Only a year ago, Cisco Systems Inc. (CSCO ) was widely hailed as the shining exemplar of the New Economy. Management gurus viewed Cisco as the prototype of the 21st century corporation, an organization where information technology linked suppliers and customers in ways that allowed the company to nimbly respond to every market nuance.
Cisco had flattened the corporate pyramid, outsourced capital-intensive manufacturing, and forged strategic alliances with suppliers that were supposed to eliminate inventory almost entirely. Sophisticated information systems gave its managers real-time data, allowing them to detect the slightest change in current market conditions and to forecast with precision. If anyone had the "vision thing" nailed for the new digital era, it was supposed to be Cisco CEO John T. Chambers.
Oops! The surprising abruptness and severity of Cisco's downturn--marked by a shocking $2.2 billion write-off of inventory in April--showed that it was just as vulnerable as any other company to an economic slowdown. And it wasn't just Cisco. Motorola (MOT ), Lucent Technologies (LU ), and Oracle (ORCL ) also failed to see the downturn until it had all but engulfed their operations. Indeed, some of the hardest hit have been companies that were supposed to show Corporate America the way into the new century. What happened?
Blindsided by hubris, too much success, and an overreliance on their world-class computer systems, some of the most successful companies of the dot-com era discovered that even the best information is only as good as the minds interpreting it. The revolution in computer technology has put near-perfect financial information within the reach of almost every manager. Yet no amount of information can ever offset simple human judgment. When that judgment is clouded, as it was in some cases by the irrational exuberance of the high-tech bubble, it may not be possible to see what the numbers are telling you. "A bubble is like a gravitational field," says Adrian J. Slywotzky, a partner with Mercer Management Consulting. "It robs you of your ability to think clearly."
So perhaps it is not surprising that some industry stalwarts, from DuPont (DD ) and Eaton Corp. (ETN ) to the Big Three carmakers, responded more quickly to the slowing economy. Months before Silicon Valley stars realized that they, too, could fall prey to the business cycle, these Old Economy players were downshifting to tough out a manufacturing recession that is still idling assembly lines all across Corporate America.
Their market-intelligence systems flashed warning signals that their executives--veterans of past downturns--were savvy enough to recognize. "In those kinds of organizations, memories are pretty long," says Harvard Business School professor David A. Garvin. "You have to have a jaundiced eye and say, `Wait. This can't last forever."'
CHEAP FINANCING. Many managers insist that the falloff in business was so sudden and so severe that no one could have foreseen the depth of the abyss. "We saw the same signs as everybody else, but we viewed it as an inflection point and an opportunity to gain market share," says Claudia Ceniceros, a Cisco spokesperson. "We were ultimately wrong about the severity of the downturn." Indeed the order flow did screech into reverse with astonishing speed. And it's always harder to gauge the market in young, fast-growing industries. Still, once you look beyond the order book, there were signs that in a less giddy period might have signalled trouble ahead.
Lessons from the fallout abound. In some cases, cheap financing distorted real demand, underscoring the need to look to the end users, particularly if a company is not directly selling to them. How much are they actually consuming of your product? And no company, no matter how high-flying, is divorced from the larger economy. If earnings are dropping elsewhere, falling demand is likely to follow. Above all, forecasting is hard. No one does it perfectly. So companies have to constantly plan for less likely scenarios. It's also important to remember that even highly accurate forecasts have their limitations. Knowing that your business is about to get slammed by a falling market doesn't shield you from the blow; it only helps you limit the damage.
Perhaps the biggest lesson is to look behind the numbers. Companies such as Cisco, Nortel Networks (NT ), and JDS Uniphase (JDSU ), which sell telecom equipment, clearly misread the market, in part because they weren't analyzing their customers' businesses. A good chunk of their sales came from alternative phone companies that had relied on venture capitalists and Wall Street for funding equipment purchases rather than using cash flow from their own operations. When that money abruptly dried up, the shortfall drove many of the new phone companies out of business.
Some of the artificial demand should have been even easier to spot. In the case of both Cisco and Nortel, significant funding came from the equipment makers themselves, skewing real demand. Both companies looked at their 40%-plus revenue gains, took a gander at their mountains of customer orders, and figured the good times would just keep on rolling.
Simply projecting the past gains forward cost the telecom equipment giants dearly. JDS Uniphase reported in July a $7.9 billion fourth-quarter loss and a $50.6 billion loss for its 2001 fiscal year, the largest in corporate history. A month earlier, Nortel announced one of the biggest quarterly losses ever--a $19.4 billion hit--and ratcheted up its planned layoffs to 30,000. Nortel Chief Executive John Roth warned that "the market is contracting at an alarming rate," and he cautioned that Internet traffic, still thought to be in its infancy, was even declining. While these Goliaths still refuse to fault their forecasting systems for missing the slide, some critics argue that they failed to do the most basic homework. The customer orders they relied on to measure demand, for instance, could have been checked to screen out the double- and triple-ordering that had become common because of past scarcities.
What's more, they should have taken careful measure of the health of their customers to see whether they really were in a position to buy. "People started to believe too much in technology vs. their common sense," says Wojtek Uzdelewicz of Bear Stearns & Co. "You can't look at what your customer is telling you, you have to look at what is going on with the customer's revenues."
