The Future of Purchasing and Supply: Demand-Pull Possibilities

#7 Take the Initiative: Continuing coverage of the trends shaping the profession

By Roberta J. Duffy, writer for Purchasing Today®.
January 2000 Purchasing Today®, page 48.

Driven by customer requirements and aided by technology, demand-pull purchasing concepts allow material and information to traverse the supply chain in a seamless fashion.

A sale becomes final and the customer is happy. It sounds like the end of a transaction, but in fact it's just the beginning. One simple action triggers a whole chain of subsequent events - that's the idea behind demand-pull purchasing. The process begins as a need is recognized to build or replenish a sold product. From there, the various links in the supply chain receive the information - perhaps through the Internet - and the materials are subsequently "pulled" by the demand, down through the chain. According to "The Future of Purchasing and Supply: A Five- and Ten-Year Forecast," these concepts, which are in place to some extent in certain capacities, may be used more prevalently in the coming years as purchasing and supply organizations search for ways to create seamless links along the supply chain.

The evolution of demand-pull purchasing concepts derives from goals aimed to reduce inventory and streamline processes. Traditionally, if a forecast were given to build 50 units, perhaps varying in color or some other customized factor, a schedule would be sent to the production facility and kits would be put together to build those units. However, if orders came in that didn't match the forecast - perhaps all the orders were for a single color - then more materials were needed for the demanded color and kits for the other colors sat idle or were "robbed" for parts. One possible solution is to wait and have the needed materials or components pulled only when a specific trigger gives the signal. The trigger could be a specific order, which prompts the pull to build that order, or a sale that prompts the build to replenish the sold item.


Organizations that have these types of systems in place realize benefits such as inventory reductions, labor cost savings, and more efficient processes because purchasing and supply management professionals spend more time procuring those materials that are actually being sold.

Take the example of a large equipment manufacturer that must coordinate the activities of suppliers, manufacturing locations, dealer locations, and final customers. Demand-driven procurement systems could be used so that final product sales are pulled through the supply chain. Dealer orders would always match what is in assembly if the system were used as a build-to-order model, but could also be used for automatic replenishment of component parts within service facilities. Ideally, as orders are electronically entered, assembly line orders are automatically produced, and the individual component demand is determined. This information is then electronically transmitted to suppliers, who pull accordingly. The demand-pull system helps to maximize material velocity, which minimizes the amount of redundancy in the process. Manufacturing based on demand also means that organizations don't have to reconfigure product to meet specialized needs of customers. This improves quality and ultimately provides a great deal of customer satisfaction.

The Prime Environment

In what type of environment does this model thrive? The first factor for success is that demand-pull purchasing systems should operate in an environment with relatively steady and robust demand. Demand-pull allows for flexibility to accommodate customization and special requirements, but it's optimized if at least a reliable volume of product or activity is anticipated. For this reason, many organizations have found success with demand-pull concepts, but have come to realize that they can't be used for every assembly line. They need to have some predictable degree of volume, then the pull is triggered for a particular component or feature.

Although demand-pull purchasing models evolved as a contrast to traditionally forecasted models, the two concepts must work in conjunction with each other. For demand-pull to be successful, organizations must have some reliable forecasting information with which to prepare suppliers. "Forecasting should be used, not to produce a work order, but to give suppliers an accurate estimate of potential," says Drew Curtis, C.P.M., CPIM. Curtis is manager of strategic sourcing programs for Northeastern United States and Canada, for TTI, Inc., a distributor of electronic components. "Providing the most accurate forecast as possible lays the framework. No matter how automated and streamlined the system, if a supplier had no forecast or history whatsoever and then got a trigger for a year's worth of supply, it wouldn't be prepared." This reinforces the need for integrated relationships and solid two-way communication between supply partners.

Another factor to consider is that demand-pull systems rarely exist as an isolated model. The organizations that are employing them successfully do so in conjunction with other inventory and production methods. Certain product lines, such as those with consistent demand, might work well in the demand-pull model, while other commodities, where need isn't as consistent, can rely more on traditional forecasting. In certain situations, consigned inventory or on-site stores might be the solution. The key to remember is that demand-pull - alone - is typically not an absolute answer to all inventory issues.

