Inside Andy Grove's Latest Crusade Intel's chairman is out to change the way companies are governed. His first job: Change himself.
By Brent Schlender

(FORTUNE Magazine) – Andy Grove and I became co-conspirators on a January afternoon in 2003 about 20,000 feet above the Sierra Nevada. We were sitting opposite each other, knee to knee, on a business jet en route to Las Vegas, where later that night I was to interview him and the surprise guest, Steve Jobs, onstage before a throng of revved-up Intel sales and marketing people. Noshing on crackers and crudites and making small talk as we streaked along, I asked Grove what else was on his agenda in Sin City, other than taking in a Cirque du Soleil show the next night. "We've got a board meeting all day tomorrow that I'm really looking forward to," he said. Referring to CEO Craig Barrett, he went on, "We'll be doing Craig's performance review, which should be interesting because we're doing it right in front of him." Grove went on to describe how he had revamped the board's CEO evaluation process to mirror the rigorous 360-degree reviews used throughout Intel. He alluded to other strategic issues the 11-person board would discuss and boasted about how lively the meetings could be. "Too bad you can't be a fly on the wall," he said.

Having dealt with Grove on stories for 20 years, I knew he had slyly left ajar the door to the Intel boardroom. "Well, all you have to do is invite me," I replied. He paused to peer out at the crumpled mass of snow-capped peaks below, and I could almost hear the gears turning in his brain. Since becoming chairman in 1997, Grove has, in his words, "gotten religion" about corporate governance, an eye-glazing term if ever there was one. He has become a "self-anointed governance guru," writing op-ed pieces and lecturing at universities and conferences. He feels strongly that independent directors should be truly independent, that they should make up a clear majority of any board, that the roles of CEO and chairman should be separated, and that requiring companies to expense stock options is unfair--stances that have caused serious friction in his friendship with Warren Buffett. He also believes many boards are poorly managed and falling down on the job. By the time we flew together that afternoon, Grove had worked for six years to make Intel's board exemplary. Emerging from his reverie, he leaned forward. "There could be some pedagogical value in that," he said with a puckish smile. "But I'll have to see what the board thinks first."

I didn't get the chance to eavesdrop on the board meeting in Las Vegas. True to his word, Grove asked the directors how they felt about being the subject of an article that would show how a board operates in the post-Enron era. "I won't say that they laughed me out of the room," he told me later. "But they clearly did not like the idea." Besides, upon further rumination Grove had decided that having a journalist in the boardroom all day would distort the group's behavior. "It's like the Heisenberg uncertainty principle in subatomic physics. Some of them might start playing to the audience or clamming up, and you wouldn't get a true picture of what we do," he said. Then, sounding for the world like the magazine editor he had wanted to become as a youth, he added, "That makes this a story that is almost impossible to directly report and tell, which is why no one has done it before. But it's a story that needs to be told the right way. So let's keep talking about it."

The challenge was pure Grove, who has used variations of this tactic for decades to motivate the people around him. "If Andy wants something, you're not going to say no," says Jane Shaw, CEO of biotech startup Aerogen, who has been on the Intel board since 1993 and now runs the audit committee. "He has this uncanny ability to instill in you the desire to do things right and to the best of your ability, because you simply don't want to disappoint him." In this case, as Grove well knew, I couldn't say no. Despite the clamor for boards to be more accountable to shareholders, directors remain the most secretive and mysterious players in what are ostensibly public companies. What journalist wouldn't want to peer behind that veil?

There are good reasons for board secrecy, of course: For one thing, the group is privy to a company's most sensitive long-term plans. But with so much happening behind closed doors, it is difficult for shareholders and the public to assess whether a board is doing its job. Ratings services such as GovernanceMetrics International use rudimentary statistical methods--checklists, really--to rate governance at public companies. (Intel is one of only 22 companies--out of 2,121 rated by GMI--to earn a perfect score of ten.) Yet businesses differ so widely that it's hard to generalize about what makes a good board. All shareholders really can go by is whether the enterprise thrives over time.

The $30-billion-a-year Intel certainly has thrived during Grove's years at the top. The value of Intel shares multiplied more than 25-fold during his 11 years as CEO. And while the company has faced nettlesome challenges since the dot-com crash--flagging growth in PC microprocessor sales; peskier competition from its chief rival, AMD; a struggle to broaden its business beyond PC microprocessors into chips for telecom and consumer electronics; the risky allure of investing big-time in China--Intel remains a bastion of IT.

