Home | About | Recent Issue | Archives | Events | Jobs | Subscribe | ContactBookmark The Sterling Report


Will the enterprise market spend significant IT budget on Windows Vista in 2007?



Entrepreneur Spotlight: Mike Stonebraker, StreamBase

By Angel Mehta, Managing Director, Sterling-Hoffman Executive Search

Angel Mehta: How did you evolve from academic to entrepreneur? Tell me about the first series of events that led to you leaving academia and becoming a ‘businessman’.

Mike Stonebraker: Between 1972 and 1977, my research group at Berkeley wrote the relational database management system called “INGRES”; currently owned by Computer Associates, which they just decided to open source.

In 1977 Arizona State University considered using INGRES for their student record system, all 35,000 students worth. They could get by the fact that one had to run Unix, which was, at that time, an unsupported operating system from AT&T in North Carolina. They could get by the fact that one had to get this relational database system from these goofy guys at Berkeley, which was, of course, also unsupported. The project came crashing down when they realized there was no COBOL available for Unix at the time and that they themselves were a COBOL shop. Most serious application of our software got hung up on unsupported operating system, unsupported database system or no COBOL. So I founded INGRES Corporation in 1980 to commercialize our academic prototype.

It was a perfect situation because, at that time, the University of California encouraged faculty to consult on the outside with IT companies. So, I consulted for my own company one day a week for the next eight years. I continued to be on the faculty full-time at Berkeley and was CTO of INGRES on the side.

Angel Mehta: So, you had one foot in the academic world and one in the business world?

Mike Stonebraker: Yes. And during that eight-year period at Berkeley, we wrote another database system called “POSTGRES”. In 1992, I started a second company, “Illustra” which commercialized POSTGRES and I was CTO of Illustra part-time. In fact, StreamBase was the fourth company I started and I’ve never worked full-time for any of the companies. I’ve been CTO part-time during this entire span, and my main job has always been some academic position elsewhere.

Angel Mehta: Did you have any business background that helped in your decision to start INGRES?

Mike Stonebraker: Well, it turns out that at the time I started INGRES I had no clue what was going on. This was all learned by doing; the motivation for forming INGRES Corporation was not financial in the least. It was to prove that relational databases were a good idea by providing the infrastructure that would let people like Arizona State University run it.

However, as a faculty member at Berkeley, you’re forced to be an entrepreneur in terms of hustling research funding, getting tenure, supporting graduate students, and writing papers. Every faculty member is really a CEO of a small company that has to bring in enough money to pay the research team, publish papers, get famous and say wise things. You’re forced to be very entrepreneurial and then you only need to add some business acumen. You just have to pay special attention to what customers will actually buy.

The INGRES experience made me realize that there’s a ton of really good ideas in universities. In fact, right now I’m on the faculty of computer science at MIT and it is really alive, really vibrant. There’s a ton of really neat ideas but technology transfer is the main challenge. I firmly believe that having academic spin-offs like INGRES, Illustra and StreamBase is a terrific way to do technology transfer.

Angel Mehta: Let’s talk about learning by doing. What were some important lessons you had to learn as an academic adapting to life as an entrepreneur? Were there any aspects of running a company that were counterintuitive for you?

Mike Stonebraker: Oh yes. First of all, I think that if you are fundamentally a research-type of person, there is a tendency to not listen carefully to what customers say. And eventually, you will fail if you don’t listen to customers. They’ll tell you either why they like your product or why they don’t. Why they don’t like it is probably more important because it gives you ideas on how to fix it and then gives you additional research problems. A terrific focus for research problems is customers’ dissatisfaction with either your stuff or somebody else’s. I think one of the things that motivates me to spend time with customers as I do is the ability to get exposure to their thinking as grist for new research problems.

In terms of running a company, it took a while for me to realize you have to act like a CTO and that requires re-treading your brain from the way academic institutions are typically run. University departments are very collegial, which is to say that not everybody makes exactly the same amount of money but there aren’t huge differences in salary. If you have extra money, you help the guy next door and vice versa. Decisions on hiring new faculty members into your department are made very collegially and more-or-less democratically, which is a very ineffective way to run a company. You cannot run a company using democracy.

Another issue… it was a real shock to me how sales actually happen. I was horrified at the way sales are done in IT, in general, and how expensive it is. You would think in the best of all possible worlds that companies would be intelligent enough to purchase the best product fit for the application. It was a real eye-opener to see how political the sales process is, how much process management, how much policy there is. I’m surprised by how technically unsophisticated top management in a lot of major companies is, even companies that are unbelievably heavy users of IT. Because of this, senior management at a company that consumes high-tech products is unable to judge the IT products to decide on their merits and make the right decision.

Angel Mehta: Can you give me an example?

Mike Stonebraker: Without naming names…one major Eastern money-center bank that has 4,000 technical IT programmers: basically I would characterize them as an “IT shop disguised as a bank”. Senior management thinks of the company as a bank, not as a company that uses IT as a major differentiator in the marketplace. They have a lot of Unix in this particular institution but they wouldn’t hire one of my best people at Berkeley who wanted to live in New York. He was the best Unix programmer I knew; however, he had a ponytail and didn’t wear shoes. I can picture their bankers looking at this guy saying, “Oh my God”. I think this particular institution would be way better off if senior management would get a brain transplant and become much more sophisticated in IT.

Angel Mehta: I was talking to Jo Tango at Highland Capital Partners about you and he mentioned that, in most cases, he tends to question whether entrepreneurs who have ‘succeeded’ before will still be hungry to do the next one… but with you, there is just something odd that keeps you going and going… I have to ask… what motivates you to keep starting companies?

Mike Stonebraker: I have plenty of money, so it’s certainly isn’t the money. I believe if the U.S. is going to remain a world leader for the next 50 years, you’re going to have to win IT. Basically, IT is going to be like oil; if we don’t stay ahead I don’t see how we’re going to remain competitive in the world economy. As near as I can tell, the major IT companies, when they get big enough, they get very cautious, very sales-driven. They’re getting very good at paying attention to their customers but that isn’t necessarily what makes them long-term competitive.


  Home | About | Recent Issue | Archives | Events | Jobs | Subscribe | Contact | Terms of Agreement
© 2006 The Sterling Report. All rights reserved.