Maxing the Gain

The paint department of your local hardware store may not seem like the place to go for strategic insights, but on display is one strategy that offers the potential to streamline offshore manufacturing operations, create markets and even generate new jobs.

The strategy is called postponement, or delayed differentiation, and the paint blending service in hardware stores is an everyday example. Instead of trying to anticipate what colors customers will want to buy and producing batches of finished paints, the manufacturer supplies base colors to stores. The retailer then mixes the base with various pigments according to a computerized recipe, giving the customer the exact hue he or she is looking for.

In addition to providing great customer service on the front end, this strategy offers big benefits on the back end. First, it drastically reduces expensive inventory for both manufacturer and retailer, since the store needs only stocks of basic pigments rather than much larger quantities of the finished paint. Shipping the more compact pigments instead of bulky paint product also saves on transportation costs. And returns are minimized because the buyer specifies the product configuration.

That last part is crucial: The product is finished at the point of sale when the consumer has decided exactly what he wants. Accurately forecasting customer demand becomes easier the closer you are to the actual sale.  In this case, demand is known because the final product configuration has been postponed until the point of purchase, the last stage in the supply chain.

It is this feature-the ability to refine demand information by delaying the finished product-that makes postponement such a powerful tool. Forecasting customer demand with precision is increasingly difficult in today’s fast-changing markets. The migration of manufacturing to low-cost centers in countries such as China has stretched supply chains internationally, increasing lead times and making demand forecasting even more precarious. Further, customers have become choosier about the products they buy, and are demanding more customization. As ongoing research at the MIT Center for Transportation & Logistics shows, if companies can learn how to apply postponement effectively in this environment, they can meet the more stringent market demands and discover better ways to deliver the product variants that buyers want.

Since the paint mixing service was first introduced by manufacturer Sherwin-Williams decades ago, postponement has become an increasingly important tool. Today, delayed differentiation is used in numerous industries from aerospace to textiles. In technology, Dell is one of the most obvious examples. The company built a global business on configuring PCs when orders are received, rather than stockpiling finished product on the basis of demand forecasts. It postpones final assembly until an order arrives via its online retailing network.  

Delayed differentiation also is embedded in the manufacturing processes of Hewlett-Packard. The company’s Deskjet and Deskwriter printers are made in its Vancouver and Singapore plants and distributed to the U.S., Europe and Asia. Selling printers in Europe means following each country’s requirements for printer configurations: different decals, a country-specific power plug and language-specific manuals. Six HP printer models and 23 different country configurations added up to 138 versions of the finished printers. In the past, the company forecast demand for each European country and then manufactured the appropriate numbers of printers for each one.

But forecast errors caused frequent product shortages. To increase product availability without increasing inventory costs, HP switched to pan-European forecasting and opted to postpone printer customization. Instead, HP began shipping generic printers to its European distribution center in Holland, where the units were configured to each country once local demand was known.

Shifting demand patterns pose major problems for the fashion industry, which must cater to multiple buying seasons and fickle consumers. Clothing manufacturers typically can’t supply the product just when the customer is ready to buy it, which results in lost sales and markdowns that cut into profit margins. Benetton, the Italian clothing manufacturer and retailer, addressed the problem by changing its manufacturing and distribution process. Most clothing makers dye the yarn first, weave it into fabric, and then cut and sew the fabric to create the finished garments. Under this traditional system, the manufacturer must decide how much to make in each color six to nine months in advance of replenishing stores. 

Benetton redesigned its manufacturing process to make some items of apparel-those with difficult-to-forecast demand for colors-in an un-dyed, generic state called greige. A test batch of each new garment was dyed and sent to a set of carefully chosen stores where its sales were monitored closely to gauge consumer preference for colors. With this more accurate demand information in hand, the company quickly dyes the greige garments and ships the items to retail outlets. Such postponement of dying increased Benetton’s manufacturing costs per garment by about 10 percent, but it increased profits by a far higher margin. Among other things, it cut the expense of overstocking and the associated costs of discounts and merchandise liquidations.

Postponement is by no means confined to the actual end product. Shaving products manufacturer Gillette applies the strategy to its packaging processes, for example. Retailers of the manufacturer’s shaving products were demanding broader selections of packaging and shorter delivery times, so Gillette used postponement as a way to achieve higher service levels without large inventory stockpiles of blades. It uses a specially equipped Pack Center to package product as late as possible in the ordering cycle.   

