Shearer Talks Up the Strengths of the System i
October 2, 2006 Dan Burger and Alex Woodie
Mark Shearer has been general manager of the i5/OS and OS/400 platform through two name changes and four COMMON events, which makes him one of the longest-running GMs in recent times. He is the face of the System i for thousands of organizations and millions of users. Shearer sat down with IT Jungle at the COMMON conference in Miami to give us an update on how the System i is doing, and where it is going.
Being the GM of any of IBM’s server divisions is a bit of a challenge. When IBM’s system and server units were separate fiefdoms within the IBM empire, GMs controlled their own development. That is how IBM Rochester invented commercialized relational databases, 3.5-inch disk drives with magneto-resistive heads, high-density memory, and 64-bit RISC processors, just to name four prominent technologies that changed IBM’s overall systems business. And to a large extent, more than a decade ago, the GMs were able to control their own marketing budgets. Today, GMs are not in control of development or marketing–at least not in the same way. This makes IBM present a united front, which is good. But it also means that the power of a GM is greatly diminished, which is bad.
Because marketing is such a hot-button topic in the System i community, we jumped right in with that.
Dan Burger: So, what about the marketing complaints? What has been done since the last COMMON?
Mark Shearer: Let me just mention a couple of the more interesting things that we’ve done. I do believe that it is a multiyear effort to materially move the needle in System i marketing. Since the last COMMON, there are a couple of dimensions to the System i marketing program that I really am getting excited about.
First of all, we are going back to the universities. One of the problems in our marketing is the fact that there is an entire generation of IT professionals that doesn’t know what System i is. I’m really excited about the university program around the world. The element that is the most innovative is that we are not taking a theoretical approach to teaching System i technology. We are taking a commerce approach to find and develop new skills that they need and to help universities place the students in jobs. We are taking the approach of partnering with clients, colleges and universities together. I’m seeing great promise. It’s a strategic message. It’s not going to move the marketing needle on Day One. But I think it is necessary, not sufficient, but necessary to introduce business computing concepts and systemized technology back in the universities.
The other thing that is positive is supporting initiatives like iSociety, where we are trying to help the community in exchanging information and experiences about the platform in a broad community. Wherever I go, clients often tell me they can never find the application to run on the platform. A lot of times these applications exist, but they don’t know where to go to find them. Or there are professionals that can’t find jobs or companies that need but can’t find people. iSociety is a community-led initiative, not an IBM-led initiative, but I think it is one of the most powerful things we could do to change the perceptions in the minds of the industry.
We call it marketing, but I think it’s maybe a little broader than that. The real challenge is taking the System i platform and making it relevant to the industry today. Doing things like enabling PHP. You know, there are more PHP Web script programs than any other language in the world. Taking the core strength of i5/OS and letting PHP people leverage it is just fundamental. That’s drawing attention to the platform in a community that wouldn’t have known how to spell System i before. That 3,400 of our current customers have downloaded PHP in the beta version is evidence that there is something there. And I am really committed to try to make this platform more relevant to what the world cares about today.
The IP telephony move is another example. IP telephony represents one of the fastest-growing elements of IT budgets today. No one ever thought of virtualizing IT telephony for small and medium businesses. Today, an SMB client using any of the popular, industry-standard approaches to IP telephony would have to have a gaggle of Wintel servers in a rack, and the idea of having a single tower to support hundreds or thousands of hand-sets and their business applications is a very powerful idea. It makes the platform more relevant to today.
So we often talk about advertising and why we don’t advertise on every NFL game, but I think what is behind the concern, which is completely valid, is “What’s IBM doing to make System i relevant today?” And then the related question is “What is IBM doing to tell the industry about it?”
I sincerely believe it is a multiyear journey. I have to continue to enhance the platform, re-engage with the community, and tell the story in simple, more business-oriented terms.
Alex Woodie: What would you count as your biggest success after almost two years on the job?
MS: The thing I am most proud of is that IBM is listening more intently to the community today. Every one of the good ideas I’ve been talking about like PHP and IP telephony was inspired by one of our customers. Were getting much more structured in how we get future requirements from user groups, like COMMON. We spent a couple of days prior to the COMMON conference with the advisory council on our future requirements. We do a similar thing with every user group around the world. I think we are listening more closely and making changes more quickly–never quickly enough–but more quickly. And we are headed on better trajectory of being relevant for more workloads.
The other thing that I have started–and we are a long way from where I want to be–is the process of recentering our business around solutions instead of technology. We have about 2,500 modernized or new applications on the platform in the past 18 months, which is really encouraging to me. There are almost 900 i5/OS applications that are new or modernized to the current platform in the past year and a half. But . . . clients still can’t find the right one when they need it, so we have more work to do. I am pleased with the Initiative for Innovation.
DB: You mentioned in the general session that you are involved in a “Work Stream” project to get a more competitive entry-level iSeries. Can you tell us more about that?
MS: I am not announcing anything. Let me tell you the kinds of things customers have asked for. We are sort of determining the path we are going to take. At the low end, one set of clients would like me to deliver System i capability as a network service. So they would not even have to install the system to get the first five users. This is a System i pay-as-you-go, by-the-user, and on the network.
DB: So you see that as a possibility?
MS: Let me tell you what I’ve observed from some of my customers. There is a set of clients that are very technically proficient. They have good-sized System i boxes. But are still choosing to embellish their own infrastructure with offerings like Salesforce.com CRM or the ERP equivalents of salesforce.com. They are still transferring data to their core systems.
