The Importance of Differentiating Push and Pull Supply Chains for Creating Unique Business Models

Author: Dr Keith McNeil

Introduction

There is a significant amount of supply chain management literature highlighting the business success that can occur as a result of creating customer or demand pull supply chains.

Defining Push versus Pull Supply Chains

Definitions are very important to create context for this article.  A supply chain is a minimum of a network of a business, its suppliers and customers.  Thus, supply chain management by a particular business is the management of human capital, processes, materials and information between that business, its suppliers and its customers that ensures maximum customer service at maximum margin to that business.  Importantly, while all participants in a supply chain can benefit from improvements to the functioning of the supply chain as a whole, rarely do they benefit equally.

Customer or demand push is usually defined as a business response in anticipation of customer demand and customer or demand pull as a response resulting from customer demand.  However, from a whole supply chain viewpoint, deciding whether a particular supply chain is push or pull is often difficult and generally depends on the perspective of what constitutes the supply chain and where particular participants are placed in the chain.  For example, the manufacture of Toyota automobiles is heralded as a leading example of a demand driven supply chain.

However, the mining of the iron ore or operation of blast furnaces that process the iron ore for ultimate manufacture of automobiles is not.  At some point in most supply chains, in their widest sense, demand push meets demand pull, and at this point inventory accumulates.  This point is referred to as the push-pull interface or as the supply chain decoupling point.

What is Best?

The advantages of operating to a pull model are very compelling for businesses that are able to do so, principally because the planned level of production and/or service delivery is not dependent on forecasts with their inherent inaccuracies.  The inevitable over-compensation for forecasting inaccuracy is inventory.  One of the five basic tenets of the “Lean Thinking” business model, founded on the Toyota manufacturing model, is to “Only make what is pulled from the customer at the rate of their requirement”.  The closer a business can move to real-time customer demand fulfilment, the greater is the reduction in risk of producing or locating in the wrong place what is not wanted by the customer.  In addition, a company operating in a demand pull mode can be expected to be better positioned to innovatively respond to changes in customer tastes and expectations.  This theme was one noted in a recent article published in the McKinsey Quarterly (October 2005) that was titled “From Push to Pull; The Next Frontier of Innovation”.

However, as suggested above, not all participants can rationally or easily apply a demand pull model.  A business operating in demand pull mode will strive to move the inventory decoupling point further upstream to its supplier or its supplier’s supplier.  For example, consider a manufacturing company (A) in Australia that exercises a demand pull model and uses components manufactured off-shore by another company (B).  This company, in turn, acquires its raw materials from another Australian bulk-exporter (C).  Company A will be rightly concerned at its risk of supplies if they are procured direct from overseas and may insist that B maintains inventory in Australia to manage that risk.  Company B because of concerns with respect to shipping lead times is likely to supply to forecast and becomes in respect to its Australian customer, a push supplier although, for its own domestic customers, it may too be able to operate a pull model.  Company C, because of even longer lead times will likely operate as a push model either because of long lead times to the ultimate end customer or because of a prime business imperative to capture economies of scale in its processing infrastructure.

There are in addition, a range of industries that typically require a hybrid push and pull model, notably resource and processed product supply chains, including forestry, pulp and paper, meat, dairying, mining and oil products.  Resource and processed product supply chains usually involve producing a large variety of high value and lower value end-product stock keeping units (SKU’s) from an original raw material of variable quality.  Whereas it is often possible to create demand pull supply chains for the higher value end products, the lower value end products or by-products are often variable in output and typically have to be pushed to end customers, particularly by using pricing to stimulate off-take.  The original raw material may be pushed continuously into the supply chain by a blast furnace or petrochemical plant or something far less sophisticated such as a cow being milked twice a day.

