Enough of ‘ready, fire, aim’

Colorado legislators get involved in trying to prevent IT projects that misfire and waste millions.

A series of failed information technology projects in Colorado in the past two years have generated front page headlines and wasted millions of taxpayer dollars. Resolving the debacle required high-level involvement by the state’s legislative, executive and judicial branches.

Farther west, Washington state had a better situation. IT officials there created the country’s first statewide electronic archives and largely avoided the setbacks that plague many IT projects.

Although no IT project is perfectly managed or entirely mangled, the disparate experiences of Colorado and Washington reflect many of the factors that experts say push projects toward failure or success. As states’ IT portfolios grow in volume, cost and complexity, governments are eager to understand why some projects finish on time, meet their budget and deliver promised functions — something that rarely happens — and why many more do not.

Two of every three IT projects do not meet state expectations, a statistic that translates into billions of dollars of wasted effort. Cautionary tales have increased faster than new sources of revenue in cash-strapped states, and governments are eager to adopt policies and best practices that will increase their odds of IT success.

“This is definitely a hot topic in the states,” said Doug Robinson, executive director of the National Association of State Chief Information Officers.

An assessment of state-level IT projects that NASCIO released last fall confirms that rigor is a key component of IT success. NASCIO’s survey, “Discipline Succeeds,” found value in enterprisewide IT solutions, more stringent governance structure, attention to change management and the professionalization of IT project managers.

However, in the freewheeling atmosphere that dominates much of the IT culture, rigorous oversight of projects, including risk management, has been in short supply. In comparison, oversight of engineering projects, such as in aerospace research and development has been much more rigorous.

“Lockheed Martin and Northrop Grumman have invested years to develop a higher level of maturity in program management,” said Jim Rogers, director of product and industry marketing at Primavera Systems, a project and portfolio management software company. “Historically, IT has had a culture of less rigor around project maturity and standards. They didn’t have the same tools and processes in place because of a culture of fast turnarounds and very little investment to support the IT staff.”

“With the investment in IT growing in more and more government initiatives, any hiccup in those services creates tremendous public awareness,” Rogers added.

The public is aware of Colorado’s missteps. The state’s $200 million welfare-benefits system went online two years ago and misfired right away. It cut off some welfare recipients, improperly denied food stamps and medical assistance, and overpaid others by an estimated $10 million. Advocates for the poor went to court to force the state to fix the system. Colorado spent $20 million to comply with a subsequent court order and help the state’s counties master administrative complexities. Yet, the EDS-designed system continues to fall short of the state’s expectations.

Colorado also cancelled a pair of IT contracts with Accenture that were worth about $50 million. The state determined that those contracts, for a tax-and-benefits system and a voting system necessary for compliance with the federal Help America Vote Act, were doomed, despite payments running into the tens of millions.

Colorado Gov. Bill Owens intervened and the legislature drafted a law that requires the state’s Office of IT to certify best practices of significant IT projects and verify qualifications of project managers and analysts. The oversight requirement applies to projects with budgets of more than $5 million, multiyear life spans, high risk levels or collaboration among multiple agencies.

“The legislature continuously saw difficulty with some segments of project execution,” said John Picanso, Colorado’s CIO. “It was difficult for them to keep hearing from agencies that the projects were struggling or that they needed more funding or that they needed more time.”

The legislative solution, however, gets mixed reviews among IT experts.

“The problem with the government is that they have so much oversight that it gets in the way — paralysis from analysis,” said James Johnson, chairman of the Standish Group, which conducts research on the IT industry. “Rather than more oversight and more bureaucracy,” he said, “what you need is people taking responsibility and making it happen and getting results at the end of the day.”

Kyle Young, director of operations at Sire Technologies, who handles technology projects for state and local agencies, said “the idea of passing legislation to enforce certain criteria and increases in education and experience isn’t a bad idea. Do I think it will drive the success of a project? No. It’s more important to have an IT person who can see the vision of what an agency needs.”

Peter Aiken, a professor at Virginia Commonwealth University and founder of Data Blueprint, a data management consulting firm, said Colorado’s effort to stiffen IT requirements might not be a silver bullet, but it indicates that the state is “certainly heading in the right direction.”

