The Devil Is In The Details

One federal IT project promises to thwart terror in the air by using database searches to pinpoint high-risk plane passengers. Another boasts it will overhaul the way the FBI tackles fraud, kidnapping, bomb conspiracies and a wealth of other cases. A third intends to give the Department of Homeland Security (DHS) a much-needed financial tech tune-up.

The results for all three of these high-profile projects have amounted to nothing. Despite years of planning and execution, neither project has produced anything, and two have shut down entirely. But did they have to fail? That's a subject of debate among integrators and other IT players in the federal marketplace. One conclusion seems universal: It won't be the last time that literally hundreds of millions of dollars of opportunity will be squandered.

"Failures such as these are not isolated cases," says Yaneer Bar-Yam, Ph.D., president of Cambridge, Mass.-based New England Complex Systems Institute (NECSI), an internationally recognized nonprofit research institution that consults with federal industry IT companies as well as U.S. government agencies. "A large percentage of big government engineering projects fail. Of the ones that are not completely scrapped, most are substantially over budget. Or they don't deliver."

Here's a recap on the three projects that have followed that very course:

* CAPPS II, the second iteration of the Computer Assisted Passenger Prescreening System, was a highly controversial project proposed by the U.S. Transportation Security Administration (TSA) shortly after the terrorism attacks of Sept. 11, 2001. It would require that passengers provide their birth dates, home phone numbers and home addresses before boarding a flight. Transportation authorities would then run the personal information through databases. The result would be a color-coded label on each passenger to flag what level of security risk he or she posed. But in February 2004, the General Accounting Office (GAO) released a scathing report on the program, saying it was behind schedule in testing and development because of difficulties in obtaining passenger data. There was also considerable public outcry over CAPPS II, citing privacy concerns. The TSA also failed to adequately address inquiries from Congress regarding how the system would be managed and monitored. And the agency couldn't sufficiently demonstrate how the system would be secured from unauthorized access, or whether the data used would even be accurate. By summer 2004, CAPPS II was dead in the water.

* In January, a Justice Department Inspector General report stated that the $170 million FBI Virtual Case File project was a failure. The project was intended to vastly improve the FBI's case-management system. Published reports cited the difficulty in tracking 100,000 cases annually as part of the problem, as well as the fact that the project's managers apparently got more ambitious than they were originally supposed to, by attempting to expand Virtual Case File into noncore areas, such as evidence management.

* For unspecified reasons, in June, the DHS put its $229 million Emerge2 program on hold. Its future remains uncertain. The project, formally known as the Electronically Managing Enterprise Resources for Government Effectiveness and Efficiency program, was intended to create a new financial IT system for the DHS.

Could these failures have been avoided? Maybe, maybe not. But what is clear among integrators and industry experts is this: There are proven, time-tested steps that the integrator can take to effectively approach and manage a federal IT project. In many ways, working on a fed-tech contract is a vastly unique experience. In others, it mimics the ways of the private sector. Integrators who meet with success in the federal IT marketplace are well aware of the dynamics involved, and plan accordingly.

As fallout from the projects' shortcomings is sorted out, a consistent theme has come up that has proved poisonous in any business relationship--a failure to communicate. Expectations are inadequately stated. Progression toward goals isn't spelled out. In the end, this often results in a contentious "push me/pull you" relationship between the integrator and the government customer.

"Reports from the projects seem to scream about the scope of the projects' changing multiple times," says Bill Beeson, vice president of federal sector sales for Computer Application Specialists, a Laurel, Md.-based reseller and IBM Business Partner specializing in data and information-management systems. "It appears that the contractor and the customer didn't set baseline definitions of objectives and control changes to the objectives. As projects progress, the customer always discovers new features they want implemented. Without strong controls in place, these desires become realities that are completely outside the scope of the original contract."

Bidding for federal jobs can build into a frenzy when many competitors vy for the same job. But Beeson advises integrators to make it clear from the beginning as to what resources its team will provide. "In their zeal to secure a federal contract, too many contractors fail to carefully analyze the details of the project," he says. "The result is that the project is underbid and fails in execution."

Getting it right from the bidding process to the end can lead to payoffs--both with the immediate project and future work. And there's much work out there to be had. President George W. Bush has proposed a fiscal year 2006 IT budget of $65.2 billion, up 7 percent from fiscal year 2005, according to McLean, Va.-based Federal Sources Inc. Over four years, the federal IT budget has grown 5.5 percent every year, on average.