January 2001

Level 5 Leadership
(Sidebar: One Question, Five Years, Eleven Companies)
Harvard Business Review

by Jim Collins

The Level 5 discovery derives from a research project that began in 1996, when my research team and I set out to answer one simple question: “Can a good company become a great company and, if so, how?” Most great companies grew up with superb parents—people like George Merck, David Packard, and Walt Disney—who instilled greatness early on. But what about the vast majority of companies that wake up partway through life and realize that they’re good, but not great?

In the original Harvard Business Review publication there is a chart that is a graphic illustration of the overall research project that appears at this point in the article. This particular chart is incompatible with this web site’s format. For this chart, please refer to the original publication of this article.

To answer that question, we looked for companies that had shifted from good performance to great performance—and sustained it. We identified comparison companies that had failed to make a sustained shift from good to great. We then studied the contrast between the two groups to discover common variables that distinguish those who make and sustain a shift from those who could have but didn’t.

More precisely, we searched for a specific pattern: cumulative stock returns at or below the general stock market for 15 years, punctuated by a transition point, then cumulative returns at least three times the market over the next 15 years. (We used data from the University of Chicago Center for Research in Security Prices, adjusted for stock splits, and all dividends reinvested.) The shift had to be distinct from the industry; if the whole industry showed the same shift, we’d drop the company. We began with 1,435 companies that appeared on the Fortune 500 from 1965 to 1995; we found 11 good-to-great examples. That’s not a sample; that’s the total number that jumped all our hurdles and passed into the study.

Those that made the cut averaged cumulative stock returns 6.9 times the general stock market for the 15 years after the point of transition. To put that in perspective, General Electric under Jack Welch outperformed the general stock market by only 2.8 to 1 during his stellar 15 years from 1986 to 2000. A dollar invested in a mutual fund of the good-to-great companies in 1965 grew to $470 by 2000—compared to $56 in the general stock market. These are remarkable numbers, made all the more so by the fact they came from previously unremarkable companies.

For each good-to-great example, we selected the best direct comparison, based on similarity of business, size, age, customers, and performance leading up to the transition. We also constructed a set of six “unsustained” comparisons (companies that showed a short-lived shift but then fell off) to address the question of sustainability. To be conservative, we consistently picked comparison companies that, if anything, were in better shape than the good-to-great companies in the years just before the transition.

With 22 research associates working in groups of four to six at a time from 1996 to 2000, our study involved a wide range of both qualitative and quantitative analyses. On the qualitative front, we collected nearly 6,000 articles, conducted 87 interviews with key executives, analyzed companies’ internal strategy documents, and culled through analyst reports. On the quantitative front, we ran all financial metrics, examined executive compensation, compared patterns of management turnover, quantified company layoffs and restructurings, and calculated the effect of acquisitions and divestitures on companies’ stocks. We then synthesized the results to identify the drivers of good-to-great transformations. One was Level 5 leadership. (The others are described in the article “Not by Level 5 Alone.”)

Since only 11 companies qualified as good-to-great, a research finding had to meet a stiff standard before we would deem it significant. Every component in the final framework showed up in all 11 good-to-great companies during the transition era, regardless of industry (ranging from steel to banking), transition decade (from the 1950s to the 1990s), circumstances (from plodding along to dire crisis), or size (from tens of millions to tens of billions). Additionally, every component had to show up in less than 30% of the comparison companies during the relevant years. Level 5 easily made it into the framework as one of the strongest, most consistent contrasts between the good-to-great and the comparison companies.