How Intel Grooms Its Leaders The succession formula is downright revolutionary: It picks CEOs years in advance, without drama or surprise.
By G. Pascal Zachary
(Business 2.0) – Next May, Paul Otellini will likely become the new CEO of chip powerhouse Intel--but for now, earning the lofty title means submitting to a humble exercise: hitting the books. As the first Intel chief executive without a degree in science or engineering, the soft-spoken 53-year-old doesn't yet have the technical expertise that mentors like CEO Craig Barrett and chairman Andy Grove possess. Which is why Otellini, the company's current president and COO, has already crammed in more than 50 tutorials over the past year, on everything from next-generation wireless networks to microprocessor design, with many more to come. His main tutor? Intel chief technology officer Pat Gelsinger. "Paul will say, 'Teach me about something,'" Gelsinger says. The hour-long sessions, he adds, are "light on slides, long on discussion." Otellini says his plan for the months ahead is simple: "Work harder."
The training regimen isn't some chore handed down by the HR department. It's part of a little-known but deliberate philosophy at Intel to grow and groom its own CEOs and--in defiance of conventional wisdom--turn loyal company "lifers" into successful leaders.
In an era of corporate headhunters, celebrity CEOs, and management by "creative destruction," succession at Intel, one of America's most profitable manufacturers, is a rare model of discipline. The company plans orderly regime changes years in advance, without enervating gossip, infighting, or drama over the identity of the new boss. "By knowing succession is assured," says Richard Doherty, research director of the Envisioneering Group, a consultancy that closely follows Intel, "resources expended on wondering who the next favored son is instead go to customer relations or better understanding of the divisions." And the company has consistently turned out some of the most respected CEOs in corporate America.
That Otellini will become the fifth homegrown CEO to run the company since its launch in 1968 suggests that there's an "Intel inside" aspect to its management formulas as well as its high-performance chips. The first two chiefs, Robert Noyce and Gordon Moore, weren't just founders but legends in their industry. The third was Grove--one of Intel's original employees and considered one of the best executives of the 20th century. Intel's current CEO, Barrett, a renowned manufacturing guru, taught materials science at Stanford before joining Intel in 1974.
Next May, Barrett will give up his job soon after turning 65--his board-mandated retirement age. At most companies such an imminent switch might trigger alarms. Not so at Intel, which anointed Otellini as heir apparent two and a half years ago--the equivalent of an epoch in an industry that lives and dies by Moore's Law. (So ho-hum is the anticipated hire that the question of succession never came up at the company's annual meeting in May.)
The long lead times are a hallmark of Intel CEO transitions, mainly because the company's board of directors insists on them. "We discuss executive changes 10 years out to identify gaps," says David Yoffie, an Intel director since 1989 and a professor at Harvard Business School. Every January, he says, the board receives rankings of two dozen or so senior managers. Then it devotes portions of two or three more board meetings to combing through the list. Choosing the CEO, Yoffie says, "is the single most important role of the board."
The Long Handoff
The board's obsession with the future helps foster another crucial element of the system--a gradual shift in duties from one CEO to the next. Moore set the example in the mid-1980s, when he allowed Andy Grove, then his second in command, to gradually assume CEO chores; likewise, in the mid-'90s, Grove steadily ceded his authority to Barrett. In effect, says Les Vadasz, an original employee and former director who has witnessed every CEO hire, "the successor gets the job before he gets the title."
Training successors by these tacit fraternal codes is all but unheard of in Silicon Valley, where founders can hold on too long and where talk of life without the chief is often heretical. Steve Jobs, Apple's CEO, remains synonymous with the company he founded--and returned to save in 1997. But how many more years will that be possible? Scott McNealy, CEO at Sun Microsystems, has suffered high turnover in his senior ranks in part because he refuses to step aside after two decades at the top.
"Two in a Box"
At Intel, though, CEOs and their apprentices swap roles not just to burnish skills but to streamline performance where it's needed. The practice flows out of a wider Intel ideal, known internally as "two in a box." By encouraging overlapping duties and responsibilities, the thinking goes, Intel managers can better support one another in a crisis.
