The Education Of Andrew GroveRobert D. Hof
Cramped, crowded with management and technical books, and festooned with Post-it notes, Andrew S. Grove's cubicle at Intel Corp. doesn't look much different than those surrounding it. More than anything else at the company, though, it symbolizes the open, proactive culture that has led Intel to sales of $11.5 billion and absolute dominance of the market for microprocessors--the chips that lie at the heart of the world's millions of personal computers.
Grove sits in a cubicle precisely to eliminate barriers between chief executive and staff. When he is around, anybody can stop by and bend his ear, even to tell him he screwed up on a decision. There's no fear of getting fired--Grove encourages what he calls "constructive confrontation." In the ceaseless race to put more transistors on ever-smaller slivers of silicon, there's no time for anything as inefficient as bureaucratic protocol.
Such an atmosphere has propelled Intel to the top of its market and installed Grove as one of the business world's most admired gurus. His two books on management are standard reading in business schools. So why, when news surfaced in late November of a minor flaw in Intel's flagship Pentium chip, did Grove make one of the biggest blunders of his career by not offering to replace the chips, no questions asked? Because even the best high-tech managers--those most skilled at wringing profits from technological breakthroughs--can be babes in the woods when it comes to selling technology to savvy consumers. And Grove himself, at times, may be more stubborn and less open to criticism than the office architecture suggests.
Much has been made of the notion that Intel's engineers simply couldn't grasp consumers' demands that the Pentiums be replaced. But the problem is really more complicated than that. Sure, Grove is a tightly wound engineering PhD who has molded the company in his image. But both the secret of his success and the source of his current dilemma is an anxious management philosophy built around the motto "Only the paranoid survive."
Intel is so thoroughly focused on the Herculean task of doubling the number of transistors on its chips every 18 months that it has trouble seeing much else. Its vision is further obscured by a certain macho arrogance born of its market dominance. "The culture, the centralized power, the tremendous profitability--it's a blinding light," says venture capitalist Jack C. Carsten, former Intel senior vice-president. Concedes Grove in his crisp Hungarian accent: "Sometimes you need a real jolt to realize a reality has kind of happened around you. We have to learn some skills that are second nature to others."
"HEART AND SOUL." Intel's response to the Pentium problem was classic Andy Grove--by turns relentlessly aggressive, coldly analytical, and angrily defiant. He calmly explained that, by Intel's calculations, the problem occurs too rarely to concern any but the most demanding scientists. When critics came up with examples of more frequent errors, Grove shot back: "If you know where a meteor will land, you can go there and get hit." He finally acquiesced with a replacement policy, but not until IBM launched an embarrassing broadside: Big Blue announced it would stop selling Pentium PCs after releasing its own test results claiming the chip failed more often than Intel was saying.
While Pentium's Christmas sales and Intel's stock price (chart, page 62) apparently survived the public-relations disaster, the episode shows just how intransigent Intel's culture can be. And it's a bad time for that. After years of selling mostly to the business market, PC makers are hell-bent on courting regular consumers. PC sales passed car sales in 1993 and are gaining on those of TVs. Meantime, Intel rivals such as Advanced Micro Devices and the Apple-IBM-Motorola consortium developing the PowerPC are mounting credible alternatives to Pentium. And the Apple Macintosh still has a strong nonbusiness franchise.
Most suppliers of a component leave advertising and marketing to the manufacturer of the finished product. But Grove feels that since the microprocessor is an increasingly integral part of the computer, the company must develop a brand of its own to differentiate itself from its rivals. Its huge "Intel Inside" and Pentium advertising campaigns have thrust the company into the consumer arena. "The microprocessor has become a consumer product," insists Grove. "Its role in contemporary society is not a passing phenomenon."
Consequently, many believe Grove is going to have to reexamine his own stubbornly analytical style. In the Pentium debacle, "we shouldn't have taken the arrogant attitude Andy did," says Intel board member Arthur Rock, an initial investor in the company. "He's going to have to change a bit." Says William W. Lattin, a former Intel vice-president and now a top executive at Synopsys Inc.: "Andy's analytical approach is his tremendous strength--and at times a weakness. Not everything with human beings is done with the accuracy of electrons."
Of course, there is no arguing with Grove's results. With more than 70% of the $11 billion microprocessor market, Intel's Pentium and 486 chips utterly define IBM-compatible PCs--far more than IBM itself. And the company's 56% gross profit margin is the envy of the industry. That dominance stems directly from Grove's signature combativeness and the culture that has grown up around it. "Andy is the heart and soul of the company," says Chief Operating Officer Craig R. Barrett.
Grove's feistiness has a long history. Born Andras Grof in Budapest in 1936, he was the son of a dairyman who was forced into labor on the Russian front during World War II. The young Grof, who liked to sing opera, fancied becoming a journalist. But when his articles were rejected after a relative was arrested, he decided to pursue science instead. Science was absolute; it couldn't be colored by politics as easily.
After the 1956 Soviet invasion, Grove left Hungary for New York. He earned degrees in chemical engineering before joining Fairchild Semiconductor in 1963, where Robert N. Noyce had invented the integrated circuit. When Noyce and Gordon E. Moore left Fairchild to found Intel in 1968, Grove joined them as a researcher.
MEMORY LOSS. He made his mark early. In 1971, Grove instituted a late list to note employees arriving after 8 a.m. Outside Intel, he's even more of a heavy. He has spent millions to sue AMD, Cyrix Corp., and others at every turn for alleged copyright and patent violations (although AMD recently won a key decision in the California Supreme Court). He once sent a violin to former White House budget director Richard Darman to protest the government's "fiddling" with Japan trade policy while the U.S. chip industry burned.
