Q&A with IBM’s Mark Shearer: Still Mister System i
March 3, 2008 Timothy Prickett Morgan
Q&A with IBM’s Mark Shearer: Still Mister System i
When IBM reorganized its Systems and Technology Group once last July and then again this January, the System i product line seemed to get a bit lost in the shuffle. In July, the System i went from being a division to being a product line split between high-end and low-end divisions, Power Systems and Business Systems, and then in January IBM finished redrawing the organization chart and cut product development and manufacturing from sales and marketing.
More than a few people have had the impression that the System i product line was without a leader, regardless of what was going on inside IBM, mainly because there has always been a head of the AS/400, iSeries, and System i Division and we all have known for two decades exactly and precisely what that meant–one person to hold accountable for the product line. Three years ago, much to the relief of the iSeries division (its name at the time), Mark Shearer was named general manager of the unit. Then, after the July 2007 reorganization, all of development of Power-based systems was pulled into the Power Systems division under general manager Ross Mauri and suddenly a division set up quietly in January 2007 by IBM to chase sales among small and medium business shops, called Business Systems and run by general manager Marc Dupaquier, was in charge of anything smaller than a System i 570. Shearer was no longer a general manager of a product line, and was technically a vice president and business line executive in IBM’s Power Systems division. As of the January 2008 final step in the reorganization, which converted Business Systems to a customer-facing sales and marketing organization and pushed all product development back into Power Systems and the newly created Modular Systems division (formerly known as the System x division), Shearer was moved from one side of Systems and Technology Group to the other, and is now vice president of marketing and offerings for the Business Systems division.
Basically, Shearer is back in charge of peddling and promoting the System i product line as well as any other products this SMB-focused division ends up selling. He does not have to worry directly about product development and manufacturing any more, but as you will see, he does have to be involved with these processes in his new role at IBM. To explain it all, Shearer took some time to chat on the phone last week.
Timothy Prickett Morgan: So, where to begin? What does it mean to be the System i at this point, given all the changes in the Systems and Technology Group, and what does it mean to be in control of it, managing it, speaking for it? That’s what I am trying to get my brain wrapped around. I have been told that it is you, again, who is in charge of the System i, and I think people will be good with that.
Mark Shearer: I think as a starting point, I would say that one of the most significant things that IBM has done in the seven years that I have been in STG is what we announced back in January, and that is shifting from having a product-oriented go-to-market model and business model to having a more client-oriented business model. This is really part of a multi-year transition in Systems Group that you have followed. You saw a few years ago when we consolidated our approach to development, and about a year ago we consolidated our approach to marketing, and we took the final big step in January when we shifted our management of the business and our sales/marketing go-to-market model from multiple, independent products to this client orientation.
As you know, as we moved towards this client model, we divided the world into large enterprises, the general business market, and industry-specific solutions. I probably should comment on why we did it, and starting from the System i vantage point, the average company that has a System i in its data center has an awful lot of Intel servers, an awful lot of Unix, and many of the clients have some form of storage area network installed as well. And we literally were approaching those clients focused on our technologies rather than being focused on what they were trying to accomplish for their businesses. We got a lot of feedback from both our large enterprise clients and our small and medium business clients, and what I found sort of stunning in the System i space is that we have all talked about System i loyalty to the platform–that clients love the platform and are loyal to it–but what I discovered as general manager is that although they love the simplicity and integration of i5/OS, these companies were purchasing an awful lot of competitive Wintel and Unix systems.
TPM: This is not necessarily the kind of love a general manager wants to have. . . . Not when you are trying to make your revenue numbers.
MS: Exactly. And if you think about it, we really were missing the boat not having a more holistic view of what our clients were trying to accomplish. I know you and I have discussed this in the past.
TPM: Yup. We’ve gone over this, the numbers of how many servers are installed and how much budget goes to Wintel versus System i.
MS: If a client was buying a new ERP system, they could have a System i rep, a System p rep, and a System z rep all recommending SAP on their particular platform, and our clients were looking for a more integrated approach, one that was more focused on what they were trying to accomplish than on what a given sales person was trying to change. So this is really a significant shift from product centricity to client centricity.
TPM: And I am hoping that as part of that someone can actually sell the System i. This is something that we have all crabbed about because this has gone on for years and years: different reps come in pushing different platforms, different partners coming in, and because in the early 1990s, application software became increasingly cross platform, customers would get into a position and ask IBM, “What should I do?” And IBM’s effect response was, “How much money you got?”
