Not Good to Great

I’ve mentioned the book Good to Great before, as it was where I read about The Stockdale Paradox. It’s worth quoting again, as great philosophy.

You must never confuse faith that you will prevail in the end — which you can never afford to lose — with the discipline to confront the most brutal facts of your current reality, whatever they might be.

Given my first exposure to the book was this section, I had very high hopes for the remainder of the book. This is the 30th book to get a rating on my book rating system introduced in October 2007. There’s been one top score of 5 in this series, Gang Leader for a Day. Of course, there are other 5’s on my bookcase, like Fahrenheit 451 and The Four Loves. I just read them before this series started.

At the low end, several books have gotten a 1 out of 5, but there have been no zeroes. Good to Great came close, closer than any book beforehand. In the end, though, it squeaks by with a 1 out of 5, primarily because of James Stockdale. I was severely disappointed.

So, what went wrong? Let’s start with the obvious, the decline of mentioned companies. This book was published in 2001, which dates the research to 1999 and 2000. Mr. Collins and his team focused on eleven companies that appeared successful at that time: Abbott, Circuit City, Fannie Mae, Gillette, Kimberly-Clark, Kroger, Nucor, Philip Morris, Pitney Bowes, Walgreens, and Wells Fargo. Let’s look at their performance over the last ten years. I’ll use the same metric, stock price. As a comparison, the S&P 500 index was at about 1328 ten years ago. Unlike the NASDAQ, Dow Jones, and house prices, the broader market had less of a boom. It got up about 1500 in 2000, declined into the 800s in 2003, and rebounded into the 1500s last year, before the crash. Right now the index is in the low 900s. In other words, a broader comparative market has lost about 30%, all in the last year.

  1. Abbott: from 43 to 45. Above average, not great.
  2. Circuit City: bankrupt.
  3. Fannie Mae: from 70 to 0.51. Bailed out by the government.
  4. Gillette: from mid-40s to about 50 in 2005, when bought by Proctor and Gamble. Average.
  5. Kimberly-Clark: from 54 to 52. Above average, not great.
  6. Kroger: from 31 to 21. Average. It made the 2000 decline than never really recovered.
  7. Nucor: from 11 to 41. Great, taking off in 2004 along with world steel demand.
  8. Philip Morris: from 9 to 16. Very good, EXCEPT that they had to rename their company to Altria because of negative press.
  9. Pitney Bowes: from 63 to 20. 1999 was their high point, never reattained.
  10. Walgreens: from 28 to 28. Another above average consumer products company like Abbott and Kimberly-Clark.
  11. Wells Fargo: from 22 to 22. Above average performance. It took some bailout money, but relatively little. It’s considered strong.

Of the 11 companies, three (Nucor, Philip Morris, Wells Fargo) have excellent performance for their industry group. But one of those three had to rename itself and another took government money. Four have been above average but not great, two average, and two went bankrupt. I wouldn’t call this Great Performance. Thus, we’ve hit the first problem. What they found didn’t transfer. It’s more like Good to Great, temporarily, then basically Average.

Searching deeper, there were two huge problems with their approach. The first, more obvious, one is relying on a flawed metric. The research team used Stock Price, likely because it’s the only thing with sufficient history. Unfortunately, Stock Price makes the ultimate judge the Kapitalist Fundamentalists that I detest. Maximizing shareholder value is not what corporations should do. For a team that conducted dozens of interviews with each “winner”, failing to consider the definition of “win” is a shocking flaw.

The second one is subtler and statistical. Mr. Collins and his team defined success numerically, then tried to find explanations for success. A statistician like myself would call this proof by exploration. They used an exploratory technique, looking backward in an observational study, and then tried to prove things from observation. All my students in Math 200 and 205 should know better than that. Unfortunately, this happens so often in Business books that it gets a name: Survivor Bias. The Freakonomics fellows noticed the problem. Nassim Taleb wrote a book about it. There’s an entire website,, on this problem.
Survivor Bias is a major failing.

Is most of the advice even complicated? Well, no. Let’s summarize: Make sure leaders delegate and focus on company succession, not personal success. Quality people matter more than product. Understand the Stockdale Paradox (the hardest one). Keep to your plan, slow and steady, with a culture of discipline. Outside high tech, technology accelerates but does not transform. Successes and failures turn a flywheel, not a jumpstart. Like most business books, this is not difficult stuff. Then again, it’s business.

Overall, this was a very disappointing use of my money, even with a coupon. I had started with the best three pages, true philosophy from a Hero. And I don’t mean Mr. Collins. My suggestion is to just listen to the Stockdale Paradox piece and ignore the rest. If you want to learn about people for business, read Gang Leader for a Day. Street boss J. T. will teach you more.

About Adam

My quest is a world where calling someone "virtuous like a fairy tale hero" is routine, not fantastic or ironic. My vocation is the teaching and learning of statistics. My dream is a long happy life with a wonderful wife and kids. Who knows if any will become true? More information is at my homepage on the twelvefruits network:
This entry was posted in Book Reviews. Bookmark the permalink.

One Response to Not Good to Great

  1. Ron Barr says:

    Your review is disappointing and misses the mark. Collins never claimed to be predicting success – he was trying to explain the success that had already occurred.

    To claim that Good to Great is flawed because the companies it profiles did not do well a decade later is wrong-headed. Nowhere in the book did Collins claim that the next 15 years were going to be great.

    You can fault his methodology for its anecdotal nature, but he lays it out in the book and there is wisdom there. To dismiss it it out of hand is like claiming that Moneyball is a bad book because the Oakland A’s are having a losing season this year.

    The point of the book is that there were a set of companies that had performed well over a 15 year period. He looked at that set and identified common attributes in their leadership. He made a case that those attributes are good things for a company. It’s very self-evident.

    One of the major points of the book is that good CEOs (by his definition) is that they work consciously and in good faith on succession. I would argue that CEOs are the worst people to worry about succession. They are too close to the job to identify a successor, and because people will falsely attribute the wrong traits to their success. If my assertion is correct, then CEOs will fail at creating a succession plan, and the company’s success will revert to the mean.

Replies are Welcome.