The silent killer of innovation

Technology changes fast, while the federal government struggles—and usually fails—to update its procurement policies to match. It's not a new problem, and it continues to be an ever-present challenge.

If the way feds procure IT – spending some $80 billion each year, 80 percent of it to maintain legacy technology and old investments – does not change, industry and government acquisition experts fear a decline in the development of new IT solutions and innovations that could otherwise improve government.

"Procurement is the silent killer of innovation," said Michael Byrne, geographic information officer at the Federal Communications Commission, in what was certainly the line of the day at the Aug. 22 Lowering the Cost of Government with IT Summit in Washington, D.C.

Byrne's line summed up in one sentence the arguments made by several speakers, including the day's keynote given by the White House's top contracting policy executive Joseph Jordan, to promote strategic sourcing initiatives and push for agile development over "waterfall" approaches to IT projects when applicable.

Agile development – breaking up larger projects into modular or incremental pieces – has been used successfully in a number of federal agencies already, including the United States Postal Service. The goal, Jordan said, should be for agencies to get away from large contracts locked in over several years in favor of a modular approach in which an agency works closely with its vendor partners on a series of shorter projects with closely defined goals and metrics to measure progress by.

Jordan said this approach can help agencies avoid expensive projects that go nowhere, and there is plenty of room for the government's 36,000 contracting officers and total acquisition workforce of 200,000 people to improve. Over the past decade, the federal government has blown at least $9.2 billion on failed IT investments, according to the Government Accountability Office.

"Too frequently, we try to throw the Hail Mary pass in every IT development," Jordan said, using a football analogy. "But we've only got a few Tom Bradys running these things, so let's make the process work."

Yet the savings are there to be had.

The Department of Homeland Security, for example, used strategic sourcing to save $200 million on office supplies and shipping in fiscal 2010 alone, according to Dr. Michael Valivullah, chief technology officer at the Department of Agriculture's National Agricultural Statistics Service.

Jordan called on feds to take more risks in IT acquisition, part of a "cultural mindset shift" he hopes feds will adopt. Fearful of retribution from oversight bodies like GAO or inspectors general, Jordan said, procurement officials take the "lowest risk" solution instead of the one that has the biggest benefit to the agency's mission and taxpayers.

If federal acquisition regulations do not expressly forbid a decision and if it "comports with sound business judgment, you can do it," Jordan said.

Projects will still bottom out sometimes, and some failures might even merit a dreaded explanatory hearing before Congress or an unsightly audit, Jordan said, but if the project was based properly on analytics and a thought-out path, "we'll learn and do better the next time." He argued that failing fast and small is better than one huge, drawn-out failure.

"I get the oversight culture, but we've got to get past that, and got to get a focus on what is the best outcome for the agency mission and taxpayers," Jordan said.

Acquisition officials stressed the government is at a crossroads in dealing with how it purchases IT.

Big jumps in mission performance come from innovation, but innovation often requires investment dollars that most agencies do not have.

Keith Trippie, executive director for the Enterprise System Development Office at DHS, said if agencies want faster projects, they may have to accept a tradeoff in terms of a project's completeness. Looking toward the future at DHS, Trippie said nobody wants to spend 18 months procuring technology that might be obsolete by completion, but is rolling out an incomplete project faster actually a better option?

"Are we willing to do tradeoffs, where an 80-percent solution quickly is better than a 100-percent solution that takes 2-3 years? Are we willing to?" Trippie said. "That's a fundamental question. We [feds] need help on the business side."

Mark Day, acting deputy assistant commissioner of Integrated Technology Services at GSA, which oversees a large portion of federal IT investment, said his department is forecasting possible IT procurement shifts years in advance to stay ahead of the curve in acquisition.

"We're looking at where the puck is going to be in 2-3 years, not where it is today," Day said. "We all know acquisition lead times are less than desirable, they are not what we'd like in in order to be real-time."