IBM Brags About Power Systems Takeouts in Q4

January 24, 2011 Timothy Prickett Morgan

In discussing IBM‘s fourth quarter and full year financial results last week with Wall Street, Mark Loughridge took a few moments, as he has done for several years now, to brag about the competitive wins that the Power Systems lineup is getting against rivals Oracle and Hewlett-Packard. While takeouts had been losing steam earlier in 2010, they heated up as the year ended and IBM got the entire Power7 lineup shipping.

In the fourth quarter of 2010, Loughridge said that IBM had over 280 competitive displacements, which accounted for around $325 million in revenues for Big Blue. (Those revenues include sales of Power Systems servers, plus storage, networking, and any other needed hardware plus software and services fees that can be booked during the quarter.) This, said Loughridge, was the best quarter in the multi-year history of the Migration Factory effort for competitive takeouts for the Power family of machines. For the full year, IBM did over 1,000 competitive replacements and these generated nearly $1 billion in revenues during 2010; there will be incremental hardware sales in future years as customers upgrade their boxes, plus revenue recognition for multi-year services and software contracts that customers purchased with their machines on top of this. Of these Q4 takeouts, about 60 percent came at the expense of Oracle and its acquired Sparc/Solaris business, with another 30 percent coming from HP accounts (probably a mix of HP-UX, OpenVMS, and NonStop Itanium-boxes with a smattering of Windows and Linux on them).

This time last year, Loughridge was telling Wall Street that in the fourth quarter of 2009, IBM had nearly 200 competitive wins accounting for nearly $200 million in revenues, and that for all of 2009, IBM did more than 500 deals comprising more than $600 million in revenues.

What IBM doesn’t talk about, and very likely never will, is how many new customers it brought into the IBM i fold during the quarter. At a run rate of around 2,500 new customers per year, it should have been on the order of 600 to 650, and if you assume the 30 percent bump in entry Power Systems sales in Q4 as reported in the lead story in this newsletter means that the IBM i platform did comparatively better than the average, then maybe it was as many as 1,000 new customers. (That is a wild guess on my part, but IBM i is popular in niche industries and in fast-growing and emerging economies because of the applications that run on it.) If the numbers for the IBM i platform do indeed have a shape like that, it sure would have been nice for Loughridge to have mentioned it.

Of course, bringing up IBM i will only make people pause and think about how many deals IBM is losing out to HP, Dell, and others as companies move their software to Microsoft Windows platforms running .NET or Java. As I explained back in October, for the OS/400-i installed base to decline from around 220,000 in 2005 to the 100,000+ number IBM was talking about last fall, you’d have to bleed off 25,000 customers for every 2,500 you added to the fold. That is not, obviously, a good ratio.

Still, someone at IBM should be held accountable for whatever decisions led to the defections of so many loyal IBM customers. How many billions did IBM lose by not retaining some 175,000 customers who have left the OS/400 fold since the turn of the century? For every relatively easy competitive win against Sun/Oracle, HP, and Fujitsu in the Unix and proprietary systems racket or every new customer brought to the IBM fold, Big Blue seems to have lost seven OS/400-i shops. That is not good business, especially not when it takes 10 times as much energy to get a new customer as it does to keep an existing one. In fact, that is a money-losing proposition, if you look at it hard enough.

This is what happens when a great systems company focuses only on the short term. I could be all philosophical about it if it were not for the fact that my livelihood has been adversely affected by the bad decisions that IBM has made. As has yours. But, we persevere, and there is plenty of good work still to be done. So let’s get on with it.