Competition adds value to IT contracting
- By Steve Kelman
- Nov 21, 1999
About a year ago, the Federal Computer Acquisition Center (FEDCAC), which is part of the General Service Administration's Federal Technology Service, recompeted a contract for information technology disaster recovery services. The center originally had awarded the contract in 1993, prior to the growth of multiple-award task order contracting under the 1994 Federal Acquisition Streamlining Act.
As was common at the time, the contract was a single award, in this case to Comdisco Inc., one of the leading commercial providers of disaster recovery services. Consistent with the approach towards disaster recovery at the time, the contract was aimed primarily at providing backup coverage for mainframe data centers in the event of floods, earthquakes, bombs or other catastrophes that may hit a data center. It enabled agencies throughout the federal government to place task orders for recovery services.
The scope of the new contract, consistent with changes in the IT world since 1993, has been expanded to include desktops, servers, midrange systems and disruption from hacker attacks. It enables people working at desktops shutdown or destroyed by a disaster to move into local disaster recovery centers or to work out of mobile offices set up at their work sites. It also includes risk reduction consulting services. Task orders include performance metrics, mostly commitments by vendors to get replacement services up and running after a disaster strikes.
In line with the development of other governmentwide acquisition contracts (GWAC) since 1994, the new contract was awarded to a group of vendors: Sungard Data Systems Inc., IBM Corp. and Comdisco. Each company has been highly successful in selling disaster recovery services to the commercial marketplace, and, together, the companies account for 90 percent of the commercial disaster recovery market.
In the past year, 28 task orders for existing FEDCAC customers have been re-awarded, and the contract has attracted nine new customers, including the Census Bureau and units of the departments of Agriculture and Energy. The contract is managed for FEDCAC by Dave Krohmal, who also had managed the previous single-award contract.
FEDCAC is showing the rest of us how to do GWACs right. Two features characterize GSA's Disaster Recovery contract and make it a model for other GWACs.
The first is its devotion to streamlined competition. Proposals generally are only 10 to 15 pages long. Much of the communication between the agency and vendors is electronic. Typically, task orders are awarded within three weeks of issuance of the task order solicitation.
There is no such thing as a sole-source, one-bid award. Except for one $5,000 job, every task order awarded under this contract has had more than one bid; for the vast majority of task orders, all three vendors have put in bids. Streamlining and competition can go hand in hand. In the competitive environment, the government is getting astounding price breaks. To win the contract, the vendors bid fixed ceiling prices for various services that, on an apples-to-apples basis, are considerably lower than the prices Comdisco charged under the previous contract.
However, task order competition lowers prices. Typically, agencies are paying 50 percent less under the new contract than they did under the old one; some prices are as much as 80 percent lower. Competition has heated up during the year under the contract, and prices have dropped. Comdisco, the incumbent under the single-award contract, has won about half the recompetes, while the other two vendors have won the other half.
Although price naturally plays an important role in vendor selection, the vendors have somewhat different technical approaches as well as strong and weak points that also weigh into the best-value selection process. The vendors have off-site center locations, which, given the center's particular specialty, may be better or worse at meeting a customer needs.
The companies also have somewhat different philosophies about how to deal with massive disasters - like a nuclear bomb destroying a major city - that would tax any of their capacities.
Past performance generally has been excellent for all three on the more than 400 recovery test exercises that have been conducted under the contract. The results were excellent in the one time a disaster was declared - after a hurricane that hit USDA's National Finance Center in New Orleans.
The second key to the success of this contract is its specialized focus.
It would be very hard for an agency to develop much expertise in a specialized IT area such as this one. FEDCAC provides some help to agencies in developing statements of work and evaluation criteria for task orders. But Krohmal said he believes that FEDCAC's biggest added value comes in evaluating proposals and negotiating final deals with vendors.
"I've been doing this work for six years now, and I know this very specialized commercial marketplace very well," Krohmal said. "We know enough about the market to help an agency go over the vendors' proposals and differentiate vendor technical approaches and costs. And after the proposals have been evaluated, we can negotiate on the agencies' behalf.
"We know what's a good deal and what's a bad deal."
Any individual customer for those services would be a relatively small part of a vendor's business. FEDCAC, with almost 40 disaster recovery customers, is an enormous, very important customer for each of those vendors. If there are any problems during contract administration, FEDCAC can grab the attention of vendor management.
The experience under this contract is not fully comparable to other kinds of IT services. Because the contracts essentially are insurance policies, vendors have more freedom to lower prices and thus increase their risks in the event of multiple disasters befalling several of their customers. If they actually had to provide people or equipment to accomplish a requirement, their risk would not be so high.
Nonetheless, the case of the FEDCAC Disaster Recovery contract has some pretty clear lessons to all GWAC contract managers - and their customers - in government.
Competition works, as you would expect, in getting a good deal. Specialization produces the expert knowledge that provides a crucial justification for having governmentwide contract vehicles in the first place. Let's hope that the government learns the lessons that Krohmal is so deftly teaching.
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.