Demystifying Supply Chain Management
Paradigms in the supply chain industry are like the ball in a kids' soccer game:
where ever the ball goes, it is pursued by the majority of participants despite
the urgent direction from coaches to stay in position.
Vendors flock to these models because they are heralded with claims of "breakthrough technology," and the "leading provider of ..." but when supply chain event management (SCEM) is examined, we must ask, "provider of what?" The key issues for the SCEM system buyer are the features and functions they are acquiring, and how closely they relate to their business needs.
When the buyer decides to purchase a warehouse management system, he knows that the basic functions include receiving, putting away, moving, picking, shipping, and accounting for goods. When he buys a transportation management system, he know that the functions include receiving and managing orders for shipment, planning shipments (consolidations, carrier assignments, routes, and schedules), and shipment management from booking through freight payment. The same is true when purchasing enterprise requirements planning systems, though the application functions are more extensive.
SCEM has no similar, clearly-defined functional area. Event management implies the proper detection, reporting, and reaction to issues that arise in the supply chain. Since no two supply chains are identical, then by implication, no two SCEM systems should be identical. Since companies have disparate priorities and issues, the events that are a priority for individual companies to manage vary. Ideally, the SCEM system would contain an abstract set of features and functions that can be profiled as needed to the specific situation being monitored.
The SCEM system should act like an intensive care monitor in a hospital. To use an intensive care monitor, the doctor places probes at strategic points on the patient's body; each measures a discrete and different function—temperature, respiration rate, blood pressure. The monitor is programmed with separate upper and lower control limits for each probe and for each patient. If any of the watched bodily functions go above or below the defined tolerance, the monitor sets off an alarm to the doctor for immediate follow-up and corrective action. The SCEM application should act in the same manner.
The company determines its unique measurement points along its supply chain and installs probes. The company then programs the SCEM application to monitor the plan-to-actual supply chain progress, and establishes upper and lower control limits. If any of the control limits are exceeded, or if anomalies occur, the application publishes alerts or alarms so that the functional manager can take appropriate corrective action.
For example, if a key issue for a company is on-time delivery of inbound shipments to its receiving facility, it might compare inbound shipment status from carriers and vendors to their own purchase orders' required delivery dates. They might establish control limits of one day early as acceptable, but no more than one hour late as acceptable. Any status report that did not meet that criteria would cause an alert to be sent to the distribution center manager (and others).
Figure 1—The SCEM application is an umbrella system that covers the entire supply chain.
SCEM is an umbrella application, encompassing the entire supply chain and all of its processes, but owning none of them. It is a parasitic application because it "borrows" documents, key events, and issue data from other applications. It analyzes the content of these data and determines what is important, and to whom. When critical and timely filtered information is available, businesses can increase profits and reduce operating costs, inventories, and order/ship cycle times. When management has the right information, it can analyze each link in the supply chain and change any logistics process that increases costs, limits customer satisfaction or profits.
Business Issues: Data Acquisition and Analysis
The SCEM application requires a comprehensive database of information. In order to monitor the global supply chain, the company must synthesize a significant number of disparate documents and status reports from various internal company systems and from various trading partners (particularly suppliers and logistics providers). This can represent hundreds of thousands of messages in varying formats and media.
The company must analyze each function and system, and decide what data SCEM must receive and what decisions it should make. This implies defining the rules to both establish the plan and then monitor the execution of the plan. The rules must analyze the performance-to-plan, discover critical business events, and react to them. The SCEM application must inform proper individuals in a timely manner—both when critical activities happen and when they don't. The company must also decide who are the information customers and determine their needs.
The SCEM application should allow a company to detect anomalies sooner, providing more time to react and correct the issue. Often, supply chain managers find that their ability to correct a problem goes up with the amount of time they have to address it. The SCEM application should push critical information to interested parties—it should not wait for them to discover it. Individuals can subscribe to relevant events. The system only tells them when something important happens (or doesn't happen).
The SCEM application should also provide information and visibility for customers and other trading partners to better manage their processes (and, in the case of providers, to better serve the company). SCEM provides information to manage change. Since SCEM collects a complete supply chain picture, it can provide holistic business intelligence and performance metrics (report cards) on both internal and trading partner processes. The company must decide how to report performance over time to improve the process and trading partner relationships.
A successful SCEM system requires strategically incremental continuous improvement. The most critical issue is data compliance and quality. The company will need to monitor data quality (content accuracy) and compliance (providing data on a timely and complete basis). This implies the need for data performance metrics and trading partner performance reporting.
Finally, the company will want to use the SCEM information and feedback to create improvements to plans, processes and procedures, ways-of-working, and metrics. By analyzing metrics and considering best practices, the company can use accurate and timely supply chain information to reduce inventory and improve operational efficiency (especially to reduce lead time). The SCEM system can become a source of competitive advantage.
Successful SCEM Applications
These issues are synthesized from nearly 15 years of experience in building SCEM applications, and observing others do the same. In this time, we have developed key observations that can be very beneficial when a company evaluates or builds SCEM systems.
Recently, many companies' application systems concentrated on forecasting and planning systems. Notably absent was the capability to monitor any plans developed once executed. Without the ability to analyze performance-to-plan while the execution is in progress, the supply chain manager cannot determine the overall compliance of the execution to the plan. Without status reports and variance analysis, the supply chain manager can neither determine where issues have occurred nor analyze trends.
