Over the last half century, supply chains have moved from the domain of a single company (such as the vertically integrated Ford River Rouge plant) to a web of suppliers and their suppliers. At the same time, the type of product bought from suppliers has changed—from commodity materials and parts to highly engineered specialized systems and subassemblies, requiring close cooperation between the trading partners. Outsourcing the manufacturing of complete products is only the next logical step in this progression, a step that has led to the rise of contract manufacturing companies in many industries. The result is that companies have become increasingly dependent on their suppliers’ networks.
More and more processes that were traditionally performed by OEMs, have now been outsourced, many of them offshore. The original reasons for outsourcing were two-fold: first, to reduce costs and second to focus on core competencies. Outsourcing offers economies of scale and the expertise of the suppliers, while offshore manufacturing offers low labor costs and a flexible labor force. The focus on core competencies allows companies to increase their competitive advantage in the areas in which they are strong.
As more work was moved to suppliers, their manufacturing expertise and engineering prowess developed substantially. In fact, access to the suppliers’ innovation in materials, products, and processes became one of the most important reasons for outsourcing.
This trend resulted in consolidation of the supply base. During the 1980s and 1990s most companies have cut significantly the number of suppliers they do business with. Maintaining close working relationships with suppliers meant that each supplier required significant management attention and resources. In addition, focusing a large part of a company’s business on few suppliers meant that each supplier would get a large portion of the business, making the company’s business more important for that supplier. This created incentives for the suppliers to invest in these relationships; an important consideration when companies depend more on their suppliers. For example, Lucent went from 1,700 transportation carriers to 167 carriers in 2000, while Sony spent 2003 weeding out nearly 80 percent of its 4,700 suppliers.1
The combination of outsourcing with supply base consolidation—which has been implemented, to varying degrees, by most companies in Europe and the United States—has, in many ways, altered the nature of competition. Instead of competing on the basis of their own capabilities, companies are now competing on the basis of their supply chain capabilities.2 This new basis of competition has necessitated new types of contracts and relationships with suppliers, ranging from flexible contracts to risk-sharing arrangements to various types of partnerships.
Strong trading partnerships offer resilience when disaster strikes. In general, suppliers are likely to aid a customer in trouble because doing so will help solidify future markets. This is particularly true when the customer represents a large part of the business and the relationships are expected to last for a long time. These kinds of relationships are especially strong with the Japanese keiretsu and the Korean chaebols. Both structures involve supply chain “ecosystems,” including groups of companies with cross-ownerships and long-term trading relationships.3 Such ecosystems create resilience and accelerate disaster recovery.
Companies that are not part of a formal ecosystem can still create strong alliances with suppliers who will back them up when the need arises. In many cases, this results from single-sourcing and strong partnerships. Such partnerships, however, involve significant investment in the relationships, and single-sourcing creates its own vulnerabilities. Consequently, some companies choose to have multiple sources of supply. Both approaches: singleand multiple-sourcing, yield different costs and benefits, and either one is viable when implemented carefully.
Just after 4 A.M. on Saturday February 1, 1997, sparks from a broken drill ignited several wooden platforms at the Aisin Seiki Co. Factory No. 1 in Kariya, Japan. The fire spread quickly, swept through an air duct, and ignited the roof. As the fire swelled, it destroyed most of the 506 precision machine tools used by the company to manufacture P-valves—the small cigarette pack-sized proportioning valves used to prevent skidding on the rear brakes of cars. By 9 a.m. that day, much of the factory lay in smoking ruins.4
Although Aisin’s P-valves cost only ¥770 to ¥1400 each (about $8 to $14), the brake systems of Toyota’s cars depended on them. Aisin’s ability to churn out the small precision-machined parts at low cost made the company a Toyota favorite. Over the years, Toyota had come to rely on Aisin for 99 percent of the P-valves used in Toyota’s entire car production. Nisshin Kogyo Co., the second source supplier, could not boost production 100-fold to make up for the loss of Toyota’s main supplier.5
Toyota Motor Corp. produces a quarter billion dollars’ worth of cars per day. It is the largest company in Japan and the most valuable car maker in the world. Shutting down production would have idled tens of thousands of Toyota workers in Japan and hundreds of thousands of workers in Toyota’s network of suppliers and suppliers’ suppliers.
