System iNetwork

September 2007

September 25, 2007 8:24 AM

Two Minutes with Kevin Fratzke on IBM's Reorg

I connected with Kevin Fratzke, vice president of the iSeries Solutions Practice for MSI Systems Integrators, and asked him about his thoughts on IBM's reorganization -- the new Business Systems unit and the Power Systems unit. Here's a quick Q&A:

CM: Does IBM's change make sense to you? Does the rationale behind it, as publicly stated, that IBM's very large System i customers are very different from IBM's smaller System i customers, appear to reflect the real world?

KF: Yes, this change makes sense to us. For several years, we have been wondering when IBM would make a change like this. The customers are different from several perspectives -- smaller customers tend to stick with ISV packaged solutions with less customization around them. Also, the technical skills and investments required to sell to these two customer bases, from an IBM Business Partner's perspective, are very different.

CM: About the possible ramifications of this change -- what areas do you think have the potential to become much better or worse?

KF: I think IBM's exposure is that fewer traditional business partners may market to the Business Systems space, leaving IBM to rely more on the ISVs packaging hardware with their software.

CM: Can you elaborate just a bit? Meaning, an ISV would create a bundle with IBM hardware and then sell the package -- both hardware and software?

KF: Yes, I think IBM will have to rely on more ISVs bundling hardware with their solution to be effective in this space -- either by signing more ISVs as resellers or having traditional partners reach out and partner with the ISVs via the Value Net Solutions Provider (VNSP) program, which allows traditional partners to sell through ISVs to create a total solution. The challenge here is that most ISVs are not primarily interested in hardware, so once it's installed, it's hard to keep people focused on keeping the hardware current. This is why the teaming with traditional partners (via VNSP) is a better option in my view.

CM: How does this change affect MSI's approach to the System i market?

KF: No change yet, but we are looking at several scenarios. We want to give the IBM change some time to mature before we align with it. I would expect we might make some changes at calendar year end or at our fiscal year end in June of 2008.

CM: Oh, one more thing: are you seeing any other particularly hot topics of interest these days?

KF: Right now it's WebQuery replacing Query/400 and trying to figure out where it will really play and what adoption rate we will see.

Posted by cmaxcer on September 25, 2007 at 8:24 AM | Comments (18)

September 24, 2007 8:11 AM

Higher Ed, IT Industry, and Students Looking Elsewhere

Nate Viall, a System i-focused recruiter who has been serving the midrange market for 25 years, sent me a note last week about an Out of Context quote in NEWS Daily which resulted in some interesting points about IT-focused education that I'd like to pass along. Here's the Out of Context quote:

"America isn't producing enough new qualified technology workers. And the main culprit is our failing higher education system." --Jim Rapoza, in his eWeek blog, "U.S. Higher Ed Fails for Future Technologists"

And Nate Viall:

Some follow-up conversations over the last five weeks with university educators here in the Midwest would seem to indicate that the demand for new IT graduates is strong and that the class of 2007 is placed. Equally important, enrollments for the fall for technology majors appears to be up slightly, reversing the 5- to 7-year decline.

However, for the typical MIS or CIS-type program which blends business and computer science classes, there is more to the story. As I have pointed out before, these programs are typically a part of the business school or business college of the university. Enrollments in the business college have been on a long, slow increase over the past 10 years. The percentage of females in the business college is increasing at an even faster rate. The MIS/CIS programs are the ones in decline. So, it is NOT a failed higher education system. The kids have simply picked different majors. This is also the big reason that the percentage of females in technology has dropped so sharply. They are in the business school, and at ever increasing numbers and by percentage, just not in the technology majors.

The better question is, why? We know most of that answer. The top students can get better starting offers, have a faster career track, and have more career security in other fields not as likely to be outsourced or go offshore.

It is industry that is the main culprit, not higher education. They are reaping what they sowed, beginning in 1998 and accelerated in 2000 with the H-1b visa increases.

If Congress can resist the industry pressure to increase the H-1b professional visa quota, students will return to the technology majors . . . but it will take about three years of solid demand and appropriate salary offers to get the supply back.

I asked Nate if much of the difficulties in attracting and keeping System i-focused curriculum and students could be attributed to these issues that have been affecting the greater IT industry at large:

The big reason for the decline in System i education is that companies slowed the hiring of entry level staff after the summer of 1998 (NOT 1999). After the fall of 1999, the hiring basically stopped and has not returned. The kids weren't dumb. They went to Wintel-type jobs that were available. Keep in mind that we were already into the dotcom bubble. By late 2001 and into 2002, there was very little of any type of IT hiring, except for H-1bs! In fiscal 2001, industry used up nearly all of its new, expanded quota. 2002 was a lower number. But, we had tens of thousands of unemployed IT workers in the aftermath of the 9-11 recession. The bigger question was, "Why was there any H-1b hiring at all?"

