The Big Fix

Government Leader home > June 2005 issue

June 2005; Vol. 1 No. 2

The Big Fix

By Richard W. Walker

Another story on this topic is Six Ways to Turn a Project Around.

Big government can’t avoid big projects, but how do you keep them on track? What leaders have learned from The BIG FIX.

In mid-2001, the Defense Department’s Standard Procurement System project was pretty much a mess. The half-billion-dollar, multiyear program was way over budget. It was more three years behind schedule.

The Government Accountability Office ripped the project in a report, saying the department had failed to justify further investment in SPS. GAO said the department had taken a “a monolithic approach” to the project, failing to apply incremental investment decisions to SPS. Indeed, GAO concluded that SPS lacked a formal business case.

But just three years later, SPS is back on track. It’s the largest standard business system in the Defense Department, and procurement officials at more than 800 sites around the world use it to buy goods and services for warfighters.

SPS’ stunning turnaround underscores how important skilled program management is to the success of big, complex projects.

To revive SPS, DOD brought in Col. Jacob Haynes, a program manager with years of program and project leadership experience and a reputation for reviving moribund projects.



Program managers really should be program leaders. — Col. Jacob Haynes

Haynes took the bull by the horns. He reorganized the SPS, breaking software development into manageable deliverables and implementing a new requirements process to evaluate user needs and establish a clear baseline for the program. He also took the scathing GAO report as constructive criticism and used it to the program’s advantage by implementing GAO’s recommendations.

“Program managers really should be program leaders, because I think leadership is the paramount attribute [necessary] for them to be successful,” Haynes said.

Big projects are nothing new in government, but they have become more numerous as agencies move increasingly toward collaboration. The Army alone has no fewer than 17 current projects valued at more than $1 billion—some of them significantly more, according to Input Inc.

Getting a big project on track and keeping it there depends on taking control of its components.

For Randolph Hite, director of IT architecture systems issues for GAO, the human capital component of big-project performance can’t be underestimated.

“You have to make sure you have the right people on the program to do the job and execute the processes that you put in place for the program,” Hite said. “They have to be qualified and well-trained people. And that’s not just the program and project managers; it’s all of the supporting positions, too.”

Leadership skills, such as those demonstrated by Haynes, aren’t the only qualification for program executives. It’s clear that experience in overseeing large projects is also crucial.

“Once you get over $750 miilion and you get into the billions, it’s a whole different set of skills,” said Stephen Hawald, executive consultant for process improvement management at Robbins-Gioia LLC of Alexandria, Va.

That was the case at the Internal Revenue Service, where new leadership at the program management level has helped one of the government’s most conspicuously troubled programs, the Business Systems Modernization project, move into calmer waters.

Seeking experience. Two years ago, IRS officials began hiring executives with extensive private-sector experience in running large IT projects to augment its BSM management staff.

In the past, BSM components were mostly run by IRS managers whose only experience was in tax administration, said IRS CIO Todd Grams.

“In hindsight, that wasn’t fair to them as individuals, and it wasn’t fair to the agency, because they weren’t versed in the skills of running large IT programs,” he said.

The IRS now has five managers from industry and three IRS careerists on the program, creating a solid blend of skills and experience, Grams said.

“We’ve got folks from outside who know this business,” he said. “These are people who have worked on large, successful IT projects in the private sector. When they are running these projects and sit down with contractors, they speak the same language.

“But we also have IRS executives in the program who know tax administration,” he added. “They know our legacy systems and know how to get things done in the IRS. You put those two things together, and you really come up with a powerful team.”

But big, long-term programs have to be sturdy enough to survive disruptions in the leadership ranks, such as the departure of top managers, Hite said.

“The program can’t be dependent on the heroics of a few people,” he said. “It has to have stability in and of itself. There have to be decision- making protocols and a governance structure in place.”

Institutionalizing effective management procedures and processes mitigates the impact of executive turnover—which is inevitable as administrations and agency priorities change, and as managers retire or leave for other work.

“Managing turnover goes back to program management,” said Louis Samenfink, executive director of the Office of Modernization for the Customs and Border Protection’s $1.3 billion Automated Commercial Environment (ACE) project, which will digitize the tracking of U.S. imports and contribute to border security.

“A well-run program that has a documented history on what happened and why decisions were made helps new managers get engaged,” he said.

It also bolsters continuity to have a solid, knowledgeable program management staff, said Samenfink, a 26-year Customs veteran who was named ACE executive director last October.

“Having a program management staff that can get up to speed quickly really helps,” he said.

A large program also has to be limber enough to adapt to changes in priorities, said Rob MacDonald, acting CIO and acting assistant commissioner for CBP’s Office of Information Technology.

