Systems — True Pull Based on Rate-based techniques. The most destructive element in today's producer organization is the use of old MRP logic for inventory management which is embedded unerringly in all ERP systems and the many support mechanisms associated with or integrated into ERP, including customer relationship management (CRM), supply chain management (SCM), product lifecycle management (PLM) and other newer ones. ERP systems universally have scheduling and inventory replenishment logic that is essentially the same as it has been since the 1960s when manual processes were the rule. Designers of MRP and then MRP II and then ERP, etc., have all been like lemmings, re-using the same logic that Ollie Wight developed in the 1970s — Time Phased Requirements Planning. It has never worked well due to the overwhelming amount of highly detailed data that defeats rational analysis, the rigid inventory safety stock floors and the high level of change orders.
The ERP logic is at its core a push system, which will fail miserably in the new demand economy. It is not surprising that the Asian manufacturers have rejected ERP systems almost universally. Some think this is because they tend to be small and have low cost overhead. My contacts in Asia simply do not trust ERP systems and would rather use direct communications and visual signals, which were used successfully by Toyota. Unfortunately, there are very few companies who have the years' long product life cycles, fixed factory operating rates and ultra-low complexity found in Toyota and other auto manufacturing plants. So, simple manual and visual processes for planning, scheduling and control will not work in the demand economy.
The new approach is demand pull balancing technology. You don't need to throw out all of ERP, just replace the logic streams and keep the communications and data handling structures.
- Demand — Demand drives replenishment, we make what is selling or will very likely sell in the next time period, like next week.
- Pull — The rate of customer demand pulls inventory and resets schedules, all the re-order points and kanbans and min-max systems are removed. Rate-based systems will be used.
- Balancing — The key is to bring elements into balance. Risk of a service failure must be balanced against supplier realities, production resource economics and response rewards.
- Technology — The new tools are computer based and replace the inventory and scheduling logic inside ERP.
The new balancing concept will replace the logic inside logistics, warehousing and procurement systems, as well. Most inventories will be kept only in those items that are guaranteed to sell and sell very soon. Making C items frequently or on demand does not constitute a pull technique. C items should either be non-stock with long replenishment or, if they are more important, kept in inventory and made only infrequently.
External Relationships. The collaborative approach must be spread outside the organization. Management by hammering suppliers will give way to management by collaboration. Decision-making processes throughout the network should be linked, and probabilistic modeling will be relied upon for replenishment and demand balancing. A complete stream of success must be forged with a win-win approach.
Further, the new economy may require a reverse rationalization of a company's supplier base. More rather than fewer suppliers are probably needed to meet the anticipated variability in demand promised by the new economy. A company's sources will depend on its customer value proposition; the more variable the needs, the more flexible the network that will be required. Suppliers should be grouped around value spaces, with different suppliers by space. Overseas sourcing is not necessarily the right tactic, as a longer supply chain is less flexible and may not be able to meet the needs of the new economy.
Physical Network. To meet demand economically and more flexibly, a company should challenge where it holds raw, semi-finished and finished materials, where it sources raw materials and components and the number of distribution centers, among other physical handling factors.
The operations research (OR) professors have long known that making these determinations requires a dynamic modeling technique that deals with demand variability. Dynamic simulation modeling is the correct OR technique. Simulation has been in limited use for network design because the relationships of items to sources to stocking points to customers are so numerous in larger companies. Simulation engines capable of handling millions of relationships have not been available until 2005 when an Ann Arbor software company pulled a partially developed network simulation model out of a failed dot-com business and developed a new engine.