When IBM announced its customer-centric reorganization last summer, the news wasn't overly surprising. Few large companies have successfully penetrated the SMB marketplace, and yet the SMB space is where most of them hope to develop new revenue streams. Most System i value-added resellers also peddle the System p, so the restructuring wasn't expected to introduce very many new problems for IBM's business partner network.
A healthy percentage of Fortune 1000-sized businesses already use IBM solutions. IBM pays attention to its big customers because they write big checks to the company each year. Money talks. Solutions follow. Managing growth in enterprise accounts is different from trying to attract new buyers or enticing SMB customers to spend more with Big Blue.
Still, IBM wasn't likely to capture a lot of very small business customers as a result of the reorganization. Part of that is due to the premium price IBM charges over the competition for lower-end systems, and the other part is a result of the associations that small and price-sensitive businesses make when it comes to IBM — basically that IBM is too big, too expensive, hard to contact, and "not for me." As I see it, IBM is still more likely to have success in the bigger end of the "small" pool and, of course, with medium-sized businesses.
If most medium-sized enterprises have IT departments that are managing a growing pool of servers and applications, it stands to reason that those shops would also be open to simplification of their environment. Let me put it another way: Once complexity is a problem, simplification becomes important — even if the cost of acquisition is a bit larger than the existing budget. In this case, an i5/OS-focused marketing message would seemingly be met with open ears.
However, there's a problem, a big problem, and I believe it's such a fundamental issue that IBM may not have the willpower to overcome it. I hope I'm wrong. Hang with me as I explore this core dilemma.
The System i — and i5/OS if we're talking about running V6R1 on a BladeCenter — doesn't have any direct competition. There's no other manufacturer of i5/OS. There's tangential competition, of course, in the form of Unix, Windows, and Linux, but no matter how people compare these disparate systems, they will always be looking at different kinds of fruit. So what's better — an apple or a banana?
Bananas are great — lots of potassium — but apples can grow in more places and last longer. Pretend that you own a fruit stand. Imagine having this conversation with a customer. Odds are that you're going to try to sell whatever the person is looking for. Here's where it gets interesting. Say that a consumer comes over to your fruit stand and says that apples cost a bit less at the stand down the road. If that happens, you're going to point out that your apples are bigger, shinier, crispier, tastier, and much better than the competitor's apples, which just happen to have a few worms in them.
Now it's time to promote your fruit, and the fruit-selling competition is having a sale on apples. You've got two basic options for the marketing message:
So which is the better message? Should you try to lure an apple buyer with a banana or push the best apples? Let's take a look at how this decision has been shaking out at IBM.
In May 2007 when IBM announced the screaming-fast POWER6 processors, I noticed and noted in my Maxed Out blog that there was something happening that could become a disturbing trend. The company delivered the POWER6 on the System p before it offered the processor on the System i — and promoted it too. IBM's press release at the time was "IBM Unleashes World's Fastest Chip in Powerful New Computer." This notification and the accompanying analyst briefings set the stage for IT news outlets that covered the POWER6 processor and System p 570. They reported on the chip's design, speed, and comparisons to HP and Sun Unix servers.
For IBM, this was mission accomplished: massive dissemination of POWER6 news. Why, however, was the System p crowned ahead of the System i?
Competition. Perhaps IBM isn't interested in promoting i5/OS if there's any chance that it might detract from the System p. The bottom line is that IBM has achieved great success in gaining Unix market shares with the System p, often at the cost of HP and Sun. IBM has been able to create better, faster, more cost-effective solutions and to market them effectively. In fact, IBM goes to great pains to "mention" HP and Sun in its press releases. Here's a snip from one:
"The processor speed of the POWER6 chip is nearly three times faster than the latest HP Itanium processor that runs HP's server line. Even more impressive, the processor bandwidth of the POWER6 chip — 300 gigabytes per second — could download the entire iTunes catalog in about 60 seconds — 30 times faster than HP's Itanium."
