IBM, HP and BEA results point to software infrastructure resurgence

A world where all software is open source and free, right up and across the stack, including the data and the business apps ... not in any of our lifetimes.

Software infrastructure is again proving to be the "Little Train that Could" when it comes to chugging along -- even when the chips are down. My prediction is that the vendors that can sustain software infrastructure profit growth will be the biggest winners -- of any IT vendors -- over the next five years. And this is not just because I'm a software bigot.

To get to this conclusion, you have to admit that there are some really weird dynamics afoot in the enterprise server and infrastructure businesses these days. On one hand, more software and tools are open source and "free," meaning they can provide a loss-leader role for hardware and professional services.

On the other hand, growth in infrastructure software license and support profits at IBM, HP, and BEA from the most recent financial results are showing renewed strength during mid-2005, even as hardware cost-efficiency leader Dell is seeing profit growth ratchet back. This indicates that some software -- particularly proprietary software -- is where the margin is, as hardware continues to level-set into commodity status. I think these are trends, not exceptions.

Yet, Alice, infrastructure software has to be both be free and where the profit margin growth is. Hmmm ... this is the conundrum that IT vendors have been, and will continue to be, grappling with. How Tweedle-Dee and Tweedle-Dum manage in this topsy-turvy landscape, and how well all the major players maneuver within it, will determine which IT vendors are dominant and which are submissive throughout the current and pending vendor consolidation.

Some early hardship cases from the conundrum: Borland, Sun, BEA, Novell, and BMC. Some beneficiaries: SAP, Oracle, Apple, Red Hat, EMC, and IBM. On the fence, but with a sense of renewed optimism under new CEO Mark Hurd, is HP. CA, with new CEO John Swainson, is also in this category. So both CA and HP must make the enterprise software conundrum work to their favor, which means they need the right mix of proprietary software. Management software would be a good guess, uh-huh, but is it enough?

Sun has, like the Mad Hatter, been feverishly working this conundrum assignment, and is showing some signs of growth. Sun's answer at present, as a spoiler of sorts, seems to be to confuse the distinction between free software and high-margin software. Microsoft, the oblivious Caterpillar, remains in its own hazy world. The relationship between Sun and Microsoft bears close watching, though you can bet that Microsoft has no confusion over the distinction between between free software and high-margin software.

As for a world where all software is open source and free, right up and across the stack, including the data and the business apps ... not in any of our lifetimes. But this is a discussion for another day, and there will certainly be a large and growing role for open source, so hold your hate mail.

An added complexity in peering through the IT vendor looking glass is that those few vendors that possess both sizable hardware and software businesses should do better, conventional wisdom holds, as they can play one off of the other. But this is a perplexing business, and this hardware/software synergy has not always panned out (say for Sun), nor has a lack of it handicapped (as with Oracle and SAP).

No, I think what is more important now -- a point that Sun missed for a long time, and Microsoft still misses -- is that playing "free," open source software properly off of high-margin proprietary software is the clincher, the ultimate synergy. Add in just the right amount of professional services, and viola -- a long-term, viable IT vendor. Oracle has done this so well, you just have to admire them. IBM, too. SAP is coming around (fast).

But this leads to another dark hole in the ground, Alice: The role of professional services, whether it is better to work the synergy between product and service internally or via partnerships and channels? Obviously there is a wide range of current approaches on this one. And the jury remains out, but a hint of the future might be in the following new axiom: When it comes to SOA, the 80-20 rule gives way to the 60-40 rule, wherein any meaningful deployment is 40% out of the box, and 60% customized. Services may hold a trump card.

So winning long-term vendor growth boils down to which software a vendor owns and newly develops -- and then how well it plays off of the open source affect (that code which no one owns and is newly developed). Taking this relationship to town on the best professional services platform seals the deal.

Therefore, in conclusion, the biggest winners in proprietary software should ultimately be the biggest IT vendor winner overall, at least in terms of profit margins and Wall Street affections. Because, if all things are equal when it comes to the hardware -- and for mainstream server computing we're just about there, and virtualization will hasten the trend -- a high-margin software business is the one essential ingredient for long-term IT vendor success.