Early February 1997 was an especially bad time for a major disruption. Toyota expected a near-term sales boom in Japan, driven by an approaching increase in the Japanese sales tax from 3 to 5 percent, which was expected to spur consumers to buy cars ahead of the April 1 hike.6 Thus, Toyota and its suppliers were already running at 115 percent of normal production volumes. Toyota had even taken the unusual step of hiring temporary workers in anticipation of the pre-hike boom in car sales. And then Aisin’s plant caught fire.
As a just-in-time manufacturer, Toyota had assembly plants that held only a few hours’ worth of P-valves. Toyota-bound P-valves, on board trucks dispatched before the fire, provided another couple of days’ worth of stock. But on Tuesday, February 4, Toyota had to shut down 20 of its 30 Japanese assembly lines because it had no more P-valves.
The recovery of Aisin would take months; it required buying replacement machinery and creating alternative production sites. Salvaged and hastily acquired equipment let the company produce some P-valves about two weeks after the fire, but only 10 percent of the required amount. Six weeks after the fire, the company was still producing only 60 percent of demand. It would be two months before Aisin approached 100 percent production. Meanwhile, with too few P-valves, Toyota stood to lose the 15,500 vehicles per day that it had planned to make.
Finding alternative suppliers meant mobilizing the supply base of Aisin and Toyota. Even as the fire raged, Aisin started this process, organizing an “emergency response unit” and alerting Toyota immediately. Aisin started calling suppliers for help, as did Toyota. The calls started with the Toyota keiretsu and the Aisin keiretsu. Sixty-five suppliers responded—22 Aisin suppliers and 36 Toyota suppliers, along with a few independent suppliers and suppliers belonging to other keiretsu groups.7 By Saturday afternoon, Toyota and Aisin held a war-room conference with all potential P-valve makers and hurried engineers divvied up blueprints and valve-making assignments.8
Although conceptually simple, the P-valves contain precisiontapered holes and orifices that regulate the flow of brake fluid. To help the new suppliers, Aisin provided technical and engineering assistance; it also served as quality control for the impromptu supply chain of valve makers.
The backup suppliers ranged from a six-employee prototyping shop to Denso, the giant $24-billion a year Tier One supplier of automotive components, which ended up making thousands of Pvalves per day. In fact, Denso outsourced manufacturing of some of its own products in order to focus its factories on making Pvalves for Toyota.
Few companies had equipment similar to Aisin’s specialized machinery. Because some of the makeshift suppliers used manufacturing techniques foreign to Aisin, Aisin could give them little technical assistance. These suppliers created their own collaborative processes to help each other learn how to make the valves. Denso helped translate Aisin’s specs into designs suited to alternative machining practices. Suppliers also coordinated with each other on the purchasing of scarce equipment, ensuring that each supplier received what it needed.
Taiho, an engine parts manufacturer (and a Toyota supplier since 1944), involved eleven of its own suppliers in the effort, and Kayaba Industries, a manufacturer of hydraulic equipment, outsourced the P-valves to three of its suppliers. Brother Industries, the manufacturer of sewing and fax machines, spent 500 manhours to convert a milling machine to make just 40 P-valves a day—its first experience with making auto parts. Some 150 other companies provided machinery and fixtures to help make P-valves and replace the equipment that Aisin had lost.
The effort entailed neither legal nor financial negotiations; suppliers simply went to work. Aisin provided engineering drawings and technical assistance to all participants, including more than 500 engineers from Toyota and Aisin’s suppliers who were familiar with different machines and processes that could be used to make P-valves. Suppliers never asked Toyota or Aisin what they would be paid for rushing out the valves. “We trusted them,” said Masakazu Ishikawa, executive vice president at Somic Ishikawa, a manufacturer of automobile parts and components who first became a Toyota supplier in 1937.9
The suppliers even turned the disaster into an informal race. The first prototype P-valves came from Kyoritsu Sangyo Okayama, a tiny welding equipment supplier to Aisin, and were delivered only two days after the fire. Aisin inspected the P-valves and gave Kyoritsu Sangyo approval to make more. The following day (Tuesday), Kyoritsu Sangyo shipped the first batch of P-valves. Dozens of other suppliers raced to provide prototypes and ramp up production in the days following the fire.10
By Thursday, February 6, Toyota’s Tahara plant was reopened. The other Toyota plants opened with a single shift each on the next day. Nine days after the fire started, on Monday, February 10, all Toyota-group Japanese assembly plants were back to normal with production volumes of 13,000 to 14,000 vehicles per day. Toyota reached its originally planned goal of making 15,500 cars per day about one week later.