There's more. IBM doesn't pull any punches, proving that Big Blue can smack the competition, including Dell, if it wants to:
"IBM calculates that 30 SunFire v890s can be consolidated into a single rack of the new IBM machine, saving more than $100,000 per year on energy costs. According to IDC, IBM has gained 10.4 points of Unix revenue shares in the past five years versus HP's loss of 5.3 points and Sun's loss of 1.4 points. IBM will use the new machine to target customers with less-efficient HP, Sun, and Dell servers."
Talk about a slap in the face! IBM is saying that not only are HP and Sun slow but also that they're losing market shares.
In early November, IBM published a lengthy press release touting new energy-saving software, the first POWER6 chip in a blade, and more easily accessible "best practices" information. In the release, IBM covered i5/OS in addition to Unix and Linux as part of the Power Systems Software initiative that includes virtualization, operating systems and integration, high availability, security, energy management, and platform management.
Initially, IBM noted that AIX and Linux would run on the new JS22 blade for the BladeCenter H chassis, with support for i5/OS coming November 30. IBM later corrected the press release by removing the information about i5/OS availability and then called me (I had covered the news online at SystemiNetwork.com) with an update, attributing the initial statement to a media relations snafu and noting that i5/OS would be available on a blade along with V6R1 sometime in 2008. Still, in that release, IBM couldn't resist throwing more jabs at HP and Sun:
"Calculations show that one rack of IBM's new POWER6 processor-based blades is so powerful when virtualized that it can replace 23 non-virtualized racks of Sun's latest V490 servers, potentially saving more than $200,000 per year in energy costs alone."
"At a media event in San Francisco yesterday, IBM shared its BladeCenter S demo with the press for the first time. This is a comparison of BladeCenter S with HP's c3000 BladeSystem for mid-sized clients that demonstrates that the energy-efficient BladeCenter S may be less expensive and better designed for the needs of small and medium businesses."
The point? New hardware development efforts are easier to implement with AIX and Linux, AIX and Linux are more important than i5/OS, AIX and Linux customers are more likely to jump on the biggest and fastest offerings, or competition in the Unix and Linux space requires that IBM be, well, more proactive and competitive. All these statements may be true.
Remember IBM's Vertical Industry Program — the one that paired IBM with an ISV who had a strong product and used the product to lead sales in highly focused vertical niches based on geographies? Last fall I inquired about VIP and was told that a press release was in the works. It was eventually published, and it bragged about the success of the VIP program — great news for the System i world. One problem: It didn't offer details. I asked for more information and didn't get it. I asked for soft details — more specifics without giving away proprietary information, such as hard numbers on sales. I didn't get that, either. I did have an invitation to talk to one of the participating vendors, but I ultimately passed on it because I didn't see a good story that would offer our readers additional hard evidence about VIP. Meanwhile, IBM was still working on its massive reorganization, so perhaps my requests for more facts were lost in the executive and employee shuffle. Perhaps IBM just didn't want to provide more clues.
It's common for most reasonably open companies to prebrief industry journalists and analysts before major new product launches and introductions. It's a good way for a company to inexpensively promote a particular item, teach journalists and analysts about new features and strategies, and drum up interest in the offering. Briefings are also a great tool for spin control. When you're on the receiving end of a news conference, you're busy trying to grok all the new stuff, which means that the company doing the presentation is essentially talking to a fairly dry sponge that's attempting to suck up all the information it can so as not to miss an important point. When it's time to write the article and squeeze out the facts, some spin points will inevitably ooze through the writer and find their way into the report — lest the author miss something and be shamed by other journalists who paid closer attention. A few days later, though, the reporters either move on or have the time to really consider the news.
IBM prebriefed System iNEWS and other industry publications on the V6R1 announcements in late February. When there is information of such magnitude to our industry, System iNEWS covers it in various levels of detail with several editors and writers delving into the specifics and picking out the most interesting and useful pieces. Although I focused my coverage on i5/OS on a blade, thinking that IBM's BladeCenter will affect our world more than any other development, there are dozens of great new features either built into i5/OS or that will become available to System i customers because of i5/OS. I was excited by the briefing. Then I noticed IBM's press release.