Grove's challenge to me was irresistible for another reason. His years as chairman have coincided with a harrowing period for American boards. The drip, drip, drip of financial chicanery, excessive-compensation scandals, botched CEO successions, and more have shown the august boards of some of America's most prominent enterprises to have been at best asleep at the switch and at worst incompetent. I wondered how Intel's board would compare.

It took a full year, but Grove's lobbying paid off. All but one of the 11 directors, plus several who have retired or who now hold emeritus status, agreed to lengthy interviews for this story. Director Reed Hundt invited me to tag along on an impromptu visit to Intel offices and labs in Britain. And the board finally permitted a photographer and me to attend portions of two meetings in January and March. (We'll get to that, although being a fly on the wall in this case was a bit anticlimactic, like sneaking into your parents' bedroom and catching them ... reading.)

All this access, along with Grove's willingness to share some of his own secrets, did pay off, however. It turned up not one story but two. First, Grove has quietly and methodically shifted the primary mission of Intel's board from pure oversight to helping the company reignite growth--an achievement doubly difficult for having come at a time when the board's supervisory job has grown tougher and its regulatory burdens more onerous. And second, being chairman has changed Grove himself into a different kind of leader. He wasn't just waxing melodramatic when he told me, "The most important tinkering I had to do was with me."

Throughout his storied life as a Hungarian refugee, straight-A student, chemical engineer, manufacturing mastermind, manager, CEO, author, professor, cancer-research gadfly, and raconteur, Grove has for the most part made his own luck. In business he has shown time and again an uncanny ability to clarify and optimize complex organizations, business processes, and strategies by applying exacting standards and setting high, unambiguous expectations. Yet the real secret of his success is his ability to reinvent himself.

"When Andy takes on a new role, he almost abandons what he did previously and throws himself into the new one and puts his stamp on it," says Gordon Moore, Intel's co-founder and chairman emeritus. "Not many ex-CEOs make the transition to managing the board from managing the company." Grove will tell you that becoming chairman was one of the most wrenching transitions in his life. As CEO he had never paid much attention to governance issues. But in 1997, as the board was considering giving him the additional title of chairman, one of his directors, Harvard professor David Yoffie, pulled him aside. Holding both titles is common practice at many companies--in fact Moore had been chairman and CEO in the 1980s--yet the idea made Yoffie squirm. Being an academic, he was more attuned to the nascent concerns among institutional investors and business scholars over the domination of boards by management. (Indeed, since 2000 splitting the chairman and CEO roles has become something of a trend, epitomized earlier this year when Michael Eisner relinquished the chairman's title at Disney.) "David didn't make a big stink about it," Grove recalls, "but he said it was kind of the wrong thing to do."

The affable professor, who had been elected to the board in 1989 at the tender age of 34 (Grove nominated him), says some directors had a more pointed objection as well: They worried that giving Grove both titles would make it harder for him to cede power to Barrett, who was slated to succeed him as CEO in 1998. "Our biggest qualm was that Andy would continue to be a micromanager, an interventionist, and potentially not allow the new CEO to become a real CEO," Yoffie says.

Grove learned of the board's graver reservations not from Yoffie but from another director, Arthur Rock. The pioneering Silicon Valley venture capitalist, who had been Intel's first chairman in the late 1960s, told Grove bluntly that he wasn't sure he was right for the job. Says Grove: "He hit me right in the gut. But I chose to look at it as a challenge. Their concern was my tendency to speak up too much, to suck all of the oxygen out of the room. Arthur was really telling me that if I wanted the board to be more helpful, I would have to learn how to shut up. So I looked at him and said, 'I can change. I can submerge myself. Give me a chance to prove it.'"

Grove got his chance the following spring, during the board meeting at which Barrett was formally elected CEO. The other big agenda item that day was Intel's biggest acquisition ever--a proposed multibillion-dollar deal to buy Fore Systems, a maker of telecom gear that competed with Cisco Systems. The purchase had been planned during Grove's watch as CEO, but it was Barrett's baby, part of an ambitious strategy to jump into new chip markets. And on the surface, at least, the directors seemed receptive. One even says privately that many board members felt that Grove had grown too risk-averse in his last few years as CEO, and thus they looked forward to Barrett taking over and being more of a "cowboy."

