Stating the Problem

As summer gave way to fall in 1991, I grew increasingly exasperated with the inability of Autodesk to come to terms with the challenges facing it. Since I moved to Neuchâtel in May of 1991, I'd had impromptu visits from John Lynch and K.C. Jones, Al Green, and Malcolm Davies, all seemingly willing to listen, but for one reason or another, unable to act upon the issues we all agreed were vital to the survival and continued success of Autodesk.

When I wrote Information Letter 14 (see page ), the thought never entered my mind that we'd need to replace the senior management of Autodesk—all they needed was some education in the realities of the contemporary software market, and they'd be able to continue their five-year unbroken success—streak which was, after all, longer than my stint as Autodesk CEO.

As the months passed and the only evidence of change at Autodesk was a “reorganisation” that re-drew the TO but didn't seem to change the priorities of the people in the boxes, I began to think more and more about whether the only way Autodesk could escape the downward spiral I believed it to be in was to find new leadership at the top. On September 26, 1991, I wrote the following letter and circulated it to a very short list of people whose discretion and judgement I trusted entirely, to see if I'd gone off the deep end in reaching this conclusion. I was not alone.

After deciding that a change a the top was required, I started working out how it might best be accomplished. Unexpectedly, as I was pondering ways and means to bring about change, Al Green contacted me to say that he'd decided to retire and initiate a search for a new CEO. The coincidence may be difficult to accept, and conspiracy buffs will never believe it, but that's what happened—I guess Al and I reached the same conclusion at the same time.

My respect for Al Green and what he accomplished at Autodesk (see page ) was reinforced when he, as I had in 1986, concluded that Autodesk would be best served by handing the tiller to a new skipper.

Stating the Problem

by John Walker — September 26, 1991

Back when I was teaching myself engineering out of the Junior Woodchuck Manual, one of the first things I learned was that often the hardest single step in solving a problem is simply stating it correctly.

Recently we've all been involved in discussions about the state of Autodesk, post re-org, and I think it's safe to summarise them by saying that we all believe that serious problems remain unaddressed. Before we start considering what might or might not be done about them, let me see if I can state the problem as clearly as possible, and then let's see if we agree on it. If so, we'll sweep aside lots of irrelevant stuff that might otherwise occupy our attention—matters which are symptoms of the problem rather than the problem itself, and which can best be addressed by fixing the cause, not palliating the effects. I find that my reaction to the seemingly accelerating flood of “zingers” which create in me such states of anger and frustration that I am barely able to do my work has a tendency to obscure the problem which I now believe is the First Cause of the individual events. Over the past three weeks I've been trying to get a handle on the nature of the underlying problem.

So here goes.

The problem is executive management.

Autodesk doesn't have any.

The people who form what is now called the Executive Review Board are genial, well meaning, and for the most part hard working individuals. But whether from lack of capacity or failure of will, they are not behaving as executives. By failing to provide the leadership Autodesk needs, they are depriving our company of a resource which is generally recognised to be one of the principal factors which determine the future of any business.

Autodesk's executives are behaving as supervisors or managers, not executives. There's a difference, and it's crucial.

A manager carries out plans, accomplishes projects, and implements processes that are defined for him by his superiors. A good manager will show initiative in identifying opportunities in his area worthy of pursuing, but will present such proposals to the executives to whom he reports, then defer to their judgement as to the merits of each proposal and whether it is consonant with the overall strategy of the organisation. A manager who consistently identifies new ventures which eventually succeed is spoken of as “executive material.”

Autodesk's executives are not executive material. Present them with a clear description of a task: “Find a campus site,” “Reorganise around product lines,” “Diversify into molecular modeling,” and they will usually get the job done. That is management, not leadership. It is not what executives do. Executives must identify the tasks which the company should undertake, select the managers to accomplish them, and then provide the leadership and support to ensure that the tasks are carried out as they envisioned.

When have Autodesk's executives ever done this?

Even the name “Executive Review Board” betrays a passive, reactive view of the rôle of the executives. I think it's perceived as something more like a congressional appropriations committee than the leadership team of a company which intends to extend and expand the enviable record of growth in its first decade into the second and beyond.

Their mistaken conception of the rôle of an executive is not something Autodesk's management is inclined to hide—in fact Malcolm Davies' recent “refutation” of my Information Letter 14 in Cadence vividly demonstrates it. “ ‘The Windows project was already on,’ says Davies, one of those at Autodesk who does call the shots.” Indeed, and who would doubt that it is the most important strategic software development project in the company. Who initiated it? Which executive first identified the opportunity in Windows, then later the risk of being too late with too little in the market? Which executive decided to divert resources to that central strategic project, or to hire additional people if necessary to get it done? Who “called the shot?”

Allow me one more illustrative quote, which I found flabbergasting yet entirely consistent with the outlook I've found in lengthy discussions with all of the executives.