The IBM V6R1 announcements were folded into a larger media report touting virtualization and leading with the System p. Next, IBM hyped two new POWER6 processor-based servers, the entry-level System p 520 Express and the entry-level System p 550 Express, "both expected to entice additional Solaris and HP/UX customers to switch to IBM."
The System i got one dedicated description: "A new release of the i5/OS operating system, IBM's flagship environment for small- and medium-sized clients using the System i platform, with new support for the IBM BladeCenter POWER6-based JS22 blade server." Immediately following that information, which was really just a bullet point, IBM was back to noting that support for the JS22 and JS21 Express blade servers with AIX or Linux would be available for the BladeCenter S. IBM included links to photos of the 520 and 550, which journalists could use in their stories.
This was the worldwide public foundation of the V6R1 IBM announcements. Basically, IBM marginalized i5/OS as just another OS rather than recognizing it as a cost-effective solution and "philosophy" for running a business. I was devastated.
V5R4 at least got its own press release. However, IBM sent out a media notice in July that celebrated the POWER6-based System i 570 — which wasn't going to be available until September — and also promoted V6R1, which wasn't expected until 2008. What gives? Competition. It's not the only answer, but it's a big one.
Direct competition in the Unix and Linux space forces IBM to get specific, to innovate, and to aggressively market its solutions. If it doesn't, it will lose, which brings up the notion of a contest: all contests are tracked by points, and these events include competitors that can vie for those points. Market share points, for example, are up for grabs by HP, Sun, and Dell. Competition gives the players something to measure themselves against, a fact not lost on IBM executives. Competition helps IBM measure its own success. It also forces IBM to compete on price, and it provides direct justification for lowering a price and attempting to increase sales to make up the difference in profit per unit. With the System i, there's no direct one-to-one price pressure for lowering the cost of i5/OS and the System i; hence, the only call to increase sales comes internally from IBM.
Is the System i worth the money? The question is irrelevant in a competitive world. It's only pertinent when a champion rises up to change the game entirely and to acknowledge that it's important to talk about the box to people other than those who already own one. I don't see a System i knight standing tall right now.
Why the focus on press releases? Why not use advertisements and TV commercials and radio spots? Why not involve contests and colleges and VIP programs? Why not use viral marketing, social media, and technology evangelists?
First, press releases are inexpensive. Second, they generally speak in facts, and the words they use tend to describe what a company thinks is important. Third, press releases talk to people who are connectors — the kind of people who take information and push it out to others. Journalists are indeed connectors, and it's critical for IBM to remain in the consciousness of reporters lest the company lose an important link to its customers.
Perhaps most important, the information contained in press releases is timeless. The documents are points of record for the company releasing them, and for everyone else, they're cumulative. The more news you see, hear, or read about, the more it builds up in your awareness. When the information stops coming, it slowly trickles out of your mind.
How hard would it be for IBM, a public company that can certainly afford the cheapest marketing method in the world — the press release — to create a press release dedicated to V6R1? The investment would be minimal, yet IBM seems to think that even a nominal gesture isn't worth the effort. What does that say about IBM's view of i5/OS?
The System i indeed seems to be taking a back seat at the new IBM. Actually, that statement might be a little soft. A more accurate interpretation might be that IBM has stowed the System i in the trunk — right next to a box of bananas.
Chris Maxcer is the news editor for System iNEWS. Interestingly, if it weren’t for the competition factor, he thinks IBM would rename its POWER6-based solutions entirely, getting rid of the System p and System i in favor of a generic piece of hardware that would run any operating system a customer wanted. "Still, it wouldn’t surprise me at all if IBM moves to a more generic naming scheme," Chris says, which means i5/OS would lose much of its differentiated power, which is also its Achilles Heel. Names aside, will IBM seriously push i5/OS as the basis for a solution to SMB clients? "We’ll see," Chris says. "I hope a press release isn’t the equivalent of a canary in a mine shaft."
Links:
[1] http://systeminetwork.com/author/chris-maxcer