Our experience indicates that effective management of the supply chain can be achieved by implementing a three-part program that plans, monitors execution, and analyzes activities on a cyclical basis. This is not a new idea, but effective implementation has eluded many organizations. A SCEM system would span both the "Execute" and "Assess" components (Figure 2).
Figure 2—The cycle of SCEM planning.
The essential issue is process interdependency. For example, advanced planning and scheduling (APS) produces the plan that SCEM uses in making plan-to-actual comparisons. SCEM produces status reports and logs anomalies, which are used by a business intelligence module to assess compliance, analyze and evaluate trends, and pinpoint service failures and supply chain bottlenecks. If the results are then fed back into the APS system to refine constraints, relieve bottlenecks, and address service issues, a more accurate plan results. The presence of all three components allows a company to establish a continuous improvement program and optimize the entire process over time.
The importance of proper business process analysis cannot be stressed enough. Before starting any systems evaluation or design, a company must know:
• What their key issues are
• What supply chain processes do they want to monitor and what meaningful
information they want from those processes
• Who the information customers are
• What is the necessary timing of the information—can it be acquired,
evaluated, and communicated in time to be meaningful
• What is the quantifiable impact of making supply chain information more
visible (better customer service, lower inventory, reduced operating costs,
etc.)
When these business case tenets have been determined, the company can then proceed to the business systems analysis: what are the information sources; how are they available (internal/trading partner), when is it available (what are the key events that trigger the information creation), what the media and information format is, what is necessary to give the information meaning (plan or context data), and so on. Once both business process and business systems analysis are complete, the company can evaluate if a commercial SCEM system is viable for their specific case or if they must create their own from components.
The most frequent issue we have observed with commercial software is that software designers tend to develop fixed event models. They do this by defining the commonly observed milestones and events in the supply chain and assume that these satisfy the majority of the monitoring requirements.
Our experience dictates otherwise—the most valuable feature an SCEM system can have is the ability to define the event model dynamically (recognizing that no two supply chains are the same and no two companies want to monitor exactly the same events).
The most common example is verification of on-time delivery. Commercial software packages tend to use either the delivery appointment (from dock schedule) or the shipment plan as the argument for comparison with shipment status from carriers. This evaluation is valid if the evaluation to be made is carrier on-time delivery performance. If, however, the true information customer is elsewhere in the organization, this evaluation may not be valid.
A number of years ago, we were asked to design an SCEM system for a major wholesale/ retail pharmaceutical distributor. The information they wanted was notification to the buyer about whether he would get his orders on time. The key issue that caused the company to require an SCEM system was that buyers were scheduling delivery with their vendors with sufficient lead time to meet redistribution schedules to their stores. The distribution center, however, was blind to the buyer's needs and scheduled delivery appointments based on next available dock slots and space available in the DC to put the goods away. Too often, buyers experienced out-of-stocks of critical items and goods for promotion because the delivery schedules set by their own DC personnel were too late to meet the required cut-off for distribution to the stores. Evaluation of carrier on-time delivery to the distribution center was of interest, but was not critical. The key evaluation criteria was comparing both delivery appointment and carrier delivery to buyer's want date. The event model was established to evaluate both criteria and to notify completely separate sets of information customers, depending on the issue detected.
In another application developed for a major manufacturer, their replenishment supplies to the manufacturing floor were not being issued in a timely fashion (or not at all). This caused extremely costly and wasteful production line shutdowns and nearly resulted in the de-installation of a new warehouse management and shop floor control system. The solution involved monitoring the material replenishment work queue for a response to the request from the shop floor for materials. If the event was not responded to in a timely fashion, an expediter was notified who was empowered to issue stock from reserves to meet the needs. The key event was acting on the absence of information (positive confirmation of an internal "shipment"), by setting expectations and alerting on the failure to satisfy the expectation.
Both examples speak to the critical need to be able to define events and consequences dynamically. A static event model will not satisfy many supply chain requirements.
Finally, those anticipating installation of SCEM applications must include plans to insure data quality and compliance. Any application that is dependent on other applications and outside trading partners for its data will go through a period of refinement.
In early stages of implementation, most companies find that both data quality and compliance is poor. They must institute active programs to insure that all partners comply with their data requirements for content and timeliness (regardless of media and format that the data is provided in). Lack of data quality and timeliness must be reported and compliance reviewed with partners.
Various levels of incentive programs have been established. One of the most effective (and onerous) programs we have seen was established by a major retailer. In order to plan re-distribution to their stores, they required their vendors to send advanced shipment notices (ASNs) for every shipment being sent to the retailer. The requirement was that the ASN must be received prior to the shipment or the vendor's invoice would be debited $5/case for the shipment. The program got vendors' attention.
There is a growing body of cases that suggest that an effective SCEM system can significantly reduce operating costs and improve service quality. It can help to deliver the promise of planning systems by monitoring performance-to-plan and providing data for effective continuous improvement programs.
If an SCEM program is approached with a holistic vision and is designed and built to address key issues and serve specific information needs, it becomes an extremely effective tool in optimizing the supply chain.