The usual prime determinants of whether a business operates in push or pull modes are economies of scale and the uncertainty of demand as shown below.  For example, an architect-designed custom built home is made to order in response to specific known customer demand. Clearly inventory is very low.  However, a housing developer builds a large number of homes to capture scale advantages but at the risk of uncertain timing of purchase and potentially will need to discount some for sale.  A “dial-up” pizza shop produces also to specific demand pull to supply specific customised pizzas.  However, there is an element of push in this instance.  To enable responsiveness, the pizza bases and ingredients are built as inventory in advance of the customer order.  The final assembly is delayed until there is a specific order.  This is an example of postponement and is an excellent example of push and pull operating in the same business.

Determinants of Push and Pull Business Model

Thus, “best” is creating, where possible, demand pull but “best” also means operating the most effective push or hybrid model if that is what is most applicable to a particular business or industry.  The following table lists a variety of differentiators that can be used to develop a best practice business model.  Primarily, the demand pull business should be developing business processes and infrastructure that allows it to produce products or services in direct response to customer demand in the most minimal time.  The demand pull business emphasises technologies and processes and relationships that reduce time to respond to customer demand.  The demand push business focuses on having tight business processes and a capability to respond to uncertain customer demand.  They include demand planning and Sales and Operations Planning (S&OP), planning technologies, excellent warehouse management systems and safety stock and customer service policies.  These businesses may also emphasise reserve production capacity and blending processes to manage uncertainty.  Supply relationships tend to be most critical because often production is highly dependent on the quality and continuity of raw materials whereas inventory buffers variability on the demand side.

These forms of comparisons, of necessity, make generalisations.  For example, while it is noted that application of sales and operations planning is critical for push business models, pull type businesses may be advantaged by S&OP as well.  Further, push business models can incorporate many smart things to gain some of the advantages that apply to demand pull models and also to manage uncertainty in customer demand.  These include:

• Eliminating non-value adding time, to reduce lead-times and dependency on forecasting;
• Application of collaborative planning to have more certainty with respect to customer demand;
• Segmenting those parts of the business for which demand pull can be created from those for which it cannot, and creating separate supply chain channels and even different business units to manage them;
• Moving the business to demand pull but buying incremental requirements of finished product through an associated supply and trading business unit; and
• Blending processes in production and applications of principles of postponement.

Conclusions

This article has explained the differences between demand pull and push.  Clearly demand pull supply chains reduce waste and inventory and are intrinsically more attuned to meeting end customer needs.  However, some supply chain participants almost inevitably will shoulder a burden of inventory, particularly if remotely located from end customers.  Therefore, these businesses should strive to create best practice business models appropriate to their circumstances, including if possible one with managed elements of push and pull.

Differentiators of Push and Pull Business Models

Differentiator	Push Model	Pull Model
Prime business driver	Maximise utilisation of resources, raw materials or infrastructure at least cost.	High levels of customer service through responsiveness and flexibility to meet uncertain customer demand.
Supply chain strategy	Operate strict processes in anticipation of demand. Emphasise demand forecasting and S&OP. Explore postponement. Attempt to separate those parts of the business for which a demand pull channel can be created and if necessary create alternative business units.	Operate in response to customer demand. Emphasise lean principles.
Lead times	Long and a prime focus should be to reduce them.	Short.
Pricing strategy	Pricing is a key means for balancing supply and demand.	Pricing does not normally impact short-term demand.
Manufacturing strategy	Long production runs.  Engineering development should focus on reducing economies of scale issues and assisting in moving more to a pull model.	Short and flexible production runs.
Inventory	Typically high. Emphasise inventory planning, safety stock policies and ABC classification. Push as close to customer as possible.	Typically low. Pull as remotely from the customer as possible.
Third party relationships	Supplier relationships tend to be most critical. Establish collaborative relationships with customers to minimise forecasting error.	Customer relationships critical. Supplier relationship criticality varies according to situation.
Most critical technology applications	Sales forecasting, inventory management, network optimisation, advanced planning, WMS	Order fulfilment, e-commerce, advanced scheduling, point of sale data capture.