After hearing the new CEO and his team make their case to buy Fore Systems, Grove, mindful of his vow to shut up, said nothing even though he favored the plan. Instead, he asked chairman emeritus Moore for his opinion--and promptly got blindsided. "It just didn't smell right," recalls Moore. "The presentation wasn't clean. And so I said it didn't make any sense." That opened the floodgates to criticisms from all around the table. After what director John Browne calls a "very, very tense meeting," Barrett's inaugural initiative was killed. "We could've afforded them. That wasn't the issue," says Barrett, who still winces when he talks about it. "The question was: Was there value to justify the acquisition? And we didn't have good enough answers."

Grove brooded about the incident. The board's rejection stalled a much-needed broadening of Intel's business. He was doubly troubled that the way he'd conducted the meeting had undercut and humiliated Barrett on his first day as CEO. The board didn't know it then, but their job was about to get a whole lot harder. The Fore Systems veto had exposed an embarrassing degree of ignorance and helplessness: The directors had balked in large part because no one understood networking or telecom well enough even to know what questions to ask. With the dot-com economy booming, Grove wanted this elite group to contribute to Intel's strategy, not just stand in the way. And chastened though he was, Grove wasn't about to subside into being a merely ceremonial chairman either. Like any team, he felt, the board needed to be managed. "We're talking about getting a bunch of ego-driven people to subordinate their egos for a common good," he says. Encouraging directors to be more vocal wasn't enough. They needed a mission--and so did their new chairman.

Barrett's wish to drive Intel into new businesses represented a major strategic shift. The explosion of the Net, which held the potential to transform almost anything electronic into a telecommunications device, was a huge opportunity. At the same time the company was facing big challenges. Much of the market for PCs and servers was becoming saturated, so to remain a growth stock, the company had to develop new markets. Moreover, much of the growth in semiconductors was shifting abroad, a trend that would accelerate as China and India emerged as players.

Intel had already begun recruiting directors with expertise beyond Silicon Valley--people like John Browne, CEO of BP, now the world's largest oil company, and David Pottruck, until recently CEO of brokerage Charles Schwab. (Ironically, it was the chairman and namesake of the firm who, acting in a very un-Grove-like way, ousted Pottruck and reclaimed the CEO title for himself in July.) More seats were due to come open as directors turned 72--a milestone Moore jokingly calls "mandatory senility"--and would have to retire. It would take a few years, but Grove was confident the board could add members with knowledge of the markets Intel was eyeing.

But because none of the new directors knew much about chips, they faced a steep learning curve. Charlene Barshefsky, the former U.S. Trade Representative, who joined the board in January, likens the experience to "walking in on the middle of a conversation. It takes a while to figure out where you fit in. And when you read the materials they send to you, there are ten things that come to your mind that you don't know for each one that you do." Intel's corporate culture of "constructive confrontation" also takes getting used to. As CEO, Grove fostered combativeness as one of the company's core values, and that feistiness extends to the board. "It can be kind of shocking at first," says John Thornton, the former Goldman Sachs president, who joined the board last year. "You realize quickly that they practice a form of honesty that borders on brutality."

Grove decided to formalize the board's education. He instituted annual seminars to teach directors the basics in key technical or regulatory issues. This year, for example, each director attended a session with a team of Intel engineers who explained the company's software strategy. Attendance was mandatory. Grove also ordered the board's support staff to build special Internet sites and beef up the briefing books that directors receive before each meeting with background information and refresher courses on particular technologies.

Grove also decided that directors needed to get out of the boardroom and see firsthand how Intel operates. Each director is now required to visit at least one, and preferably two or three, Intel sites each year. Afterward he is expected to report back to the board. "These aren't 'what I did on my summer vacation' reports," says Thornton. "We're expected to tell the board, and even management, something they don't know."

In April, I accompanied Reed Hundt on such an excursion--an eight-hour swoop through two Intel facilities in Britain. Grove's first recruit as chairman (he joined the board in 2001), Hundt is by far the most engaged director and the one Grove communicates with most--usually daily by e-mail. The ex-FCC commissioner brings a bevy of skills relevant to Intel's expansion plans. He knows intimately how public policy intertwines with telecommunications, and as a venture capitalist and McKinsey consultant, he works with the kinds of emerging telecom technologies and companies Intel might want to invest in or acquire.