“I've been planning in-house support for the last three years…. We have actually initiated the first in-house support contract with a major account, but we're still wrestling with the actual format for it. I think John's memo has probably accelerated that and maybe reduced some of the opposition to it in Autodesk. He's made a number of changes we wanted to make a lot easier for us.” There's an attitude, Davies said, of “if John says it, he's God, therefore it must be true.”

These are not the words of an executive. If an executive wishes his company to provide in-house support, it does not take three years and wielding the bludgeon of a memo from a founder to implement his plan. An executive will draw up the plan, present it to his managers, and direct them to implement it to his satisfaction. If they cannot or will not, he is empowered to replace them with others who will. Japanese companies design an entire automobile from scratch in three years. And by the way, which executive was it who initiated the plan for support contracts and aggressively sold it throughout the organisation?

Autodesk's so-called executives see themselves as the prisoners of the circumstances of the company's history and current events. Autodesk needs leaders who understand the immense power they hold to define the company's future in any way they judge to be wise. One Autodesk executive told me, face to face, than he had been focusing on diversifying into other product lines and away from AutoCAD because he had given up on getting the software development department to produce future versions of AutoCAD which would remain competitive, and therefore growth must be sought in other areas.

Whether you agree with the diagnosis or not, the cure is not the approach of an executive. An executive defines the products which software development must develop and charges the development managers with building them. He assigns whatever resources are needed, terminating less important activities if required, and then supports his managers with consistent and vigorous leadership to aid them in completing the undertakings he has entrusted to them. If they fail consistently, then he removes them and gives the task to others who can succeed. If they also fail, he removes them and tries once again. But an executive cannot give up on the future of the product which generates essentially all of his company's sales and profits. If repeated frustration causes him to lose hope, he must not resign himself to the situation, he must resign. Any other course violates his duty to the company, its shareholders, and its employees to lead them into the future.

In addition to a can-do philosophy and initiative, an executive must be engaged in the industry in which he operates. As a leader of an important part of that industry, he is a statesman for the industry as a whole and his company's rôle within it. He is aware of what is happening in the marketplace, what problems and opportunities developments in technology may create for the company, and is constantly seeking ways in which the company's assets can be multiplied by deploying them in ways overlooked by competitors: creating, for example, new markets for the company's existing products.

I don't think Autodesk's executives know very much about the CAD business. Frankly I don't think they're very interested in it. Joel Orr put it like this, as best as I can recall, when I met with him recently, “When I talk to the people in Sausalito about what's going on in the industry, it's as if they're hearing these things from me for the first time.” It was abundantly clear from our discussion that this situation is unique to Autodesk, at least among companies he deems successful. Joel does not encounter this ignorance within the management teams of the many other CAD and computer graphics firms with whom he meets in the course of his consulting practice.

You don't have to be an all-knowing guru to be an executive. John Sculley probably knew very little about computers or the computer business when he became president of Apple. But by the time I met him a couple of years later, he had certainly learned a Hell of a lot about it in the course of doing his job. You might not agree with his perceptions and you may consider the strategy he adopts to be a prescription for disaster, but there's no difficulty in ascertaining what his strategy is or what assumptions it's based upon. The same could certainly be said for companies like Microsoft, Borland, Intergraph, DEC, and IBM. Having a strategy doesn't guarantee success; you have to have the right strategy, to be sure. But having no strategy, reacting to events in a totally passive manner, almost certainly guarantees failure. Who would have predicted, or indeed could have conceived even a few years ago that Ashton-Tate would eventually be acquired by Borland? I would suggest that anybody who noted the lack of leadership and strategy at Ashton-Tate could have envisioned such an event. Companies which are vigorously growing, adding new products, and broadening the market for their existing offerings end up devouring, one way or another, others who view themselves as custodians of a single success, operating in a defensive rôle rather than as leaders.

What is Autodesk's strategy for the CAD business? Which member of the executive team has, can, or will articulate it?

To look at the future of a company, you have to see beyond current operating margins to rate of growth. Autodesk's rate of growth is falling with every passing year precisely as you'd expect when the market for its only product is maturing. And if, as Malcolm Davies said in Cadence, “We still believe the CAD business has a tremendous upside for us. We don't believe that here in the United States, the most saturated market in the world, that we're anywhere beyond 25 percent saturation of the CAD market…”, then why is Autodesk's rate of growth falling, not rising? And what strategy do our executives have to reverse this situation? And which executive will develop it and lead the company to implement it?

In the almost ten years of Autodesk's history, only a single diversification has been proposed by a member of the present executive team, and that was to buy Generic, another company in our own business, and was presented as largely a defensive move. Not a single diversification has succeeded. In the “World Class” awards in the current PC World, the companies below were represented by the following number of products, counting winners and runners up.

All of Autodesk's products were in the CAD category, of course, and to some extent therefore compete with one another. Most of the other vendors are broadly diversified across many different areas.

