For years, IBM’s massive technology services arm has hogged the limelight, the symbol of its shift away from its old reliance on the “big iron” of mainframes and other hardware.
It is software, though, that has quietly taken up the running. With little fanfare, Big Blue’s software arm, second only to Microsoft in size, has emerged to become the driving force behind the latest overhaul at IBM, capped in the final months of last year with growth of more than 10 per cent.
Given the upheaval underway in the software industry itself, that makes IBM’s future increasingly dependent on technology shifts and new competitive rivalries, including one with Microsoft, which are very different drivers from those that traditionally shaped its business.
Two things have propelled software to the forefront of the IBM story in recent months. At the start of this decade, the software division was growing at a negligible rate, held back by the slow decline of the operating system business. Little more than a year ago, though, revenue from faster-growing “middleware” – software that acts a layer between operating systems and applications, making it possible to run the highly complex IT systems of modern companies – finally grew to account for more than half of the total.
The shift “has given us leverage on the rate of growth” and will mean consistently higher growth rates in future, says Steve Mills, the executive in charge of the division.
The second, related event has been a notable increase in the pace of acquisitions. While IBM has a long track record of purchases (including the Lotus productivity software business and Tivoli, a systems management company, in the mid-1990s) it has become more acquisitive of late, buying 22 software companies over the past two years. Mr Mills plays down suggestions that this represents a new departure for the company, but it points to the growing range of opportunities that IBM sees as it extends its middleware reach.
Behind this growth lies a strategy that sounds starkly different from rivals Microsoft, Oracle and SAP. While those companies have been racing to create vertically integrated “stacks” of corporate software, extending all the way up to applications used by individual workers, IBM has concentrated on creating a “horizontal” layer of middleware that lies in the guts of IT systems, where most workers will never encounter it.
That strategy relies on a single belief: that legacy corporate IT systems are “measured in the trillions of dollars”, says Mr Mills, requiring extensive work as companies try to integrate them better, build on to them and adapt them to new business purposes. IBM has built five middleware brands – Lotus, Tivoli, Websphere, Rational and the DB2 database business – which as standalone businesses, would each rank among the world’s 25 biggest software companies, says Ian Finley, research director at AMR, a technology research firm.
Mr Finley says that IBM’s middleware strategy has positioned it well for one of the biggest shifts currently underway in software: the rise of so-called “service-
oriented architectures”, or broader and more flexible software platforms on which companies can build more adaptable technology capable of changing with their business needs.
This trend has vindicated IBM’s decision, made a decade ago, to move away from the applications business, instead partnering with other software companies, while it builds broader platforms. Some other broader software trends, however, may pose a bigger long-term challenge.
One is the rise of open source software, a movement that IBM has itself championed by supporting the use of the Linux operating system.
“They’re enthusiastic today, because it’s hurting Microsoft,” says Mr Finley. “But they may be less so in future,” as low-margin open source software moves into other parts of middleware.
“The [open source] model is certainly viable,” argues Mr Mills. IBM has itself started to offer open source versions of some of its middleware, starting with application server software, for the low end of the technology market. But Mr Mills admits that to avoid seeing the more profitable parts of its business commoditised, “we have to keep moving up” to higher-value areas of software.
A second potential challenge comes from the emergence of “software as a service” – the business of providing applications online as a service to companies, something pioneered by companies such as Salesforce.com and increasingly attracting the interest of Google.
Customers that turn to these services will no longer need to buy the hardware systems and IT integration sold by other parts of IBM, says Mr Finley, putting pressure on IBM itself to step up to become a full service provider. To do that, though, would challenge the vow of denial that has kept it out of the applications business, a strategy that has underpinned its highly profitable partnership with other software producers. “That may be one area where they hesitate, and hesitate too long,” the analyst warns.