Steve Will Keeps His i On The Prize

January 21, 2013 Dan Burger

Steve Will was in junior high school when Frank Soltis and a small group of Rochester, Minnesota-based brainiacs were rolling up their sleeves and burning the midnight oil while drawing the blueprints to what would become the Application System/400, the business computer that defined midrange computing. Soltis is an IBM legend–retired but still actively involved in representing IBM and the IBM i-Power Systems platform. Will is considered by many to be a modern day Soltis, but he is having none of it.

Replacing a legend is never easy. And it’s not really fair to Will to ask him to carry that load. Being the chief architect for the IBM i platform is enough of a load all by itself.

In Will’s own words, there are two things you can do as a technologist: you can go deep or you can go broad.

Soltis is a guy who went deep, like a rocket scientist who built business computers. Will is a guy who went broad, a smart guy whose overall knowledge of the IBM i operating system has him well prepared for the inward- and outward-facing duties of his job. He knows there are IBMers who have deeper knowledge about specific aspects of the system. He relies on them for advice regarding system capabilities and feeds them questions he gathers from “listening to clients explain what it is they need their systems to do.”

Actually, there is a lot more to it than that. Will is responsible for the IBM i part of IBM’s overall strategy. He decides where the operating system is going to go and how that fits into the way IBM runs its business. That includes making a budget that allows enhancements from the customer suggestion box–as long as there’s a business reason for making them–and his boss, Dave Nelson, director of development, will sign off on them.

Being a good listener when making customer contacts is a valuable skill, but I would imagine when you are fighting for budgets and defending what you think is the right direction for the IBM i platform, there are some valuable speaking skills that come into play as well. Will isn’t chief architect just because he’s a good listener . . . or just because he’s a team player who considers the advice of deep thinkers . . . or because he’s a good talker, for that matter.

That said, people come out to hear Will talk, like they did in Costa Mesa, California, last Tuesday night at the OCEAN user group meeting.

Attendance at the OCEAN user group meeting was three times the size of the average crowd as people came to hear Steve Will discuss the future of the IBM i platform.
(Click graphic to enlarge. Photo credit: Eamon Musallam.)

The OCEAN user group is based out of Orange County, California, and is one of the most successful local user groups in the country. Its annual one-day tech conference each July brings together 150-plus IBM i enthusiasts for educational sessions and a vendor expo. The regular monthly meetings typically draw 30 to 40 people. With Steve Will as the guest speaker, the attendance spiked to 110, an excellent turnout for a local user group. For those of you involved in local user groups, scheduling Will for an event isn’t easy. Getting several big IBM i customers in the area to set up meetings with IBM’s chief architect can help. In this case, Will parlayed a couple of IBM i customer visits into this West Coast swing, which kept him busy for a few days and diminished the travel and lodging costs that nicked the OCEAN user group budget.

Prior to the dinner meeting, Will sat down with about two dozen executives for a roundtable discussion where he handled questions and provided his perspective on what’s taken place in the past five years or so (post System i/System p convergence) and what might be on the horizon and beyond.

One of the topics he chose to talk about was how the attitude of the customers has changed. Seldom do customers ask whether IBM is committed to the platform anymore. There were many questions after the convergence, but now, Will says, there is very little questioning whether the platform is on a short leash. There is a recognition that IBM has the long-term commitment and an awareness that customers need to stay on the platform because they can’t replicate what they do on IBM i on any other platform for less money and get the same reliability.

It was important to convince customers, particularly the large users, that the platform was not dead. As that occurred, customers returned to a conversation with Will about what they need this platform to look like going forward. The enterprise-level customers get it, but Will admits there is more work to do in the small to midsize customer category to prove the reinvesting strategy has done really great things and that running an SMB on IBM i is still a strategic advantage. The capabilities are there. Now is the time to put modern technology to work.

How the IBM i cornucopia of technology is applied by SMBs is yet to be seen. Those customers, an IBM i strength and an IBM green field of opportunity, are looking for new technologies to become less complex and Will says that is just around the corner.

The enterprise customers have always had considerable clout and IBM keeps them feeling warm and wanted for the most part. I had always heard, and no one from IBM ever denied it, that the enterprise customers represented about 20 percent of IBM i customers and that they were responsible for about 80 percent of the revenue, but Will said, “from a business point of view only 15 percent of sales are to the large clients (the LUG or enterprise companies), and about half of IBM i revenue comes from those 15 percent.”

It is a little bit less than the 80-20 rule, it seems, which is known as Pareto’s principle. There are many ways to count this revenue, so some fuzziness about the numbers is to be expected. But the difference between 50 percent and 80 percent accounts for a lot of fuzziness. IBM rarely (I mean like Haley’s Comet kind of rare) talks about specific numbers, so fuzzy is the name of the game. But it is safe to assume the large users carry a pretty big stick.

