categories

HOT TOPICS

Simulating The Brain: Baynote CEO Jack Jia (Part 3)

Posted on Friday, Oct 9th 2009

SM: How long were you with Interwoven?

JJ: I was there for eight years. I started to realize that when companies become bigger innovation becomes very difficult. I really felt that during the last four years. Even with over 1,000 employees, of whom at least 300 were technical as engineers or product managers, it was hard to innovate. As the CTO I sort of ran product management and all the engineering. I also ran marketing for about a year for no particular reason other than to facilitate transition between CMOs.

We just were not able to innovate. That was a problem I had myself, and I think I am a pretty innovative guy. In that environment I could not get anything to happen at the speed that I wanted. I would say that 90% of the innovation happened during the first two years by the first ten guys. After that it was really status quo.

We could easily spend months debating the merits of doing something with dozens of people in the conference room. Early on we would have just gotten one guy to code that thing in five days, and we would have known exactly whether it would work or not.

SM: A lot more pontification happens as companies mature.

JJ: I realized that is just the nature of larger groups. We all know the mythical man-month problem. As an engineer I tried to figure out the science behind that. I don’t know if this is true or not but I think I sort of figured it out. It is a pretty simple combination game. When you have two developers developing software the communication is one line, from A to B. When you have 10 people you then have a 10 factorial of some number. There are hundreds of connections. When you have 300 people it gets even more difficult.

SM: I think the trick there is that somebody has to make decisions. I am not a big believer in consensus management. If everybody has an opinion, nothing ever gets done.

JJ: Mediocrity will win. You regress to the mean, which is not good.

SM: So you basically moved on around 2004?

JJ: That’s right. I basically wanted to go back and do another startup for a while. I was stuck at Interwoven from 2002 to 2004 to help the company a while in its post-IPO phase. However, I really wanted to start a company and I had been thinking of a lot of different ideas for quite some time.

SM: When you were at Interwoven, analyzing startup ideas in your head, what was your process? How did you evaluate those ideas?

JJ: At this time I had essentially worked in my third startup. I had accumulated my own insights and experiences. I realized that the startup is a process rather than a moment of genius. My market research process is a very simple one. I prepare a few thousand dollars and buy people breakfasts. I get people to potentially buy the solutions of the ideas I have.

I this particular case I was not going to start Baynote, at least not the Baynote that we know of today. I wanted to start a different company. I was really annoyed, day in and day out, by email spam. Why can’t we do a better job with the filters? It takes a human less than one second to know what a spam is. Why can’t a computer do it? I figured that would be an interesting problem to solve.

This segment is part 3 in the series : Simulating The Brain: Baynote CEO Jack Jia
1 2 3 4 5 6 7

Hacker News
() Comments

Featured Videos