The Best-Laid Plans

here's a long-standing belief that well-made plans can nearly assure project success. A finely crafted plan is supposed to include a clearly stated project summary spelling out objectives, assumptions and risks. It details project milestones and includes charts breaking out project tasks and showing how each task relates to the others. It spells out resource needs, budget details, operating procedures, assessment and review procedures, and the organization of the project and project team. With such a plan in hand, the project team is supposed to be able to move smoothly from initiating the effort through executing, monitoring, adjusting, controlling, closing and evaluating it.

Government's first foray into systematic project planning came with the Navy's Polaris missile program in the late 1950s. Members of the Polaris staff devised a method called the Program Evaluation and Review Technique (PERT) to track project progress, the accuracy of the plan and schedule and the effects of proposed changes. PERT users fashioned network diagrams showing all the parts of a project and whether or how they depended on other parts.

PERT soon was paired with the critical path method (CPM), a technique for improving project performance by identifying the project tasks that, if delayed, were guaranteed to disrupt the whole effort. In the 1960s, DoD adopted the Cost/Schedule Control Systems Criteria (C/SCS) to get weapons system contractors to use effective cost and schedule control systems to produce data the government could use to monitor the status of contracts.

As with so many management techniques, each of these systems rapidly degenerated into diagramming and chart-making exercises. They came to be ruled by a subculture of nitpicking zealots speaking acronym-laden language that was virtually untranslatable by program managers who were supposed to be using the information generated by the techniques. A number of useful tools-charts to help break down projects into their constituent tasks, network drawings showing which tasks depended on the completion of others, schedules showing how tasks overlapped-were damned in the process.

For example, work breakdown charts for large projects or programs can be imposing. But the charts help project managers assign tasks, build budgets and sketch out schedules. Breaking down projects into tasks isn't much help in scheduling, however, because work breakdowns don't show the sequencing relationships of the tasks involved. For that, project managers use network diagrams. The diagrams show the connections between tasks and which ones depend on others. Gantt charts further develop the picture of the project in time by showing when each task begins and ends, which tasks overlap, and which tasks have "float"-flexible start and finish dates-and which don't.

Other charts have grown up over time, as well. For example, the Software Program Managers Network, funded directly by Congress to assist the military services with large software projects, has developed a computer tool for program managers. The "Control Panel for Program Managers" displays project status on a cockpit-like panel of gauges, bar charts and fever charts.

Of course, charts and diagrams can't guarantee projects will flow smoothly. As it turns out, good project planning is as much about getting a handle on the real scope of a project-what it is that sponsors and customers really want and what they are willing to settle for-as it is about figuring out what's needed to get work accomplished. After all, the best-laid plans will run aground when the project changes shape after it has begun.

The art of project management is adjusting to uncertainty. In his study of successful projects in industry for his book Simultaneous Management: Managing Projects in a Complex Environment, Alex Laufer found that most managers actually begin their projects with significant uncertainty about the end product. Instead of foolproof plans, they use a "plan a little, do a little, learn a little" approach, postponing planning for uncertain tasks, speedily accomplishing more certain ones, and using lessons from those to apply to the next set. Planning continues throughout successful projects, Laufer found.

The incremental approach also can help solve project managers' No. 1 problem: changes in sponsors' and customers' requirements. "The idea that the customer should clearly know, in total and final detail, what he or she wants before briefing the designers is expecting too much," Laufer writes. "Insight into possible solutions and available alternatives influences the customer's ideas of what he or she really wants." Learning from the products of incremental planning provides just such insight.

'Begging, Borrowing and Stealing'

Successful government project managers are strongly resistant to formulaic approaches. "You have to get away from the mechanics and the prescriptive things," says Air Force program manager Terry Little. "If work breakdown structures and PERT and the other tools worked so well, we wouldn't have had so many catastrophes."

Good program managers often sneak around the system, says John Gioia, a former Air Force program manager and now president of Alexandria, Va.-based program management services firm Robbins-Gioia. For example, to hedge against uncertainty, they hide pots of money and build in extra time. "You take part of the program and overcost it and bank that amount," Gioia says. "In other words, if it's going to cost $100 million you add $50 million and the expectation is based on a $150 million cost. You tell your senior manager, but not the resource supplier, especially about the schedule reserve. If not, for every change you'll have to say you need more money and time."

Planning tools can be useful, but most program managers say they learned the real tricks of the trade by watching those who went before them. "Most of what I've learned that has been useful I learned from my supervisors and peers," Little says. "I incorporated the good things I saw and the not so good I incorporated as a red flag in my brain. There's not a course to go to in begging, borrowing and stealing 101."