Among the pitfalls:
A minefield of "legacy" applications. The project uncovered more old software than expected, which must either be approved for use on NMCI, discarded or maintained separately.
Cultural resistance and distrust. Independent-minded base commanders and local information-systems specialists werent necessarily quick to cooperate with the transition to a centralized system.
The few, the proud, the Marines. Though organized as a branch of the Navy, the Marine Corps wanted to opt out and continue managing its own networks.
Congressional meddling. Congress initially capped NMCI deployment at 60,000 seats and kept the Marines from being included. The last of those restrictions was lifted in 2003, after EDS demonstrated it had met the minimum requirements for 28 key service-level goals, ranging from user satisfaction to network availability.
Coordination. Until the Navy established a central project office in 2002, EDS was forced to work out deployment plans separately with different organizations within the Navy, such as NAVAIR (Naval Air Systems Command) and NAVSUP (supply management).
Military planners had been thinking about a centralized, standardized network since the early 1990s. Outsourcing promised a way of getting a vendor to make the capital investment and contribute the skilled workforce, while turning ongoing upgrades and system management into a predictable expense.
The contract EDS won in 2000 was valued at up to $6.9 billion and in 2002 was expanded to an estimated $8.8 billion over 10 years (seven years, with a three-year option). The contract set maximum-pricing and minimum-performance requirements, while leaving room for schedule changes. By the beginning of this year, EDS had assumed network-management responsibilities for 303,000 seats, but as of January only about half of those154,000had been upgraded and certified as meeting the new network standards. In February, the Navy projected the conversion of all 346,133 seats would be done this year, but EDS has already said it will need until 2005.
EDS says the push to deploy lots of seats quickly at many locations inflated expenses and hurt cash flow because of the outlay for new computers. Now EDS wants to finish the job one base at a time.
One ex-sailor, who is currently working on NMCI for one of EDSs subcontractors, says the new strategy should work better, even if it takes longer. EDSs insistence on "trying to push as many seats as possible" created its own problems, he says, meaning technicians would often have to return to correct work that wasnt done right the first time.
The Navy has been largely insulated from EDSs cost overruns by the fixed-price contract. But EDS has signaled it will get more aggressive about billing extra for work it considers outside the bounds of the contract.
EDS stands to earn more by winning performance-incentive bonuses written into the contract for customer satisfaction. Those bonuses start when the user-satisfaction rate hits 85% and, if EDS could get the satisfaction rating above 95%, could be worth as much as $146 million annually. But so far that rating hasnt even broken 70%.
The users who express the lowest satisfaction tend to be those who had the best computer equipment and support prior to the introduction of NMCI. For example, only a little over 50% of naval aviators give NMCI a positive rating, and they have more complaints about the new restrictions on computer use. On the other hand, more than 75% of reservists, who were relatively poorly equipped in the past, rate NMCI favorably.
Next Page: Taking command.