Soon after becoming president and COO in 1979, Grove launched Operation Crush, an all-out campaign to wrest 2,000 new customers away from Motorola Inc. within a year. Intel beat the goal by 500, and one of those customers, IBM, chose Intel's 8088 chip as the brain of its first PC.
But then came the biggest test of Grove's career. By 1985, Japanese chipmakers had driven prices so low that they had crushed their U.S. rivals in the huge market for DRAM memory chips--a technology Intel had invented. Several other U.S. memory makers eventually ended up in joint ventures with the Japanese. But Grove put his foot down. Over the objection of several executives, he axed the DRAM business and thousands of employees. Unwilling to let the company come to such a pass again, Grove focused Intel on microprocessors with a paranoia and manic competitiveness that informs much of what the company does today.
It comes as no surprise that Intel has never been very effective on the consumer front. As far back as 1972, when the company jumped into the new digital-watch business, it failed to realize it was selling jewelry, not just a timepiece. Then, in 1984, a retail business it launched to sell networking add-ons and modems to PC owners did reasonably well, but never traded its expertise with the rest of Intel. Since 1993, the company has spent some $100 million to develop and sell a videoconferencing system. But it has yet to fly because it's too expensive. Concedes board member Rock: "It has been terrible on the retail and consumer side."
"DIFFERENT SKILLS." By some accounts, the Pentium debacle is a sign that Grove's vaunted management system isn't as open as it purports to be. True, there was an "all hands" meeting at Intel headquarters five days before the general Pentium replacement guarantee was announced, and Grove says many of the 2,000 employees expressed displeasure with the harsh initial policy. He also received critical E-mail from inside Intel. But his circle of close advisers--about a half dozen vice-presidents--disagreed with the critics, and they held sway.
Indeed, some say Grove is hard to budge once he makes up his mind. "As confrontational as Andy is, there's not a lot of people who can take him on," says Lattin, a Grove admirer, nonetheless. "The people who are intuitive about a market before it develops get driven out." Grove doesn't buy that. "Intuition is not going to get you a 3 million-transistor microprocessor," he growls. "We are fundamentally a technology company. We don't make toothpaste."
True enough, but the company's great challenge will be to keep making the startling technological advances the market takes for granted while learning to track the vagaries of its customers. Grove concedes Intel must change. But he insists it will take incremental adjustment, not a cultural revolution. "Whatever transformation this involves," he says, "it is going to require different skills. Not a different culture."
Critics aren't so sure. A few years ago, Intel tried to better its relations with its manufacturing customers by including their input in the chip-design process. But in 1994, Grove soured things with his biggest customer--Compaq Computer Corp. CEO Eckhard Pfeiffer--by aggressively advertising the Pentium when Compaq chose to push its cheaper 486 machines during the crucial Christmas selling season. The message: Intel's agenda was more important than Compaq's. "They've had [PC makers] by the short hairs," says management consultant Rich Bader, who ran Intel's small retail operation in Oregon in the mid-1980s. "They treat the consumers with a very similar philosophy."
Some observers suggest that Intel might benefit by recruiting an executive from a consumer company such as Procter & Gamble, at least as a board member. Grove is in no hurry. "I'm not persuaded that people who come in from a consumer background are necessarily all that equipped for this, either," he says. Besides, he adds, Intel has always worked best by cultivating its own talent; nearly all top Intel executives have worked there at least 15 years. "It's very hard to adopt teenagers into a close-knit family," he says.
HOT LINE. Grove thinks what Intel needs to do is open lines of communication to the consumer and analyze the data that flow in. Intel has set up a new hot line to answer Pentium questions, and Grove intends to make it permanent. He also aims to cut through the layers between chip designers and consumers by having engineers regularly staff the hot line, as they have for the past few weeks. Much of the reason the Pentium situation blew up was because customers "flamed" Intel on the Internet. So Grove plans to expand the company's communications in cyberspace as well. "We're going to have tentacles into the PC-buying community that are going to get us closer and closer to the right answers," he says.
Whether that will be enough remains to be seen. But it's a good bet that Grove and his culture will ultimately make a more dramatic response if need be. Grove recalls that when he moved from research to operations at Intel, some people said: "`Andy Grove in manufacturing? Ha! Give me a break.' And I did pretty well with that." The momentous decision to flee the memory-chip business likewise shows Grove's ability to adapt. Hal Feeney, a former Intel manager who's now a chip analyst at Pathfinder Research Inc. insists that "Andy has made dramatic changes before. He can certainly make them again." Given the growth in home computing, now might be the time.
-- Born Andras Grof in Budapest, 1936
-- Left Hungary for the U.S. a year after the 1956 Soviet
-- While working as a waiter, earned bachelor's degree in chemical engineering, City College of New York, in 1960, finishing first in his class
-- Received a PhD in chemical engineering from the University of California at Berkeley, in 1963
-- Hired by Fairchild Semiconductor in 1963 as chief of research and development
-- Joined Robert Noyce and Gordon Moore when they left Fairchild to found Intel in 1968
-- Became president and chief operating officer in 1979 and CEO in 1987
-- Teaches management at Stanford University
-- Married for 36 years, two daughters
MOTTO "Only the paranoid survive"
FAVORITE MANAGEMENT GURU Peter Drucker
HOBBIES Tennis, skiing, sailing
ABANDONED TEEN PURSUIT Opera singing
BIGGEST LIE Said he would retire at age 55