The main effort when recovering from a disruption is to isolate the firm’s customers from the disruption’s effects or otherwise to help them recover as quickly as possible. To this end, the firm’s customerfacing functions—in particular sales, marketing, distribution, and public relations—can play a pivotal role in preserving customer relationships, by communicating with customers (and other stake holders), and prioritizing service during a disruption. Moreover, an effective response presents a chance for the firm to win customers’ hearts with exceptional remediation and strong efforts.
The Taiwan earthquake of September 21, 1999, struck near the heart of the country’s high-tech manufacturing core. The Richter Scale magnitude 7.6 quake killed more than 2,300 people but did little structural damage to the 28 semiconductor fabrication plants concentrated in the Hsin-Chu Science-based Park, located about 80 miles north of the epicenter. Although the fabs seemed undamaged, they lost electrical power for nearly 24 hours, and when the lights went out on all the wafer-processing equipment, the fabs lost tens of thousands of wafers—wafers that were destined to provide chips for computers, communications gear, set-top boxes, and cellphones. It took weeks for the fabs to bring all the finely tuned systems back online, recheck the delicate equipment, and regain their pre-quake yields on chip making.
Taiwan accounted for 10 to 15 percent of the entire world’s dynamic random access memory (DRAM) chips, 80 percent of graphics chips, and 60 percent of computer motherboards.1 It also served as home to many liquid crystal display (LCD) suppliers and assemblers. With such a concentration, the disruption caused by the Taiwan quake rippled out quickly to the chip makers’ customers. They included numerous leading computer makers, such as Hewlett-Packard, Compaq, IBM, Dell, Siemens, Fujitsu, and Apple. From there, disruptions in PC production and delivery affected end customers in different ways, depending on how each PC maker made its computers and how it structured its customer relations.
Hardest hit were computer companies using lean production systems—those carrying very little inventory and relying on lastminute supply of components such as memory chips, circuit boards, and laptop screens. Two of the leading firms in this business use lean build-to-order systems, but they fared very differently. The company that perfected the build-to-order business model is Dell, which takes possession of standardized components at the last minute, assembles them, and ships out completed computers. Less well known is that Apple Computer Inc. runs a similar build-to-order supply and manufacturing operation. “Last quarter, we ended with less than a day of inventory—15 hours. As a matter of fact, we’ve beat Dell now for the last four quarters,” declared Steve Jobs at the August 31, 1999, Seybold conference keynote speech.
Dell and Apple run similar lean supply and manufacturing operations. They differ, however, in the sales and distribution side of their supply chains. Apple emphasizes leading-edge design and a narrow portfolio of models. Apple’s launch of the new iBook and G4 Power Macintosh followed a characteristic pattern for the company. It announced the eagerly awaited products at the New York Macworld Expo on July 21, 1999, published the configurations and prices, and allowed customers to preorder the machines with expected shipment dates starting in mid-September.
The much-anticipated iBook was intended to be like a mobile version of Apple’s popular iMac—stylish, inexpensive, and available in multiple colors. Knowing that some early adopters would jump at being the first to get the new machine, Apple used preorders to gauge market interest. These data helped Apple estimate demand and commit appropriate resources to manufacturing. Apple received orders for more than 160,000 iBooks before it began shipping them on September 15, 1999.
Apple had delivered only a tiny fraction of the preordered iBooks before the quake halted production in Taiwan and disrupted component supplies only a week later. “We lost one week of production on both PowerBooks and iBooks during the September quarter. Some component supplies were also impacted,” said Fred Anderson, Apple’s chief financial officer.2 Frustrated customers saw weeks added to the expected delivery dates.
Adding to Apple’s woes, another supplier disruption hit the company at the same time as the earthquake. On the same day that the earthquake hit, Apple announced that Motorola had not been shipping the needed volumes of the speediest G4 processors that Apple used in its just-announced high-end desktop machines. The shortage of high-end chips from Motorola—it was having problems with new chip-making technologies licensed from IBM—meant that Apple would have to downgrade customers (send them a slower-than-ordered computer) or have an indefinitely delayed shipping date for the faster processor.
