System iNetwork

September 20, 2007

Competition and the Trouble with "Legacy" Revenue

In my recent BladeCenter post, Steve Richter and Trevor Perry brought up some issues that sparked this post. IBM's System i management team sometimes takes heat for its perceived lack of sales success over the IBM System p's perceived rise to glory.

I can't say that I understand the inner workings of IBM, but the System p has benefited by having relatively direct competition in the market. Because other Unix options were and are available, IBM management had little choice but to attempt to compete on price and performance, and competition usually tends to drive prices down. So while the hardware, development, and so on at IBM appear similar on the surface, the market dynamics are very different. IBM System p management's ability to lower prices and increase performance has a direct tie to their success in the market. (Sure, there are other reasons to go with IBM over HP or Sun, but that's just one factor.)

With the IBM System i, the relative value of a lower price and increased performance isn't so clear. With a relatively static and astoundingly loyal customer base, lowering prices doesn't necessarily result in any sort of increase in IBM's profit. It might result in a handful of additional sales, but it requires an increasing number of customers. The same goes for performance. Just because IBM could deliver increased System i performance, it doesn't mean that most SMB customers have workloads that actually need it.

There's definitely a chicken-egg issue going on here. On one hand, if IBM had lower prices and more powerful System i offerings, it could more easily compete against other operating systems, which would result in increased sales. On the other hand, what if it didn't result in increased sales? IBM would have lowered the prices in a division that's basically profitable (just because sales are down doesn't necessarily mean there's no profit) . . . and may have put it in a position where it could never recover.

On the other hand, if IBM keeps prices relatively high, guaranteeing a profit even in the face of declining sales, the company is doing its job for its shareholders.

Of course, IBM is a public company, and public companies sometimes have a tendency to make decisions based on fear rather than vision and risk. So although our System i management team may not have any fear, IBM's upper management team might sleep with the lights on. And although Mark Shearer might blissfully dream about ruling the world with cheap and powerful System i boxes, Sam Palmisano may have other dreams and nightmares. In fact, if it came down to numbers, and IBM's number crunchers predicted that a world ruled by System i boxes would earn a 50 percent return on investment over 10 years while some other plan would earn a 51 percent return on investment over 10 years, that tiny difference might result in major decisions. It's hard to say, and we're not even talking about any kind of perceived risk associated with truly innovative solutions that are hard to explain to the IT buying public.

So, Where Does This Put Us Now?

Enter IBM's System i and System p reorganization. There are a bunch of ways that large System i customers and small System i customers are different. Large customers who are paying six or seven figures for their System i machines are paying for performance in a variety of ways -- not just processor or memory-based performance. There are disk arms and I/O issues and so on which make it easier for IBM to ratchet up the cost of the System i. Plus, these bigger customers have more opportunity to gain from all the benefits of i5/OS so that they're more likely to see the benefits of IBM's superior ROI. The System i's ROI comes from a variety of factors -- it's never just based on price/performance. But SMB customers have a tendency to think in terms of cost of acquisition, and who can blame them? If my budget for a new car is $14,000, and all I have in the bank is $14,000, I'm sure as heck not going to buy the $20,000 car instead, even though I might be well aware of the fact that it might be more secure, more reliable, and more flexible. My budget, right this minute, forces me to make a $14,000 decision, in which case the superior value proposition of the System i loses out most every time.

IBM's reorganization brings up lots of questions, but by separating out the big money -- the big customer revenue -- IBM can argue that it's providing a different product with different levels of service and so on to big customers, which will help them retain that profit and arguably provide better overall account management to its biggest customers. We're talking about massive amounts of dollars from clients who buy a lot of other IBM solutions. And these clients aren't likely to move from their System i investments any time in the future. They are stable customers, and IBM must protect that "legacy" revenue (even though these customers are often the ones using the most cutting edge features, the account is old).

By separating out the small customers, IBM can go to market with a revised product line, which may in turn result in more competitive cost of acquisition prices -- or pricing models better tweaked to reflect the realities of the small and medium business market. We've already seen some of this with the 515 and 525, with user-based pricing and so on. IBM had a few missteps there, sure, but they corrected the model in response to customer and BP feedback rather quickly.

So what I see coming here is a wide open field of i5/OS possibilities. I think IBM will find new form factors and new SMB bundles of solutions that it will attempt to bring to market. Sure, IBM can try to compete 100 percent on a price/performance model, but man alive, that's tough. Look at AMD and Intel . . . they've been driving each other's prices down for years, and I've seen arguments made that that's actually bad for the IT market at large because it also drives unrealistic expectations for pricing across the entire IT budget. Maybe, maybe not. The flip side is, AMD came out and busted Intel in the chin with its innovative Opteron processors. It took Intel three years to catch up to the pesky upstart AMD, and the net result is vastly improved x86 processors.

The Next Two years

I think the next two years will determine the fate of the System i and i5/OS. In a worse-case scenario, IBM will slowly allow i5/OS to fade but work hard to retain its "IBM" customers and ensure that every application a customer has offers some sort of reasonable path forward. IBM does a great job of this. Look at the history of the System i and compare it to the history of the Microsoft Office products. Compatibility with previous generations of technology has been one of IBM's System i strengths. So, in a worst-case scenario, i5/OS faces a slow death.

In a best-case scenario, IBM finds new and innovative solutions for SMB customers. Business Partners will step up to help sell these new solutions, and they'll help drive innovative bundles that get to the heart of the strength of i5/OS and the System i. IBM's Vertical Industry Program is a case in point, and it comes from a pretty darn good business model and go-to-market strategy.

IBM has issues, no doubt, and although I can't see the future, I'm confident that IBM's System i and i5/OS moves will give our industry more choices and options than ever before.

Posted by cmaxcer at September 20, 2007 8:39 AM