Aiken said large projects, such as the $200 million Colorado Benefits Management System, have a better chance of staying on track if the state uses a gated approach, which means paying contractors as they complete each phase of a project with acceptable functionality.

“It’s like building a house. Give the contractor some money after he’s poured the foundation,” Aiken said. “At the beginning of the project, when we know the least about it, we’re supposed to tell everybody how long it is going to take, how much it’s going to cost, and what functionality it’s going to deliver to them.”

Despite not fully understanding projects at the beginning of the development cycle, agencies are often in too much of a hurry to reach the finish line, experts say.

“You get a ready, fire, aim approach,” said Bill Weber, vice president of professional services at GTSI, a government IT provider. To mitigate against that tendency, Weber said, organizations are likely to embrace adoption of quality-control systems, such as earned value management, which is mandated for federal IT projects of $25 million and more.

“I think [EVM] will become a de facto standard,” Weber said. “It does require more overhead and more participation on the part of the customer, but it creates much more benefit down the road. We see the states starting to adopt it as well.”

Pulley is a freelance writer based in Arlington, Va.

Getting a handle on IT projectsStudies show that most information technology projects fail to reach their goals. Some are abandoned, but even those that reach completion are often late or over budget — or both. Perhaps even worse, users ignore most of the functions incorporated into successful IT projects. A weak process and lack of oversight contribute to many problems, experts say. Federal Computer Week has compiled the following list of steps managers can take to minimize the risk of failure.

  • Put the necessary resources into understanding a project’s requirements before you start so you can avoid the ready, fire, aim syndrome.

  • Involve users from the early stages of a project, but don’t give them carte blanche to determine functionality.

  • Identify projects’ risk factors, monitor them with vigilance, and rely on independent verification and validation of progress.

  • Select the most qualified vendor for a project.

  • Use performance-based contracts and quality-control systems such as earned value management.

  • Break large projects into manageable segments, and insist on regular progress reports from project partners.

  • Eschew mostly linear development processes in favor of more iterative models, such as spiral development.

  • Expend resources to train and certify your organization’s project managers.

  • Designate a project champion among top managers who can shepherd an initiative through difficult phases.

  • Don’t let politics trump good IT practices.
How Cheney, Wash., achieved success

Eager to see a digital archives project done right, visitors from China, Australia and Singapore have been drawn to a small town in Washington state. They have come from the British Library, the U.S. Government Printing Office, and half a dozen state archives.

They visit Cheney, just outside Spokane, to learn more about the Washington State Digital Archives, the first statewide repository for storing electronic records in perpetuity. They come to see an information technology rarity: a challenging project that delivered what it promised, on time and under budget.

The challenge was to build a system for preserving records from various sources, including the state’s 39 counties, that had different ways of creating electronic records and making those documents easily accessible to users anytime and anywhere.

“There was no business model for doing this,” said Adam Jansen, the state’s digital archivist. “A lot of it was learning on the job.” A few critical decisions proved invaluable to the project’s success, he added.

First, the project had a cross-functional team that included an executive sponsor — the assistant secretary of state, who is a lawyer and a technologist. Steven Excell was the proverbial project champion, without whom the project might have failed, Jansen said.

In addition, caution was a constant watchword. Realizing that it is difficult to form fixed requirements when creating something new, the team produced a 160-page feasibility study before moving forward.

Rather than attempt to create a system from scratch, the team modified commercial products, adopting newer technologies, such as a storage-area network, but avoiding untested products.

To ensure that its partners were well-qualified, the team issued detailed requests for proposals, “telling vendors exactly what we needed as opposed to asking vendors what they had,” said Bill Peldzus, director of storage architecture at GlassHouse Technologies, a consulting firm that assisted with the project. “It put us in better control of the process.”

Every Friday for two and a half years, the team met for three hours to review potential risks and devise strategies for mitigating them. “The higher risks were actively monitored, managed and reported on throughout the project,” Peldzus said.

The project also adopted a “documented, formalized change-control process,” Jansen said. By implementing the project in phases, the team further ensured that it didn’t spin out of control. Initially, the project archived a single type of record. The team later added more records and functions, a process that continues today.

“We didn’t shoot for the moon on the first try,” Jansen said. “Failure was not an option.”

Since going online two years ago, the archive has served 600 researchers daily.