Peter Gallagher | System re-use: acquisition allusion
- By Peter Gallagher
- May 06, 2008
Government is in favor of re-using system and software assets
, so one would expect to find strong procurement guidance reflecting a preference for software products and services that create re-usable assets. But this is not always the case.
For example, the Office of Management and Budget reminded agencies
in 2004 that decisions should be broadly based on total cost of ownership (TCO), mentioning life cycle maintenance, risks and security. The memo notes that investments must be made consistent with the federal enterprise architecture (FEA) designed to advance the re-use of IT assets, but there is no specific mention of preference for re-usable systems. For the most part, acquisition guidance has not stressed re-use as a preferential attribute.
What then is preventing the acquisition of re-usable system assets for the government and making it a preferred default outcome? Why is it that the re-use opportunity, of such obvious value, is not even mentioned in most information technology procurements? How is it that a key objective of FEA makes its way to the procurement table only indirectly? The answer lies in large part with timeworn assumptions about TCO and the commercial off-the-shelf model.
Regardless of whether TCO is explicitly calculated, the evaluation paradigm tends to be bound by the scope of the procuring office. For instance, if Agency X is buying a particular solution, the frame of reference is the direct cost to their unit. Agency X typically does not try to determine what the economic costs and benefits are to the government or, even more broadly, the taxpayer. Let's say solutions A and B are functionally equivalent. Solution A costs $1,000 and can be used as needed by Agency X whereas B costs $1,100 and can be used governmentwide. Agency X is better off with Solution A, but the government might benefit greatly via solution B. And we are talking here about real dollars, not fuzzy social benefits. Agency X has a laser focus on their budget. The larger economic equation for the government is simply not within Agency X's purview.
Governmentwide procurements, schedules, shared services and common architecture initiatives are a few of the mechanisms for dealing with this big-picture challenge. Governmentwide acquisition contracts and shared services have undoubtedly resulted in huge savings and can meet many of Agency X's requirements, but never all of them. The re-use and centralization of contracting mechanisms can reduce transaction costs and lead to lower TCO. But this is not the same as re-using system assets.
A key assumption about COTS also distorts efforts to value re-usability for IT assets. COTS products are simply things that are commercially available, sold or licensed to the public. Yet despite plenty of official guidance to the contrary ' such as the OMB memo cited previously and guidance
) ' there remains a general misunderstanding that the more flexible re-use opportunity presented by open-source software is somehow different than COTS. Clearly, COTS can be proprietary or open depending on the commercial business model of the vendor. One of the key differences between these competing COTS models is the opportunity for re-use, yet acquisitions rarely focus on the distinction.
The acquisition workforce can jump-start the re-use of software systems simply by agreeing to value re-use as a factor in competitive acquisitions. Let's say Agency X wants to procure a COTS solution that must be configured and customized to meet workflow and other requirements. One vendor responds with a solution based on an open-source COTS. The software package and modifications to that package would also be released as commercial open source. The improved product could be re-used by any other federal, state or local government agency and ' perhaps even more important 'any contractor. This does not mean the re-use would be without cost, but the opportunity is clearly of some value.
The distinction between proprietary and open COTS is not well-understood and is generally not being factored into TCO. The commercial open-source model has now matured to the degree that governmental entities should fully explore the opportunity. Re-use incentives for sharing and publishing work need only be stated in request-for-proposals evaluation criteria, and the market will respond creatively and quickly. Competitive industry players, who can ill afford to inflate prices by re-writing software that already exists, will eventually adopt efforts to re-use assets funded by the government.
The aphorism 'You can't manage what you don't measure' is often too simple, but it holds true when thinking about TCO. The fact that the immense value of re-use cannot easily be measured does not mean it should not be valued at all. Re-usability should be a key decision factor. By calibrating re-use measures ' in both cost and technical terms ' acquisition professionals have the chance to transform the value proposition. Until explicit valuation of governmentwide re-use is included in acquisitions, the best value determination will remain, too often, merely an illusion.Peter Gallagher, now a partner at Unisys in the Federal Civilian group, has been working to promote the use of open source software and open standards for government since the late 1990s.