“These issues go through the entire chain, from ship to shelf. That’s why we’re not just working with the ports. It’s the truckers, the rail companies, the operators and also those retail companies that are at the other end of those supply chains.” -Pete Buttigieg, United States Secretary of Transportation
Consumers, retailers, and transportation/logistics organizations are looking at an unsteady few years ahead, with some shipping executives expecting the current supply chain disruption to extend well into 2024. For consumers, the mass disruption to supply chains has resulted in online purchases taking longer to arrive, store shelves being noticeably vacant, and why that sofa you ordered is taking 16 months instead of six weeks to arrive.
It’s also wreaked havoc on infrastructure around the world. The backlog of cargo ships has created bottlenecks at various ports, with the Port of Los Angeles experiencing record gridlock of container ships — each carrying billions of dollars worth of merchandise — that could leave holiday shelves bare. It’s also potentially the cause of an oil spill that covered beaches and endangered wildlife after a container ship's anchor hit an undersea pipeline.
Many of these events could have been avoided or alleviated — and their underlying issues could largely be solved in the future with better data transparency that would give organizations more levers to pull during times of crisis. The challenge in this space is to figure out what data is relevant and how to put it to good use.
How do companies work through supply chain pain points? How do they get to the “right” data; what’s the first step? From our experience in this space, we’ve compiled some actions organizations can take to get more transparency out of their data and ease the current pains of the supply chain.
Data is only useful to organizations when it is organized, accessible, and easy to share. Currently, organizations are struggling with having too much data to be able to understand things, and getting out of a sea of data with the right insights can be difficult. What about the quality of the data? Just because you’re collecting data, is it accurate?
This is why deciding on a few high-value answers that your organization would like to have is the best place to start. Do you want to minimize waste? Do you want your high-margin items or your loss leaders off the truck first?
Remember, data doesn't necessarily need to be "big” to be valuable – getting the data to support these smaller answers can eventually lead to having a better answer to the question of, “How can we improve our response to crises like the one we are currently experiencing?”
Third-party data can be very valuable when used correctly, and being able to correlate third-party data with industry- or retailer-specific metrics can be especially helpful. It allows retailers to discover deeper insights into how one space is affected by external events and changing trends. The trouble is some organizations aggregate multiple third-party data sets, but aren’t using them to dive deeper into their products for further transparency. Those sources – data from suppliers, vendors, manufacturers, even environmental conditions — have to be compared against your own.
The ideal scenario is obviously for all retailers to play nice and share their data, but is this a reality? The truth is it’s a bit of a utopian view. Some organizations struggle with the idea, instead saying, “No, this is my data,” and hesitate when it comes to robust data sharing. There are a lot of nuances, competing initiatives, and there can be a resource crunch when payoff or value isn’t immediately visible.
Additionally, the worst time to mitigate a crisis is in the midst of it. Companies need to prepare for the uncertain times when they are in the midst of the good ones, not just reactively in the middle of a 100-year event. That means collaborating with vendors as closely as possible to create a robust data ecosystem so you can pull better insights out of the data in the future.
If you’re seeking better, faster data transparency, then making use of machine learning and AI is essential. Machine learning uses many of the same ideas and algorithms that have been used by analysts for decades, but it can enhance AI particularly by allowing scenarios to be run at a greater scale, and self-learn to predict future insights.
What would have taken an analyst a very long time to find, and even longer to test against real world examples, can now be tried, tested, and applied to new data sets in a relatively short time frame using tools like CueAI. This is especially valuable when needing to pivot and make quick (but informed) decisions amidst the supply chain crunch, when correlations would previously have taken weeks or months to surface in the wild.
Machine learning has the power to automate processes (like recording data from your Packing List and comparing it to a Bill of Lading), understand unstructured data (like predicting maintenance from images of equipment), and personalize recommendations for action (like suggesting a real-time assortment change due to changing supply levels).
Suites with these ML and AI capabilities, like Google Cloud Platform for example, can be extremely helpful in this space. With these solutions, you don’t have to be a data scientist to build a ML model – these tools are built by engineers for engineers. When they’re paired with a multi-cloud data warehouse solution like BigQuery, companies can get AI up and running, turning their data into business insights quickly for a relatively low cost.
With these integrated toolsets, your team is equipped with a better data ecosystem and better foresight from the start — foresight that’s critical to detecting trends and helping organizations make better, proactive decisions. For example, if damaging storms are expected in the Northeast, organizations will have the data to potentially take steps to change their mix of products, find alternative suppliers, or change focus to critical protection and recovery supplies. Or, on a smaller scale, retailers can prioritize unloading a truck full of produce over non-perishable items, or getting high-margin items to the shelves first.
Being able to look at multiple data points and come up with these correlations can also be applied when trying to understand future data. The trick is to find the important measures to look at when comparing products, or understanding the impact of other events in the supply chain.
With the extenuating circumstances of global supply chain issues, we understand undertaking large-scale changes or seeking out a solution all on your own isn’t possible for smaller organizations. Not every retailer can charter their own container ships to avoid the headaches and soaring costs associated with shipping like Walmart, Target, Costco, or Home Depot have.
Still, retail organizations large and small can start planning for the future and setting expectations with their goods and the data behind them so they know what levers to pull, and when. Businesses need to demand more transparency out of their supply chain, full stop. They need to choose who they work with and choose alternatives when they feel they aren’t transparent enough in their processes or their data in the supply chain.
This will not be our last supply chain crisis, so preparing and working with the right vendors is essential.
In the meantime, companies can take steps to optimize their current supply chains by talking to vendors. Major brands like Apple, Cisco, Sony, Dell, and HP have all subcontracted with original equipment manufacturers which has allowed them to fare relatively well in terms of supply over the last few months.
Retailers bear the brunt of the fallout from lack of supply chain visibility — the industry’s lifeblood is dependent on having the right items on the shelves, or having items on the shelves at all. As such, retailers need to demand more of themselves and of the other actors within the supply chain if the needs of their customers are going unmet. But without clearly defined answers to look for on the road to greater transparency, organizations will find they have plenty of data, but will be able to gain little foresight from it – no matter how many scenarios they run.