In all, Toyota lost only four and a half days of production, even as Aisin lost at least five weeks of production. Without the selforganizing cooperation of its supply base, Toyota’s lost production would have been much higher. Yet even these curtailed losses were high—Aisin lost ¥7.8 billion ($76 million) and Toyota lost ¥160 billion ($1.55 billion) in revenues. Although Toyota claims to have made up most of the lost production later, it did incur between ¥20 billion to ¥30 billion ($195 to $290 million) in extra costs.11
Eventually Aisin reimbursed all firms for the direct expenses incurred in P-valve production, including labor costs, special equipment, and material costs. Toyota settled the issue differently. It gave each of its participating Tier One suppliers a payment equivalent to 1 percent of their respective sales to Toyota from January to March 1997. This amounted to overall payments of more than ¥15 billion ($146 million). Many of the firms viewed the payment as a reward for cooperation rather than as compensation. Most of the Tier One firms receiving the payment followed Toyota’s example and rewarded their Tier Two suppliers, who followed the example with their Tier Three suppliers.
Maybe the most surprising aspect of this corporate recovery story is that Toyota never exerted any pressure on Aisin to act, nor did it pressure Aisin to supply its P-valves at the expense of Aisin’s other customers. Aisin had customers other than Toyota and Toyota respected Aisin’s need to help everybody. Hino (a $7 billion Japanese heavy trucks manufacturer) had enough replacement P-valves coming on February 6 to reopen its own plant in Hamura.
The case of Aisin and Toyota illustrates why many companies explicitly avoid sole source procurement, fearing the resulting exposure inherent in using a single supplier.
The fire and its aftermath, however, have left Toyota executives convinced that they have the right balance of efficiency and risk. “Many people say you might need to scatter production to different suppliers and plants, but then you have to think of the costs ...of setting up expensive milling machines at each site,” said Kosuke Ikebuchi, a Toyota senior managing director. “We relearned that our system works.”
Working with its suppliers for many years and with crossownership and involvement in other businesses with its suppliers through the Keiretsu system, Toyota has very deep relationships with them. It is difficult, however, to develop such deep relationships with multiple suppliers of each part or raw material. The choices faced by most firms can, therefore, be summarized in figure 13.1.
The upper-right quadrant of this figure depicts the situation of deep relationships with multiple suppliers. This is too expensive and it is difficult to find examples of firms who practice such strategies. The relationship between Toyota and Aisin is an example of deep single sourcing belonging in the top left quadrant of figure 13.1. The next two sections give examples and discuss supplier relationships that fit in the two lower quadrants of the figure.
In late October 1998, the warm Caribbean waters off the coast of Central America brewed the fourth-largest Atlantic hurricane on record. A category 5 hurricane, Mitch lashed Honduras and its expansive banana plantations with sustained winds of 180 mph and dumped 2 to 4 feet of rain on parts of the country.12
Hurricane Mitch devastated much of Honduras, killing at least 6,500 people, destroying 70 to 80 percent of the country’s transportation infrastructure, and causing more than $4 billion in damage. The banana plantations never had a chance. The fastgrowing soft-stemmed banana plants, with their large leaves, were toppled by the winds or washed away in the floods. Honduras lost 80 percent of its banana crop to Mitch.13 Other bananaproducing countries, such as Nicaragua and Guatemala, also felt the fury of the storm. In total, Mitch destroyed 10 percent of the world’s banana crop.
Both Dole and Chiquita relied on the tropical climate of Central American plantations for banana supplies, but they fared very differently. Dole lost 25 percent of its global banana supply. Chiquita suffered $200 million in damages but was less dependent on the affected region, losing only 15 percent of its capacity. In addition, Chiquita had secondary sources for Central American bananas in Mexico, Panama, and Colombia, which were less affected by the hurricane, as well as in the Ivory Coast, Martinique, and even Australia. Unlike Dole, it moved quickly to leverage its alternative sources, increasing production in other locales and purchasing the crops of associate producers.
In contrast, Dole had no plans for alternative sourcing and suffered more than one year of shortages in its banana supplies. The financial result was that Chiquita generated a 4 percent increase in revenues that year while Dole registered a 4 percent decline in revenues.
Many other companies also rely on multi-sourcing to moderate the consequences of supply disruptions. For example, following the 1999 Taiwan earthquake that disrupted the island’s chip manufacturers, Motorola was planning to rely on second and third sources in case of a disruption. “We have a program called 1-2-3 for wafers and chips. If for some reason the primary supply is disrupted, we have secondary and tertiary sources,” explained a Motorola spokeswoman. Even so, regional disruptions can affect multiple providers; Motorola was sourcing from TSMC and UMC in Taiwan when both suppliers were hit by the Taiwan earthquake, leaving Motorola with fewer supply alternatives than planned.