Hence, another aspect of Otellini's grooming: A 30-year Intel veteran who made his reputation running the company's flagship microprocessor line, Otellini has increasingly taken charge of Intel's worldwide manufacturing operations and its enormous budget for capital projects--including chip factories that typically cost $3 billion apiece. His growing influence over internal operations has freed Barrett to focus on vital external tasks, such as lobbying Washington on R&D policy and dropping in on top customers. Likewise, Grove's public profile got a boost in the 1990s when Barrett, during his grooming phase, started taking over Intel's day-to-day operations.
Noyce established another important principle of CEO succession: Intel chiefs don't overstay their welcome. After tiring of the position, Noyce stepped aside in 1975, handing off his title to co-founder Moore. Similarly, Grove ceded power in 1998, while only 61 and at the height of his reputation.
So how do you get CEOs like Grove, who coveted executive power, to so easily walk away? Here again, Intel defies management gurus by letting outgoing chiefs wield authority long after they've left the corner suite. While Lou Gerstner vanished shortly after stepping down as IBM's CEO in 2002, leaving his successor, Samuel Palmisano, as both chairman and CEO, departing Intel CEOs are groomed to simply slide over to the boardroom. Grove became chairman shortly before Barrett's ascension, and Barrett is expected to do the same when Otellini takes over. The process promotes continuity and adds incentive for the outgoing CEO to give his successor a strong hand.
Like any management credo, of course, Intel's approach has its drawbacks. The most obvious is that it turns current CEOs into lame ducks sooner than at other companies. Barrett copes with the situation through the strength of his personality; aggressive and decisive, he can bully his decisions to the finish line. "He's still a force to be reckoned with," says Eugene Meieran, one of Intel's most senior scientists.
Still, the clock is ticking on Barrett, and some know they can wait him out. When Barrett tangled recently with CTO Gelsinger over a technical decision, he ended the debate by backing his own view. "That's OK, Craig," Gelsinger shot back. "I'll change it when you're gone."
Intel struggles with another familiar trade-off of succession. At a company that broadcasts its succession plans years in advance, talented, loyal managers who don't see a path to the top aren't likely to stick around. In May, one of Intel's most valued execs--Mike Fister, head of its server processor division--left Intel to take the top job at Cadence Design Systems, a longtime supplier of chip-design software. Dave House, another highly regarded Intel executive who was considered CEO material, left in 1996 when he saw himself losing not only to Barrett but to Otellini, who had the inside track for the next CEO opening.
But as Otellini knows, he's on a leash. Intel's board recently declared 60 the mandatory retirement age for CEOs. That means Otellini, who will be 54 when he takes the top slot, would step down after six years. While the board could extend Otellini's reign, the odds are against it. Succession at Intel, after all, goes according to plan.
G. Pascal Zachary is a senior writer at Business 2.0.
Case | HBS Case Collection | June 2002 (Revised October 2005)
Inside Intel Inside
by Youngme E. Moon and Christina L. Darwall
In early 2002, Pamela Pollace, vice president and director of Intel's worldwide marketing operations, is debating whether the company should extend its "Intel Inside" branding campaign to non-PC product categories, such as cell phones and PDAs. The "Intel Inside" campaign has been one of the most successful branding campaigns in history. However, the campaign is more than ten years old, and growth in the PC market appears to be stagnating. In contrast, sales of portable digital devices--such as PDAs and cell phones--appear to be growing at a healthy rate. Pollace is debating whether the "Intel Inside" campaign will work in these other product categories, even though Intel doesn't dominate these other markets like it does the PC market, and it isn't clear that consumers will associate Intel with these other markets.
Keywords: Advertising Campaigns; Growth and Development; Brands and Branding; Marketing Strategy; Product Positioning; Sales; Expansion; Competitive Advantage; Semiconductor Industry; Manufacturing Industry; California;