The news from Cadence: goodbye to some of that

I’ve heard many takes on the news this week that Mike Fister and a substantial portion of his management team have left Cadence, and on the rumors of a substantial layoff in the near future. Let’s see … This is the first casualty of the gathering recession. Or it is an inevitable result of the shrinking chip design market. Or it is a bunch of Intel guys getting their comeuppance in an industry they didn’t understand. All plausible, and perhaps all have a grain of truth. But I don’t entirely agree with any of them.

To begin with, I think we need to be careful to separate the revenue and profit picture at Cadence from the contraction that is just starting in the semiconductor industry. There really hasn’t been time for the impact of the global political/financial calamity to percolate back to the EDA industry yet. Unfortunately. I wish we were further along in that process and looking at the light at the end of the tunnel, instead of just fumbling for our flashlights. But people I speak to in the industry tell me the problems at Cadence are from a different source, and will only be compounded by the recession.

Second, there is little question that the EDA market is maturing. What growth in seats there is appears to be happening primarily in Asia. But that doesn’t necessarily spell doom for EDA companies. In fact, it should be an opportunity for companies that manage consolidation well to improve their profitability, if the experience of other maturing industries over the years can be educational. So I don’t think we can look there for the source of Cadence’s problem, either.

Theories about the fitness of Mike Fister and his team I will leave alone, simply for want of unbiased data. It is impossible to be an effective executive without making both some mistakes and some enemies. This is a simple result of the fact that the executive has to make decisions with inadequate data, and then depend upon others to implement them. When I hear of executives who are said to make consistently correct decisions, I generally suspect either a string of very good luck or an excellent public relations campaign. See for instance Apple Computer. So I will leave this question with the observation that even in retrospect it is nearly impossible to reconstruct the actual context of any one executive decision or estimate its real consequences. That makes evaluating CEOs rather too complex for amateurs such as your correspondent.

But I think there are some pertinent observations from Cadence’s situation, none the less. The first, and one that I have heard repeatedly from industry executives, is that the all-you-can-eat license concept has been ruinous to the income statements of the large companies. By in effect capping the revenue from major accounts, the big EDA companies attempted to trade their opportunity for revenue growth per account for a shot at greater market share. But with the close competition between the companies, major accounts quickly saw that they could exploit this idea to their own advantage, and what was supposed to be a superweapon became mutually assured destruction. There is some reason to believe that Cadence may have been more deeply harmed by this than the other companies.

Second, I would conjecture that Cadence took the concept of all you can eat one crucial step further: to all you could ever want to eat. If the company has in fact bartered away future revenue opportunities as well as current ones, there could be a very acute problem on future revenues that no amount of clever management could solve.

Third, I think there is a question of size. The late Peter Drucker argued that for a given market situation there is a correct company size. If the company is too small, it will be unable to fund both growth and the risk of doing business. If it is too large, it will be unable to fund operations. Note that this concept does not allow for companies to naturally grow larger with age. History is full of examples of companies whose success has allowed them to grow too large for their market situations, leading to earnings stagnation and eventual dismemberment. Nor does this theory support the idea that the solution to problems in an industry is necessarily consolidation. Putting together several too-large failing companies merely creates one failing company with a lot more lobbying clout. See the US automobile industry.

I think this idea may capture the situation in which Cadence finds itself. The presence of every individual in an organization must be justified, either by the work necessary to produce current income or by the investments necessary for future income. Growth simply because it is allowed by current earnings is always risky, and often malignant. So perhaps it’s time at Cadence, not to do a reflexive layoff, but to carefully sort through the operations, figure out who is necessary to sustain the existing business, what if any strategy for revenue growth there is, and who is strictly necessary to keep that growth strategy on track today. That used to be called zero-based budgeting, but it proved so painful that middle managers quickly learned to subvert it, and the phrase withered away. It is hard work. It requires executive management that can force a consensus on the real state of the company’s revenue streams today and on its realistic growth strategies. And it is very different from a blanket layoff or even a directed lopping off of programs and products. It is, in a word, a turn-around, not an adjustment.

But the semiconductor industry needs that much of Cadence’s technology be available and supported. So it is in all our interests that this nasty job get done thoughtfully, yet quickly. It’s a tall order, but our thoughts are with the Cadence team as they take it on.