A few issues ago I wrote about Terry Little, an innovative program manager at Eglin Air Force Base, Fla., who used various principles of acquisition reform to achieve better than a 50 percent cost savings on a 'smart bomb' kit called joint direct attack munitions (JDAM) [FCW, Sept. 7]. Well, Terry
A few issues ago I wrote about Terry Little, an innovative program manager at Eglin Air Force Base, Fla., who used various principles of acquisition reform to achieve better than a 50 percent cost savings on a "smart bomb" kit called joint direct attack munitions (JDAM) [FCW, Sept. 7].
Well, Terry is back, and this time he has delivered another acquisition reform success story with relevance for the information technology community. After the success of JDAM— in which the Defense Department saved $2 billion in purchasing smart bombs that were more accurate than the Pentagon had specified— skeptics in the DOD weapons community said that Little's approach may work for a relatively simple item such as JDAM, but it could never work for a complex weapon that required a significant developmental effort. Enter the Joint Air-to-Surface Standoff Missile (JASSM), another Little program.
DOD planned for JASSM to be an affordable cruise missile that would have the "stealth" properties of existing cruise missiles but would be more accurate and less expensive. DOD hoped JASSM would reduce the unit cost of cruise missiles from the $1 million to $3 million range to $700,000 a pop or less.
DOD bought JASSM in two phases: an initial down-select to two contractors from five, followed by a phase in which those two contractors were paid $330 million to develop a working system that would meet the performance requirements and be fixed-price for most of the production run. The bottom line: a product meeting the performance requirements and coming in at about $300,000 a unit— less than half DOD's original price target.
There are many lessons here that have relevance for IT systems development:
* Look for opportunities to fulfill requirements with commercial items in developmental projects. JASSM is not a commercial item, and neither are many of the complex IT systems the government develops. But JASSM made much more use of commercial components and subsystems— analogous to off-the-shelf software solutions in the case of IT— than traditional weapons, and this was a key to reducing the price.
* Signal clearly that affordability is an issue. If vendors really believe the message that the government is not just in the business of paying any and all bills submitted but that affordability counts, it is amazing what contractors can accomplish.
* Take past performance seriously. JASSM significantly used past performance in down-selecting contractors and in making the final choice. Past performance counted 50 percent in the down-selection. "We spent more time on our past-performance evaluation than we did on evaluating the whole rest of the proposal," Little said. "We substituted past performance for process evaluation. Instead of asking for a plan for software, systems engineering or cost/schedule control, we looked seriously at how they had done." When Hughes didn't make the down-select because of past performance, it protested to the General Accounting Office, producing the leading case upholding governmentwide discretion in applying past-performance information. The last chapter of that story is that the contractor that was chosen using past performance has delivered on behalf of the government.
* Consider whether parallel development of a part of a big project makes sense. In this situation, DOD spent about $150 million extra to support a second contractor through a program-definition and risk-reduction stage. On a $2 billion program— which is cheaper than originally expected because the unit cost is down— that seems like a prudent investment for the returns obtained. Government ought to examine on a case-by-case basis where a similar approach might be worth it for major IT systems development.
The ultimate test of the success of the reform efforts will not be whether we have shaved some months (or even years) off cycle times. Streamlining is one of the means toward the real goal, which is achieving better mission support and value from contractors. Making the system faster and less resource-intensive contributes to that by allowing new technology to get out quicker, by freeing up time for activities that add greater value than the value (if any) of wallowing in procurement paperwork and, importantly, by sending the federal work force a signal that we care enough about our missions to want them promptly supported.
Streamlining is but one means toward the larger goal. Little continues to show what disciplined attention to the larger goal— and to all the tools of acquisition reform— can achieve for agencies and for taxpayers.
-- Kelman was the administrator of the Office of Federal Procurement Policy from 1993 to 1997. He is now Weatherhead Professor of Public Management at Harvard's Kennedy School of Government.
NEXT STORY: iMac lives up to the buzz