Apple chose to ship computers with a less powerful chip (but the fastest available from Motorola) without lowering prices.3 Only after an avalanche of customer complaints did the company relent and offer price reductions.4 Because Apple could not modify preordered computers, the problem persisted into the fourth quarter of 1999. Many customers responded by either ordering lower-margin machines or canceling their Apple orders altogether.5
Dell also faced component shortages, such as memory and laptop computer screens.6 And like Apple, Dell had a second, simultaneous supply disruption that exacerbated the consequences of the Taiwan earthquake. Dell was backing a new memory technology supported by new chips from Intel. On September 24, 1999, three days before the official launch of the chip, Intel canceled the production release because of serious performance defects in the design. Dell had expected to use the new technology on one-third of its computers in the fourth quarter of 1999. The indefinite delay from Intel forced Dell to switch back to the old memory standard and scramble to buy older-technology memory chips in the quake-constricted market.
Dell’s sales process, which is part of its supply chain design, afforded it much more flexibility compared to Apple. While Apple was locked into rigid preordered commitments, Dell used a much shorter order-to-ship cycle. Apple had months of pent-up demand in its preorder backlog, but Dell usually operates with only three days of orders in the pipeline, and consequently had no commitment to specific models, configurations, and prices beyond these three days. Dell sold only what it could make based on components on hand.
Dell’s basic build-to-order model of postponed assembly means that it orders components based on forecast and then assembles the computers based on its customers’ orders. This supply chain design affords Dell unique day-to-day flexibility. Since Dell is dealing with the customers directly through its web site and telephone, it can use a “sell-what-you-have” strategy.7 The company steers its customers toward configurations that it can supply easily and profitably. It raises prices on configurations that are based on components in short supply and lowers the prices on configurations based on components in excess supply. It also bundles certain available components with popular configurations and runs promotions on them, highlighting these configurations on its web site. Dell’s normal operations include continuous adjustments of prices and promotions daily and sometimes even hourly, based on component availability and prices, competitors’ pricing, and expected market conditions.8
Apple, on the other hand, can be locked into agreements with retailers and into customers’ preorders that hamper its ability to manipulate the demand for its products. Consequently, Apple does not change configurations and prices frequently. When memory chip prices spiked after the quake, Dell announced that it would halve the amount of memory in base configurations to hold prices constant and charge customers who wanted more memory.9 On the web and on the phone, Dell guided customers toward configurations that it could build and ship.
In spite of the disruption, Dell increased its market share and increased its third-quarter earnings over the previous year by 41 percent. In contrast, Apple lost market share.10
Caterpillar Inc. is the world’s leading manufacturer of construction and mining equipment. It has created a unique flexibility in its relationships with its dealers in order to increase its resilience: To respond to an emergency order from a customer, Caterpillar can buy back and resell parts that have already been sold to dealers, and are in those dealers’ warehouses.
Customers pay $50,000 to $1 million per piece of Caterpillar equipment for use at mining sites, putting out oil fires, digging foundations for skyscrapers, building roads, and moving earth for dam projects all over the world. Not only is the equipment expensive, but downtime can halt a multimillion-dollar construction project, significantly increasing customers’ costs. For example, mine operators depend on massive Caterpillar earthmovers and trucks to extract valuable ore. “If a truck isn’t available to a mine, it could cost that mine anywhere from $20,000 to $30,000 per hour,” said Larry Newbanks.11 As head of inventory research management for Caterpillar, Newbanks understands the costs of downtime to Caterpillar’s customers. Consequently, one of Caterpillar’s main selling points is the reliability of its products and the speed with which it can get spare parts to a stranded piece of equipment anywhere around the globe.
To minimize disruption to its customers, Caterpillar must stock copious quantities of spare parts, ready at a moment’s notice. Because Caterpillar supports both its new equipment and its older models, it has over half a million different kinds of spare parts (stock keeping units or SKUs) that it could carry, and over 25 parts distribution warehouses in which to carry them. The problem is that most of these parts—more than 350,000 of them—sell in very low volumes of less than one per month worldwide.
In order to minimize spare parts inventory and maximize level of service, Caterpillar created a unique partnership with its 220 independently owned dealers. Caterpillar’s dealers have nearly 2,000 outlets and are, collectively, larger than Caterpillar itself. The partnership focuses on joint management of Caterpillar’s and its dealers’ inventories of parts, providing flexibility both to the company and to its dealers.