Even within those subsystems where demand-pull purchasing is employed, mechanisms should be in place that trigger occasional human intervention or deviation from the automated processes. For example, boundaries can be set for the automated system to send flags when a request is made that falls above or below certain parameters. These flags would prompt human intervention to determine the order's validity or special requirements. "Internal controls are also required if specific time periods warrant the demand-pull system to be turned 'on' or 'off,'" says Curtis. "In an environment such as contract manufacturing, customers and their requirements change frequently. If a system is not in place to identify these changes, there is a potential to signal for unneeded replenishment. Automated systems do not replace sound materials and supply management principles or responsibilities."

Though it has its roots in Just-In-Time (JIT) and kanban philosophies, demand-pull purchasing is not limited to the manufacturing industry. The concept can be applied to retail locations, with the point of consumer purchase becoming the trigger for replenishment up through the supply chain. Also, non-manufacturing organizations that use capital equipment might use demand-pull models to monitor and respond to wear, usage, and supply replenishment. Imagine a commercial fleet of vehicles or airplanes being constantly monitored for fuel consumption. As fuel is consumed, more is automatically purchased. Purchasing and supply management professionals must be aware of their total cost of ownership for their products in order to determine what automated system is best for them, according to Curtis. He warns that a purchaser must start to examine the process costs. "If you have a very low-value product and hundreds or thousands of part numbers to manage, is the time investment to manage the order points and order quantities inherent in demand-pull systems worth it?" These are the types of questions organizations must ask themselves and the area in which the purchasing and supply expertise in evaluation is the most critical.


Initially, a first reaction to this type of automated system might be a fear of losing control. After all, orders are being released for materials with little or no human intervention. Suppliers and other business partners have access to order, forecast, inventory, material, and sales information. Upon closer examination, the control issue is not the hurdle, but the information systems. "Throughout so many industries, people have been working with these value-added programs and then those people move on to other organizations and industries, becoming more and more familiar with the concepts. So, as this happens, the inclination of 'I don't want to give up control' isn't as prevalent," says Curtis.

The issues that do present a challenge revolve around information systems. With an electronic interface, suppliers are able to digest information and upload it into their system. However, it's not the technology behind these systems that is the big concern; it is more a matter of each supply chain partner having the systems. "The Future of Purchasing and Supply: A Five- and Ten-Year Forecast" says that the Internet will provide the common platform by which to transmit this information in the future. Some organizations use an MRP system which can interface with the first point of demand, such as the point-of-sale or point of order at a sales location. That system must then be able to convert the order into schedules and requirements for the manufacturing and supplier locations.

As these processes evolve even further in coming years, opportunity exists for more refinement. For example, as manufacturers and suppliers await the "pull" trigger, there might be inventory located in various locations. In the future, organizations should look to maximize inventory ownership along the supply chain. Based on customer demands, they will be able to determine what parts, components, materials, and finished products should be in which locations at which times, for the best service and cost efficiency.

C.P.M. Alert Module 2

Initiative #7: Demand-Pull Purchasing

"Box page 48"

The 1998 study, "The Future of Purchasing and Supply: A Five- and Ten-Year Forecast," by NAPM, the Center for Advanced Purchasing Studies, A.T. Kearney, Arizona State University, and Michigan State University, details 18 initiatives that will shape the purchasing and supply management function in coming years. This month, Purchasing Today® provides commentary on one aspect of the study: Initiative #7, Demand-Pull Purchasing. The study offers the following predictions.

  • In instances where suppliers are becoming more integrated, demand-pull systems could become fully implemented.
  • The Internet is going to provide the information for these models.
  • Many organizations are currently investing large sums in demand-pull systems to provide a seamless link between suppliers and purchasers.
  • Organizations are trying to become more sophisticated with their customers, on a part-by-part basis around the world.
  • The challenge will be getting the systems to work together.
  • The required systems will be enabled by enterprise-based software and Internet technology.