In 1971, the board of Kimberly-Clark, an old paper company whose stock had fallen 36 percent behind the general market over the previous 20 years, named a new chief executive. He was Darwin E. Smith, the company’s mild-mannered in-house lawyer. By Jim Collins’s account, a director pulled him aside to remind him that he lacked some of the qualifications for the position.
Two months into the job, doctors diagnosed nose and throat cancer and told him he had less than a year to live. Smith informed the board, Collins wrote, but said he had no plans to die anytime soon. He kept his schedule while commuting weekly from Wisconsin to Houston for radiation therapy.
Then he went after the company’s heart. Coated paper, made in its own mills, was Kimberly-Clark’s traditional core business. Smith and his team concluded it was “doomed to mediocrity,” and Smith announced that the company would sell its mills, even the one in Kimberly, Wisconsin, and put the money into consumer brands like Huggies and Kleenex, against Procter & Gamble.
The business media called the move stupid. Wall Street analysts downgraded the stock.
“Eccentric”
The man who did this was, in Collins’s words, shy, unpretentious, even awkward. When a journalist asked him to describe his management style, Smith stared back through thick black-rimmed glasses. After a long, uncomfortable silence, he said: “Eccentric.”
Humble about themselves, relentless for the company. That is Collins’s portrait of a great chief executive.
He called it Level 5 Leadership, in Harvard Business Review in 2001 and in his book Good to Great the same year. Level 4, in his hierarchy, is the familiar effective leader who rallies people behind a compelling vision. Level 5 adds a paradox: “personal humility plus professional will.” His team started from 1,435 companies that had appeared on the Fortune 500 and found 11 that met his test of a jump from good results to great ones. Collins says he told his researchers to downplay the top executives, to avoid the usual “credit the leader” story. They kept coming back to them anyway. Every one of the 11, he wrote, had a Level 5 leader at the time of transition.
The comparison companies supplied the contrast. Scott Paper, the rival he matched against Kimberly-Clark, hired Al Dunlap, who by Collins’s count took about $100 million for 603 days of work and later called himself “Rambo in pinstripes” in his autobiography. In July 1995, Kimberly-Clark announced a $9.4 billion merger with Scott Paper.
Collins noticed a habit he called the window and the mirror. When things went well, his Level 5 leaders looked out the window and credited other people, outside factors and good luck. When things went badly, they looked in the mirror. It gives a rough test anyone can use: listen to how someone explains a success.
Luck, said the man from Circuit City
One of his interviews makes the point almost too well. Alan Wurtzel had turned Circuit City from a ramshackle company on the edge of bankruptcy into one of America’s most successful electronics retailers; in the 15 years after its 1982 transition it beat the market 18.5 to 1, by Collins’s measure. Asked for the top factor in the turnaround, Wurtzel said: luck.
Collins took the answer as modesty. The leaders really were responsible, he wrote; they just wouldn’t admit it.
On January 16, 2009, Circuit City said it would liquidate all 567 of its U.S. stores.
That collapse came years after Collins’s 15-year window, so it doesn’t overturn what he measured. The deeper problem is in how the traits were found. In 2007 Phil Rosenzweig, a strategy professor at IMD, took aim at best sellers that promise “the way to go from good to great.” Studies like these, he wrote, lean on retrospective interviews and articles from the business press, sources colored by the halo effect: once a company is winning, people describe its leader, culture and strategy as excellent. Collins’s team had collected nearly 6,000 articles and conducted 87 interviews. Those were exactly the kinds of sources Rosenzweig meant. “Outcomes can be mistaken for inputs,” he wrote.
A humble chief executive and a great stock chart can travel together without the first causing the second. Collins himself was candid about one limit: asked how to become Level 5, he wrote that his team had “no solid research data that would support a credible list” of steps.
The two candidates
Imagine you are choosing a successor for your engineering team. One candidate gives the all-hands talk about last year’s migration and says “I” a lot. The other, asked about the same migration, names three colleagues and a quiet quarter with few emergencies. Asked about last spring’s outage, she says the rollout plan was her call.
Collins’s window and mirror points to the second. Rosenzweig’s warning says to check how much of that impression comes from knowing the migration went well. So read what each of them wrote before the result was known: the design proposals and the risk notes. Look for who named the hard problem early and kept pushing on it. Then you are judging the decisions, not the halo.
Smith retired in 1991, after 20 years as chief executive. Looking back on his record, by Collins’s account, he said only this: “I never stopped trying to become qualified for the job.”