INTERVIEW: John R. Garman, OAO's space man

Really interesting, although it was different only in perspective: same kind of issues, same kind of debates, money, people, overhead, trying to get people to agree on requirements and metrics. It's very much like Through the Looking Glass. Everything's the same, but different. I've enjoyed it, and I haven't felt strange at all.I do feel one difference. I have a lot more flexibility, a feeling of being untied from a lot of the internal oversight and regulatory restrictions. I didn't have a problem with it'you learn to work within the system that you're in. But coming over to the private sector, I feel more unencumbered, and that feels good. ODIN is certainly the largest one we have. It's 23,000 computer seats and maybe another 40,000 telephones across about half the NASA centers.Our initial foray into fixed-price, we-own-the-assets desktop outsourcing was a contract at the Jet Propulsion Laboratory. NASA owns it, but JPL employees aren't government employees. That one has 7,000 or 8,000 seats and preceded ODIN by a year or two.When we talk about seat contracting or desktop outsourcing, a lot of folks think we just provide the desktop PCs or maybe the help desk. That's not the case.Desktop outsourcing means you provide the desktop system and everything connected to it to make it work. It's a bit more complicated than simply putting computers on people's desks and helping them with office applications. When you're on the government side, you see the difficulty in getting people into the work force, but you also see that your contractors keep either having to pay more or have trouble holding onto skills. On the contracting side, it's as serious or more so than I envisioned.Even in this struggling economy, we find it difficult to find good skills at costs commensurate with other technical disciplines. There's a premium you have to pay because the demand is so high.I'm amazed at how many young folks are really into IT, but they're not really educated in it. Like a lot of high-tech companies, we're working with academic and economic development organizations to try to promote IT education.The high-demand skills we need aren't necessarily high-end things like software engineering and chip design. What we need are the skills in the middle, like systems administration, network management, computer maintenance, training and consultation. There is a lot of room to give young people good paying jobs if they get some education in those areas. From what I see and hear, I think it's doing very well at NASA. But if I look back on the difficulties that NASA had in getting this thing started, there were the same two things that have been impediments to our rapid growth. The first is the sudden change in contracting style, and the second is the notion of outsourcing by function instead of by project.In most government IT contracts, outsourcing significantly impedes the government manager's ability to direct specific workflow. To me, it's similar to the changes, some of which are cultural, of moving to performance-based contracting. Fixed-price seat contracting is probably the extreme of performance-based contracting.Let me give you an example of the second one, outsourcing by function instead of by project. A lot of federal installations outsource their base support operations, building maintenance, security and roads. If they outsource the IT as well, you end up with the IT outsourcer providing services to the outsourced base operations. In other words, you get overlap. Success in one agency often is a pathfinder for the government as a whole. But if it's not a success, will it slow things down? Yes, I think it could.I'm surprised, given NASA's relative success with ODIN and all the pressure I see from the government to outsource commercial services in general, that ODIN-like contracts haven't become more common. I don't mean to sound biased here'I guess I am'but ODIN is a natural.It's easy for management in an agency to let outsourcing fail. On the other hand, it's a challenge to make it work. I can't tell you if 5 percent is the right figure.But it's so easy to predict failure because I think it will take a pretty strong commitment from all levels of decision-makers to make it happen. That's tough in any organization, be it government or private sector. From my perspective, absolutely. I'm not trying to say that the rank and file would come close to calling ODIN the best thing that ever happened. They wouldn't. But I think seat management is gaining acceptance, especially among those employees who have taken time to understand what NASA's giving them.Interestingly, I don't think cost or service has much to do with rank-and-file opinions. The greatest gain I've seen at NASA is the raising of the lowest common denominator. They all now have machines that are no older than 3 years old and on average only 18 months old. That's current technology in terms of features and software.There are always those who had expectations of their needs far beyond what NASA management was willing to provide. I would say they are surely less happy than before. In government, and even more so now, I have appreciated the need for the government to be a smart customer. Retaining the skills to deal with this slow but sure shift of IT responsibilities over to the contracting side is tough.The government folks who are supposed to understand all this technology have got to feel like they have a career that's more than overseeing the outsourcer. So I think there's logic for declaring some of what would otherwise be commercial IT services as inherently governmental, simply to retain the skills to be a smart customer.