The next organizational strategy we’ll look at is the demand-driven enterprise. The bull-whip effect is driven by demand forecasts; the solution is to replace the forecasts with actual demand information. This isn’t necessarily a simple matter either, but supply chain professionals have evolved techniques for letting actual orders (not forecasts) drive production and distribution. In the demand-driven chain, supply management is focused on customer demand. Instead of manufacturers planning their operations based on factory capacity and asset utilization, the demand-driven supply model operates on a customer-centric approach that allows demand to drive supply chain planning and execution—moving the “push-pull frontier,” as it’s called, back up the chain at least to the factory. Instead of producing to the forecast and sending finished products to inventory, the production process is based on sales information. There is, in other words, no fixed production schedule in a strictly demand-driven supply chain. Product is turned out only in response to actual orders, “on demand,” in other words. (Note, however, that on the supplier side of the plant, forecasts still determine delivery of raw material. The art of forecasting remains crucial, even in a demand-driven chain.)
This is also known as a “pull system,” and it entails the following:
- In production, the production of items only as demanded for use or to replace those taken for use.
- In material control, the withdrawal of inventory as demanded by the using operations. Material is not issued until a signal comes from the user.
- In distribution, a system for replenishing field warehouse inventories where replenishment decisions are made at the field warehouse itself, not at the central warehouse or plant.
- When a supply chain works in response to forecasts, it’s called a “push” chain, and it entails the following:
- In production, the production of items at required times based on a given schedule planned in advance.
- In material control, the issuing of material according to a given schedule or issuing material to a job order at its start time.
- In distribution, a system for replenishing field warehouse inventories where replenishment decision making is centralized, usually at the manufacturing site or central supply facility.
Everything in a push system is pushed downstream from one point to the next according to schedules based on the forecasts. The supplier delivers components in the amounts determined by the schedule to inventory, where they await use in manufacturing. The plant turns them into finished products and pushes the products to the distribution center or the retailer, where they await an order from downstream.
The challenge in changing from forecast-driven (push) to demand-driven (pull) systems is in reducing inventory without also lowering customer satisfaction. When a demand-driven system is set up and managed properly, it can actually enhance customer service while reducing costs. But stock outs are a risk. As always with supply chains, the decision to switch to a demand-pull process trades one type of risk for another:
- In the forecast-push process, the risk is related to the build-up of inventory all along the chain. Not only does inventory cost money while it sits in a retail stockroom, distribution center, or preproduction storage area; it runs the risk of becoming Obsolete or irrelevant for a number of reasons. In a world of sapid innovation, inventory obsolescence is a very real threat. (For example, Cisco Systems, for years an exemplar of successful and innovative supply chain management, had to dispose of US$2.25 billion worth of useless inventory when the dot-coin bubble burst at the beginning of this millennium. All those season close-out sales you see in clothing and department stores are a way of clearing out the overstock. Bookstore remainder tables (which are much less in evidence than they were a decade or two in the past) are a sign of inventory overhang caused by failed forecasting.
- Magazine distributors used supply chain strategy to destroy huge quantities of monthly magazines 12 times a year when they came back from retail outlets. (Since magazines are inexpensive to produce and destroy compared to their retail price, the distributors would rather destroy ten copies than miss one sale.) Those are the results of producing to forecasts that no one trusts and purposely overstocking to be sure of meeting unexpectedly high demand.
- In the demand-pull, make-to-order model, on the other hand, the risk is that orders will begin to come in above capacity and al/ along the chain there will be expensive activity to run the plant overtime, buy more and faster transportation, or sweet-talk customers into waiting for their orders to be filled or substituting a different product. (Running short of stock is also a risk in the forecast-driven chain. Forecasts can be wrong in either direction. That’s why the safety stock builds up at each point where orders come in.) One technique to prepare for uncertain demand is kitting, which is preparing (making/purchasing) components in advance, grouping them together in a “kit,” and having them available to assemble or complete when an order is placed.
In Gartner’s annual supply chain report, they rank the top 25 demand-driven supply chains, thereby underscoring the importance of this strategy. In fact, the companies that gain a position on this list have all applied demand-driven principles to coordinate supply, demand, and product management to better respond to market demand. If you would like additional information about this report, a link is provided in the online Information Center.
In reality, most organizations pursue a push-pull supply chain strategy and the point where push moves to pull is the key strategic decision. Once that decision has been made, building a demand-driven enterprise can require significant changes in all supply chain processes. The following are some major steps:
- Provide access to real demand data along the chain for greater visibility of the end customer. The first requirement is to replace the forecasts with real data. The only supply chain partner with access to these data first hand is the retailer, and retailers in the past have been no more willing to share business data than any other firms. The other partners lack “visibility”—one of the main supply chain principles. They simply cannot see what’s going on with the end customer. But visibility supply chain strategy is a necessity for building a pull system, and pioneers like Wal-Mart have led the way in that regard. With point-¬of-sale scanning or radio frequency identification (RFID), a retailer can alert its suppliers to customer activity instantaneously. Instead of producing to the monthly forecast, manufacturers with that kind of immediate signal from the front lines can plan one day’s production runs at the end of the preceding day. They produce just enough to replace the sold items.
- Establish trust and promote collaboration among supply chain partners. Collaboration is implied in the sharing of information. But more is at stake than simply sharing sales information. Partners may have to invest in new technology and develop new systems to be able to use the real-time data. With orders going out without a schedule, all processes will have to be altered—warehousing (storage no longer needed), packaging, shipping, and planning will all be handled differently in the new system. In return for receiving real-time data that allow reduction of inventory; suppliers and distributors have to agree to change their processes in whatever ways may be necessary to make the new system function without disrupting customer service.
- Increase agility of trade partners. Because the inventory buffers will not exist or will be much reduced in this demand-driven supply chain, the trade partners need to develop agility—the ability to respond to the variability in the flow of orders based on sales. The plant, for example, may have to undergo considerable change if it has to produce several different kinds of products under the new circumstances. When making to forecast, a plant can run a larger volume of each product to send to inventory. But when making to order, the plant may have to produce several different types of products in a day. There will be no room for long changeover times between runs of different products; therefore, equipment, processes, work center layouts, staffing,—or all these things—may have to change to create the capacity required to handle the new system.