But what I am hearing in the marketplace is time to value is of great interest to people. I have a striking memory of a very sophisticated client telling me that there was no value to him in upgrading his operating system and hardware platform to enable new applications when he could be live one week from making the decision. I believe the market is shifting. Everyone has been talking about software as a service for decades. I was in our Global Services division in the 1990s and I literally had 100 different experiments going around delivering applications as a service. But enough variables have changed that that might be one of the delivery paradigms going forward that is broadly accepted. It is within range of happening. Time to value is driving this.
IBM and its partners already do a little bit of this. Its not marketed in a broad programmatic way, but we do deliver i5/OS capacity to some of our Global Services clients today. We do manage System i5s at client locations today. We outsource i5s today. We do application as a service selectively for some applications today.
The question some of our clients have asked us is to consider is to make it more of a standard option so that it would be easier to buy without negotiating terms and conditions and all that kind of stuff.
I think in the true small business space, it’s likely to be one of the ways small businesses acquire technology. I think IBM is well positioned because we have an awesome services infrastructure.
Some clients have asked us to consider–a lot of the solution providers tell me that we’d be a lot more competitive on the low end if we packaged our offering and charged by user rather than a fixed fee for a model. They are interested in a per-month, per-user charge for the hardware rather than selling it in a traditional way. Others in the industry have user-based pricing schemes. If you have five users, it costs this much. If you have 100 users, it costs that much. The ISVs suggest that could make the platform a lot more attractive if they had the pricing aligned with the value they were getting from it based on the number of users.
This is one of the themes I am hearing from all sizes of clients. They want us to align the price of the offering to the value they receive. In the case of high availability and capacity backup, CIOs all over the world told me the second system for backup isn’t worth as much as the primary system. It’s important, but they don’t want to pay retail. So we announced in July a series of offerings for Capacity BackUp editions where they can save up to 60 percent on the second system because it is priced on the value they derive from it.
Across the board, we are trying to improve the price value dimension of our offerings. Once again, this is a journey and I’ll never be done. But we are trying to systematically improve the pricing value–I don’t mean by just dropping pricing–but pay per value wherever we can.
DB: So will we see more CBU-type offerings?
MS: A lot of people have 520s, so we are looking for an option on the low-end versions.
In the CBU offering, if you choose to take this engine of workload and move it over to the backup machine, you can. You move the software license from system A to system B. And you could run it there for three months, if you wanted to, and then move it back. It is more flexible than just HA or DR. If there are advantages to moving workloads around, you can do that. Clients can purchase full-time processors on the HA systems when you need them. So it becomes an upgrade path for the production machine. It allows more flexibility in how they deploy the same amount of capacity.
The Capacity BackUp for the low end is coming. I’m looking at the services paradigm as well. I’m not sure what combination. There are other customers that ask for the old rental model. So we are looking at various options, but I have heard the requirements. They want an option that allows us to start smaller, with fair price to value. We are still working on it.
DB: If what you have described has come to you from existing customers, how does it apply to acquiring new customers?
MS: Acquiring new customers is 100 percent about solutions–selling new business solutions that happen to run on the System i. That’s why the Initiative for Innovation is so important in getting 2,500 applications out to the platform. Now we are focused on improving our sales and marketing of those applications. Co-marketing with the ISVs and creating value at the local market that can come together and sell the solutions.
Although my installed base inspired IP telephony move, that’s an interesting example where as we bring that to market businesses can either upgrade your current iSeries or you can buy an IBM Express Edition for 250 handsets or 500 handsets and it comes as an IP super server in a box. It is an example of a new application sale that often will include the sale of a capacity backup model.
DB: You believe the IP telephony example would be an appealing option to a company that has never been an iSeries customer?
MS: Yes, because it’s all pre-integrated. It’s a simple implementation of IP telephony versus the industry-standard racks. Consider 20 Wintel servers and the cabling, etc.
AW: How will service oriented architecture change the way people think about platforms? Does it not matter anymore what platform people use?
MS: I think SOA is an extremely important architecture for all of IBM’s clients, including the System i clients. The System i is an extraordinarily robust SOA platform. I think it’s quite possible that companies end up with a hybrid model of computing, both on premises and off, with SOA being the enabling architecture that ties them together.
A customer I have talked with is doing ERP as a network service, plus Salesforce.com, and they are all SOA enabled. The SOA capability on the System i brings all the data in and makes it look like one system. The customer is a little more sophisticated than average. I don’t think all of our customers will necessarily know all the SOA plumbing inside, but a lot of our base, over time, will take advantage of it. We have an excellent relationship with Software Group to make sure System i remains leveraged.
AW: What changes are ahead for System i?
MS: We are going to evolve the Power architecture with Power6 and Power7 and our work on Cell processors. There will be new elements introduced into the Power architecture, but it will be evolutionary. We have a multi-release road map for i5/OS. At a broad level, you will see us try to simplify the total operating environment. This year’s availability of the BladeCenter attachment is an example of how we are trying to simplify the supporting architecture around i5/OS. BladeCenter integration with System i is likely to be in the cards because so many of our clients have an investment in Wintel as well and by virtualizing the storage we can help manage Wintel more effectively.
As we go after new business we will continue to work on a contemporary user interface and contemporary front end from the application standpoint. A lot of the modernized applications were about user interface.
We do have a client base that likes green screen. So, obviously, that will remain for them.