The visa was initially pitched as a way to cover for shortages. In 2002, there were no shortages. In the end, it has always been about cheap labor.

The "higher education" issue is just a smoke screen. The YouTube video* back this spring finally put a face to that point. The panel moderator can be seen and heard saying that that particular portion of the lawyer's immigration seminar is about how to not hire an American worker.

Let's hope that our Congressional members and their staff have not forgotten.

* InformationWeek published a few articles that cover the episode if you're interested in more detail: Story 1, 2, 3.

Posted by cmaxcer on September 24, 2007 at 8:11 AM | Comments (9)

September 20, 2007 8:39 AM

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 on September 20, 2007 at 8:39 AM | Comments (3)

September 17, 2007 10:22 AM

PHP and the Competition Catalyst

After 15 months of work, System i application development tool provider BCD has officially released its WebSmart PHP development environment. Although BCD is a long-time System i-focused ISV with thousands of customers and toolsets that have stood the test of time, the company's entry into the PHP space isn't important because its toolset is better than the competition's — which it may or may not be. No, BCD's entry, in and of itself, is what's most important here: BCD is a competitor, and Zend, the company that has been the leader in PHP on System i, could use a good competitor.

You don't necessarily need to spend a lot of money on development environments to utilize PHP applications, but they certainly help. You can build a house with a hammer, but who wants to mess around with a hammer if he or she can get the job done far faster, and arguably better, by using a nail gun? So let's just say that WebSmart PHP and Zend for i5/OS are both integrated development environments that are more like nail guns than hammers. Their features may be different, but their basic aim remains: to help System i shops build houses with PHP nail guns, which makes them competitors. Of course, WebSmart PHP runs over Zend Core, so we might want to tweak the metaphor a bit. WebSmart PHP is an IDE nail gun, and Zend Studio for i5/OS is an IDE nail gun, but both use Zend's PHP nails or a Zend-powered air compressor. Whatever. In any event, what I'm getting at is that even though they are competitors, their competition will ultimately result in a bigger market for each of them than either could have gotten on its own.

More Noise

First of all, there's only so much noise a single company can make in a market. There's only so much advertising, only so many campaigns, and only so many events a business can be involved with. Sure, a single company can run multiple ads, have a massive booth at COMMON, or drive 18-wheeler road shows through every major city, but inevitably, the noise will cost far more than the return. A few good competitors will start making their own noise, and when all added up, the resulting cacophony produces an increased awareness. In this case, the knowledge could be that maybe PHP on i5/OS is a viable option, in and of itself.

As this PHP activity percolates, customers will give it a shot. If they're successful, word of their success will trickle into the market through peers, COMMON sessions, case studies, and the like.

Those who are slower to adopt new IDEs and new technology will consider their options, and because they actually have options, spend more time learning about PHP so they'll end up making the best decision for their companies. Because both Zend and BCD want to capture and retain these potential customers, they'll promote PHP and provide lots of collateral materials to help educate the market. A case in point is BCD's free 28-page White Paper on WebSmart PHP. Sure, it describes what's great about WebSmart PHP, but it also discusses how PHP can be used with i5/OS.

Publications like System iNEWS, SystemiNetwork.com, and our own competitors will write more about PHP and PHP solutions than we ever would have in the past if there is only a single solution provider, and we'll do it both from a news perspective as well as from a how-to perspective.

Will PHP's use on i5/OS ever become widespread? There are a lot of factors, of course, but having two solid competitors in the market is, in my opinion, the biggest catalyst of all.

Posted by cmaxcer on September 17, 2007 at 10:22 AM | Comments (1)

September 10, 2007 8:52 AM

Possibly Flawed i5 Possibilities

I've got a premise that could use a little kicking around.

I'm assuming you're aware of some of the rumors surrounding new form factors for i5/OS — IBM's Elaine Lennox, System i marketing vice president, hinted at it during the OCEAN user group conference earlier this year, right before IBM made its reorganization announcement, and Mark Shearer has made some similar hints. The most likely candidate for the rumors appears to be a POWER-based blade that can run i5/OS. Oh, here's an upfront disclosure — I do not have any inside information, so keep that in mind.

My premise starts with this question: If IBM did indeed deliver i5/OS on a POWER-based blade, which could then slide into a BladeCenter, what might this be particularly good for? For large System i customers, who have massive System i machines that are powerful enough to rival mainframes, I'm not seeing how an i5/OS blade would be particularly helpful. If such a company is running a large, mission-critical application, ERP or otherwise, I'm not noticing a rush to buy i5 blades.

Sure, the POWER6 processors are wicked-fast, but there's more to overall performance than just a fast processor. Would such a blade, in a BladeCenter, scale? And if not at the high-end of IBM's Power Systems group customer set, what about at the small and medium end, the Business Systems group of customers?

