Many companies have developed very efficient supply chains, but are they smart? What makes an intelligent supply chain? Or is it a silly question to begin with?
Well, actually I tried to answer the question a few years ago. As some of you know, I took a brief spin as an analyst at META Group (recently acquired by Gartner) a number of years ago, and in 1999 I wrote a piece on “Crafting the Intelligent Supply Chain” that garnered some reasonable attention at the time. I thought it would be interesting to review that piece almost seven years later to see how the model played out, and to serve as a catalyst to thinking abut intelligent supply chains today.
Here are the capabilities I said we needed to build a “smarter” supply chain:
True visibility to actual demand: My take then: “Understanding true customer demand is essential for a pull-driven approach to supply chain management. A company’s own internal processes and technologies must also flow true demand information freely to all parts of the organization.”
My take now: We still haven’t really solved the bullwhip effect after all these years, though our understanding of the problem and the opportunity has increased substantially. Many companies are making solid progress to be more “demand-driven.” Better visibility to actual demand and a supply chain that can respond appropriately to those signals across functions seems a fundamental requirement of an intelligent supply chain. We also have examples such as retailer Canadian Tire, which has used granular forecasting data to align its internal and external supply chain in a way I would argue is highly intelligent.
Web-enabled global visibility: My take then: “Near-real-time track and trace of inventory at the SKU-level must be established for both international and domestic logistics activities.”
My take now: In 1999, visibility was a vague concept that hadn’t really gained a lot of traction. It’s gone mainstream now, though it still means different things to different companies. The “global” part is essential to manage offshoring and globalization efforts, and we’ve reported on these pages on the visibility efforts and results of companies such as Cisco, HP, Payless Shoes, and others. “Visibility” is clearly part of an intelligent supply chain, though I believe the definition has expanded from what I wrote then. For many supply chains, visibility to multi-levels of their supply chains is also a key element of intelligence, for example.
Componentized application architectures: My take then: “Componentry at increasingly granular levels will enable tighter systems integration, deployment of specific capabilities as rapidly as needed in company-owned and third-party facilities, faster customization of functionality without destroying upgrade paths, and event-driven architectures in which awareness of events triggers intelligent reaction and processing.”
My take now: This sounds like analyst gobbledygook, frankly, but contains buried within it the principles of Service Oriented Architectures (SOA) that are all the rage now from a technology perspective. SOA enables flexibility, which I suppose is related to “intelligence” but isn’t quite the same. What SOA can do, however, is enable more true event-based processing, which can improve how information flows and enables things like alerts for exceptions or potential problems.
Real-time planning/execution linkage: My take then: “Linkages between supply chain planning and execution systems must occur both at the data and process levels.”
My take now: The walls that existed then between planning and execution still need to come down, and we need more “smarts” embedded in our transactionally-focused execution systems. This one is easy to say, but harder to get specific on, but “real-time optimization” is partially here and will expand.
Reporting and analytics: My take then: “Deployment of supply chain analytic applications that compare planned results to actual performance. This will give rise to the new concept of "logistics statistical process control,” using concepts widely deployed in manufacturing environments to inform supply chain operations around predefined tolerance bands and reduced performance and service variability.”
My take now: I don’t know if we’ll ever really develop the concept of supply chain or logistics SPC, but we have certainly seen broad and growing use of Six Sigma methodologies within our supply chain processes. There’s also no question the use and sophistication of supply chain metrics has increased substantially since 1999. Metrics are critical to the intelligent supply chain, and we are likely to see them become increasingly “real-time.”
Common messaging-alert system backbone: My take then: “Messaging and alert systems that cross multiple applications should be deployed to achieve supply-chain-wide intelligence.”
My take now: This basically meant we should have “event management systems” that ultimately covered the entire supply chain. Very few companies have moved much beyond very basic event management/ exception systems. There are some exceptions, such as Best Buy tying inbound logistics visibility with merchandising and store delivery programs. More intelligent event management will be an important attribute of a smart supply chain, but I don’t think we’ll have or even need an “uber” event management systems across all supply chain processes.
I think I didn’t do too bad back in 1999. As was common at the time, the analysis was heavy on technology and to a lesser degree process, and sort of ignored the people part. Clearly part of an intelligent supply chain today would also be how we organized and incented our people, but that’s for another day. I think this also shows how long it really takes things to change. Seven years later, we’ve made progress, but most companies and technology vendors are still working on these same issues.
What do you think makes an “intelligent supply chain?” Is it just another word for responsive? What would you add or subtract from this model? |