Caterpillar’s Dealer Parts Inventory Search (DPIS) system features an automated search-and-fill capability that finds spare parts anywhere within Caterpillar or its dealer network. Every day, dealers inform Caterpillar how many parts they have and how many they are willing to share. Caterpillar feeds that information into its order management system, so that the system knows the inventory level, by part number, at every dealer store. (This jointly managed spare parts inventory includes only the inventory that each dealer is willing to share, not the dealer’s entire inventory.)12
A request for a spare part usually comes from a dealer who is notified by a customer of a stranded piece of equipment. If the dealer does not have the part in its own inventory, it can turn to the DPIS. The system first looks at the nearest Caterpillar-owned distribution facility. If the needed part is not there, it searches dealers in that region and then expands the search to Caterpillar locations outside the region. If DPIS finds the part at a dealer store, it routes the orders for that part to the dealer. “We treat that inventory as our own, no matter where it is,” Newbanks said. “That allows us to direct orders to wherever the inventory is, out in the world.”13
Caterpillar buys the part back from the dealer who has it, at the original (wholesale) price plus 10 percent. Caterpillar pays for the transport and sells the part at the regular (wholesale) price to the dealer who needs it. (All of this is done automatically by the DPIS system.) Caterpillar absorbs the 10 percent charge in the name of service and avoids the costs of excessive inventories on rarely needed spare parts.14
DPIS provides inventory risk pooling that helps both Caterpillar and its dealers reduce inventories; at the same time Caterpillar maintains high availability of emergency spare parts. Many companies pool the risks of finished goods inventory by having a single replenishment system with access to their entire inventory in all of their warehouses. Similarly, many retail outlets can direct consumers to a different store of the same retailer, where a wanted item can be found. But what is unique about Caterpillar’s system is that it takes such inventory risk pooling a step further by creating a single virtual inventory involving its customers—the independent network of dealers—in order to serve the ultimate customer, the equipment owner. The result: “We operate 24 hours a day, every day of the year. We’ll ship 99.7 percent of all those items the same day,” Newbanks said.15
Caterpillar and its dealers’ inventory risk pooling strategy does not only offer flexibility in responding to emergency orders; it also means that its parts inventory is physically decentralized, and is therefore less vulnerable to regional disruptions.
One of the most important elements in helping customers recover from a disruption is giving them as much information as possible. The inevitable confusion and uncertainty that is part of most largescale disruptions can be mitigated by open and proactive communications with trading partners and, in particular, customers. Such communications help trading partners focus their own recovery efforts based on up-to-date and accurate information.
As a supplier to the fast-paced electronics industry, Taiwan Semiconductor Manufacturing Company (TSMC) needed to inform its customers of the impact of the 1999 Taiwan earthquake on production and shipment of chips. Within minutes of the quake, the company reestablished communications between TSMC in Taiwan and its sales and marketing headquarters in the United States and immediately contacted major customers with the latest information. By 9:00 a.m. local time, seven hours after the quake struck, the company had set up a 24-hour hotline for customer inquiries. Top executives broadcast two letters within 48 hours of the earthquake. Twelve press releases rounded out the stream of customer communications.16
To avoid negative reports on the company’s post-quake health or its progress in combating the disruption, TSMC’s post-quake customer relations initiative also focused on the press and analysts. The company conducted more than 100 interviews in the first five days.
As a result of its dominant position in the industry, TSMC knew that it would receive more than its share of media and analyst attention. Deadline-obsessed reporters often tell the story of only one or two leading companies. Thus, TSMC received 57 percent more coverage than its largest competitor. The challenge facing TSMC was that more obscure competitors, who were not covered in these press reports, could easily have manipulated any negative news on TSMC to their own advantage.17
With an audience that spanned 15 time zones and four languages, TSMC hosted events for the press and analysts in both Asia and North America so that it could provide accurate information and avoid morbid speculations. As a result, 86 percent of reports that mentioned TSMC did so in a favorable or neutral light. Finally, TSMC used the event as an opportunity to talk about other, nonquake-related issues, and 85 percent of the reports mentioned these issues. TSMC won a Silver Anvil Award from the Public Relations Society of America for its exemplary communications campaign.
