I spent an hour with Matthew Szulik this morning, wanting to get his input for my Open Source CEO Series. Matthew isn't the sort of person to seek the limelight for himself, so it was actually hard to convince him to answer questions. As became evident in his answers, though, Matthew firmly believes in the open-source model and the culture of personal excellence that makes it fruitful.
Red Hat is the defining technology company of the 21st century. Love it or hate it, this is the company to watch. If Red Hat succeeds, an entirely new way of developing, delivering and supporting software will come to dominate the IT industry. If it fails, well, what a long, strange trip it's been.
How did you get involved in open source? What's your background?
If you were to speak to Bill Kaiser, who recruited me, I always had an interest in what (Richard) Stallman and the FSF were doing in free software. I followed his work while I was in Boston. The Mass High Tech Council fostered a lot of the early interest in that world, and I tuned into that.
I was interested in the free software movement because of the increasingly evident problem with software quality. I was a Unix product manager, during which time I saw the problem firsthand and met Bob Young. Bill remembered this and reached out to me.
Importantly, I didn't grow up in one particular aspect of the software business. I wore many hats, and therefore had broad exposure to disparate functional responsibilities within an organization, and that helped me. We're trying to do the same thing here: to retain and motivate bright people, we've focused on internal progression, which...often comes by exposing them to a range of different opportunities.
This brings up an interesting point. Red Hat employs a range of people with strong opinions and the opportunity to daily express these opinions publicly (blogs, mailing lists, etc.). Some of these people might not fit well in other companies. How do you manage the "unmanageable" people?
We've done well and poorly in this area. We just try to focus on reinforcing corporate values, and not try to manage the individuals, as it were. People manage themselves according to the goals and values of the company.
We constantly try to provide good examples of these values in other organizations, so we've had our board members talk with our company to show how one can be successful but also humble, as well as accountable in their personal and professional lives. Internally, we've tended to reinforce the success of the individual and not title and hierarchy. In fact, we grew to 800 people without titles. We strive to have one manager for every 20 associates.
In sum, our belief is that the best management is the peer process, just as in open source. If you measure up to your peers at Red Hat, you thrive. If you don't, you either change or self-select out. When you find people that can do things in an "honest way," without a mercenary view of their assignment, you win. A lot of people don't like this approach, and they leave.
I had a great conversation with a senior developer who was going to visit a customer during his nine-year wedding anniversary. He didn't do it because the company required it of him. He did it because he wanted to show support for his peers by helping the customer even when inconvenient for him. That respect and consideration for one's peers is what drives us as a company.
This is how we reward people, too. We don't target awards for titles. We recognize performance as it relates to the customer, regardless of title or hierarchy. This is why my cubicle or office has always been in the midst of Engineering or Services. I want to sit closest to those that work with our customers. When we meet, as in all-hands meetings, our gatherings are designed to remove title and hierarchy and recognize performance. You don't have to be a manager or a VP to speak up. Everyone's voice is valuable.
This sounds great, but how do you accomplish this? What tools have you found successful for building transparency and peer review?
One of my great fears is that as we grow the organization will lose its tolerance for the outspokenness of individuals. People joining Red Hat from the outside may want to bring their traditional ways to Red Hat. While we respect how other companies operate, we're breaking new ground. We don't want to replicate the 20th century. We want to push forward and find new ways to deliver value to customers.
Coming back to your question, our greatest tool is recruitment. Starting at the recruiting process we try to hire self-directed people. The modern manager is there to be an outstanding recruiter, and someone who can prioritize and motivate, then terminate or promote as performance dictates. We have to conduct ourselves with speed on a global basis. We can't afford an organization filled with order-takers. We need people to be motivated to be creative and driven to solve customer problems.
There's a certain amount of certification that one has to go through in this company to become a first-line manager, otherwise you end up with managers who really don't know how to motivate and lead. You have to prove yourself before you can be a manager of people here. That said, we don't want to build our organization around the charisma of individuals. We don't want single-point-of-failure dependencies. We're a community.
Ten years into this, what would you do differently if you could do it over again?
Ah, that's a long list...[He pauses for 10-15 seconds.] The easy answer would be that I've made hires that were wrong in hindsight, and acquisitions that were premature. But the biggest thing would be to build up our information systems sooner to handle our growth and aspirations. The complexity of building global information systems is tough. Google and others have had the foresight and the competency to build that infrastructure early and quickly. We're now at nearly 2,500 people, but the complexity of building finance, service and support, etc., systems is much more difficult than I had imagined. I would have invested more aggressively and hired much stronger, at an earlier date, in this area.
But how do you know you need this until you need it?
That's the problem. We went public in 1999 and had all sorts of governance and compliance, international taxation, etc., issues. You instantly have to build that competence. There was no net: $300 million on the balance sheet, already public, and the internal fights were GNOME or KDE. But all I could think about was that we desperately had to figure out solutions for an international tax system!
So, what three pieces of advice would you offer to those aspiring to be in your shoes?
- Culture is much more important than technology. Values and culture drive the best companies. It's easy to become intoxicated by technology in our business, but technology doesn't win. People do.
- Invest in a strong board. We've been very lucky to have an outstanding group of directors. I've been the beneficiary of excellent advisers. They don't have to have technology experience. Some of ours are just great on culture and organizational behavior. We have great leaders of people on our team.
- Courage. There were times when we asked ourselves, "Should we get rid of the GPL?" when there was no clear economic model. I had one manager ask "When are you going to get rid of this open-source gimmick?" It takes courage to stick it out and commit to freedom and value for customers. We have to build customer satisfaction around this model. You don't get this with halfway commitment.
Red Hat is now the industry's greatest open-source success story. How do you avoid the hangers-on in your hiring?
Management like Michael Cunningham, Paul Cormier, etc., buy into this, and hire and promote other people that buy into it. You need leaders who will hire other leaders. Interestingly, our success in finding "believers" seems to be higher outside the United States. Of the last 1,000 people we've hired over the last 12 months, two-thirds have been outside the U.S. This the hardest part of my job: hiring good people. I take it very seriously.
[Matthew then quoted a former Netscape employee whose name I didn't catch.] "There are many people who want to work for a great company, but few that want to build one." Here's an example. Michael Chin graduated from NC State, then went to UNC. When I asked him if he'd be interested in running Red Hat China. It took him 2.5 hours to come back to me and say "yes." His wife and infant child were happy and situated here in Raleigh. Michael is now returning after three years. Finding people like Michael has been motivating for me and for others here. People like Michael build great companies.
Matthew's sincere appreciation for individuals and their ability to productively contribute to communities is inspiring. In fact, it's hard to listen to Matthew without being inspired. He tries hard to deflect attention from himself, but this only serves to underscore how important he is to Red Hat in setting its cultural direction and reinforcing its best attributes. I'm sure there are ex-Red Hatters who will disagree with me, but there are 2,500 current ones that likely agree.