You’ve busted the budget. The project is late. And it still didn’t deliver half the things it was supposed to. Sound familiar? Then you may need an information-technology portfolio management system to help housebreak that wild pack of projects.
Three years ago, Agere Systems, a communications chip manufacturer in Allentown, Pa., couldn’t even be sure how far off-schedule its information-technology projects were. Nor could the department’s managers easily figure out whether they had staff available to complete a project by a given date. Projects kicked off without much regard for their business value.
“Whoever had the best connections in the I.T. department,” recalls Chris Morris, manager of Agere’s program management office, “got things done.”
And the information-technology group was overextended: It was agreeing to take on more projects than it could realistically complete on time. “Sometimes I.T. was overly customer-service-oriented,” Morris says. “And that’s not necessarily the best thing.”
In 2002, Agere rolled out Primavera’s portfolio management system. The software lets managers create detailed project plans and assign work to staffers—whose salaries, prorated per hour, are factored into the system—as well as determine whether people with specific skills crucial to a project are available. The company requires everyone who works on a technology project to enter the actual amount of time they devoted to it, so results can be compared with the original plan.
Now Agere has a much clearer idea of when a project could be finished and how much it will cost. That’s fundamentally changed the way the company makes decisions about which projects to pursue, Morris says: “If someone finds out a new feature is going to cost $20,000, all of a sudden that ‘critical report’ isn’t so critical.”
The Primavera tool, according to Morris, has primarily paid off on big, complex projects—the biggest being an upgrade to its Oracle enterprise resource planning system, which has involved 130 full-time employees in the U.S., Singapore and Thailand. The 15-month project, scheduled to go live on Oct. 2, was budgeted at $32 million. Morris, based on data from the Primavera system, predicts the final price tag will be just under $30 million and that the system will go online on the exact day his team originally planned.
How? By and large, Morris claims, it was Agere’s ability to anticipate staffing shortfalls early enough to keep things on track and identify cost-effective alternatives; for example, outsourcing some work to an Indian software development firm. “We had complete visibility into this project from day one,” he says.
Goodness knows, most companies could improve their project management competencies. Consulting firm Standish Group’s notorious 2004 analysis of 9,236 information-technology projects found that just 29% were delivered both on time and on budget. Meanwhile, only 13% of project managers say they use portfolio management software, according to a survey by Project Management Solutions’ Center for Business Practices.
So is using such software really worth the effort? It was for Hallmark Cards, which put in PlanView’s project management system three years ago. The greeting card company recently analyzed data from completed projects, and found that on those that included a “requirements architect”—someone trained to determine what functions an application must include—21% of the project’s total hours were spent in the requirements-gathering stage, compared with 32% for projects with no requirements architect.
As a result, Hallmark has mandated that all projects with more than 1,000 hours of implementation time include a requirements architect. “The only way we were able to reach that decision was because we’re collecting metrics on discrete tasks,” says Cheryl Randle, manager of the company’s information-technology project office.
Often, the simple fact of having everyone on the same page—showing managers where information-technology spending and staff resources are committed—can foster a greater sense of accountability. “When the CIO has visibility into the projects, things get more attention much more quickly,” says Jeff McIntyre, assistant vice president of technology services at railroad company BNSF Railway.
Putting projects under the microscope sometimes yields unexpected, and ugly, revelations. When human-resources services provider Ceridian cataloged its information-technology project portfolio in 2002, “executives thought we had about 25 projects,” says Ben Stivers, director of the company’s project management office. “We actually had 200, and six were working toward the same goal.”
Information-technology executives for Oakland County, Mich., conducted a similar review in the late 1990s and discovered its staff of 160 was overcommitted by—no joke—33,000 hours for the year. The county’s technology department adopted portfolio management software from Niku, since acquired by Computer Associates, to handle project planning and execution from start to finish.
“We’ve been able to manage customer expectations with this tool,” says Janette McKenna, the county’s chief of information-technology internal services. The software has allowed her group to prioritize projects and pull the plug on less critical ones, though these days fewer than 5% of planned projects get killed. “Customers are smart enough now that they don’t propose a project that doesn’t have a payback,” she notes.
But be prepared for pushback. Both project managers and their teams may resent the idea that senior executives are hovering over their shoulders. “We don’t micromanage anyone,” says Bob Green, director of information technology at Health Net, a health maintenance organization. “But at the end of the day, we need to know: How much time did you spend on that task?”
For Mike Covalt, project support manager at Cabela’s, an outdoor sporting-goods retailer, the No. 1 challenge is: “Junk in, junk out.” What he means is that unless employees accurately enter hours worked and status of their tasks into the company’s Primavera system, the data is incomplete and, therefore, not terribly useful. “We have to explain to our folks that we’re making serious management decisions based on the data,” he says, “so it’s critical to have that data right.”
