For IBM, Size Does Matter

Shareholders might demand that Palmisano break up IBM.

Timing is everything, and Lou Gerstner has it. Gerstner, of course, took over IBM at its nadir, brought it back to prosperity and then left, just months before massive layoffs and what could be the first quarterly loss in eight years. (OK, Lou is still chairman through the end of the year, but you can bet that the buck has not stopped at his desk since March 1, when Sam Palmisano took over as president and CEO.)

Poor Sam. He gets the job, and practically the first order of business is to carry out layoffs not seen since, well, early in the Gerstner era. In an interview earlier this year, Palmisano told me that IBM would remain a systems, software and services company because thats what customers want. While theres a logic to having all those components working in harmony, it ignores the issue of sheer size.

One interesting fact is that IBMs total number of employees is just a little more than what it was just before Gerstner took over. In 1992, the work force was about 302,000, and last year, it reached about 320,000. Gerstner oversaw a drop to about 220,000 in 1994 and then a gradual climb back. Could it be theres a law of nature at work here? That a company that does what IBM does can get to a certain size—the neighborhood of 300,000 employees—but no larger?

In the dark days, Gerstner resisted breaking up IBM into smaller, more agile companies, but Palmisano might have to do just that. Shareholders might demand it. Wouldnt it be better to own stock in three or four smaller, growing companies than to own a single stagnant stock? Breakup or no, my advice to Sam is that IBM will be a different company five years from now and that trying to maintain it as the company it has been is a strategy for failure: Think as far out of the box as possible.

And another thing: Why did HP think it needed to become as big as IBM by acquiring Compaq, anyway?

And why are no animals larger than the blue whale? Tell me at