by
Louis J. De Rose, CMC
DE ROSE AND ASSOCIATES

Xerxes (a real company, fictitious name) is a midsize producer of semiconductors for the networking and telecommunications industries. The company is a subsidiary of a larger parent, which is headquartered in another state, but which does business worldwide through multiple divisions and affiliates. Xerxes is a fabless operation, doing, in-house, only product design and final test. It contracts out wafer production and testing, as well as component assembly and packaging to outsourcing providers in Asia and Latin America. The company’s product mix is extensive and steadily expanding, due to fast growth in developing markets. Given the prospects for continuing growth, and the complex structure of its supply base, Xerxes decided to initiate a Supply Chain Management project.

An SCM task force comprising representative from Engineering, Marketing, Operations, Procurement-Logistics, and Finance was assembled and met on a weekly basis. Its mission was to study the supply chain situation, and come up with an action plan in three months time. An outside consulting firm was retained to assist the task force, and to guide it through the study phase to the point of implementation. Three months stretched out to five months and "Supply Chain Management" remained more a theoretical concept, and less an implementable initiative.

As a consultant retained to assess the company’s Procurement and Logistics operations, I saw first-hand the reasons for that initiative to fail. To cite only a few of those reasons:

Item: There was little real commitment from senior management to make the project succeed. Lip service was paid to the SCM concept, but management’s prime concerns were increasing market share, speeding up new product introduction, and meeting the profit projections made to the parent company for the next fiscal quarters.

Item: At the task force level, there was continuing debate as to the scope and extent of the Supply Chain model and structure. Operations insisted it began when wafers were tested and received from foundries. Procurement-Logistics argued that it must begin at the product design stage. Clearly, it contended, once Engineering qualifies wafer, test, and assembly sources, they’re locked into the company’s supply chain, for better or for worse. Hence to ignore Product Design as an integral factor in the Supply Chain paradigm, would be to invite failure.

Item: In Xerxes’ period of fast growth and high capacity utilization, management’s priority was to build to total demand, regardless of distinctions in product mix, market segments, or specific customer demand patterns. Considerations of revenue contribution, high versus low volume demand, standard or unique application, susceptibility to engineering change, or early obsolescence were ignored in the planning and scheduling process. The result was that when business fell off, revenue and inventories were badly misaligned. This led to write-downs on excess inventory, price-cutting and reduced margins on salable product, and serious budgetary and cash flow problems.

Item: Xerxes’ information systems included software from multiple vendors, each addressing one or another element of the supply chain process. Further, with no common platform to integrate them, there were serious disconnects between one element and another. For example, one system generated forecast information, inventory levels, projected demand by product and customer, all in quantitative terms. Another system reported inventory in terms of dollars, enabling management to monitor bookings, billings, and inventory levels from a macro-financial level. Needless to say, this disconnect made for problems at the operational level in prioritizing resources and capacity, as well as scheduling day-to-day production. Do we build what the forecast and experience tell us to build, or do we react to Marketing’s or Finance’s sense of priority?

Item: Much of the planning and scheduling at Xerxes was manual. Although there was computer-generated information on billings, forecast demand, finished and in-process inventory, planners still used spread sheets and personal judgment to fill in the gaps. For example, by analyzing the forecast, planners might determine a requirement for a specific die type (a die is separated out of the wafer, and becomes the base for assembly and final product). Accordingly, an appropriate number of wafers would be ordered from the foundry. Devices A, B, or C can be built from the same die, and the forecast may show requirements of 1,000 for each device, or 3,000 in total. However, bookings may come in differently. Actual demand for device A may be 2,000, and for devices B and C, it may be 500 each. The result is that customers for device A would be shorted, while quantities for device B and C would be greater than needed.

Item: The parent company introduced a Customer Relations Management (CRM) system designed to integrate inventory information from all divisions and subsidiaries. The system was designed to enable the parent to provide real time information on availability to all customers and prospective customers, world-wide, across its total product mix. By doing so, it was believed that field sales staffs could be reduced, while demand could be more effectively determined through a centralized corporate Marketing activity. It was also believed that by centralizing demand information, planning operations at Xerxes could also be reduced. Where before, field sales could assess from personal contact, the reality of customer demand projections, and where planners could manually allow for unforeseen demand fluctuations, the only information now provided to Planning were CRM extrapolations of past demand. All too often, these proved to be unrealistically high or low.

Lessons to be Learned:

From just these few examples, there are clear and powerful lessons to be learned about contemplating and initiating a Supply Chain Management project:

1. Make absolutely certain that there is strong and continuing support for the SCM initiative from senior management before initiating a project. Without that support, the project will degenerate into a theoretical exercise, or worse, be seen simply as just another "management fad".
   
   
2. Create a cross-functional team or task force with strong, clearly identified leadership to guide and direct the conceptual and implementation phases of the project. Without that leadership, the project will flounder, go off course, or merely cease to function.
   
   
3. Establish clearly spelled-out goals as to what the SCM project will accomplish, with specific metrics for results, and clear time frames to ascertain results.
   
   
4. Make certain that the Supply Chain Management initiative is aligned with the company's business objectives and current IT technology. Although SCM demands the employment of IT, its aim is not to advance IT. It’s to achieve business and operational goals. Further, the technology it employs must be compatible with systems and programs across the entire business spectrum.
   
   
5. Finally, remember, for so complex a process as Supply Chain Management to succeed, human oversight and intervention are essential. Software systems and codes are no substitute for knowledgeable judgment and functional experience.

Customer Relations Management systems cannot supplant the information and insights on true customer demand that continuing face-to-face contact with customers can provide. MRP systems cannot always address beforehand the consequences of "what if" scenarios concerning supplier quality or delivery problems, engineering changes, market volatility, or sudden changes in customer buying pattern. And unless the Enterprise Resource System is state-of-the-art, and employed much beyond today’s current capabilities, there will be serious disconnects between engineering, financial, production, and procurement-logistics information bases and information needs.

It is a prime task of SCM project leadership to identify areas of IT weakness, and to compensate for that weakness through timely human and functional input.