Its About Time for IBM to Leave PCs Behind

Opinion: While the company's exit from the market may be historic, its impact will probably be about nil for customers-and it might even be an improvement.

Almost exactly 24 years ago, IBM got into the personal computer business. Now, its getting out, as it prepares to sell off what remains of its PC and notebook operations—which amount mostly to a team of designers and engineers and a pile of intellectual property.

And while that exit may be historic, its impact will probably be about nil on IBMs customers. In fact, if IBM finds the right foster parent, it might even be an improvement. Thats because the PC business has become anathema to IBMs core business—generating consulting revenues.

IBM may still have a channel for its PCs, but it doesnt have an effective way to turn that into the kind of relationship that lets it tell customers to send it checks regularly. But the company has been losing traction in the PC market almost since the day it created it—Compaq crushed IBM in the channel for years, and then Dell crushed the channel.

The irony of the end of IBMs run is that IBM itself created the very forces that eventually drove it out of the PC business at the very beginning. In order to succeed in bringing a product to market, the IBM PC team saw that it would have to open up the specifications to other manufacturers.

/zimages/5/28571.gifClick here for analysts take on whether Lenovo Group, Chinas biggest PC maker, is the right choice to take over IBMs PC business.

The catch, of course, was that doing that created an opportunity for other manufacturers to "clone" the PC. And by the time IBM tried to put the genie back in the bottle with the introduction of the PS/2 and its MicroChannel architecture, it was too late. Compaq and others had already taken the open architecture off in another direction, and IBM would never find its way back to the head of the pack.

Sure, IBM designs great notebook PCs. The Thinkpad line has been, by all accounts, tremendously successful for IBM; based on IDC numbers, IBMs notebooks were flying out the door at an all-time-high rate halfway through this year.

But IBMs profits from its PC business were less than 1 percent of sales—not the kind of margins that IBM CEO Sam Palmisamo likes.

And with PC sales growth slowing, those margins arent going to get any better—at least not with IBMs management. The company has already squeezed all of the cost out of manufacturing that it can by outsourcing to contract manufacturers; all thats left to cut is the internal design staff.

On the other hand, the intellectual property that IBM has amassed associated with PCs does make the business somewhat attractive for acquisition. Innovations such as the butterfly keyboard and eraser-like "isopoint" pointing device should be worth something to a buyer, even if IBM couldnt make any money off of them.

Price wars pushed IBM to outsource most of its PC and notebook manufacturing over the past decade to contract manufacturers such as Lenovo. The fact is that desktop PCs have become such a commodity that the only thing left of any value in the PC market is the support behind them and the reputation of the manufacturer.

While IBM has had a reputation for quality, it hasnt been able to capitalize on its brand the same way Dell has, for example. And IBM has no real direct relationship with its desktop PC customers like Dell does. But IBM can still make money off servers, because it has relationships with the people who buy them—and services it can still deliver on them.

The same market forces existed more than a decade ago in the printer business, and they pushed IBM into spinning off Lexmark, its printer business in 1991. But I doubt IBM will play the spinoff game again and release its PC division into the wild. With Merrill Lynch now apparently lined up to arrange a sale, it looks like IBMs management is looking for a nice, tidy transaction with which to pad its Christmas stock-option sales.

/zimages/5/28571.gifCheck out eWEEK.coms for the latest news in desktop and notebook computing.