Steve Mills, senior vice president and group executive at IBM Software, said open-source software is changing the way folks look at the market, but that startups have to play into the scheme that established vendors have established, not the other way around.
IBMs acquisition of Gluecode Software, an open-source software service provider, last May, indicated the systems giants willingness to meet open-source competition head on. Mills sat down with eWEEK senior editor Darryl K. Taft to discuss the issue.
eWEEK: How much of that move to acquire Gluecode was a defense against the low end and JBoss, and how much of it was a strategic play in services?
Mills: Well, we dont generally think that defensive plays are all that satisfying, because they tend to be somewhat short-lived. You get myopic and you decide you have to do something thats purely defensive in nature. And I think, frankly, customers, if they get the sense that youre doing something youre not committed to and youre doing it in response to something else that tends to shine through.
Our perspective on Geronimo and Gluecode is that there were very legitimate buyer groups out there—some very centered around open source-based implementations, others just in market segments in places around the world that we could reach through having Gluecode as part of IBMs portfolio.
We recognized early on that we had to drive to where the higher value was in the marketplace. The design point for WebSphere clearly is geared toward more sophisticated uses.
You then look at the market and you say, well, there are other market segments you want to go after—we want to keep growing our business. So what about serving born-on-the-Web companies?
We think WebSphere is a great piece of technology, but along with its high-level capabilities also comes a higher level of complexity. And you reach a point where you say you want to stretch it out and go down market. Weve had good success with our Express offerings … but my view is that another play that gave us legitimacy and a code base that made us more attractive to that end of the market would give us expansion capability.
So we were not in a protect mode, we were really in an expansion mode—moving into parts of the market that were not in today. Now others compete in that market space.
BEA has got a lot of the WebLogic base sitting down there at the lower end of the market. Youve got JBoss and a variety of other offerings that offer simpler capabilities for simpler uses. So we know were going to face competition in every part of the market, no matter where you go.
But the strategy is designed to be very proactive and focused on building a new customer set. Should some of them want to move up into WebSphere, great.
But we didnt acquire Gluecode because we thought this was a stepping stone into selling more WebSphere. We bought it because we wanted to sell more Gluecode.
Staying Ahead of Commoditization
Whats your strategy for keeping ahead of the Pac-Man-like commoditization that open source represents?
In some sense were driving a dual strategy—that is for part of the market were going to embrace commoditization. The Gluecode model is the classic open-source model—take a collection of open-source code that effectively is available for free, package it, offer installation and support.
You put a fee on your packaging, installation and support, and youre selling something that is an application server system but at a dramatically lower price because it didnt cost you any money to have to produce the code. Youre working on integration, enhancing usability, not working on base level code. That gives you a much lower cost base and therefore the ability to charge less money.
So it allows you to participate in parts of the market where commodity level pricing is the way the market works. And then, in our case, we take the rest of our investment and continue to drive up market. Thats what WebSphere XD is all about, thats what the WebSphere Business Integrator family is all about.
There are more than 200 products in the WebSphere family today, and theyre principally focused on up market, more complex, industry vertical-based solutions.
So we believe there are lots of places you can demonstrate value and extract value in the market. But you have to know where those places are, and you have to invest in the code base to get there.
The WebSphere initiative has always been about high-end, high-performance transaction processing. And at that end of the market, we think were the unparalleled leader; no one else comes close.
Its what we specialize in, and we have the ability to put forth a value proposition that customers will pay for.
Some people might say, “Its wonderful you can do that, IBM. Now the fact that you can run a complete mission-critical trading environment on WebSphere is great, but thats not what I do.
“Im running a retail operation and I only need 10 percent of that function in WebSphere. And since I only need 10 percent, I only want to pay 10 percent.” Now I have a product that only costs 10 percent.
Where do you think these open-source startups will wind up?
Only time will tell. You have to separate out being able to build a business versus a grand vision of a monster business. I have a particular viewpoint about business building, which is the one-brick-at-a-time, dont-get-out-in-front-of-yourself model.
And some of these guys get way full of themselves and way out in front, and that creates a set of expectations that go unfulfilled … Many tech companies ride that wave up and then fall off. And if you did an analysis of that, youd find that a few principals likely benefited dramatically in terms of their own personal net worth, but most of the employees didnt get that much out of it.
So I would look at all of these startups and say they all have that potential to flame out along the way because the over-promise, under-deliver phenomenon is all too prevalent in the tech industry.
These companies are exploring very precarious, in my view, business models. What theyve done is theyve said an element of the value Im going to essentially forgo—Im not going to write code. Im going to pick up code, Im going to leverage open source and get this big savings in investment.
So the good news is you can probably start up one of these companies with relatively little money and relatively few resources. We could start a software company around open source and we could claim were successful, were earning a living.
Now, if we let VCs come in and give us more money, then they want big returns so now the expectations rise.
But the business is more of a services model than it is a software model. And services companies have never commanded the kind of multiples that software companies have commanded. And so giving a high multiple to an open-source company I dont think makes a lot of economic sense.
Theyre trying to move in against established players that are significantly better funded, more mature, have deep customer relationships and a level of certainty and guarantee that they bring to the table because their models are more stable.
Do you think therell be more acquisitions?
Yeah. I think weve watched that pattern for a long time. These companies cant quite seem to make it fully to the escape velocity that was promised and the other exit strategy is find a buyer. And far more so than the 90s, when the IPO market was so white hot.
These days selling out to a larger company is considered a major exit strategy option for these startups.