He is also the board's champion globetrotter, having visited ten Intel sites in the past 18 months. This time he happened to be in London on other business, so Intel didn't even pay for the entire trip. Our itinerary included two stops--a research lab affiliated with Cambridge University and the company's European headquarters in Swindon, 80 miles west of London. Hundt had a different purpose in mind for each. "The Cambridge lab is a low-cost experiment for leveraging university research into new commercial technologies. I want to see if it makes any sense," he explained as we pulled up to the lab in mid-morning under gray skies. He was amused to find it housed in a building named after Bill Gates. It's a tiny lab, staffed mainly by blue-jeans-clad grad students. Although Hundt isn't an engineer, he's gregarious and insatiably curious, and peppered the half-dozen people we met with questions about how their seemingly arcane research might fit Intel's strategic goals. Much of the work involves communications software, and he became so engrossed that we left an hour late.

At Swindon, Hundt was more like Dr. Phil, holding a two-hour encounter group with a dozen or so employees, encouraging them to share whatever bothers them about Intel. Several fretted that Intel might stop issuing stock options if U.S. regulators require companies to expense them. Another complained that the company has gotten too bureaucratic and impersonal: "It just doesn't feel like Intel anymore." On the drive back to London, Hundt asked if I sensed the beginnings of a serious morale problem. To me the meeting seemed quintessentially Intel, I replied. The employees vented because they knew they could; it was just a variation on constructive confrontation. But they also seemed to believe that as a director, Hundt really could get through to management, so his visit had made them feel more connected to the top.

Over the years Grove has come to feel that making directors like Hundt more visible is as important as honing and mining their strategic thinking. "For them to own a piece of the action, they need to be visibly associated with the outcome," he says. He now has at least two directors address the annual meeting each year. This year Shaw and Hundt explained to shareholders how the board supervises financial auditing and how it arrives at executive compensation and stock-option plans. "I can't stress enough how important it was," says Grove. "They put a human face on who is really looking after the shareholders' interests."

Which brings us to Intel directors' biggest conundrum: Grove has doubled their workload by making it their job to vet corporate strategy. By law the directors are also watchdogs, charged with ensuring that shareholders don't get screwed. That's a task that's grown much bigger too. Based on figures provided by corporate secretary Cary Klafter, Intel's outside directors put in some 300 hours per year, not counting travel time--vs. an average of 200 hours for directors of a typical FORTUNE 500 board, according to governance expert Richard Koppes. Their pay, by contrast, is just average: $60,000 a year, with an extra $10,000 to $20,000 for those with demanding committee assignments. The directors also get stock options--the typical yearly grant recently has been 15,000 shares--and a handful of directors from Intel's early days grew fabulously wealthy on the options alone. Still, the compensation isn't much by executive standards.

"You have to wonder, Why does anyone even want to be on a board?" says corporate governance expert Joe Grundfest, a law professor at Stanford and former SEC commissioner. "The pay is lousy, there's potential legal liability, the workload has gotten much heavier, and your reputation can be tarnished by something you had no way of knowing about."

It's not as if Grove balks at the heavier compliance burden. On the contrary, he thinks the Sarbanes-Oxley Act of 2002, which set strict requirements for improving corporate governance, minimizing conflicts of interest, and ensuring financial transparency, is a good thing. "We might even need a Sarbanes-Oxley II," he jokes. "We should require directors to attest to a certain level of knowledge and involvement in strategic decisions, so they'll act responsibly when approving something like, say, the AOL/Time Warner merger. After all, good strategy is good governance." That still doesn't explain how directors are supposed to find the time to be both master strategists and fierce watchdogs. Grove's solution comes from Management 101: He has delegated virtually all the compliance work to the board committees. Each director, of course, has to sign off on the audit review and compensation plan each year, but the full board now devotes far less time to reviewing committee work.

Grove insists there's one other crucial element that truly effective boards share--a rapport he calls the "x factor." He means a spirit of intellectual and social give-and-take--not an easy state of grace to achieve among a diverse group of overachievers who sit together in the same room only six or eight times a year.

I finally got my chance to see Grove and his x factor in action during a board meeting in late March. They let me watch an hourlong segment during which the board discussed issues on the agenda for the upcoming annual meeting, and Yoffie and Shaw reported back on site visits to Intel facilities in Korea and Argentina. But first, let's set the scene: Room 3 in the Executive Business Center at Intel Corp.'s headquarters in Santa Clara, Calif., where the directors meet, is the boardroom equivalent of a Dilbert cubicle. No sumptuous wood paneling, no Persian carpets, no oil paintings of the founders. Just a utilitarian, U-shaped maple table rimmed by 16 ordinary black-leather desk chairs that look as if they came straight out of an Office Depot catalog. Grove sits at the bottom of the U and usually asks the newer directors to sit near him so that he can prod them to speak up if they seem a little shy. Today Barshefsky, who is attending only her second meeting, sits to his left.