It's well to remember that Microsoft was a one-product company in the late 1970s, and that Borland was a one-product company for quite a while after Autodesk was founded. Neither is today. They have outgrown Autodesk. Why? Because Bill Gates and Philippe Kahn are engaged in the industry, developing strategies to further their companies' position in the marketplace, and providing leadership to their organisations to carry out their plans—they are acting as executives. Ours aren't.

The reorganisation has been presented, even by the executives, as leading to a more effective Autodesk by removing the executives from positions of leadership in relation to the products. While this can, and I believe is, helping AutoCAD by replacing individuals who took a custodial approach to the product with others who will focus on the still unexploited opportunities it holds, it doesn't address the future of Autodesk as a company. If the same people who, by reorganising, acknowledged their own ineffectiveness in leading AutoCAD are still running the company as a whole, why should we expect them to do any better with it than they did with AutoCAD?

I'm not even sure they're interested. Again, Malcolm Davies told Cadence, “We really feel the whole future of computing is multimedia computing. Everything we learn in that effort, in the multimedia side, is going to directly benefit every other business we're in. If it ever becomes a giant application business in its own right, that remains to be seen, and if it does, that's going to take a long time.”

Well, if it's “the whole future of computing” then it's pretty likely there's some money to be made there, somehow or other, and it would seem to me that an executive team who wants to see their company prosper in that future should be aggressively creating that market so when it finally arrives they end up owning it. Just like we did with AutoCAD. Instead, they'd apparently rather wait and see, and increasingly disappear from sight in a market they could have utterly owned by now, while Microsoft, IBM, and others are aggressively investing in developing multimedia and guiding its emergence in their own interests. No, instead they cut the multimedia budget because “it isn't paying off,” and justify what we are spending by its benefits to AutoCAD.

Imagine if Bill Gates had viewed his plunging into operating systems based on its benefits to sales of BASIC. Or if Philippe had positioned Sidekick to pump the sales of Turbo Pascal. What is missed if you focus on AutoCAD's numbers is the number of opportunities Autodesk has missed or bungled which could have, by now, grown us far beyond our present success. There is not a single thing Borland has done which we could not have done, given the will and the competence to carry it out.

What drives me nuts about the overt symptoms of this problem, which include, for example, the “Cyberspace” issue of Scientific American where we finally coughed up the money for an ad for…Chaos(!), and the accelerating disappearance of Autodesk from consideration as a player in things we pioneered, often years ago, is that I've seen it all before. Precisely.

Around 1980, it became clear to me that I was losing it with Marinchip. In '78 and '79, I had staked out a position as a technology leader—first true 16 bitter, first 64K dynamic RAM board that worked, Unix clone operating system, Pascal compiler, etc., etc., but by the time '80 arrived almost nobody knew who Marinchip was. I didn't have the time or the money to promote and, furthermore, I didn't have the time or money for engineering to port from the dead-end 9900 chip I was using to the 8086 or 68K. I knew I had to do it to survive, but I just didn't have the resources.

It's a uniquely maddening kind of helplessness to see others become the “inventors” of things you were doing 2 or 3 years ago, and it hurts even more when they begin to really rake in the dough and become able to fund their consolidation of the market from cash flow while you can't even maintain your old stuff. This is what is happening at this very moment to Autodesk in the fields of desktop video and virtual reality, and it may be beginning to occur in CAD.

What's different between Marinchip and Autodesk isn't the state of the slide, it's that Marinchip had an excuse—lack of money—that Autodesk doesn't. Most of the things that Autodesk needs to be doing but isn't can be done by telling people to do them or by writing checks. And we have lots of talented people and plenty of money.

Another telling parallel was that Marinchip's sales and profits continued to increase from 1980 until I deliberately beat it to death in 1982 in order to focus on AutoCAD. It continued to grow from momentum but it had lost the opportunity to broaden its market and ever be a real success.

What Marinchip lacked in the key period when it had its shot at success was executive leadership. I failed my company at that time, and eventually my company failed.

Although I'm the one stating the problem here, I'm hardly the first to point it out. Several years ago, Jim Stocker was telling me exactly what I'm telling you now, and not that long ago I supported his removal from the board when he continued insisting on addressing this very problem. For that matter, nine years ago Dick Elkis identified the problem precisely, but I wouldn't listen because he was talking about me. He was absolutely right. You can get incredibly lucky, but luck won't carry you as far as Microsoft has gone.

Finally, in order to succeed in implementing the strategies he develops, an executive must lead. In order to lead, one must have the confidence of the people who will follow. Today, among the employees of Autodesk who think about such things, I doubt you could find five people who have confidence in Autodesk's executives or who could begin to describe what strategy they have for the company's future. I can not name one single person who has that confidence.

Engagement, strategy, and leadership. This is what an executive must provide his company. It is what a company must demand of its executives. Autodesk's executives are managing our company but they are not leading it. Autodesk, and everybody associated with it, is suffering the consequences.

That's the problem.