Incentives, Deterrence, and Defence     European Software Centre

Searching for a New CEO

After Al Green announced his intention to retire as CEO, a search firm was retained to manage the process of finding a new CEO for Autodesk. The Board of Directors appointed a search committee, chaired by Jim Warren, which appealed for input, company-wide, on which criteria should govern our search. Here's my response.

Profile of a CEO

by John Walker — October 24th, 1991

This paper is an attempt to identify the characteristics Autodesk should look for in candidates to become the new CEO. As you'll see in reading it, I don't think there's any single precise picture that fits the job; there are essential prerequisites, but not a unique correct profile. I use male pronouns for concision and to avoid sounding like a politically correct Marin County weenie, not to specify the gender of candidates acceptable to me.[Footnote]


My first assumption is that Autodesk is looking for an individual who will be a leader and a strong, highly-competent manager. Autodesk is not searching for a super-hero, nor are such characteristics needed to succeed as CEO, nor would we likely to find such a person if we were foolish enough to seek one.

Second, we cannot expect to find candidates who have already demonstrated success in running organisations of Autodesk's size and complexity. People who are currently CEOs of multinational S&P 500 companies with a billion dollar market cap, quarter billion revenue, rapid growth, gut busting margins, and more money than Scrooge McDuck are rarely out on the job market, and if they are it's probably an indication we don't want 'em. Whoever we pick will almost certainly be assuming the largest challenge so far in their life, and that's even more reason to focus on the fundamental capabilities that indicate the potential for growth into and in the job. If you find Bill Gates or Philippe Kahn standing in line at the unemployment office…hire them, but don't waste a lot of time looking.



Your reaction to at this point is probably, “That's it? What about experience in high-tech, or international companies, or CAD, or in public companies. What about the size and nature of the organisations the person has run?”

I believe we must be interested in every aspect of each candidate's background, weighing them in terms of the extent they demonstrate his qualifications or lack thereof to meet the challenges that will be waiting on the desk the morning he sits down at it the first time. But as far as I'm concerned, the only requirements are the ones I listed.

I believe people with a wide variety of backgrounds could succeed as CEO of Autodesk, which is not to say the job is easy; only that being a CAD guru, a numbers whiz, or a super salesman isn't the single background that qualifies one to do it.

The most essential part of the process of selecting a CEO will be the interviews the candidate has with the operating managers of the company, the people he will, if chosen, have to rely on to implement the policies and strategy he develops, or develop and/or replace if they do not. If we select a candidate who is enthusiastically endorsed by Ruth Connolly, John Lynch, Pete O'Dell, Lew Goldklang, Richard Cuneo, Frank Balinson, Tom Mahood, Volker Kleinn, Scott Davis, and Marc LeBrun, I'm not going to lose any sleep over whether the person is likely to succeed in the job, even if I'd never met the guy. In the process of interviewing candidates, one of the most important indicators I expect to use is the individual's reaction to the various people he has met already within Autodesk. If the person has the right sense of people, he's probably going to be able to get the job done.[Footnote]

As unlikely as it may seem that the list of people I reeled off would ever agree on anything, it is essential that any candidate not only pass the scrutiny of the overwhelming majority of these people, not only as somebody who's “OK”, but as a person they're excited about reporting to. And if one or two individuals dissent, then it's likely their careers at Autodesk may not continue far into the term of the new CEO, as the first priority of any new CEO must be assembling a team he can work with effectively.

Although I certainly believe that choosing the next CEO is the most important decision Autodesk faces at this time, I'm not as worried about our choosing unwisely as most of the people I speak to seem to be. I believe there is a tremendous pool of wisdom in the operating and middle managers at Autodesk, the people who now feel so frustrated at not being permitted to carry out the plans that seem so obvious to them. I am confident that when they see a person for whom they could work, a person who would support them in turning Autodesk around, they will recognise that talent, endorse the person, and urge the board to select him. We're probably going to have to interview more people than we'd like to, but I think we'll be able to conclude the search with satisfaction and confidence, not in frustration and acceptance of a “least-bad” alternative.


If all of this seems too amorphous, I think in part it's because we won't get a real feel for what we're looking for until we do a little looking. Just as in buying a house, the best way to understand what you're looking for is seeing several alternatives; it helps you sort the truly important from the merely desirable.

If we can agree on a set of fundamental requirements, whether these or others, whether a short list or a long one, it'll help sort out people who simply haven't demonstrated the qualifications for the job. Then it's up to us, through interviews, to decide, among a series of qualified people, who is most likely to succeed.

Incentives, Deterrence, and Defence     European Software Centre