Managers can't stop with their own customers, either. No company exists in a vacuum. Until early last year, it was all too easy to believe that New Economy outfits were different, somehow outside the normal forces of the economy. Managers who fell into that trap learned the hard way that ultimately the same forces that drive down demand for old-line companies curb the sales of everything from fiber-optic cable to PCs.
PANIC-ORDERING. DuPont--whose products are used in everything from medical equipment to clothing, carpeting, and paint--has been through countless economic swings. Managers there have learned to pay close attention to such macroeconomic factors as oil prices, currency fluctuations, inflation, and Federal Reserve policies. That's how DuPont was able to figure out as far back as July, 2000, that business was deteriorating. DuPont cut back early, keeping layoffs and write-offs to a minimum. The main warning signals then: high energy prices, lackluster apparel sales, and government-reported declines in new factory orders.
Even for the most experienced hands, forecasting demand is an inexact science. At Massachusetts Institute of Technology's Sloan School of Management, students and executives for years have played "the beer game," where they take the roles of brewer, distributor, wholesaler, and retailer and try to estimate demand across economic cycles. Nearly everybody gets it wrong--especially those who are farthest from the customer. The trick, says MIT Professor John D. Sterman, is to figure out how the ultimate consumers are behaving and produce to suit them, not the manufacturers, distributors, or retailers in-between. So-called panic-ordering and other distortions only confuse everyone. Says Sterman: "If you are in the pasta business, you want to know how much pasta people are cooking and eating, not how much they're buying, and certainly not how much supermarkets and distributors are ordering from the factory."
Gathering good information, of course, doesn't guarantee acting wisely on it. Carmakers, for instance, regularly get a better-than-average fix on overall demand as their in-house economists pore over such time-tested data as household formation rates and consumer confidence surveys. "The forecasting is pretty good," says David J. Andrea of the nonprofit Center for Automotive Research. The problem, he says, is "industry behavior."
Detroit still winds up during a slowdown with at least some excess inventory--as it did in January. Why? Each carmaker bets that its models will outsell the rest and produces accordingly. Lately, too, the Big Three carmakers have sullied their profit margins and clouded forecasting by offering steep incentives to buyers. Much like the networking industry's vendor financing, such market-distorting inducements make it tough to gauge real demand--even if the carmakers can boast of better results than, say, the producers of networking equipment.
Other companies, though, were able to use a sound forecast as an intellectual framework for smart decision-making. Eaton Corp., with $8 billion in annual sales and products that range from truck transmissions to circuit breakers for homes, started getting cautious last October when Adrian T. Dillon, Eaton's chief financial and planning officer, warned of an impending "short, sharp shock" in the economy.
Dillon's insights didn't save Eaton from the consequences of the downturn; no forecast can do that. But it did allow the company to cushion against the coming drop. As demand for its products shriveled, Eaton halted a stock-buyback program to save cash and pulled the plug on a couple of planned acquisitions. Although this year's first-half earnings have plunged by 64%, the decline could have been steeper had the company not prepared its managers--and Wall Street--early on. In fact, Eaton's stock is up 11% since early January.
FOOL'S GAME. Sun Microsystems Inc. (SUNW ) doesn't rely on long-range forecasts in its purchasing but instead tries to buy supplies--often over the Internet--as needed. Its top operations and sales staffers meet weekly in supply-demand gatherings to make sure their work is balanced. And they try to ensure that their partners don't get left in the lurch when demand moves south. Twice last year, Sun asked supplier Celestica Inc. (CLS ) to delay deliveries of memory devices that Sun had ordered, but when it came time in April for Celestica to either take a charge for the goods or deliver, Sun took them. "Sun treats suppliers more decently and they remember," says Sun operations chief Marissa Peterson.
Regrettably for many in Silicon Valley, the ability to make accurate forecasts can depend on how well-established a company's products are. Young industries on steep growth curves are almost always surprised by how well their products do in the first few years, and then they're taken aback when demand ebbs. Says Stanford University business strategy professor Kathleen M. Eisenhardt, "in a highly dynamic, ambiguous, and unpredictable market people are going to make mistakes. It's inherent in the type of business."
In many corners of Silicon Valley--and elsewhere--unpredictability is inevitable. One solution: keep innovating but develop sound service businesses to sell with products. "I break the world into two segments--one is creating demand and the other is servicing demand," says Craig H. Muhlhauser, president of the cutting-edge power-storage systems maker, Exide Technologies. Building a "very robust service business," he says, smooths out the rough spots between innovations.
No matter how well companies try to forecast demand, they will almost always be off the bull's eye. Chaos is simply more the norm than orderly, predictable patterns. Indeed, Wharton School adjunct professor Paul J.H. Schoemaker argues that sticking too closely to a forecast is a fool's game. Better to use scenario planning, in which you prepare to handle different sets of circumstances. "You can't reduce the uncertainty, but you can manage it by having options," says Schoemaker.
No business, of course, is recession-proof. But managers who understand that markets go down as well as up are are a lot more likely to read the tea leaves correctly. Better to see the warning signs and ratchet down early than find yourself stuck with billions in equipment that suddenly no one wants. Just ask Cisco.
By Joseph Weber
With Ben Elgin and Peter Burrows in Silicon Valley and Michael Arndt in Chicago