Interviews by the MIT Center for Transportation and Logistics revealed that in the aftermath of the 9/11 attack, some firms started insisting on multiple sources of supply for everything they purchase. Other companies, however, decided to deepen their relationships with their main suppliers. One purchasing executive from a leading supplier to the aerospace industry summarized his company’s approach this way: “We intend to put all our eggs in one basket and watch this basket very carefully.”14 The second part of the statement—watching the basket—is key to the success of the single-sourcing approach. Deep sourcing is more than just sole sourcing.
UPF Thompson has supplied chassis to Land Rover since the 1950s. As the sole chassis supplier for Land Rover’s popular Discovery line of SUVs, it was selling Land Rover 70,000 chassis per year. Following Land Rover’s acquisition by Ford in 2000, it completed the conversion of its planning and production processes to a modern just-in-time system in the third quarter of 2001. But then, on one Friday morning in December of that year, the chassis shipment from UPF Thompson failed to arrive at Land Rover’s Solihull plant. When Land Rover managers called to inquire, they found out that as a result of losses suffered by UPF in an unrelated foreign venture, UPF Thompson was bankrupt and in receivership.15
The news got even worse.16 The receiver assigned to manage the bankruptcy, KPMG, demanded that in return for continued supply of chassis, Land Rover should assume all of UPF Thompson’s outstanding debt of £49 million. With only two days worth of inventory, Land Rover had its back to the wall. The Discovery line accounted for fully one-third of the company’s revenues. Finding a new supplier would have suspended Discovery production for up to nine months while tooling was developed, and it would have led to 1,400 layoffs at the company’s assembly plant. An additional 10,000 jobs among Land Rover’s other suppliers would also have been severely threatened. Shortly thereafter, Land Rover received written notification from KPMG, asking them to buy the business or to agree to a long-term supply contract at higher prices. Land Rover, refusing to be strong-armed, rejected the offer and went to court to try to get a temporary injunction compelling UPF to keep supplying the chassis.
KPMG’s hard-nosed stance received wide coverage in the British media. It even generated Motion 714, introduced on January 24, 2002, in the United Kingdom Parliament by Mark Tami, the labor MP representing the constituency of Alyn and Deeside. The motion called for Parliament to express its “deep concern” at the conduct of KPMG and called upon the government “to review all its department contracts with KPMG in light of their actions.”17 The dispute heated up when Land Rover chairman Robert Dover hinted in the media that if KPMG insisted on its demands, sourcing of the Discovery would move abroad.
Despite the hostile media reports, KPMG was simply doing its job in the competitive market culture of the U.K. Land Rover had no written supply agreement with UPF Thompson, and even though Land Rover paid for all the tooling that UPF Thompson installed in order to produce the chassis, it had no agreement as to ownership of the tools in case of bankruptcy. As receiver to UPF Thompson, KPMG was, in fact, obliged to recover all the money it could on behalf of UPF Thompson’s creditors, and a recent court ruling had held that in such situations receivers could exploit a customer’s vulnerability to the benefit of the creditors.18
After weeks of failed negotiations and legal maneuvers, however, England’s High Court found KPMG’s demands to be “arguably illegal” and granted the injunction guaranteeing the flow of chassis on an interim basis. After a lengthy series of appeals and additional hearings, the two parties eventually came to an agreement. To settle the dispute, Land Rover paid £16M of UPF’s £49M debt. Land Rover demanded, however, not only the continuing supply of chassis but also the replacement of KPMG as the receiver.19
The UPF Thompson bankruptcy almost stopped the production of the Discovery vehicle for many months and could have inflicted significant losses for Land Rover, as well as considerable job losses in the U.K. Land Rover clearly was not monitoring this supplier closely, despite its assertion that it had processes to “keep a careful and continuous financial reviews of all its suppliers, delving into order books and business and revenue plans.”20
But sole-source strategies do have their place. Firms cut down on their number of suppliers not only to consolidate buying power; the strong supplier relationships bring with them access to engineering prowess and innovation. This is particularly important as in many industries suppliers are providing more and more systems rather than simply parts. For example, Visteon, the $18 billion Tier One automotive supplier, offers complete all-wheel drive systems, entire climate control systems, integrated door modules, and entire audio systems, in addition to a full complement of parts and components. And Intel, the maker of processors and chips, builds complete chipsets and motherboards used by PC makers.