Grove hews strictly to the agenda throughout the day, allowing only a half-hour for a rather spartan buffet lunch. About half the meeting time is devoted to updates on Intel's performance, finances, sales and marketing efforts, and reports from board committees. There has also been a two-hour grilling of executive vice president Sean Maloney, the general manager of Intel's thrust into telecom chips and gear. Because Intel is counting on his business to be a growth engine, he faces the board at least twice a year. "This one went pretty well," he says afterward. "I didn't whack PowerPoints at them--the goal is discussion and it needs to be a vigorous experience. The board can be pretty tough with their questions because they do their homework, and you are held accountable for not living up to claims you've made in previous presentations. The only crime is to be defensive."

My chance to watch comes after Maloney's cross-examination. For the next hour much of the discussion revolves around the company's revised stock option plans, which are to be presented at the annual meeting in May. The compensation committee, chaired by Hundt, has consolidated the option plans for top management and rank-and-file employees so that their strike prices and vesting periods will be identical. The idea is to make the grants more egalitarian throughout the hierarchy, thereby helping forestall shareholder criticism. The discussion is boring, and I'm beginning to wonder what has become of Intel's vaunted combativeness when Barrett speaks. Turning to Grove, the CEO observes that it seems inconsistent for the directors to align the employee and executive options and not do the same with their own plan. "Why are we deviating here, boss?" he asks Grove. "What's so special about the board, especially given the fact that we're going to make such a big deal about it?"

A few directors shift in their seats, and there is nervous laughter, but nobody seems to want to reply. Grove is plainly irritated. "It's a little too late to fix it in time for the annual meeting, and it's going to be complicated to change, so let's table it for now," he snaps. End of discussion.

I wasn't the only one who thought this exchange odd. After the meeting Grove sought out Barrett and asked, "What was that all about?" Barrett started to laugh. "I was pulling your chain!" he said. "I didn't want our visitor to think all our meetings are as boring as that hour was."

Toward the end of my final interview with Grove, a three-hour marathon in late May, I asked him to define a great board. He answered in a heartbeat. "Continuity is the key," he said. "The goal is to ensure that the success of a company is longer lasting than any CEO's reign, than any market opportunity, than any product cycle. So you're looking for the board to act as a gyroscope, to stabilize the company over time."

That's a pretty good description of what his board has done. By this time next year, when Paul Otellini is expected to move into the CEO's corner cubicle, Intel and its board will have glided through two seamless CEO transitions in eight years. Its annual meetings have been a picture of calm, even though its opposition to expensing stock options places Intel on the unpopular side of the controversy. (At the most recent annual meeting, shareholders narrowly passed a resolution asking Intel management to begin expensing options.) Moreover, Intel's forays into telecommunications and consumer electronics chips, while not yet very profitable, are gaining traction. The company has opened several chip-assembly plants in China and is working closely with the government to set wireless standards that could favor Intel's technologies. Employees know their directors better than ever as a result of the site visits and other board outreach programs Grove has instituted. Several directors maintain running e-mail dialogues with Intel managers they've met during their peregrinations.

Just as significant is that during Grove's watch the Fore Systems problem has never recurred. Intel weathered the biggest slump in its business since the 1980s and still managed to jump a generation ahead of most of its rivals in chipmaking, largely because directors did not second-guess management's plans to spend $10 billion on plants despite the recession. The board has also resisted the urge to micromanage when Intel struggles--as it has, for example, in rolling out its Itanium microprocessor. (To be sure, such problems come up for spirited discussion at nearly every meeting.)

Grove is still five years away from "mandatory senility," but if he follows Intel custom, he'll cede the chairmanship to Barrett, who turns 65 next summer and must step down as CEO. What will happen to Grove when he is no longer chairman? After 31 years on the board and never missing a meeting, he could walk away to devote himself to causes dear to his heart, such as agitating for more effective cancer research. He could remain a rank-and-file director. Or he could, like Gordon Moore, sign on as chairman emeritus. "Now that's not much of a role to reinvent," the founder chortles. Only Grove knows what comes next, and he isn't saying.

And what will become of Intel when he leaves? There's no way to say, but its board, at least, is a well-oiled machine.