The CBP is a powerful case in point. After 9/11, the agency was moved from the Treasury Department to the new Homeland Security Department and given border security as an additional mission.

As a result, the ACE project had to be revised to meet new mission needs.

“We rebaselined our program last year to insert new sets of software focused on screening and targeting to include that border security mission,” MacDonald said.

The ACE case underlines the difference between adding requirements and changing requirements.

“Changing requirements and adding requirements are two different things,” Haynes said. “Adding requirements is normal. But if you’re changing stated requirements that have already been scoped or even coded, then you’ve got problems.”

Negative example. The FBI’s collapsed Virtual Case File project, which recently ground to a halt after five years and $104 million in expenditures, is an example. One of the factors that contributed to the project’s demise was the agency’s incessant requests for requirements changes.

Over one 18-month period, the FBI saddled its contractor, Science Applications International Corp. of San Diego, with nearly 400 requirements changes.

GAO’s Hite said that some changes in requirements for projects that run for years are inevitable, but the goal should be to keep those changes under control.

“You have to have a rigorous and disciplined management process in place that is going to control, not prevent, changes in requirements,” he said. “Changes in requirements will happen, but you want to control the extent that it happens and make your decisions based on the cost, benefit and risk of those changes.”

“If your change requests are starting to increase and have a financial or [schedule] impact, then you need to take a look and see what’s wrong,” added Robbins-Gioia’s Stephen Hawald, who is a consultant to Customs and Border Protection on the ACE project.

In long-term projects, establishing a stable baseline of requirements from the outset is integral to keeping a project on course. That starts with input and buy-in from the ultimate users of the system.

“You have to have the operational folks at the table,” Samenfink said. “We try to make sure that we have experienced field managers out there who really understand the needs of [users], the people who are going sit there and look at the screens.”

Requirements into specs. Officials at the Office of Management and Budget say that the business and IT managers need to work together through an integrated project team to identify business drivers and then turn the resulting requirements into technical specifications.

“The IT project team has to have a way to talk to the business people to translate their requirements into technical specifications that the IT team can program,” said Karen Evans, OMB’s administrator for electronic government and IT.

Keeping requirements aligned with program objectives is also critical to keeping a large project headed in the right direction.

To be sure, agency executives have to keep the project’s objectives in their sights from the beginning and not let those objectives go out of focus, which can happen in the case of big projects that take years to complete.

“You have to have clarity of outcome,” Evans said. “Once the project commences, the clarity of outcome [tends to get] a little watered down so you have to constantly come back and ask the question: What result are you trying to achieve?”

Objectives are easier to meet if a large project is broken into manageable pieces with short-term outcomes.

That’s what the IRS didn’t do early on in its project.

“We bit off more than we could chew,” Grams said about the modernization’s project’s problems. “Everybody who looked at the program thought that. The individual sizes of our projects are large; they are in a complicated environment, and we were doing too many of them.”

The modular approach to major IT acquisitions was formally espoused in the Clinger- Cohen Act of 1996, which stated that, to the maximum extent practicable, large systems should be acquired successively in interoperable increments.

“If you look at the original intent of Clinger- Cohen and all of the the OMB guidance, we have always advocated a modular approach,” Evans said. “It’s designing a small piece of the project so that it gets you to a specific outcome. You bite off small enough chunks of the project so that you get the biggest value going forward. That’s how good system development work gets done.”

Added David McClure, a director in Gartner Inc.’s global public sector research group and former director of IT management issues for GAO: “By their very nature, modernization projects are long-term and multiyear. They have to be broken into investment segments than can show a payoff in shorter-term cycles and in very measurable ways.”

A large program’s success can hinge on modular design, strong program and requirements management, and unwavering attention to final goals. But in the end, each large project is different in scope and complexity, and has its own unique challenges. While some basic principles apply, there is no one-size-fits-all method for either getting a project started on the right foot or getting a problematic project turned around.

“There a thousand stories out there in the naked city, and for one of these big projects to be successful, you’re really talking about, literally, a thousand things coming together,” said Hite, who has been monitoring big projects for 20 years.

The key to managing a large project? Be realistic.

“When you have to orchestrate a thousand things, not all of them are going to be done well, and not all them are going to go right,” he said. “So part of this is establishing realistic expectations for big projects.”

The Billion-Dollar Club
Number of contracts valued at $1 billion and up, by agency:

Office of the Secretary of Defense 19
Army 17
Navy 13
Air Force 9
Transportation 9
Homeland Security 9
Health and Human Services 6
NASA 5
Energy 4
Postal Service 3
Treasury 3
Commerce 2
Veterans Affairs 2
Agriculture 1
Education 1
Interior 1
Justice 1
State 1

Source: Input Inc.

Another story on this topic is Six Ways to Turn a Project Around.