The federated planning model does not depend upon a utopian dream that ignores the inherent conflicts between supply chain partners (such as the need to maximize returns to their separate shareholders). Rather than assuming this extended enterprise can be “optimized” as a single entity, federated planning accepts that each will ultimately optimize alone … but this need not result in suboptimal solutions. Supply chain partners can collaborate to address the trade-offs and possibly even break constraints across the extended enterprise.
For example, rather than leave the product flow to a series of one-off ordering decisions, Kroger and Unilever — or Ahold USA Inc. and Procter & Gamble Company — could agree to strategically reshape their distribution networks and eliminate some of the redundant regional facilities that each operates independently, potentially loosening the tight grip Wal-Mart Stores Inc. has on both competitors and suppliers throughout the grocery supply chain. (For more on federated planning, see “Beyond Utopia: The Realist's Guide to Internet-Enabled Supply Chain Management,” by Keith Oliver, Anne Chung, and Nick Samanich, s+b, Second Quarter 2001.)
The third principle underpinning SCM is to employ cross-functional support systems, especially in information technology. Just like the first two principles, this one focuses on breaking the functional perspective at both the strategic and tactical levels; companies need process-oriented support systems that link across functions. Having independent systems for each functional area — too often the norm — encourages suboptimal decisions.
In the early 1980s, business systems were designed to support only narrow functional decision making, usually within single departments. Companies might have a distribution requirements planning system to determine a finished-goods ordering pattern, to minimize distribution cost. Simultaneously, a production, planning, and control system would independently optimize the manufacturing plan to produce the goods — quite possibly out of sync with the distribution requirements. The procurement department would issue purchase orders to suppliers with limited insight into manufacturing plans and no understanding of supplier economics. Often these systems were developed by separate software vendors, and linked only loosely, if at all.
Thanks in part to the fears of Y2K bugs, which persuaded companies to upgrade their IT systems, most major companies in the 1990s started to think outside the box. They implemented enterprise resource planning (ERP) systems to link and coordinate their disparate systems in support of an improved process orientation. Today, software vendors tout “eERP” for linking systems among customers and suppliers over the Internet.
Though integrated systems represent an appropriate evolutionary step, we remain concerned that too many companies treat these systems as an unmanaged “black box” — a device that provides answers through an unknown process, forgetting the old computer adage of “garbage in, garbage out.”
For example, in a recent engagement for a kitchenware supplier, we found that its multimillion-dollar investment in a new, state-of-the-art ERP system was going to waste. Like many other companies, this client had fallen prey to the overhyped promises from software vendors claiming that the newest system would solve all supply chain woes, virtually automatically. More than three-quarters of the items in the purchasing system had default values for the order quantity rather than an analytically derived order size. To address the problem, the consulting team loaded the system with appropriate targets while implementing a sales and operations planning process; this drove a 20 percent reduction in inventories while simultaneously improving service levels by 5 to 10 percent.
Rather than investing in “black box” transactional systems with ever-more-sophisticated algorithms, leading SCM practitioners are turning to specialty software providers such as Viewlocity Inc. (which merged with SynQuest Inc. in 2002) and Jonova Inc. These relatively small companies have developed tools to support tactical trade-off decisions across the extended enterprise with rigorous but comprehensible analysis. In our view, a separation of the tactical tools from transactional controls offers the best long-term answer to growing software complexity.