Two Pillars Of IBM's Growth Look Shaky

A better second quarter hides challenges in mainframes and services

When IBM announced earnings on July 18, investors breathed a sigh of relief. With revenues up a relatively healthy 6%, to $21.7 billion, and the first increase in its services backlog in six quarters, it looked as if IBM (IBM ) had recovered nicely from a disappointing first quarter.

Well, maybe not. The strongest gains came from software and server computers. Services landed where they were in the first quarter: a 6% revenue gain -- tepid for Big Blue. And mainframes plunged 24%, after a 16% drop in the first quarter. That's bad news because mainframes and tech services have been two of the key pillars of IBM's success. The former has helped drive profits; the latter has helped generate sales. But now, as IBM navigates a huge transition in how it does business, both pillars are looking a bit wobbly. "I don't see a resurgence in mainframes, and they have long-term challenges in services," says A.M. Sacconaghi Jr. of Bernstein Research. (AC )

That helps explain why, the day after earnings came out, Big Blue announced a shakeup of its $45 billion global services business. Division head John R. Joyce is leaving for private-equity firm Silver Lake Partners, and the business is being split three ways. In a staff memo, CEO Samuel J. Palmisano said the moves are designed to "better focus our services growth strategy and enhance our marketplace performance." And on July 26, IBM will announce its new mainframe, called Danu. With it, IBM hopes to make the mainframe once again the hub of big corporations' computing systems.

Both moves aim to help Palmisano, who was not available for comment, reach his overarching goal: to transform IBM from a traditional computing company into a business-services provider. He's betting that by harnessing technology to improve clients' performance rather than simply hawking machines or providing routine tech services such as managing PCs, IBM won't be hobbled by the ongoing commodization of tech and tech services. The earnings report shows those plans are on track. Business-transformation services revenues rose 25%, to $900 million, and bookings rose 192%. And consulting revenues rose a healthy 9%, to $3.7 billion. But Palmisano can't relax: Mainframes must keep delivering big profits, and the much larger traditional-services portfolio must click if IBM is to shift successfully.

Services are the knottiest problem. Palmisano faces a slowdown in traditional tech services, rising competitive pressures, and changes in customer behavior. IBM, the market leader, has long depended on megadeals, yet the value of such contracts for the industry overall dropped 43% during the second quarter, to $25 billion, says market researcher Datamonitor PLC. Why? Tech buyers are increasingly splitting contracts among a handful of providers rather than signing with just one.

The reorganization should help. Palmisano is grouping related services in an effort to goose sales. Mike Daniels, who headed overall sales in the Americas, will run the traditional services businesses. Ginni Rometty, who had run business-consulting services, will oversee a new unit that includes consulting and business-process outsourcing and an army of software developers. She also heads business development. That's key. By tracking what IBM's consultants and programmers are doing, she should be able to move quickly on new opportunities, via acquisitions or building new services.


Palmisano's other move could have a more immediate impact by cutting costs and letting IBM price more aggressively. Bob Moffat, who oversees the supply chain, will reorganize and manage chunks of the 180,000-person services staff, many of whom do such commodity tasks as logistics or maintenance.

Since he took over the supply chain in 2002, Moffat has produced $5 billion to $6 billion a year in cost savings. Now he'll try to do the same for service staffing. One possible target: consolidation of IBM's 177 global call centers.

On the mainframe front, the question is whether IBM's new Danu can duplicate the stunning success of its last upgrade, T-Rex. Launched two years ago, it improved sales from $3 billion in 2002 to $3.8 billion in 2004. Thanks in part to the success of T-Rex, mainframes drive 45% of operating profits after related revenues from services, software, and financing are added, says Bernstein's Sacconaghi. Yet analysts predict that Danu won't pack the punch of T-Rex, which combined a major step up in technology with a 25% price cut.

What's more, many customers who rushed out to buy T-Rex won't be in the market again for years. "You buy a mainframe like you buy a car: You want to drive it for a bunch of years," says Peter Dougherty, CIO of Touro Infirmary in New Orleans, who bought a T-Rex last year.

IBM hopes to keep the momentum going by positioning Danu as the linchpin of corporate computing. The hardware is valued for its reliability, security, and stability. Now IBM is adding software to extend these capabilities to its customers' entire computing environments -- no matter what kind of machines and software they use. Erich Clementi, general manager for mainframes, expects concerns about security, in particular, to drive demand. "This system will be around in 50 years -- no matter what anybody says," he predicts.

But will it be as key? That's questionable. Servers from rivals offer big savings and ever-better capabilities. Guenter Bodner, CIO of five of publishing giant Bertelsmann's book clubs, says his shift from mainframes to PC servers saves almost $1 million a year. "I grew up with IBM, but now, for me, there's no future for mainframes," says Bodner.

For IBM the key, in computers or services, is to keep its old franchises vibrant even while it's building up their successors. Right now that job is mighty tough.

By Steve Hamm in New York

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