Commentary: Big Blue Should Get A Little Smaller

Big Blue is doing a bit of reshaping, and it's about time. Word leaked out on Sept. 1 that IBM has put its Global Network operation up for sale, hoping to bring in $3 billion to $4 billion from the unit that handles data communications traffic for customers in 900 cities around the world. It's the second business IBM execs have put on the block since April. That month, the company started shopping around a division that makes printers, looking to fetch $2.5 billion.

These are smart moves, but IBM should think about more. The company has too many businesses that are laggards or that simply no longer make sense for IBM to own. Should IBM be selling educational software to consumers? The answer is no, because it lacks a strong consumer brand--either in hardware or in software. The same can be said of other parts of the business where the IBM brand has faded, including networking gear, desktop computers, and Unix software--to name a few operations that should either go or be dramatically scaled back. IBM declined to discuss its plans.

OVERCAUTIOUS. The fact is, these are not realms that will make IBM a success in the Internet Age. The giant could easily shed 10% of its heft, raising some $8 billion, if it sold its noncore businesses. It should take the money--and the resources that could be freed up--and pour them into core areas, such as software and services to facilitate electronic commerce and other Web-based systems for corporate clients. A few years ago, CEO Louis V. Gerstner Jr. appeared to be ready to do just that, making bold moves to acquire software companies Lotus Development Corp. and Tivoli Systems. But lately, the pace of major acquisitions has slowed. "They have become highly risk-averse," says Howard Anderson, president of market researcher Yankee Group. Gerstner "is running the thing like a damn mutual fund."

IBM doesn't need a radical divestiture plan. But it does need an aggressive pruning. Despite doing a stunning job curbing costs and boosting earnings to double-digit levels, Gerstner hasn't done much to pump up the top line. In 1997, the computer giant's revenue hit $78.5 billion, a piddling 3% bump over 1996. Sales growth in the first six months of 1998: zippo.

That's why it's time to rethink some businesses. Take personal computers. SoundView Financial Group analyst Gary Helmig figures that IBM's latest misstep in that business--building too many PCs during the holiday season--has cost the company $1.3 billion in revenue so far this year because resellers have been discounting machines to work off excess inventory. And despite ranking within the top five suppliers of PCs worldwide, IBM's profitability has been spotty. The company lost $300 million in PCs last year.

IBM doesn't have to get completely out of PCs. It can let someone else manufacture them for IBM to resell as it already does with a few models. The savings in development and manufacturing costs alone could boost earnings. Then the company can focus on staying ahead in laptops, where it is a world-class competitor because the innards of those machines still require a lot of proprietary design and technology--something IBM excels at. The company can also invest more resources in its PC server business, an area where IBM badly needs to catch up to Compaq Computer Corp. by bringing mainframe-class reliability and services to its products.

What else should go? IBM's Unix software and workstation businesses. The company's Unix workstation business has been flat for years. "IBM should give it up," says Forrester Research analyst Inc. Stuart Woodring. He suggests that IBM resell Sun Microsystems Inc.'s Solaris version of the Unix operating system because it is much more popular with software developers, especially Web designers.

Another area from which IBM needs to stage an organized retreat: networking equipment. This $2 billion operation makes esoteric communications gear, mostly for IBM machines. Long ago, the computer giant was eclipsed in this sector by networking upstarts such as Cisco Systems Inc. and 3Com Corp. "That business is really atrophying," says SoundView's Helmig. Big Blue can't abandon its customers who rely on the equipment, but it could sell the business to an organization that would promise to support those customers. Meanwhile, IBM should concentrate on building networking hardware and software that's more geared to the World Wide Web.

SELLING EXPERTISE. For years, IBM executives have claimed that the company's broad portfolio of businesses--everything from tiny computer memory chips to massive mainframe computers--is a strength. When one part of the business is sluggish, the argument goes, other areas can make up the difference.

But treating computing like a balanced stock portfolio is the wrong approach. IBM's greatest strength is its ability to provide the high-tech services to handle complex computer jobs. That's why the company routinely wins the mega outsourcing deals. On Sept. 2, Big Blue was awarded a $3 billion, 10-year contract to maintain Cable & Wireless Communications' computer network operations.

Gerstner has a chance to reinvent the world's largest computer company. So far, he hasn't tackled the job. IBM has some outstanding businesses--particularly in computer services, a $19 billion operation that is growing by 22% a year. Overall, however, IBM might have a better chance to shine if it unloaded more of its lackluster businesses.

Before it's here, it's on the Bloomberg Terminal.