The first stage in communicating with customers and trading partners is to collect accurate information to feed the communications campaign. Many disasters strike core infrastructure and disrupt communications, especially in low-cost producer nations where the infrastructure is limited to begin with. The vulnerability of power, phone, and data lines adds to the vulnerability of the company because it disrupts management communications. Lacking contact with the affected facility means managers cannot assess the situation, implement business continuity plans, or provide reassuring status information to customers. For this reason, Intel built a regional emergency operation center in each region in which it operates, and it equipped these facilities with layers of backup communications. Landline telephones, cellphones, SSB18 communications, satellite phones, Internet connections, and even globe-spanning ham radios ensure that the company can be in continuous contact with its entire international network of facilities. Periodic drills ensure that each regional emergency operation center knows how to regain contact with headquarters and the facilities in its region. Timely communications ensure that customers do not fear the worst when news of a disruption reaches them.
Johnson & Johnson has always earned high marks for living up to its credo of highly ethical corporate behavior. After the 1982 Chicago Tylenol poisoning (see chapter 2), J&J launched an immediate nationwide recall of all Tylenol, redesigned the product’s packaging, and reintroduced the product. Using nationwide media, the company ensured both the safety and the peaceof-mind of its customers. J&J’s actions in this case remain the gold standard of corporate disruption response and responsibility. Similarly, in 2002, when someone counterfeited J&J’s anemia drug Procrit (see chapter 7), the company launched an intensive direct mail campaign that targeted health care professionals (who are the de facto customers and administrators of the injected medicine). J&J sent out more than one and a half million letters in eight mailings. The letters informed distributors, doctors, and hospital workers how to spot the counterfeit drugs, what J&J was doing the problems, and how to recognize the new anti-counterfeit features.
As these examples demonstrate, one of the most important requirements after a disruption occurs is to communicate. Communications to the public can be based on press conferences, but it is important to have a direct line of communication with stakeholders. To this end, Martin-Lockheed maintains, in a central location, all the supplies needed to communicate in writing with each of its 170,000 employees and 45,000 shareholders within two to three days.19
Disruptions can in many cases offer companies the opportunity to make a positive impression with a quick and ready response, thus enhancing the brand.
In 1992 Lexus, the luxury division of Toyota, was just getting a foothold in the United States. In April of that year it recalled its flagship LS400 model for some minor brake and hydraulic system defects. Owners of the car were initially upset, given that they had just bought an expensive luxury car from a company known for its reliability. On the phone, customer service representatives were reassuring; they asked each customer to leave the car outside the home where it would be picked up. On the appointed day, Lexus representatives left an identical car in the driveway—a loaner for the day. The owner’s car was returned by evening, fixed and cleaned, with a full tank of gas and a voucher for a free oil change.
By comparison, Audi reacted angrily to a November 1986 television exposé (aired on CBS’s popular program 60 Minutes) that documented drivers’ claims that their Audi 5000 had “sudden acceleration” problems. VW, Audi’s parent company, issued press releases of engineering explanations that demonstrated that the problem was with American drivers, not the cars (the problem was never reported in Europe). While government studies in the United States, Canada, and Japan proved conclusively that VW was right and the cause was driver error, the damage was done. As VW continued to blame its customers, Audi’s U.S. sales shrank from 74,061 units in 1985 to 12,283 units in 1991, with dealers dropping the franchise and VW considering leaving the U.S. market.
The point of this comparison is not the fairness of the allegations but rather the realization by Lexus that a problem can offer an opportunity for impressing its customers with excellent customer service. Audi “went to war” against its customers20 while Lexus looked at the recall as an opportunity to increase customer loyalty. In the final analysis, Audi management did not understand the disruption they were facing: They thought it was a technical problem that their engineers could prove did not exist. In reality, they had a customer relations problem that they hardly addressed at all and they were duly punished by the market.
