Government's approach to EC has room for improvement
- By Howard J. Stern
- Feb 18, 1996
In October 1993 President Clinton issued a landmark memorandum mandating electronic commerce (EC) for federal procurement. The response from federal agencies thus far has been mixed. The government's decision to implement EC is a good one, but utilizing an approach that is opposite the proven commercial model has led to significant implementation, and thus acceptance, problems.
The administration has specified that its EC program must present "a single face to industry." A key element of the program has been the establishment of an infrastructure known as the Federal Acquisition Computer Network. FACNET is, in essence, the "single face to industry" called for by the official policy.
The government's approach to EC can be improved in three areas. First, redefine the concept of "single face to industry." Second, start with a different set of procurement documents. And third, leverage the commercial value-added networks' core competencies.
The idea of presenting a single face to industry is a good one. Businesses have proven the advantages of a single generic electronic data interchange (EDI) standard and a single set of business processes with all trading partners. But the government has misconstrued the single-face concept, interpreting it as a systems architecture issue.
The idea of performing all translations at gateways is flawed and goes against the proven commercial model. Most companies are aware that remote gateways can become black holes where data can be lost, mangled and difficult to retrieve. As a result, commercial organizations typically perform translations at the document's point of origin or use. This simplifies EDI tracking and enhances buy-in from the functional business department involved. It empowers department staff—who have a greater vested interest than those at a remote gateway site—in the success of EDI document flow.
In the commercial world, EC implementation typically begins with such simple EDI documents as purchase orders and invoices. More complex documents, such as requests for comments, are the last EDI documents to be implemented. This occurs for a good reason. Companies want to implement EDI quickly to reap immediate benefits. They begin with EDI documents that offer the greatest operating efficiency and that are easiest to integrate into existing computer systems.
Instead of starting off its EC implementation with the EDI purchase order (or delivery order, as it is called by federal buyers), the government chose to begin at the start of the ordering cycle, with the request for quotations.
Unfortunately, the RFQ is one of the most difficult and complex EDI transactions to install and use. Unlike the invoice, which typically provides only part numbers and/or order numbers, the RFQ can carry a great quantity of diverse information, and much of that is free-form data. There is, of course, a certain logic to beginning the EDI implementation at the start of the purchasing cycle, but a systems approach would have offered more benefits and greater ease of use.
Most companies realize that their core competency lies in operating their businesses, not in running an EC network. For this reason, most have a link to a commercial VAN. Indeed, almost all EDI traffic today flows through a commercial EC network.
Unlike some commercial EC VANs, the Defense Information Systems Agency does not have years of experience assisting with EC challenges such as documents that have gone astray, system debugging, mapping, EDI errors, templates and software. The government lacks the experience needed to operate an efficient EC network.
With the government running its own EC network, issues of customer service and support arise. Customer service is a challenge even for the most experienced network providers, and it is often an area of differentiation among VANs. As its EC program grows, the government can expect to experience increasing difficulty in supporting the EC efforts of its agencies.
In addition, the government does not have the expertise or the infrastructure to offer the value-added services available from commercial EC providers. These services—ranging from implementation assistance to trading partner marketing programs, EC education and training, and user conferences—which account for a majority of the costs in implementing an EC program, enable a company to leverage its EC investment and improve its operational efficiency.
What Should Be Done?
To jumpstart its EC program, the government needs to stop fixating on RFQs and concentrate on EC implementations that will improve productivity and yield operational cost savings.
It should abolish the gateway concept and allow agencies to perform translations at the sites of origin of EDI documents. The originating agencies understand the applications involved and are better equipped to deal with problems when they occur. In addition, the "single face to industry" idea should be redefined as a uniform set of business practices and EDI standards, not as a limited number of gateways and an internal approach to EDI.
This does not mean throwing away the existing program and starting again from scratch. Instead, the government should expand the program to reach out to more vendors and, especially, to include additional EDI documents. These documents should be selected from a systems standpoint, with operating efficiency and ease of implementation being the selection criteria.
Most importantly, the federal government should follow the commercial model and stop trying to run its own EC network.
One option would be to utilize the commercial VAN infrastructure, with a VAN being selected by each agency from the General Services Administration schedules that required interconnection between the VANs. Another alternative would be to keep FACNET and let an experienced VAN operator take over FACNET management. Systems outsourcing is, after all, nothing new for the federal government.
Whichever alternative the government chooses, the time for action is now. FACNET is not yet a huge entity. Currently it only serves 21 agencies, with just over 1,300 vendors registered and certified to use it. The number of documents flowing through FACNET is also not yet vast: 49,679 RFQs and 112,420 delivery orders in the last 10 months of 1995. Considering the enormity of federal government purchasing, this is minuscule.
Clearly it is in everyone's best interest for the federal government's EC program to succeed. But for maximum success to be achieved, the government must modify its approach to EC, and it must do so soon.
Stern is vice president of marketing in Sterling Software Inc.'s Federal EC Division.