Does the Size of Your IT Supplier Really Matter?

Numbers matter to people. There are symbolic floors and ceilings, expressed as numbers, that delineate what most of us consider the proper bounds to reality or significant milestones. Think of Dow 10,000, or the turn of the millennium, or a 25th wedding anniversary, or breaking the 10-second barrier for the 100-meter dash, or pushing through the sound barrier. Moreover, disasters magnify with increasing numbers by more than you might think possible–if you weren’t human, that is. We rank the events and the things in our lives by numbers, and in many cases, the changing numbers change our behavior. Even if it is not always logical.

People are funny that way. And last week, as Hewlett-Packard was reporting its financial results for its fiscal third quarter ended in July, just about every story in the financial and the trade press made much about the fact that HP is poised to surpass IBM as the largest supplier of information technology in the world. Because I watch the IT players for a living, I didn’t find this surprising, particularly as IBM has been divesting itself of several businesses over the years–notably hard disk drives and PCs, which made up a considerable part of its annual sales, but which IBM never quite figured out how to do profitably in the past decade.

I didn’t think much more about such comments, until I started getting questions from people about whether this was an important barrier, that if somehow by being bigger than IBM, HP would accelerate the gap between itself and Big Blue and somehow make life very difficult for IBM. Once someone broke the 10-second barrier on the 100-meter dash, it was proved to be possible and a lot of other runners could then do better. The Dow at 10,000 seemed like a ridiculous idea, and if the Dow dropped 10 percent to hit 10,000 today, the financial world would be in a panic. What seemed like an impossibly high ceiling now seems like a floor. And, what might seem like no big deal could, possibly, be one of those significant numerical milestones, a psychological barrier that, once breached, opens up new possibilities.

On Friday morning last week, an iSeries and System i5 reseller with one of the largest OS/400 accounts in the world called me to ask me if HP being the largest IT supplier in the world was important or not, and I said I didn’t think so. And then he said his customer said this to him: “I always buy best of breed, and I buy from IBM because it is the dominant supplier. Should I be looking at HP?”

See what I mean about the way we think about numbers not necessarily being logical?

HP, as you might imagine, is pretty jazzed up about being the largest IT supplier in the world. And when it merged with Compaq in September 2001 for $25 billion, the company’s stated goal was to transform HP from a printer company with substantial expertise in midrange and high-end servers and operating systems into a company that could compete head-to-head, toe-to-toe, dollar-for-dollar with Big Blue. It took almost a year to get the merger done, and it has taken four years to reach that goal of besting Big Blue. HP’s own growth–after a pretty rocky merger and the ousting of its chairman and chief executive officer, Carly Fiorina (who engineered the deal and is rarely given the proper credit for seeing that this was, indeed, a smart merger)–is part of the reason why HP is going to be larger than IBM, in terms of annual worldwide revenue. IBM’s divestitures is another cause. If IBM hadn’t sold off its PC and disk businesses, I would not be writing this story right now. Then again, IBM would probably be losing money, and certainly be a lot less profitable that it is. If Dell is getting hurt by a revamped HP in the PC business, IBM would be getting hammered.

Making exact comparisons between HP and IBM are not easy for a number of reasons. For one thing, HP’s fiscal year ends in October, while IBM’s ends in December. So the two companies are out of phase by two months. So if you compare HP’s prior three quarters, it has two months that actually are pulled from November and December 2005–which is, as we all know, peak IT spending time. For the trailing three quarters of fiscal 2006, HP had sales of $67.103 billion and brought $4.501 billion to the bottom line, about 7 percent of revenue. For IBM, the trailing three quarters runs from the first day in October 2005 through the last day in June 2006, and IBM booked $66.975 billion in worldwide sales and brought $6.915 billion to the bottom line–about 10 percent of revenue. HP is bigger, and growing considerably faster, and the question now is will HP be able to improve sales enough to get the kind of profits IBM is enjoying on a slightly smaller revenue stream.

Of course, revenue and profits are only one measure of the economic might of an IT supplier. In the case of public companies, market capitalization might be a better indicator of dominant position. And here, too, the gap is not that large. After a run-up in HP’s stock following its good showing in the fiscal third quarter last week, HP’s market cap rose to $96 billion. By the way, that’s about half of HP’s market value in early 2000, it’s peak in the past decade, but that level is still double since Mark Hurd, who used to run NCR (where IBM founder Tom Watson cut his teeth in the late 1800s), came on board five quarters ago. IBM’s market capitalization is slowly trending downward, and as of Friday, the total value of IBM’s stock was $120.3 billion–considerably larger than HP’s. IBM’s stock peaked in early 1999, giving the company a value of over $200 billion.