Similarly, in late 1994, Intel did not understand that a problem with the floating-point operations of its Pentium 486DX chip was not a technical issue (for the vast majority of users) but a customer relations issue. In mid-1999 Thomas Nicely, a mathematics professor at Lynchburg College in Virginia, was computing the sum of the reciprocals of a large collection of prime numbers on his 486DX Pentium computer. Finding an error in comparison to theoretical values, he traced the error to the chip and notified Intel. After getting no answer he posted the problem on The MathWork Worldwide, a mathematics web site. Magazine and TV interviews followed.21
Intel first did not acknowledge the problem and then tried to minimize it and belittle those who complained. It announced that “an average spreadsheet user could encounter this subtle flaw once in every 27,000 years of use.” This response generated an avalanche of headlines such as “Intel . . . the Exxon of the Chip Industry,” as well as Internet board messages such as “It’s close enough ...we say so” and “You don’t need to know what’s inside.” As a result of the problem IBM stopped shipping computers with 486DX chips.22 When Intel finally acknowledged the problem, it announced that it would replace the chip only if users could demonstrate their need for floating point operations. It took a strong public outcry for Intel to relent and recall all the 486DX chips23—however grudgingly—and as a result, Intel took not only a $475 million charge to earning, but also tarnished its reputation and lost customer confidence.
Intel’s problem was, in a sense, more similar to Audi’s than to Lexus’s. The floating-point-operation defect of the 486DX almost never showed up, and affected a very small fraction of the customers a very small fraction of the time. Intel, however, had a customer-relations problem that it took too long to acknowledge and correct.
Good long-standing customer and community relationships can help a company recover from a disruption by making customers willing to help the company directly or to be flexible in accommodating the company’s needs during the recovery. The story of Cantor Fitzgerald is a case in point.
The 105th floor of the World Trade Center in New York City was an enviable office location. More than 1,200 feet above the bustling financial capital of the world, it was an excellent aerie for running a financial trading powerhouse. Cantor Fitzgerald, with more than 1,000 employees in its New York offices, occupied the 101st, 103rd, 104th, and 105th floors of the North Tower of the WTC.
Cantor brokered trades worth a staggering $1 trillion every week. The company played an important role in the U.S. Government Treasuries market, specializing in brokering billion-dollar trades between major Wall Street institutions. Because of Cantor’s massive trading volumes, the company also supplied valuable pricing information to Wall Street firms, selling real-time data on the latest prices of the financial instruments that Cantor handled. The events of September 11, 2001, would severely test the company’s ability to survive a major disruption.
At 8:45 A.M. on that Tuesday morning, nearly two-thirds of Cantor’s New York contingent of traders, analysts, and other employees were already at the top of the North Tower, at their desks. They were busy checking the latest financial news, analyzing the overnight numbers from the European and Asian markets, talking to customers, and getting ready for another hectic day on Wall Street. Howard Lutnick, the firm’s CEO, was late to work because he was dropping his son off on his first day in kindergarten. At 8:45 a.m., a time-stamped photograph shows the CEO with his son at the school.
At 8:46 A.M., a hijacked fuel-laden wide-body Boeing 767 struck the WTC about 10 floors below Cantor’s offices in the North Tower. While at his son’s school, Howard Lutnick heard the news that a plane had hit the WTC.24 A second plane struck the South Tower at 9:03. Lutnick rushed to the site and tried to find Cantor employees among those streaming from the burning skyscraper, but nobody above the 91st floor ever came out. When the South Tower collapsed at 10:05, Lutnick fled and was knocked down by the enveloping tornado of debris and choking dust. At that moment, he was convinced his company and his life were over. As the dust started clearing, Lutnick realized he was not dead. But he couldn’t be sure about his company.
At 10:28 A.M., the North Tower of the WTC collapsed, killing 658 of Cantor’s 1,050 U.S. workers—everyone who was in Cantor’s offices when the first plane hit. The dead included most of the superstar traders who had made the company successful, most of the people who had hard-won personal relationships with Cantor’s customers, and most of the support personnel who kept the high-energy operation going. The dead even included Howard Lutnick’s younger brother, Gary.
Without computers, phones, and a place to work, Cantor could not survive. Despite the emotional devastation, the remaining employees in the United States, as well as employees from Cantor’s European offices, worked around the clock to recreate its systems in temporary offices in New York, New Jersey, and Connecticut.
Less than 48 hours after the first plane struck the WTC, Cantor was ready to trade. Its eSpeed electronic trading systems were back online in time for the September 17 reopening of the U.S. Treasuries markets.25
Less than two months later, Cantor was handling its usual volume of bond trading and 80 percent of its usual level of stock trading, ramping back up to its previous $1 trillion-a-week rate. The company also continued its recovery process, looking for permanent space on the lower floors of another office building.
