By Robert Handfield, Bank of America University Professor of Supply Chain Management
Everyone seems to be developing a new AI platform that will forever change the world of supply chains. But will it really? Let’s review a few facts about AI and what it means for retail business strategies. Artificial Intelligence is the study of the “general principles of rational agents and components for constructing them” (Russell and Norvig 2016). On the other hand, agents are systems that can take actions based on the inputs they perceive. A rational agent is thus “one that acts so as to achieve the best outcome or, when there is uncertainty, the best expected outcome” (Russell and Norvig 2016). Business AI, therefore, is the development of artificial agents which, given the information they have about consumers, competitors, and the focal company, suggest and/or take actions to achieve the best business outcome—in our case, for the entire supply chain.
It is important to note that AI can only make suggestions about what might achieve the best outcome, but ultimately, it is humans who must make the decision. Traditionally in supply chain management, when we use the word “model,” we are describing a descriptive model, i.e., a model that describes the observed data. An optimization model will use existing data and then describe what is an optimal solution, which may or may not be practical in the real world to implement. On the other hand, machine learning tends to focus on predictive modeling—models that predict future data points. Predictive models are aimed at helping computational agents improve their performance. We can also use predictive models on historical data to determine how well the model is able to predict the future. In most cases, there will still be a lot of errors, as retail demand is always shifting and changing.
For this reason, I believe the emergence of AI is going to change how managers work, how they plan, and how they manage supply chains. But one thing it will not do is eliminate uncertainty. Let’s come to the realization, after having an honest and hard look, that these supply chain challenges are not going to go away—even with the use of AI!
In the retail supply chain world, this rings more true than ever. The reality is that even though we are in the post-COVID era, we will be living in this world of instability for another five years. Consider the following:
The conflict in Ukraine shows no sign of slowing down.
The Middle East is in turmoil. Hamas and Israel will be in conflict for years to come, and the Red Sea Houthi attacks funded by Iran have no end in sight.
The presidential election in the US has resulted in a Trump Administration, which is threatening 25% tariffs on Mexico and China, disruptions to the Panama Canal, and further tariffs on China.
Left-wing governments in Latin America are refusing to honor the fair election process, causing more economic turmoil.
In this environment, these problems will not resolve themselves overnight. They are not going to go away. Although the latest and greatest AI technology claims to be the solution, the reality is that the fundamentals of supply chain management remain essential. Moreover, the ability to plan and have visibility to what you have in inventory is still the problem we need to solve. When you peel all the technology back, many managers will admit that “we don’t know what we have and where it is—and we don’t know accurately what we will need in the coming year.”
The Need for A Different Solution
For too long, companies have relied on historical demand as the basis for forecasts. The typical approach when planning for next year is to take what happened last year, make a few tweaks based on what you guess you need to change, and there you have your forecast. But this isn’t working anymore. You can’t use that historical pattern anymore. Everything is so different. It changes all the time, and the volatility and direction of change are impossible to predict. The other element that is fundamental to supply chains—and critical to surviving the volatility ahead—is the ability to integrate planning with our key supply chain partners. If we look past the legal and confidentiality barriers, it helps to envision what we must do to improve our ability to react to unexpected swings in demand.
To enable this conversation, it is important for retail supply chain managers to ask themselves the following questions:
What are the biggest challenges you face today?
What types of data and technology solutions would you like in an ideal world?
If you and your supply chain partner were both a single company, what would you do differently?
How would you share information?
How would you structure your supply chain design?
How could you become more efficient?
Beyond a simple data dump, what kinds of information are required to make these changes?
In the current environment, moving the needle even 1% in improved availability on the shelf would equate to millions of dollars. This is the only way to increase margins in this environment. One possible outcome to move forward would be to host a Roundtable, with the following questions to be addressed in an open dialogue:
What are the rules to create a safe space for retailers, sellers, transport providers, and distributors to have a dialogue on how to achieve improved outcomes?
How can all parties make money, put at risk their competitive insights, and come together to solve this problem?
This is a whole new level of complexity in the supply chain. Using seven years of historical data won’t cut it anymore. Let’s figure out what the plan is.
Disclaimer: The opinions expressed in this article are those of Robert Handfield, PhD, Bank of America University Distinguished Professor of Supply Chain Management; Executive Director, Supply Chain Resource Cooperative; and Director, Ethical Apparel Index Initiative. The Supply Chainer’s Insights are submitted content. The views expressed in this column are that of the author and don’t necessarily reflect the views of The Supply Chainer.
Comments