Or worse still, IBM would in effect say, “Well, that’s not up to us. I am not going to help you make that decision.” We in the OS/400 base used to get terribly frustrated because nobody’s job at IBM was ever to say to the customer, “You really need this one, and I am speaking as IBM now. You do not need Unix–you are going to hurt yourself. This is what you know, this is what you do, and this is how you justify paying upfront costs in hardware and software for what will end up being operational costs on other platforms.”
It is easy to sell Windows when it is the thing that everybody has on their desks. It is easy to sell Unix when there is an open systems boom going on and this becomes the normal, acceptable platform for the dot-com build out. It apparently not too easy to say, “No, you are good where you are, and you need to build that AS/400 infrastructure out and here is the best way to do it.” I have been angry for a long time that there has not been an official, economic justification for the AS/400, iSeries, and System i platform beyond the same old “ease of use, ease of administration, reliability, total cost of ownership, don’t think about it” mantra. Which doesn’t really go down well with me, as you know from knowing me all these years. I think IBM can and should have made a case for that box, that it should be the platform that more companies should be deploying more and more things on. You have done the software and systems engineering to justify it.
MS: I think that by having a product and technology-oriented go-to-market that not only were we not doing that sufficiently well, but we were literally competing against it at the same time. So for a variety of reasons, I actually think our industry is needing to go more towards this client-oriented approach, where the end result becomes more important than the components. But I think for a company like IBM, it is our strong suit that we happen to have a portfolio of capabilities, but we just have not represented it well in the marketplace. I think that this approach will eliminate a lot of the contention and internal competition.
What Erich Clementi and I are focused on in this Business Systems space is really driving innovation for small and medium companies. And over 90 percent of the System i customers, as you know, would fit into what we call the general business space, and that is why I continue to be extremely active in the System i community. I will be at my fourth consecutive COMMON meeting in Nashville at the end of March, and I think I will continue to be quite visible to the System i community, it is such an important product in this general business space. What we are focused on is driving new innovation into how we go to market with small and medium businesses, new innovation in the products that we introduce for the marketplace, and also new innovation in terms of a new paradigm of computing, driving the new agenda for IT for small and medium businesses with some of our colleagues across IBM in Software Group and in Global Services.
I think having a dedicated focus on the total market and the entire portfolio of products will let us focus on the unique characteristics of general business and I think it will ultimately allow IBM to grow its share in what is the fastest-growing segment of the systems market.
TPM: But what is the effect of these changes on the System i as a platform? Bill Zeitler has made statements about how Business Systems will emphasize i5/OS and Linux, and Windows was not part of his statements. [Editor’s note: See IBM Aims for Server Expansion in 2008, Including System i Reincarnation.] I am assuming that eventually Business Systems does Windows, and I assume Modular Systems will create blade products for Business Systems, and I am hoping beyond hope for a low-priced entry blade that can sit side-by-side with Linux and Windows.
There are things that you can do, obviously. But what I am trying to figure out is who does what? Does Power Systems and Modular Systems engineer this stuff and then Business Systems and Enterprise Systems cherry picks the technologies they want?
MS: At COMMON, you will see Ross Mauri, the general manager of Power Systems, and myself, together. I will be talking about driving the market requirements and the general business characteristics of the Power Systems family. I have got a counterpart in Enterprise Systems driving the high-end product requirements, as driven by our large user group. We have separated making products from marketing and selling products–and that is a statement across Systems and Technology Group.
TPM: Who makes the decision about what products get made? If you decided to have a whole new packaging of an entry server, how does that work? If you need a certain thing, how do you get it? In the past, if the System i division wanted to build beige machines based on Motorola chips, it could have done it, and then the division would market it, sell it, and see how it played out. How does it work now?
MS: We have this integrated development process, that I know you are familiar with, that continues and the go-to-market units provide input. I sit on an advisory committee for Power Systems and we have a major say in what low-end offerings are created, how they are packaged and priced, and how they are brought to market. It is really a team effort. Ross has to factor in the high-end requirements and the low-end requirements.
TPM: Just like the iSeries and pSeries divisions, and then the System i and System p divisions, had to do in the past. What I am wondering is just how strong of a feedback loop is there back into Power Systems. If you are the customer-facing organization, as Business Systems is, then you should be driving the next set of requirements, it seems to me.
MS: Exactly. There’s one process at IBM, and we are heavy on the requirements and go-to-market side, while Ross’s team is heavy on the technology development and product management side. In that particular example, since Ross and I have worked together since 2000, he and I are very well linked–literally, for the interesting part of COMMON coming up, we will be there together because for our past three jobs, we have been working on the things that are coming to fruition now.