“We played the cards the terrorists dealt us,” said CEO Lutnick.26 Fortunately, Cantor had redundant systems that allowed it to recover. Its backbone trading system, eSpeed, was built on a dual architecture that replicated all machines, connections, and functionality at the World Trade Center and at a Rochelle Park, New Jersey, site, with a third facility in London. In addition to data, the backup systems stored critical corporate knowledge, including customer relationship information. Systems redundancy was crucial to Cantor’s recovery. Just as crucial was the external support it got from suppliers, competitors, regulators, and customers.
It is not surprising that suppliers would help. Lending a hand to a customer in distress is good business. Thus, Compaq delivered 100 desktop computers at 2 a.m. on Wednesday, September 12, and Verizon expedited the installation of voice lines and the transfer of some of eSpeed’s digital circuits. Cisco provided a phone system based on Internet protocols and Microsoft had a Windows operating system team continuously on hand, as eSpeed’s server and desktop maintenance group had been especially hard hit. (Some of the systems’ passwords were lost with the lost people.)
Competitors helped too. ICI/ADP, another electronic trading company, offered to take care of eSpeed’s clearing and settling of transactions through its own connection to banks, and UBS PaineWebber provided temporary office space in Manhattan.
Regulators also wanted Cantor and other afflicted market participants to survive. Cantor had hundreds of billions of dollars of pre-9/11 trades that were awaiting settlement—the final exchange of the money and securities that is legally required to occur within a specified number of business days after the trade is logged. Chaos in closing these trades could have affected liquidity for financial institutions in all areas of commerce, reverberating throughout the world economy. Regulators kept the equities markets closed for four days, in part to allow companies like Cantor to recover.
But the most important help came from customers. Major financial institutions relied on Cantor as a discrete middleman. For example, when financial giants like Goldman Sachs or Merrill Lynch want to sell a billion dollars in U.S. Treasury securities based on a bet on changes in interest rates, they do not want their competitors to know that they are bearish on the bonds. Trading houses, like Cantor, broker these mega-deals, helping financial institutions keep their trading strategies private.
When asked, immediately following the 9/11 attack, what it would take for Cantor to survive, Lutnick had suggested that “if every money manager of a pension fund just gave us a little bit of business, then maybe we’ll survive.”27 Lutnick discovered that Cantor’s customers really wanted the company to survive.
When the equity markets reopened on September 17, Cantor was unsure that its systems would operate, much less handle the normal volume of trades. To ensure a “soft start,” the company instituted a rule of one trade per customer to avoid stressing the system. The response from customers to this conservative restart surprised Cantor. When Cantor’s head of equities took his first client call on the 17th, he explained to the fund manager of a large mutual fund, “We are open for business, but we’re not here to impress anyone, we just want to do one trade—as a test.” But that wasn’t enough for the client. “Our management committee got together and we’re giving you all our business today and you have to do all of our trades,” she said. The Cantor employee replied, “We’re not sure we can handle it,” to which the fund manager said, “You don’t understand. If you don’t take these trades, I’m going to lose my job. I’m faxing them to you, you do the best you can,” and then she hung up.28 Cantor used customers’ bursts of goodwill and willingness to deal with “the best you can do” to get back on its feet and worked hard to justify customers’ willingness to help the company.
While cases of customers helping suppliers may be less common in Western business, some East Asian systems are, in a sense, built on corporate mutual assistance. The Japanese keiretsu system of cross-holdings creates bi-directional support as suppliers help their customers and customers help their suppliers. Similarly, the Korean chaebols, such as Samsung, Daewoo, Goldstar and Hyundai, involve groups of companies with cross-holdings and cross-commitment to the other members of the group.
Even if a company has strong and flexible customer and distribution relationships, it may not be able to satisfy all customers during a disruption. Although no company wants its disruption to propagate to the company’s customers, limited resources can force a firm to make difficult decisions during severe disruptions.
Setting customer service priorities for post-disruption activities involves a decision process similar to the emergency medical practice of triage. In the realm of emergency medical care, doctors divide patients into three groups: those who have minor injuries not needing immediate medical attention, those who will die regardless of any medical attention, and those who will live only if they receive prompt attention. When resources are scarce, doctors treat the last group first, because that is where the incremental ounce of medical intervention can make the maximum difference.
