IBM Corp may have the most successful services business in the industry but the company needs to maintain and grow its software division if it’s to make continued profit, according to Gartner Group analysts. While the Armonk, New York based-company sees annual revenues of around $26bn a year from its services division, Thomas Bittman, research director with Gartner said Big Blue’s software sales generate the most profit margin. The margins in services are only 20%, which isn’t particularly high. But those derived from software sales are around 70%, he said, So IBM can be a successful services vendor, but if it wants to make a continued profit, the company needs to grow and maintain the software business. By 2002, over half of IBM’s profit will be driven by services and software sales, while profit in all other areas will begin to decline, Bittman said. If you strip out the services, IBM’s compound growth rate is only 3%, which is not exactly the sort of company I’d want to invest in, the analyst said. Since 1991, when it was introduced, the services division has grown at a rate of 18%, and even now, it’s increasing at a rate of around 17% to 19% per year. Everything else IBM does only grows at between 0% and 1%, Bittman said, Software is basically flat and all the rest actually goes down somewhat. Bittman admitted that the Year 2000 problem is having a knock-on effect but while all aspects of IBM’s business will take a spiral dive in the second half of 1999 as a result, he predicted services spending will continue. He said the the massive turnaround of the software division was a result of IBM’s aggressive acquisition strategy, noting particularly the purchases, in 1990, of Tivoli and Lotus. As revenues from IBM’s traditional breadwinners, mainframes and AS/400s, gradually start to decline through the next century, Bittman said IBM was poised for growth in the software market by around 2001/2002. Bittman’s comments came in a post-meltdown look at what IBM has achieved since it was hit by sales slumps as a result of bad strategic decision making in 1993. Today, having got back on its feet, IBM faces the next big challenge: competition from Redmond. Here, Big Blue’s Java strategy is crucial, Bittman said. I can’t understate this one. IBM will run dry, in terms of deep pockets, before Java fails, he said, very simply, if it loses the application interfaces to Microsoft, then IBM loses. So it is absolutely critical for IBM to make Java succeed and that means helping Sun and Netscape and other partners. IBM should prop them up financially or, he said, if necessary, buy them out. He criticized the company’s Java development environment, San Francisco, for being too late to market, too slow to develop and too complex to be commercially successful. But he added that Sun has been able to learn from IBM’s failures, which has ultimately led to improvements in Java’s performance.