10-20-2026
This September I had the privilege of giving the keynote talk at Subaru of Indiana Automotive’s MRO (Maintenance, Repair, and Operations) supplier conference, organized by Daniels School alumna K-Lynn Reynolds. Subaru supply chain leaders opened the conference, held at their training center in Lafayette, by speaking about their current business challenges.
Rising supply chain costs, driven by tariffs and fuel costs, combined with sluggish automotive sales, threaten profitability and necessitate world-class supply chain management to succeed in the current business environment. I was therefore asked to speak about how Subaru and their suppliers can work together to meet these challenges. In my talk, “From Suppliers to Strategic Partners,” I focused on supply chains’ interdependent systems, what can happen when we rely too heavily on big data and AI tools, and more.
Supply chains are best viewed as three interdependent systems: material flow, information flow and cash flow. While material flow is typically the most visible, it is enabled by well-designed information flow and cash flow. For example, in a standard supplier-customer relationship, lengthy payment terms are generally considered a “win” for the customer. However, terms that strain the supplier’s cash flow can result in late material shipments or logistics delays, and ultimately, a worse result for the customer. Suppliers and customers who can develop collaborative relationships and work together to manage the flows of material, information, and cash have the potential to create real competitive advantages in their supply chain.
In a time of unprecedented technological change and more data than ever, too many companies are rushing to implement artificial intelligence tools without a clear understanding of how they will create business value. Decision-makers don’t need more data or more technology; they need more information. Of course, artificial intelligence tools can do a lot of routine data processing, create visualizations and dashboards, and alert managers to anomalies. However, scaffolding the quantitative information with business context and applying judgment still remains the purview of people. Understanding the decisions to be made and what information is needed must be a precursor to capturing specific data streams and choosing an enabling technology.
One of the most common challenges in a supply chain is accurately forecasting demand. Often, a supplier assumes that the more data they can get from their customer, the better they can forecast demand. For example, a manufacturer selling to a retailer places a high value on the point-of-sale data gathered when the retailer sells to end consumers. This may consist of massive amounts of data, gathered item by item, location by location, and second by second. Modern AI tools can process all this in minutes to find trends and relationships among products purchased. This may be useful, but it is also rear-facing — it tells what has already happened. It does not tell the manufacturer that the retailer is running an upcoming sale or that the retailer has decided to no longer carry a competing product line. It is therefore critically important that the retailer is sharing this type of qualitative information with the manufacturer on a regular basis.
Further, when data sets get large enough, it becomes more and more likely to encounter “red herrings” during analysis. Suppose our theoretical retailer is a grocery store selling 30,000 different items, or SKUs (stock keeping units). An automated analysis of relationships among items would therefore be looking at 900 million pairings. Let’s say our software tells us that sales of mustard increase at the same time as sales of hot dogs, buns, windshield wiper fluid, and paper plates. Most humans could immediately spot that one of these is likely just coincidence. An AI tool may not. Too much reliance on big data and AI tools can easily de-emphasize the constructive relationships that ensure data is properly collected, curated and shared in a useful way.
Going beyond data shared for demand management, trustworthy supplier-buyer relationships allow for the exchange of information aimed at continuous improvement across the supply chain. An understanding of a partner’s cost drivers and major risks can lead to design or operating decisions that increase efficiency and lower cost for both sides. Typical contracts focus on purchase price, freight and tariffs, lead times, and quality standards, but can leave out many other facets of the total cost of ownership, such as repairability, disposal costs, and reputational risks. Typical suppliers do not want to lay out the exact details of their input and conversion costs to a customer, as this may reduce their negotiating leverage, but discussing the volatility of a raw material price may, for example, open up a conversation about an alternative design.
Right now, global supply chains are under increasing pressure due to rising oil prices, geopolitical uncertainty and a record number of weather disruptions caused by climate change. As manufacturers, distributors and retailers seek to respond, solutions that lower costs at the expense of another link in the supply chain are doomed to chip away at supply chain resilience. Instead, strategic partnerships and a continuous improvement mindset will allow organizations to develop their supply chains in a manner that creates long-term competitive advantage.
Amy David is a clinical professor in the Daniels School’s Supply Chain and Operations Management Department. She teaches experiential learning in operations management, and undergraduate and graduate courses in production planning and control and supply chain management. David has been recognized by Poets & Quants as a “top 50 undergraduate business professor,” and has received a Council of Supply Chain Management Professionals Teaching Innovation Award. David, along with the Daniels School’s Qi Feng, is the author of Architecting Global Supply Chains, now available from World Scientific Press.
