year-over-year comps. Shippers can either take aggregated monthly data for trailing years and average it, or use total year-over-year growth to project forward-looking comps from the current month. In either case, 2020 can be viewed as an outlier.
Sequential comps. This method of benchmarking involves looking at consecutive quarters for figures affecting freight demand, such as sales forecasting and GDP data. These comps tend to be more reliable than year-over-year data.
Stacked comps. Two- and three-year stacked comps are useful for benchmarking if shippers examine the same time frames and determining variables, and factor them in with current circumstances.
While these methods are useful for projecting comps data, they’re often complicated and cumbersome to calculate.
Machine learning provides a smart solution
This complexity has many shippers turning to machine learning and artificial intelligence (AI) to carry the calculation burden for them. A transportation management system (TMS) with features for automation and data analytics gives shippers access to prior data and the current rate environment. The result is a data-driven approach that incorporates historical data and current insights to fuel forward-looking comps.
As the pandemic continues to rock the freight industry, the need for an intelligent TMS is increasingly evident. Impact TMS from nVision Global has the advanced functionality necessary to get shippers the accurate comps they need to budget costs across shipping lanes. Visit nvisionglobal.com to learn more about AI, machine learning, and all the benefits of a smart TMS.