I turned to Charles King, principal analyst of Pund-IT, and asked him to give my line of thinking a couple quick kicks:

"You've hit on an intriguing subject. Since POWER-based blades already exist, providing i5/OS seems like a logical move for IBM at some future point. The scalability and integrated features of the BladeCenter architecture could provide System i customers remarkable flexibility in the way they deploy, upgrade, and manage their systems. The sticking point, of course, is how System i users might react to such an offering, as it constitutes an abstraction away from traditional standalone systems. Another real concern is whether BladeCenter can provide the levels of performance and reliability that System i customers are used to. I believe that it can."

So, maybe not right this minute, but in the future, the BladeCenter may provide scalability solutions that theoretically could replace a large i5, as funny as that may be on the surface. It doesn't seem like the most cost-effective method, but it certainly could be in the future. There's already a movement to directly attach BladeCenters to i5s anyway, so this shouldn't be such a big leap of the imagination given the IT world's current desire to combine and compress.

Might a BladeCenter-based solution work for small and medium businesses? Maybe less far out in the future? It would let businesses manage multiple servers, multiple workloads, and let them buy and upgrade blades, if not lease them, rather easily.

The BladeCenter S

IBM has introduced a new BladeCenter built specifically for smaller businesses — the BladeCenter S. It's small enough to fit on a desktop, and it plugs into a regular 110v power outlet. Right now, it's geared more for basic workloads, but I'd bet IBM plans to expand its capabilities. Now think of it more as an end-to-end SMB solution. . . . IBM can then market (advertise) a BladeCenter that's flexible, scalable, and works with most any solution from any Independent Software Vendor. Might the BladeCenter S evolve into the "platform" of the future for IBM SMB clients? Is it cheap enough? Small enough? Easy enough to manage?

And here's the rub for our world — could existing smaller System i clients move to a BladeCenter S-type solution or would it be too much of a leap of complexity and cost? Could it be a viable replacement for a 515? A 525? Would it allow an IBM Business Partner to bypass operating system biases and make it easier to sell i5/OS-based applications?

King notes:

"How this might play out among small businesses is an interesting point. The real question is what part of the SMB market the new system is most appropriate for. I doubt it will appeal to very small companies that are used to buying servers on a one-at-a-time/as-needed basis, but it should find a place among larger small businesses and mid-sized organizations used to considering the strategic implications of their IT purchases. The BladeCenter design makes it among the most flexible and scalable IT server solutions available, and BladeCenter S extends all those same benefits to small businesses. All in all, I believe the platform will considerably strengthen IBM's position among SMBs."

Thinking ahead, if I were IBM, I'd be wondering if I couldn't "BladeCenter S" the whole SMB market. I wouldn't have to market a System i; I could just market the BladeCenter and sell i5/OS. BladeCenter is one of IBM's more successful "brands" . . . and BladeCentering the whole world would let IBM focus on solutions, letting the company promote and sell business integration, Information on Demand, and any other new business concept without getting customers concerned about the box — as in, the box is a BladeCenter, decision made, now let's get it configured for you. . . .

And that's the thing about the BladeCenter — it doesn't have to solve business problems in the very best way because it solves them in a marketing way, and when it comes to selling new solutions to customers, marketing is at least half the battle. In the System i world, we're acutely aware of the marketing battle, are we not?

In any event, what do you think? Is this whole BladeCenter line of thinking flawed?

Posted by cmaxcer on September 10, 2007 at 8:52 AM | Comments (27)

September 6, 2007 11:15 AM

I'm Back in the Office and Gearing Up

Maxed Out has been pretty quiet lately, and that's because I was taking an extended time out to go backpacking in and around Yosemite National Park. In case you haven't been there, the terrain is awesome, and if you want to see a lot of rock, you can't go wrong.

In any event, I'm back in the office and gearing up for a wild fall in the System i world. IBM's recent System i reorganization is raising dozens of questions for businesses and System i pros alike, and we're expecting to learn a lot more this month and in October as IBM opens up a bit.

The flip side of opening up is IBM's recent decision to close the door on its plans to support NSFDB2 for i5/OS — too much work for not enough return? What if System i folks ditch Domino for Exchange or switch to another platform? You gotta believe that IBM has done at least some rudimentary forecasting here, in which case they're not seeing a return on investment, or they'd rather funnel their customers in some other direction.

Either way, some System i shops will find themselves between a rock and a hard place, which is a far cry from backpacking in Yosemite. It's a hot, sweaty job climbing thousands of feet, packing your gear miles into the mountains; but at the end of the day, the choices are relatively easy — freeze-dried spaghetti or oriental chicken?

For better or worse, I've got a feeling that overall options for System i customers are going to rapidly expand, so much so that our universe will have a much different look in a couple of years. Storm clouds or sunshine? Both, of course, and both are enjoyable — as long as you plan ahead and use the right gear.

Here are a couple of pics from my trip:



(And both were great days!)

Posted by cmaxcer on September 6, 2007 at 11:15 AM | Comments (2)