IBM’s stock is worth more because the company generates more profits. If HP acquires a bunch more companies with the $11.5 billion in cash it has left over after it acquires Merucry Interactive for $4.5 billion within the next few months, it could add considerably to its profits, and thereby its stock could rise significantly. IBM had $7.1 billion in cash and just under $3 billion in marketable securities, but has spent $1.6 billion acquiring FileNet, $740 million to buy MRO Software, and an undisclosed amount to buy Webify. That leaves the IBM cash hoard in the range of $7.5 billion. Which means that, barring share buybacks that cannot be funded through cash generated in each quarter, HP has a 50 percent bigger war chest than IBM right now. This gives HP maneuvering room to build up its enterprise business.

Which brings me to my next point, and this is the important one that none of the trade press took the time to think about last week. IBM is now, after mothballing its retail PC operations many years ago and selling off Prodigy (remember that?), almost exclusively focused on the enterprise IT market. HP has a very large consumer-oriented business. How much is not easy to say, but you can pull apart some numbers to get a good guess. And I think to make a fair comparison between IBM and HP, you need to rank their respective enterprise computing businesses. You might argue that IBM doesn’t have a consumer business to tide it over when the enterprise business softens, and that makes a certain amount of sense, too. But I still think you have to compare like for like.

The Imaging and Printing Group at HP, which makes printers, scanners, cameras, and other products, is the profit engine. In the last quarter, this group had sales of $6.2 billion, with $1.6 billion in commercial hardware and $870 million in consumer hardware. Consumables–paper and ink–make up $3.7 billion in sales, and if the consumer-business ratio is the same, then 35 percent of the money spent on supplies–a whopping $1.3 billion–does not come from businesses. So within IPG, only a little over $4 billion comes from companies, not people spending their own money for home products. (IBM’s commercial printing unit is tiny by comparison. It barely registers, in fact, in IBM’s numbers.) HP’s Personal Systems Group raked in $6.9 billion in the most recent quarter. While the company does not provide a breakdown of its PC business showing what percentage of sales come from consumer and commercial sales, respectively, it is probably a good guess that the retail channel and direct consumer sales probably account for 65 percent of HP’s PC sales. So take out $4.5 billion to do an enterprise comparison.

When you do the math, that means HP’s enterprise business comprised maybe $15.2 billion of its total $21.9 billion in sales for the third fiscal quarter from companies, not consumers. IBM is, as far as departments and data centers go, still the king of the IT hill.

If you look literally into the data centers and departments of companies, IBM still dwarfs HP. Of that $15.2 billion in enterprise spending in the latest quarter, HP sold $4.1 billion in servers and storage, $3.9 billion in services, and $318 million in software. In IBM’s last quarter, ended in June, IBM had nearly the same $21.9 billion in sales that HP had in the quarter, but it brought in $5 billion in servers, storage, and chip revenues, $11.9 billion in services revenues, and $4.2 billion in software revenues. HP is the volume leader in servers, to be sure. It sold over 2 million servers in 2005, compared to IBM’s 1.2 million. But IBM sells some pretty expensive mainframe and midrange machines, while HP tends to sell two-socket and one-socket servers. HP’s high-end Integrity server business is dwarfed by even the much-diminished (in terms of revenue) System i5 business. The i5 line is averaging somewhere around $400 million to $500 million a quarter in sales. HP’s Integrity line, which runs HP-UX Unix, OpenVMS, Windows, Linux, and NonStop operating systems, brought in $312 million in the last quarter after posting 76 percent growth. HP has a lot of growing to do to catch the System i5, and IBM’s Unix server business is a lot bigger than the Integrity line. HP’s high-end server prowess is still based on selling older NonStop, PA-RISC, and Alpha systems to the vast installed base that HP and Compaq built up over the years. And over time, HP will convert that base–don’t misunderstand me. These customers like HP, for the most part, but they are in no hurry to move. (Sound familiar?)

So, yeah, HP might be bigger than IBM overall. But unless and until HP is bigger than IBM in the enterprise, I don’t think HP has broken any psychological barrier at IT shops. However, should that happen–well, that would be another story. Such a change could signal a massive shift away from the traditional host systems that have propped up IBM for decades and to other X64 architectures and new software technologies that are platform-agnostic, and in such a world, IBM will only get its share, not more than its fair share as legacy lock-in allows it to enjoy these days.