COMMENTARY

6 keys to avoiding another 'grand design' failure

Agency executives must learn to resist the temptation of developing sweeping modernization programs

Consultant Frank A. McDonough is former deputy associate administrator of the General Services Administration’s Office of Intergovernmental Solutions.

Grand designs have been around since the Great Pyramid of Giza, and modern designs involving computer-based information systems have been around since the 1960s — and have failed regularly.

Each new generation of government managers hears the mythical Sirens calling them to the rocky shoals of grand designs. It’s hard to resist the temptation to develop programs that push to the edge of emerging technology while aggregating all possible requirements, including those that are barely understood.

Sometimes the government gets value for its money, sometimes it does not. Although numerous midlevel and low-dollar contracts provide good results, many programs at the grand-design level fail with hundreds of millions, even billions, of taxpayer dollars wasted. If you doubt me, try searching for “failed federal modernization programs” and skim the first 100 hits.

Asking why major systems fail with regularity is like asking why there are no more .400 hitters in Major League Baseball. Success is simply too difficult.

You could write a book about the Homeland Security Department’s SBInet, the Federal Aviation Administration’s efforts to modernize air traffic control systems, the Internal Revenue Service’s tax systems modernization project and other failed programs from recent decades. In fact, former IRS commissioner Charles Rossotti did just that when he wrote “Many Unhappy Returns: One Man’s Quest to Turn Around the Most Unpopular Organization in America.”

The government equivalent of a Razzie award for the worst performance in the past 30 years or so must go to the U.S. Patent and Trademark Office, although many others are in the running. In 1982, USPTO began an ambitious program to bring better, more modern tools to patent examiners. Yet in 2009, Director David Kappos said the agency had a backlog of more than 770,000 patent applications, “long waiting periods for patent review, information systems regarded as outdated and an application process in need of reform.”

Currently, Vivek Kundra, the federal CIO, is addressing troubled major systems, as is Roger Baker at the Veterans Affairs Department. But other people with a lot more power have made similar attempts in the past, including Al Gore when he was vice president and Franklin Raines when he was director of the Office of Management and Budget. Neither of them had much success.

My former organization at the General Services Administration had an effect for a few years when we published the popular report, “An Evaluation of the Grand Design Approach to Developing Computer Based Application Systems” (now available at www.frankamcdonough.com on the “Grand Design report” page).

Back in February, Kundra said, “We’re going to review investments and take decisive action so we can terminate projects that are not yielding dividends for the American people, turn around projects that can be turned around, and halt those where we believe that there’s serious issues.”

Those words could have come from a dozen or so senior government officials in the past 50 years. Will the results be different this time? The measure is how many systems Kundra cancels.

Meanwhile, here are six guidelines that could help government managers avoid ending up with a turnaround project.

  • Listen to contractors, but do not make them your only source of information. They can propose outlandish solutions to your outlandish ideas.
  • Avoid the certain failure of a grand design. Instead, insist on an incremental approach.
  • Use technologies readily available through the GSA schedules and governmentwide acquisition contracts and steer clear of beta products and other emerging technologies.
  • Recognize that government systems are huge, complex and one of a kind. If you managed an inventory system or a social networking project in a previous job, you are not ready for the grand design league.
  • Employ people on your immediate staff who understand complex systems development. Be sure they have about 10 years of true hands-on systems management experience.
  • If you choose to ignore these suggestions, polish up your résumé. You will be in a “special assistant-nothing job” or moving to the private sector in a few years.

The 2014 Federal 100

FCW is very pleased to profile the women and men who make up this year's Fed 100. 

Reader comments

Wed, Nov 3, 2010 David Virginia

Best commentary I've ever read on FCW. As a Federal IT program manager on a major system, I agree with every point he makes. I'd add a few more, but his are the top ones. He also could have given rules of thumb to tell if you're doing grand design. Mine are: Rule of thumb #1: if you're not delivering significant customer functionality in your first 6 months, then you're too grand. Rule #2: Never pay for COTS licenses or infrastructure you don't need right now. Corollary: if your business sponsorship is so weak that you cannot follow rule #2, then you're going to fail anyway, so the rule doubly applies.

Sat, Oct 30, 2010 Edmond Hennessy United States

Understand that the article is biased towards the IT challenge, however for every example Mr. Mcdonough cites that represent failings of major IT Initiatives, there are many, rich examples of successes on the Defense & Military (Programs) side of the business. The nature of the design challenge and scope of development and deployment are somewhat different from IT, however it would seem that the the design methodologies/processes and lessons learned could be useful to the IT'ers. Is there any cross-agency talk, these days?

Wed, Oct 27, 2010 Rodney Texas

Albert Einstein once said "The definition of insanity is doing the same thing over and over again and expecting different results". Think about this quote for a second and ask yourself, does this quote apply to the way the Government runs IT projects? Stop using the same young architects, stop trying to use the latest technology and start listening to the gray-haired folks (if you can find them) who wrote the old systems regarding how they should best be modernized. This seems to be a theme in this thread - maybe somebody out there in a decision making capacity is paying attention - but I doubt it! Rodney W Texas

Wed, Oct 27, 2010 DToad101

Number 7 could be, "Don't use / rely on the same 'experts' or 'leaders' who were the experts or leaders of the previous failures." Totally new blood may help immensely! But using the advisors every time gets the same poor results. They usually don’t learn from their past mistakes after making them three or 4 times already.

Tue, Oct 26, 2010 Alexandria Va

Frank has a good handle on large "red" projects, however he also hasn't been involved much in the past 10 years regarding better planning for the Government, namely the Clinger Cohen induces discipline of Enterprise Architecture being institutionalized everywhere. Most of these large projects fail because of a lack of up front planning and architecting. It sometimes seems we just don't learn our lesson over decades and what Vivek K is doing seems right but is politically expedient and, as said in the article, was attempted similarly over the past 50 years without real results. Here is a novel idea: Look at the existing Enterprise Architectures of those programs required by law and you will find out that they either were not done, were done only to satisfy OMB or were done badly by Government and contractors who learned on the job instead of putting eral professionals such as Certified Enterprise Architects on the job and thereby radically reduce their risk of failure.

Show All Comments

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