A large part of the IBM i strategy in recent years has been devoted to virtualization. Yes, it has been around for much longer, but the emphasis has been dialed up lately. Answering a question about what percentage of IBM i customers are doing virtualization, Will had this to say:

“Approximately 30 to 40 percent of our large users are implementing the Virtual I/O Server in order to do some type of storage and most of the rest (of the large users) will be there in the next 18 months. The bigger clients have to move there in order to gain the benefit of virtualization for things like partition mobility and sharing disk across multiple operating systems.

For SMB, it’s more like 10 to 15 percent. The future is virtualization, but it is too complicated right now. We’ve been working on the interface to make it simpler for the small client, but new technologies are always more complicated at the beginning. Most of virtualization was created so the customer could do any of a hundred things, this means interfaces become complicated. Five years from now 50 percent of small customers will be doing Virtual I/O-based virtualization.”

Cloud computing is another of the IBM i investment themes for the future. Commenting on the cloud developments as he sees them today, Will said the IBM i ISV community is where the action is.

“My rough guess is that approximately 10 to 15 percent of our clients are doing cloud-hosted environments. Most of that is through ISVs that have grown their businesses doing that. ISVs are recognizing they can displace Microsoft (by offering a better solution as in the case of Jack Henry in the banking industry), take over lack of skills in smaller shops, and run the i environment for their clients and it grows their business.

When IBM released a white paper on software as a service (in 2010), there were over 90 vendors who had a significant portion of revenue coming from hosted environments. It has grown since then and more business partners want to get into this business. I am working with vendors who are growing with cloud offerings.”

The relative size of AIX on Power compared to IBM i on Power was another topic of conversation for the roundtable. This tends to be a matter of pride with the IBM i community often bristling as its large user base sits in the shadow of AIX wins in the Unix War. This is how Will handled that one:

“There are more IBM i clients than clients on any other operating system. The Unix market is mostly large companies. The IBM i is in many small to midsize businesses. The only reason that group of people (the SMBs) are IBM customers is because of i. When IBM wants to sell products to the SMB, it’s the IBM i customers they are approaching.

The Power Systems divide is not a competition. I’m glad I’m winning, but it is not a competition. I want AIX to take up so much of the Unix space that they have as many customers as IBM i, but they don’t have that many customers and that is important.

I can’t talk about the revenue split, but IBM i generates a lot more software revenue for IBM than AIX does. Because when you buy the IBM i operating system you pay for the database and the Web server. In the IT industry in general, it’s software sales that are growing. Hardware is not growing so much.

IBM cares about the IBM i business because it sells software.

Unix drives more hardware sales than IBM i, because they sell big boxes to big customers and they sell smaller servers to large companies that are building server farms. AIX derives more like 80 percent of revenue from large clients.

It depends how you count. The large IBM i customers also have Unix and we have been the entry point for AIX in many large accounts. And the convergence to Power was helpful to those businesses that run both i and AIX.”

There was a time when a lot more money was being spent on IBM i than there is now. Part of this is because IBM used to be more interested in selling hardware. Today it is more interested in selling software and services. So the question was asked, how does IBM’s strategy affect the investment in the IBM i platform? Will’s response came in two parts, one that addressed IBM business goals and the other relied on the Power Systems convergence to explain doing more with less.

“What gets funded within the software development organization (which is primarily what IBM is now) is affected by more than what happens with my IBM i projects. In general, software development spends way more money when things are getting started than when it has reached a steady state, even when that steady state is supposed to drive growth.

IBM’s business goal is to increase profitability by doing more with less. In that environment, I am happy with where IBM i development is going. IBM has decided that they will not ask to do what we do with a lot less.

During the past eight years, we’ve grown a lab in China and two of the architects on my team of architects–that help me do the planning–are located in China. We have been able to develop those people with deep knowledge because they are not being rotated to other areas. We work on getting more knowledge in the componentry they work with.

As we became Power Systems, we took the team that did systems management for i, and they became the core of the team for Systems Director for everyone.

I have people who are working on i, AIX, and Linux who are i-knowledgeable people. The same is true with the virtualization team. I am confident that people who work on the other platforms are taking into account the i perspective.”

After the roundtable discussion, Will mingled with the large group that turned out for dinner and Will’s slide show presentation on trends and directions for IBM i. IT Jungle has reported on Will’s trends and directions observations and you can read about them in the Related Stories links below. These are pretty much annual outlooks, which Will says will be updated once again this spring.