The Best-Laid Plans
How project-management software can help keep your people on schedule and on track
* * *
For the first five hours of a meeting last spring, I sat silently, observing how the executive committee of a $150-million retail chain was coping with the agenda: how to restructure the chain. Early in the meeting the chairman had defined the direction in no uncertain terms. "We must reduce retail outlets and reposition the remaining ones to concentrate on higher value-added products," he said. "The mail-order houses are eating us alive." Obviously, the problem before the committee wasn't where the company should head. Instead, it was what to do, who would do it, when to do it, and what to do when the new direction didn't work as intended.
The vice-president of personnel kept notes on flip charts taped to the walls. Everybody had a different view of what was necessary. Leases were running out in the next 90 days, and the financial people didn't want to renew them. Marketing insisted on a revised product and pricing policy before divesting anything. The chief executive officer thought that nothing should be done until the outsiders on the board concurred, and for that he needed a revised cash plan. The president was concerned that premature leaks about the restructuring might cause his best people to leave.
By the end of the five hours, I had counted more than 100 actions that had been proposed without any suggestion of what should take priority. When the note-taker was asked to summarize, he pointed out that there were very few conclusions. Then I spoke up. Did the committee members realize that they were describing an overhaul that would take at least 12 months to complete? That it might take a well-orchestrated sequence of from 2,000 to 5,000 actions to get the job done? That's how the subject of project-management software came up.
When you're faced with large projects -- such as a reorganization, new-product introductions, entry into new markets, or the launch of a profit-improvement campaign -- you can manage the job more effectively with the help of project-management software. Luckily, there are many inexpensive and easily used software packages for personal computers on the market that can make even a messy situation an exemplar of superb coordination. And while you, as head of your company, should be involved personally in the early thinking stages, have no fear that you'll be chained to your computer until the project is completed. You are the architect, not the operator.
How can you get the most out of project-management software packages? Before you sit down at your computer, you will have had to think out the sequence and priorities of what must be done. For the retail chain, we took the flip charts from the executive-committee session and identified the major milestones of the reorganization -- fixed dates when we expected to reexamine the progress of the entire project. The five key milestone dates out of a total of eight included the next board meeting, the beginning of employee briefings, the completion of the pilot program for four stores, completion of personnel reassignments, and when all store divestments would be accomplished.
The milestones help to divide a complex sequence into manageable project phases -- the stretches of time between major milestones. The setting of milestones and phases is the creative and judgmental part of the exercise, for which there's no known computer solution. The hard work -- at which the computer excels -- begins when you test whether you have enough resources and time available to complete what you have decided you want to get done.
The detail of implementing even the grandest of plans is made up of dozens or even hundreds of individual tasks. From a list of events compiled during the executive-committee meeting, we entered into the computer the descriptions of what had to be done between the first two milestones. We did this by pulling onto the screen preformed rectangles and writing a brief description of each individual task within each rectangle. Once a task rectangle was filled, we could click a mouse to reveal a hidden form on which we could describe all sorts of details about the proposed tasks. This could include names of the people assigned to the work, how much effort they would need to finish the job, what would be the minimum and maximum times needed to get the task done, and so on. When I do project planning, I usually postpone inserting the details for tasks beyond the initial major milestones. I just map out the major resource-consuming future events to get an approximate fix on the feasibility of the proposed dates.
The computer functions I have described so far can be learned by anybody in about 30 minutes. Yet the effectiveness of such simple ordering is remarkable. The retail chain's executive-committee, for example, was much more comfortable once it understood what lay ahead. The program had also set up the management structure needed to keep track of events, so there was no need for staff coordinators. The computer retained everything that was entered about each task.
The next step in the process -- deciding which task must precede or follow another -- cannot be delegated to a technician either. Defining these dependencies is the essence of project management and requires experienced judgment. On the computer, this is done by simply pointing the mouse at a task rectangle and dragging a line to the task that's to follow. When all the tasks are connected, the result looks like a spiderweb, and the weaver is you, or one of your key people.
