The Death of BRIDGE: The US Government's IT Failure of the Year
BY Matthew Burton | Wednesday, January 6 2010
Back in October, the Director of National Intelligence killed a program called BRIDGE. (I've written about BRIDGE here before.) As such a vocal advocate of BRIDGE with a financial interest in its success, my bias is clear, but for whatever that biased opinion is worth, BRIDGE's death was the biggest government IT failure of 2009.
The cause of BRIDGE's death is the most frustrating aspect of it, and it's a reminder of what makes government innovation so logistically difficult: BRIDGE wasn't deemed a failure or a waste or a PR risk. Technically, it wasn't even killed; it was just put on ice. Following the presidential transition, new priorities were made at the top levels of the bureaucracy. These priorities had nothing to do with BRIDGE in particular, or any other tech-related goals. BRIDGE just got lost in the shuffle along with countless other programs that deserve attention.
It was a casualty of a paradox of government bureaucracy: in government, change comes very, very slowly. It takes a really long time for something to happen, and nothing ever happens without a lot of deliberation. Everything is slow. Everything, that is, except for its hatchet. The government can end large swaths of projects with breathtaking speed.
I used the word "hatchet" deliberately. Many programs are killed not because they were hand-picked as faulty, but rather as a simple result of changes in government administration. This happens at least once every four years, and even more given all the agency directors, cabinet secretaries, deputy secretaries, under secretaries, assistant secretaries, and principal deputy assistant secretaries whose routine replacement can cause not-so-routine changes to the way their staff do business. I stayed at the same desk throughout my three years at the Department of Defense. In that time,
five six different people sat in my boss's chair, and my office's name and mission changed twice.
New government executives come and go. Each new one sees things a bit differently than their predecessor and orders a change in strategy. Whether those changes are for better or worse, they're changes, and they're enough to nip any promising project in the bud. That promising project probably took two years of paperwork to get off the ground, but it's gone in an instant. Later, someone will realize there's a need for it…and the process begins again. The slow start/fast hatchet paradox makes it extremely difficult to succeed with technology projects that take a few years to bear fruit. They can barely get going before simple bureaucratic machinations force them to stop. That is the reason for BRIDGE's demise.
The official word on BRIDGE came in October:
Thank you for participating in the BRIDGE experiment...We are working on an updated concept which we hope to launch in the near future under a new name. We would like to see all of you back as we move to a more permanent infrastructure. As soon as we have a launch dateand URL we will post it at http://about.bridge-ic.net/.
BRIDGE died before its first birthday, long before the "next model" deserved any thought. From my seat as a BRIDGE developer, even the first BRIDGE hadn't been fully fleshed out before it got the axe; some of the technical specs were still undefined. When will this next model be live? The provided link is dead, if that's any clue. (Remember: quick to kill, slow to start: there's no time for the previous project to transition to the new one, which probably won't be announced for quite some time.) As far as I know, the next model still lacks an executive agency--an entity of the Intelligence Community like the NSA or CIA willing to oversee the project. This will be a hard sell, given that it's not apparent how an individual agency would benefit more from BRIDGE 2's success than its partner agencies would.
But there is benefit. Major recognition awaits the people who champion this project and guide it to success. BRIDGE and A-Space (a sort of private, classified instance of BRIDGE) are a lot more than a platform for intelligence analysis and information sharing. They are a model for software development that could change the entire government procurement process. Traditionally, the government has bought software--and many other products--through a buy first, evaluate later model that doles out excess money for poor work. BRIDGE provided a way to change that, creating a market for software tools that any developer could contribute to. I've detailed the mechanics of BRIDGE in the past, so I won't reiterate them here.
BRIDGE's and A-Space's incredible potential has also been their major fault: the real benefit and genius of these programs has been hard to explain to non-techies, and while government executives have sung their praises to the press, those praises don't really get to the heart of what these programs are about. (Nor does Wikipedia's A-Space entry.). They are not simply "MySpace for spies." They are so much more. If you are an agency CTO frustrated by the insane way you buy software, consider a new and better way. Adopt the next iteration of BRIDGE. While the technical details will be hard to explain to your bosses, the results--better performance and huge savings--should speak for themselves when it comes time for your performance review.
Thanks to Lewis Shepherd for his guidance on this.