Calculated Risks

Pity the poor CIOs. It falls to them to wrestle back into reality the overblown promises and overrun budgets of government information technology. At its heart, the job is calculating risks. CIOs are expected to guide agency heads in choosing IT investments. CIOs must whittle down the list of projects clamoring for funds to a hardy handful likely to deliver the most measurable bang for the buck. To do it, CIOs will have to set up a stringent system for proposal-winnowing. But even in the fabled private sector, where the bottom line is a yardstick for just about anything, calculating the potential value of technology is less a science than an art.

Until recently, most studies showed no connection at all between business spending on technology and productivity or profitability. For example, Paul Strassman, former Defense Department director of information, has collected reams of data showing no relation between computer budgets and profits. Strassman now is willing to concede IT can add value, but only in well-managed firms. And in the last couple years, research has shown IT provides a business edge, but not exactly why, and the experts can't say for sure which technologies should be applied where and to what degree.

With so much uncertainty in business, what are federal CIOs to do? Fortunately they have help in the form of capital planning, a process mandated by the 1996 Information Technology Management Reform Act. The process is outlined here based on the recommendations of the governmentwide CIO Working Group.

IT Toll Booths

Capital planning sets up toll booths along the road to IT investments. CIOs must regularly check in with solid plans and data in order to get senior executives' stamp of approval to proceed with investments. Capital planning is a continuous loop of activities that begin and end with an agency's overall strategic plan. The key is measuring and comparing the costs, benefits, risks and returns of proposed IT investments. The goal is to transform the agency's IT investment board into a team of portfolio managers constantly shuffling IT and other investments competing for scarce funds. The board is to approve projects based on their potential costs and payoffs.

To come up with a range of IT investment proposals, CIOs first must know the ag-ency mission and goals, then inventory existing IT systems and applications. CIOs also must assess why current systems exist and whether and how well they interact. Then comes the heavy lifting. Only the IT proposals most likely to improve mission-related processes should reach the top team, so a technical review team must thin out low-potential projects. To do that, the technical team weighs:

  • Past successes or failures with similar projects;
  • The potential harm to customer service and/or program performance should a project not proceed;
  • Whether the intended users agree with the estimates of system benefits;
  • The degree to which an investment makes future projects possible;
  • Whether the system is likely to deliver promised long-term benefits;
  • How well projected costs reflect declining software and hardware prices;
  • Whether the proposal improves the quality, cost, speed, accuracy or productivity of programs.

When the technical team is finished, the investment board will have a list of top IT contenders to rate according to mission relevance; design feasibility; use of commercial, off-the-shelf components; incorporation of other agencies' experiences; conformance to the agency's IT standards; and the presence of well-defined stages and decision points.

Finally, the executive review team selects IT investments, mixing improvements with maintenance to ongoing systems, new with ongoing projects, and high-risk with low-risk proposals. Each winner must have a review schedule, a plan for reducing risks and a set of performance measures. Acquisition plans should include modular, short-term procurements to be evaluated annually.

Stay up-to-date with federal news alerts and analysis — Sign up for GovExec's email newsletters.
Close [ x ] More from GovExec