If a task is out of sequence or if it skews resources, you can pick it up with the mouse and place it in a more suitable spot. [. . . T]asks on the critical path -- tasks that must be completed on time to meet the planned milestones -- are printed in bold boxes. The dates set by the CEO are underlined. Other dates are calculated by the computer from the logic of the task sequence and from information about resources and time requirements.
What I have described so far could be done by arranging paper cutouts. But what makes computerized project-management different is its capacity to show how individual tasks contribute to reaching a milestone. The computer recalculates all the start and finish dates for every task whenever you change the sequence. It adds up all the resources (in work hours or workdays) and accounts for the project costs (in daily, weekly, or monthly cash flow) whenever you shift resources from one task to another or discover another forgotten mission.
Most important, the computer flags situations for which staff, time, or money is inadequate. These appear instantly on the screen as optional pop-up windows, which represent a graphic view of the underlying calculations. For instance, you can call for a resource tabulation (resource capacity versus actual use versus time), a task assignment list (tasks sorted by responsibility for delivery), cash-flow tabulation (plan versus actual costs versus time), and critical path events (tasks that must be done on time to avoid delay in project completion) just by clicking the mouse. Based on this information, you can then poke and pull at the network of task relationships until your goals fit the reality of the situation.
You can also pass your floppy disk to others so they can reexamine their parts of the plan, fitting them into the complete scheme. In this way, the software becomes an effective means for communicating intentions and for negotiating the allocation of resources. In short, it creates the basis for sharing knowledge. Because the knowledge is presented in graphic form that can be rearranged, it can serve as the means for focusing group discussions from generalities to commitments. I think that for the retail chain reorganization, it was the easily revised project-management printouts, not the package's powerful accounting or mathematical features, that did the most good by promoting understanding and cooperation among the key executives.
Project-management systems prove their worth during implementation. Actual dates and resources used can be inserted into each task description. The computer can then generate a series of easily understood displays that show how progress to date affects the final schedule or the expected budget. It takes only two clicks of the mouse to display a graphic summary of the consequences of the latest status report. I particularly like the project alert feature: it flags all impossible-to-meet dates if a current task is allowed to slip its completion target.
From your point of view, formal project planning should be seen primarily as a communicating and monitoring tool. The flip charts and the inspirational talks are still necessary, but not adequate for guiding organizations through the slips, distortions, and omissions of even the best-intended aims. Laying out the sequence of milestones and tasks and setting the priorities for resource allocation can't easily be delegated. But once you have established the major milestones and resource commitments, you can safely delegate to others the tasks of fleshing out the details and maintaining the reporting system.
And what's happening at the retail chain? The CEO has bought a project-management software package. He told me that it took him only a few hours to master it. He is now laying out the general structure of events for the next two phases, which will be discussed at the next meeting of the executive committee.
* * *
Paul A. Strassmann managed the computerized information systems for General Foods, Kraft, and Xerox from 1960 to 1977. From 1977 until his retirement in 1985, he was vice-president of strategic planning at Xerox. His home office is in New Canaan, Conn.
The top-rated project-management software
I've selected the following easy-to-learn software packages for those of you who aren't entirely comfortable with newprograms.
For the Macintosh
MacProject II, 1.0: Claris Corp., 440 Clyde Ave., Mountain View CA 94043; (415) 962-8946. $495.
For IBM PCs and compatibles
Time Line: Symantec Corp., 10201
Torre Ave., Cupertino CA 95014; (408) 253-9600. $595.
Harvard Total Project Manager II, 2.0: Software Publishing Corp., 1901 Landings Drive, Mountain View CA 94039; (415) 962-8910. $595.
Microsoft Project 4.0: Microsoft Corp., 16011 NE 36th Way, Redmond WA 98073; (800) 426-9400. $495. n