TPM: What should System i customers be expecting in terms of announcements this year? We have got one Power6 machine, the 570, and a new Power6 blade, the JS22. Can you give us any hints about the further fleshing out of the product line in 2008?
MS: [Somewhat incredulously, with humor.] Well, we just announced the biggest version of i5/OS in years. . . .
TPM: Well, I covered that–I’m done now, I have got to move on, brother. I think we did 12 or 15 articles on it. . . .
MS: I am glad you did, and I think we gave you a lot of things to write about.
TPM: But seriously, there has been some new Power6 iron that has System p labels on it in entry and midrange machines that doesn’t have System i labels on it yet. So what is the general shape of the System i product roll out? Are we expecting the entire line to be refreshed with Power6 this year still?
MS: Absolutely, and I think you and I talked about that last year.
TPM: True, but there have been so many changes at STG I have to check. . . .
MS: I have had a few different titles and I don’t think I have changed what I do for a living in seven years. I am still working on the same things, and pretty much with the same people.
But back to your question. What we are trying to do is to clearly feature i5/OS–it’s first 20 years and its next 20 years–as the secret sauce of the System i platform. Now, all of our clients, and everybody at COMMON, and many of you have told me, “Boy, you really do not feature i5/OS enough, or talk about the unique value it creates.” And one of my personal missions has been to get i5/OS “mainstreamed,” available on our hottest platforms. This recent announcement of the BladeCenter support with i5/OS V6R1 and the BladeCenter S chassis is a big deal. We introduced the virtual BladeCenter support two years ago, and we are now giving customers the ability to put i5/OS into a BladeCenter right next to their Wintel systems. I think this is a really compelling value prop for a certain segment of the System i users.
TPM: But the JS22 is aimed at pretty high-end customers. I love it, I really do, but it is a P20-tier machine and it has got a lot of oomph in it. The JS22 is great if customers want to consolidate a lot of Linux or AIX onto the blade as well. But it is a lot of horsepower for the typical OS/400 and i5/OS shop. So I am knocking on wood for a smaller P05 and P10 class blade.
MS: It is a lot of horsepower, and I think that you can expect to see a more well-rounded Power-based blade family. That’s something that we are working pretty closely with Ross on.
Of course, you will see i5/OS V6R1 on other Power6 products, and we will roll Power6 out over the course of this year. And that will be a screaming product. As you know, the System i 570 we released last year with the Power6 processor, and that together with our new low end and end-user pricing, helped drive a significant increase in System i volumes last more–an increase of more than 20 percent. The Power6 570 was very well received, and so we are getting in position for i5/OS V6R1 across the board on Power6.
TPM: I know you have had shipment increases in the quarters throughout 2007–did you do it every quarter for the System i?
MS: Shipments started to grow beginning with the “Project Ignite” launch in April. [Editor’s note: That’s the code-name for the user-based 515 and 525 machines.] The second, third, and fourth quarters had solid volume growth.
TPM: Can you maintain that going into 2008?
MS: Well, you know I am not going to project. [Laughter]
TPM: Hey, it is my job to ask. Will we see a revenue uptick for System i sales? The 570 went off like crazy, well above expectations, but I am wondering how sustainable it is. IBM changed a lot of things with that box to make it more appealing, and my assumption is that it is not just a one-quarter phenomenon. i5/OS Application Server, more aggressive pricing, lower initial operating system costs, you don’t have to take a minimum of processors, etc. It is a good box.
MS: Those things that you just mentioned I consider all part of the “price to value” improvements that we made. I think that is why the Project Ignite low-end boxes took off as well–it is the same concept of paying for what you use instead of paying for everything. We are definitely going to continue and enhance this idea of paying for what you use as we introduce the remainder of our Power6 offerings. That is absolutely critical.
So are a few other things. As part of our go-to-market strategy last year, we launched the Vertical Industry Program solutions approach, and that is something that we are also going to continue. I think you will see more visibility on how we sell the stuff.
TPM: Do you know the effect of the VIP? Can you quantify how much incremental sales it drove, or new accounts, or what have you?
MS: The short answer is yes. The System i volume growth was the effect of the low-end product restructuring, the Vertical Industry Program, and the Power6 570. This VIP approach, which we introduced first with System p and then with System i, is showing material impact on the business and is absolutely measurable. Everything we have seen encourages us to invest more in this kind of go-to-market approach.