In the immediate aftermath of a disruption, managers face a triage-like choice about which customers to serve first. Such decisions can be based on customer vulnerability or on more internally focused criteria such as how profitable the customer is, how costly would it be to serve it, or how important the customer is in the long term. Which criterion is used may be less important than having a coherent process for setting priorities during the time-sensitive post-disruption period. The company has to be seen as using a fair allocation process so long-term relationships are damaged as little as possible.
Continental Teves, a supplier of automotive, industrial, and agricultural products, had to make tough decisions when 9/11 shut down all U.S. air freight traffic and disrupted cross-border freight flows. On the afternoon of 9/11, the company assembled a list of all outstanding customer, part, and supplier orders. Most important, it collected data on its North American customers’ inventory levels. Knowing these customers’ production rates from past order patterns, it calculated the number of days of parts supplies each customer had before their operations would run out of parts. This statistic—days of supply—was the one on which Continental Teves based its priorities.
By late afternoon on 9/11, Continental knew which shipments were the most critical to its customers and required immediate action, and it expedited many of these by rerouting them via ground transportation, leveraging its relationships with transport firms such as Emery Ground to supplement air cargo delivery.29 In emphasizing customer needs in mitigating the disruption, the company did incur some added costs, but “Cost is not the question,” said James Gill, a spokesman for Continental Teves. “It’s keeping the customers happy and making them feel comfortable.”30
Most Taiwanese semiconductor manufacturers also considered customer vulnerability to set priorities in the aftermath of the Taiwan quake. A month after the earthquake hit the island, Richard Chang, president of Worldwide Semiconductor Manufacturing Corp., reported that “certain customers are asking us for some parts that they very urgently need to be delivered. Those things we are going to hurry and ship out to them, and in one case, we’ll do it two weeks ahead of time. I think the key is to work with the customers to find which is the most urgent and support them.”31
Deciding which customer is vulnerable, and to what extent, requires a qualitative assessment of which customers’ situations are more critical than others. Practically, in many cases it will be the customers who exert the most pressure. This was the case with the Philips chip plant fire. Originally Philips chose to focus all of its efforts on serving its two highest-volume customers, Nokia and Ericsson. But Nokia’s aggressive pressure and immediate attention to the problem, including the involvement of the highest levels of the company, forced Philips’s hand. Nokia actually monopolized the limited recovery resources of Philips (in fact, overriding Philips’s allocation decisions), leaving Ericsson with a severe parts shortage.
When producing for the consumer market, companies can prioritize on more internally focused metrics. For example, when New United Motor Manufacturing, Inc., ran out of parts in the midst of the 2002 West Coast port lockout (see chapter 4), it decided to airfreight parts from Japan. But at $50,000 of added costs per container, airfreight “would increase production costs of the cars by $300 to $600 each and the trucks by $2,000 apiece,” said a NUMMI spokesman,32 because truck parts were heavier and larger than car parts. NUMMI chose to airfreight only car parts33 In effect, NUMMI has set its priorities on the basis of cost and profitability impacts. Consequently, its car customers were not disrupted, but truck customers experienced delays.
Public agencies prioritize on effectiveness, measured by helping the largest number of people in the shortest period of time. For example, snow removal after a storm is focused on major arteries and mass-transit routes first, while local streets are plowed only later. Health authorities also prioritize based on effectiveness (and vulnerability), offering limited influenza vaccines to the most vulnerable members of society first: the old, the very young, and pregnant women first, as well as to health workers and caregivers.34
Downstream processes can help flexibility in many ways. Dell achieved flexibility by structuring its entire supply chain around its make-to-order system, thus avoiding difficult-to-change longterm customer commitments.
Risk pooling provides flexibility for sending products or parts from customers who do not have an immediate need for them to customers who do. Caterpillar’s parts distribution system uses the inventory already delivered to dealers as if it were still available to be sold to other dealers, thus increasing its ability to respond to demands for rare parts anywhere in the world. Similar systems are used by retailers to move merchandise from one store to another where the merchandise is needed to fill outstanding orders.
Strong partnerships with customers are likely to help secure both help and leeway when disasters strikes. Customer reactions can be the difference between a company rebounding or not, since signs of confidence from a major customer are likely to signal to other customers, suppliers, financial institutions, and investors that the company is likely to survive.