That sort of added value was one of the reasons Toyota made the decision, even after the Aisin fire, that its sole-sourcing system worked well and that it would stay with the same arrangements. It was also why Unilever decided to stick with a sole-source strategy even after the disruption it suffered when Hurricane Mitch swept through Central America in 1998. Unilever’s Puerto Rico plant—a plant that produced roughly half of the North American supply of Q-Tip cotton swabs—was significantly damaged by Mitch; Unilever lost two weeks of production as a result of the hurricane and forced many of its customers into stock-out situations. But despite this experience, Unilever chose to relocate 100 percent of its Q-Tip production to the Puerto Rico facility after repairs were made.
To mitigate its exposure to similar disruptions, however, Unilever also increased inventory levels in North America by 10 percent and arranged contracts with barge shippers in the event that road or rail systems were made unavailable.21 Likewise, many of Toyota’s sole-source suppliers decided to increase their own resilience after the Aisin fire. Somic, which makes all of Toyota’s steering linkages, revamped its system so that it could easily shift to another site if disaster struck its main plant. Toyota itself launched an effort after the fire to trim the number of its parts variations, realizing that most of the recovery difficulty was rooted in its use of hundreds of variations of the P-valve.
As Lucent moved to cut the number of suppliers and deepen the relationships with them as part of its Supply Chain Networks (SCN) organization, it has formally instituted the role of suppliers’ relationship managers. These managers keep in close touch with all key suppliers, ensuring that information is flowing freely both ways and that neither Lucent nor its suppliers are caught unaware by the other party’s actions.
Multi-sourcing, of course, provides built-in redundancy, and therefore flexibility to switch suppliers when disruptions occur. When, for instance, the Avon Rubber and Injected Plastics plant was destroyed by fire in 1999, all but one of the plant’s customers were able to switch immediately to alternative sources for the rubber gaskets, door seals, and other automotive products made by the plant. The reason was that Avon was never the sole source supplier to any of them.
But even having multiple suppliers does not guarantee full continuity of supply, for at least three reasons:
First, regional disruptions can affect multiple suppliers. For example, the 2002 West Coast port lockout closed all Pacific ports. This motivates companies like GM to assess risks on a geographic level, not just a supplier facility level.
Second, the other suppliers may not always have available capacity; furthermore, suppliers who serve only as alternates may not be inclined to invest ahead of time or incur extraordinary costs in order to help out a customer who chose not to do business with them in the first place.
Third, the global connectedness of markets, especially for commodity materials, means that one supplier’s disruption can create price and availability problems for all customers in that industry, even for customers who do not source from the disrupted supplier. For example, the September 1999 Taiwanese earthquake disrupted dozens of semiconductor makers accounting for 40 percent of the world’s memory chip production during a period of tight supplies. As a result, the spot-price of memory chips climbed five-fold all over the world.22
Another risk associated with multi-sourcing is that secondand third-tier suppliers are often not as efficient as the first tier; Toyota sole-sourced from Aisin, for instance, largely because of Aisin’s superior cost effectiveness. Some companies consider the extra costs needed to manage an additional supplier, and the incremental cost of the parts supplied, to be analogous to an insurance premium that the enterprise spends to ensure supply in an emergency. Toyota decided that such a premium was just too high.
Given the risk-spreading advantages of multi-sourcing, some companies further insist that their suppliers have other customers. To avoid suppliers depending on it completely, Wal-Mart usually specifies that its own business can represent no more than about 25 percent of a supplier’s business (depending on the business in question). A portion of a supplier’s business above that level, a Wal-Mart spokeswoman said, “makes us nervous,” because “we don’t want to be the sole source of survival for a company.”23 This stance gives Wal-Mart more flexibility to change suppliers with lower exposure to accusations that Wal-Mart uses its dominance to abuse a highly dependent supply base. And encouraging suppliers to have multiple customers can also provide a company with surge capacity, since it may be possible to use the capacity dedicated to another customer when that customer doesn’t need it.
The question, then, is not whether to have one vs. many suppliers, but whether the sourcing strategy is internally consistent: Is the number of suppliers aligned with the nature of the relationships with those suppliers? When companies decide not to invest in close relationships with their suppliers, caution requires the use of multi-sourcing (unless the company uses commodity and off-the-shelf parts). Thus, the prudent combinations are either to have deep partnerships with a single supplier (or a small number of key suppliers) or to have multiple suppliers, as depicted in figure 13.1.