Forecasting Uncertainty

A number of factors have been propelling the strategy of delayed differentiation toward center stage.  Customer buying patterns have become less predictable, while demand forecasting has become more challenging. Product proliferation means that each product is sold in smaller quantities, complicating the forecasting challenge, and shorter product life cycles do not allow for learning from sales history. Surplus manufacturing capacity means stronger competition and thus less time to recover from supply chain failures.

The growing importance of offshoring-outsourcing manufacturing operations to contractors in low-cost countries-has created even more uncertainty. Although companies enjoy lower manufacturing costs, delivering product to end markets is more complicated because the contract manufacturers are thousands of miles away. As a result, globally stretched supply chains are more vulnerable to disruptions such as port stoppages and customs delays, and a longer product pipeline makes it more difficult for companies to respond quickly to demand changes.  Consequently, supply chains must be designed to be less sensitive to demand variations, yet deliver products at low cost and at high service levels.

That’s where postponement comes in.  Deploying this strategy means redesigning manufacturing into two distinct processes: 1) producing first a basic product with little differentiation, in accordance with a forecast, and 2) customizing the product in accordance with customer orders and delivering it.

In other words, the “basic” product can be manufactured overseas in long production runs with manufacturing quantities based on long-term forecasts, and therefore taking maximum advantage of contractors’ low operating costs. Further, the forecasting of the basic product, before it is differentiated, is based on large quantities and is less prone to extreme error. Once the basic product is made and shipped to the U.S., the manufacturer can wait until demand is better known (i.e., when all or some of the orders are at hand and the selling season is about to commence) and make the differentiated items “to order.”

A study at the MIT Center for Transportation & Logistics is looking at which markets and what products are ripe for a postponement strategy.  The study shows that  delayed differentiation is effective only in volatile markets where there are multiple product variations made to satisfy many consumers and consumer segments. Customers’ willingness to wait for a product is also critical. Dell customers agree to wait a few extra days to get a made-to-order computer product rather than making a more immediate purchase at a retail store. Not all retail customers would be so patient.

Another important factor is the modularity of the product. Products that are based on relatively simple, discrete component parts tend to be more amenable to postponement. For example, although Dell desktop computers are built to order, the company builds many of its laptop computers in China based on forecast and only minimal customization to some configurations. The reason is that laptops are more integrated products while desktops are more modular and thus easier to configure at the last minute once the orders are in. Laptops are also cheaper to ship from the Far East, tilting the trade-off in favor of more work performed overseas.

The Long-Term Potential

The growing demand for customization is just one of the many competitive pressures that companies are under today, and these pressures are likely to intensify as global markets expand and the balance of power in commercial transactions continues to move toward the customer. These changes may seem threatening, but they also offer limitless opportunities for creating new markets. Instead of struggling to forecast and fulfill demand, perhaps companies should encourage customers to be more discerning by expecting more product variations. The proliferation of products can open up new business opportunities for those companies savvy enough to recognize the potential and then deploy postponement to help them meet it. 

Customization and delayed differentiation could have a huge impact on the outsourcing debate by developing into a large-scale industry that then employs many of the workers displaced by offshore outsourcing. Customization, the fine-tuning of a product before it is purchased, is clearly best carried out in close proximity to the end customer. A contract manufacturer in China may be able to make a generic product cheaply, but it’s less likely to possess the intimate market knowledge required to customize a product for an American consumer. And in rapidly changing markets, the physical distance of offshore contractors makes it difficult for them to react quickly enough to tailor a product in tune with changing demand. Customization requires specialized skills deployed close to end markets-in this case, in the U.S.

This emergent industry will require a new generation of facilitators, and some players are already engaged. They could include logistics companies such as UPS or FedEx, which can manage the flow of the basic product inbound, store it and customize it to order, and then ship it around the country to end customers. Large e-tailers such as Amazon and Yahoo may yet step up to the plate. Alternatively, a brand new set of players may emerge, or a combination of these possibilities may prove more effective.

Whatever shape this new industry takes, it will thrive on demand variability. And it will meld the advantages of global manufacturing networks with local customization skills to help companies compete in expanding markets. 

Yossi Sheffi is director of the MIT Center for Transportation & Logistics, Massachusetts Institute of Technology. His upcoming book is The Resilient Enterprise (MIT Press, October 2005).