Can grand design be driven to extinction?

Costly, massive IT projects that aim for sweeping reinvention of agency computer systems and business processes — the so-called grand-design approach — found themselves in the Obama administration’s crosshairs this year.

The administration’s first target was fixing projects already in the pipeline. Office of Management and Budget officials worked with agency leaders to pull the plug or recast four projects that suffered from the all-too-common afflictions of being over budget, behind schedule or otherwise seriously off-course. The officials also scrutinized and then tweaked and accelerated several more projects.

Now the administration is considering structural changes in how future programs will be funded, staffed and managed. The President’s Management Council has promised the details on that reform plan in the next six months.

But observers say there are two critical elements that will make or break the effort to end the grand-design era: the ability to embrace agile development techniques and the creation of a well-trained acquisition and project management corps to oversee the new rapid delivery style. There might be reason for optimism about both objectives.

The idea of using agile development as an antidote to grand design is hardly new. Nearly 20 years ago, the General Services Administration advocated that government avoid giant, multiyear IT modernization projects and instead deliver new systems in small chunks and solicit user feedback to identify problems early and facilitate frequent course corrections.

That advice, certainly ahead of its time, went unheeded for numerous reasons, including the inertia of government culture and the long learning curve — for government and industry — of switching to a new development paradigm.

“Some people thought that if you gave up the giant monolithic development process, you lose easy program accountability, but that old approach doesn’t help improve outcomes; it just makes it easier to audit these projects,” said Thom Rubel, vice president of research at IDC Government Insights.

But attitudes are changing. Rubel said that in recent years, government has been moving toward smaller, more discrete projects for two reasons: Tighter IT budgets demand less ambitious plans, and agencies are focusing on software and services to improve business operations now that big investments in hardware and basic infrastructure are behind them.

Another shift that will likely facilitate agencies’ efforts to embrace agile development is the government's hiring of more young people whose only training and experience is with agile development, said Steve Kelman, a professor at Harvard University's Kennedy School of Government and a Federal Computer Week columnist.

The Obama administration’s IT reform plan is expected to include the creation of a special career path for program managers and acquisition professionals who will be trained in agile development skills. Rubel said agencies will have a decent chance of retaining those valuable workers given that government is bringing a lot of IT work back in-house after years of outsourcing and because job security in the private sector will remain shaky.

He said he expects that in the next few years, 30 percent of all government IT projects will use agile principles. For its part, the Veterans Affairs Department, whose effort to modernize its accounting system was one of the faltering projects shut down last summer, is making the adoption of agile development techniques a priority in 2011, according to its annual management report. Other agencies are following suit.

Finally, going agile doesn’t mean agencies should abandon big ideas and bold visions.

“You can keep the grand design, but you look at [how to get that] elephant cut into bite-sized pieces,” said Larry Albert, president of the health care division at Agilex Technologies, which specializes in agile development.

Read more of 6 ideas with some 2011 bounce or return to the 2011 Federal List.

2014 Rising Star Awards

Help us find the next generation of leaders in federal IT.

Reader comments

Mon, Dec 13, 2010 TG

The problem with "Grand Design" systems is with the culture change the new system brings, and the layers of bureaucracy. Technically, most are not hard to design and build; it's the politics and culture that gets in the way.

Mon, Dec 13, 2010 CJ

'bout time actually - let's see if it really happens. One big risk to an agile approach is cuting corners on the big picture architecture - the destination you're steering toward if you want to stay with Beck's analogy. The other is of course that wasteful spending still ends up in somebody's pocket - there'll be pushback to any approach that doesn't keep the right trough full.

Mon, Dec 13, 2010


“Grand design” déjà vu. It seems to come around every eight to ten years or so. My time with IRS’ Tax System Modernization back in the 90’s exposed me to the folly of Grand Design. Almost always resulting in failure, the information technology is almost always blamed as the failure usual culprit. One of the ballyhood fixes of the past was to acquire system modules using a modular acquisition approach. It can work, but didn’t. The real problem is the assigned people and culture. When IRS was about to award a huge contract, it was met with dismay rather that considered a success. Curious after years of work.
To give these huge efforts a better chance for success, they must be treated as acquisitions (not “procurements” or “contracts”). DoD is often criticized for its major systems development methods, but something like it (with reduced baggage) is an essential. To make that work, dedicated, supported, and accountable (and don’t kill the messenger) program managers are the key. Eben Townes, Federal Strategies, LLC

Please post your comments here. Comments are moderated, so they may not appear immediately after submitting. We will not post comments that we consider abusive or off-topic.

Please type the letters/numbers you see above