According to the U.S. Centers for Disease Control and Prevention (CDC), 36,000 American die annually as a result of the flu. In fact, the number of deaths attributed to the flu has risen four-fold between the late 1970s and the late 1990s.24 During the 2000-2001 flu season, federal officials were caught unprepared when one of the four vaccine manufacturers had to cease production as a result of FDA concerns over plant conditions. Shortages, delays, and four-fold price gouging ensued. The situation was investigated by the General Accounting Office, which reported in May 2001 to the Department of Health and Human Services that “the circumstances that led to the delay and early shortage of flu vaccine during the 2000-01 flu season could repeat themselves in the future . . . Now, a production delay or shortfall experienced by even one of the three remaining manufacturers can significantly impact overall vaccine availability.”25
A little over three years later, and despite continued warnings from public-health experts about the dangers of continued consolidation among pharmaceutical companies, the United States experienced another shortage. Chiron, one of the two companies relied on to manufacture the flu vaccine for the 2004-2005 season, had its sole vaccine plant in Bristol, U.K., shut down by British health authorities on October 5, 2004, because of contamination. As a result, half of the 100 million doses of vaccines the United States needed were not available. Rationing, long lines, panic, and price gouging followed. And again, there was no easy and quick solution since the other manufacturer was already at 100 percent capacity and no manufacturer around the world had extra capability.26
Having a small number of suppliers, with no extra capacity to cover for each other, meant that the United States was, in fact, in a single sourcing situation. Despite that, the suppliers were not carefully monitored. In June 2003, a year before the 2004-2005 crisis took place, U.S. officials learned that Chiron had reprocessed contaminated batches of flu vaccine, but the problem was not investigated vigorously. As late as September 28, 2004, one week before his plant was shut down, Chiron CEO Howard Pien assured a U.S. senate committee that his company would deliver 48 million doses of the vaccine for the 2004-2005 flu season. They took his word for it.27
Chapter 10 argued that information technology should always be backed up and redundant sites prepared, since the cost of an IT disruption is high relative to the cost of back-ups. One can easily argue that when the cost of disruption is measured in lives, public policy should ensure redundancy through the use of more than a single supplier, or a supplier with more than a single plant. In any case, if the decision was made that two suppliers would suffice, the relationships with these suppliers had to be deep and they should have been monitored early and often.
As companies depend more and more on their suppliers, it is important to examine the flexibility of the entire supply chain. Toyota’s deep and trusting relationships with suppliers were instrumental in the company’s recovery from the Aisin fire.
The MIT Supply Chain Response to Disruption study found that with sole-sourcing comes added engagement with the suppliers on the issue of resilience.28 Suppliers get involved with the company’s business continuity planning, and many companies require suppliers to develop and maintain comprehensive business continuity plans of their own. In some cases, industry groups set standards and provide security and resilience-related data—such as TAPA’s guidelines regarding carrier selection and its distribution of carrier data to its hundreds of members. At the very least, resilience has become a factor in supplier selection. For example, Electra Ltd., an Israeli elevator installation and maintenance company, uses Otis Elevator Company as its sole source. Since Otis operates a worldwide network of more than three dozen plants, Electra decided not to develop a second source, effectively relying on Otis’s network of multiple plants.29
Most companies treat suppliers of unique and/or complicated engineered parts—that is, parts that require significant supplier qualification for quality—very differently from the way they treat suppliers of near-commodities. P&G, for example, runs a training program for its top suppliers and conducts regular information exchange sessions with them. These suppliers are monitored for costs and can share in the success of the products they feed, but they generally do not have to compete annually for P&G’s business. For the most part, these are the suppliers on which P&G relies for technology and innovation. Other suppliers, on which it relies mostly for capacity, do compete for the P&G business annually.
Nowhere is this strategy more evident than in Dell’s procurement. Dell buys its processor chips and motherboards only from Intel,30 and its operating system only from Microsoft, both of whom are very resilient. For all other components—such as disk drives, memory chips, modems, keyboards, and displays—Dell keeps multiple suppliers on hand, routinely whipsawing them against each other in its quest to reduce costs. Dell even outsources the manufacturing of its laptop computers to three contract manufacturers in Taiwan: Compal Electronics, Wistron, and Quanta (which also make laptops for HP, Apple, IBM, Sony, Sharp, Siemens, and Fujitsu).
Thus, aligning the procurement strategy with the supplier relationships can ensure resilience, either reducing the likelihood of an unexpected disruption with a single supplier or providing the resilience that comes with multiple supplier relationships.