One of the questions that supply chain managers ask is if there are solutions that are both simple and efficient, approaches that they could follow in order to manage the business of supply chain operations.
In a previous interview with Gene Long, VP Industries Supply Chain at Chainalytics, EBN got invaluable insight from his expertise in supply chain resilience to help supply chain management avoid and mitigate risk through adopting new tools and forecasting capabilities. (See Resilience in Supply Chain Management: Interview With Gene Long, Jr., Part 1 and Resilience in Supply Chain Management: Interview With Gene Long Jr., Part 2 .) Now, EBN gets back to Gene Long to dig deeper into the benefits of using predictive analytics in supply chain resilience.
Long is a big advocate of understanding what is likely to happen before it happens. Citing the parallel thinking that is driven in a financial department, where most CFOs depend very heavily on economic forecasting as a basis for creating operating plans and measuring performances of their companies and supply chains, he notes that very little of the management in charge today is up on what could happen. The major focus is on what is happening now, or what just recently happened. Short-term history is the basis for managing most supply chains and most businesses today.
Driving higher risk mitigation performance in the supply chain
Supply chains are driven by data. According to Long, if you are able to capture historical data in the near term and to analyze that data in a meaningful way, you can also do what we call predictive analytics. You can analyze the historical trends and extend those trends, and basically you begin to predictively analyze what is going to happen with your business.
For Long, capturing and using predictive analytics with the data that is generated out of the operations in the supply chain is the one thing that can be done to help drive higher risk mitigation performance in the chain. “Understand what you are doing and then use that data to take you in possible new directions, support modeling of different scenarios, and then support decisions based on those facts,” he recommends.
Historical analytics vs. real-time data
Responding to threats before they cause sustained or catastrophic financial consequences is of paramount importance in today's supply chain management.
The following questions may serve as a guide to determine if the best approach is to use historical analytics or real-time data:
- Is it better to have a basic random analytic approach?
- Is it better to know what happened in the last 72 hours in order to understand what may happen in the next 72?
- Or are you concerned about trying to understand what is happening right this moment in real-time rather than trying to navigate the forecast?
According to Long, the use of historical analytics is best in this case. This is due to the fact that there are limitations in trying to use real-time data to respond to threads. “Take the example of a dashboard; you look down at the dashboard and real-time data is going to tell you how fast you're going when you're driving your car. What I'm actually after is the ability to look at the windshield and see someone pulling in the street in front of me so I can avoid hitting them, and that requires a predictive analytic, not a real-time data.”
The importance of having the right analytics performance
There are two main considerations when looking at the data collected. “The first one comes down to having the right analytics performance of large bodies of data whether it is near-term history, or real-time data,” says Long. “The second is to look at the right thing.” Looking at the wrong data from the wrong resources, or failing to be inquisitive enough to get the right data can result in wasting precious time while facing a near risk. “Both near-term history and real-time data are important and useful if you are looking at the right data.”
To be able to forecast what is going to happen to your supply chain, supply chain management needs the support of someone else's data from outside the company. This could be a trading partner or a specialized supplier that is willing to assist in getting the right data and understanding the value of data that will help the supply chain in mitigating risk efficiently.
Supply chain big-data maturity curve
If we visualize a supply chain big-data maturity curve we can see that it goes through different levels: from unprepared to having a contingency plan to a continuity plan to continuity management, and finally to resilience. The road to the supply chain using data maturity for resilience looks like this: