When asked why projects overrun schedules, people usually say the proj- ects start out fine, but somewhere along the way a snag develops that begins to push one or more deliverables later and later. Everyone knows that it takes only one task to be late on the critical path to make the whole project late. As the shift begins to hit the plan, management tries to solve the problem causing the shift, usually diverting resources and making changes in the project plan to cost more and more emphasis on the part of the plan that is slipping. The people working on the snag usually feel a lot of pressure to get their part of the project solved and therefore put in a lot of extra time and feel considerable stress. These are often the resources in most demand in the company, so putting more time on the project in trouble leads them to neglect the other projects they were supposed to be working on, causing other projects to slip as well.
When asked why that happens, people respond with two general types of answers. One type of answer focuses on the specific problem with the specific project that is most recent in memory (often still in trouble).
They usually blame it on poor performance by the group responsible for that part of the project. The second type of response is more general, blaming the problem on the tendency of stereotype task performers to underestimate or on management’s setting arbitrary completion dates.
How often do people complete activities and pass their work on early?
How often do they complete activities for less than the budgeted activity cost? You might find that occurs less frequently than you would expect, if the estimates were truly 90% probable estimates. Even with skewed dis- tributions, tasks should complete early a substantial percentage of the time. Figure 3.4 illustrates typical results for actual times that project tasks, compared to their planned duration. Figure 3.4 shows that most tasks complete exactly on the due date; often, as many as 80% complete on the due date. That is not consistent with the task completion time estimate presented earlier.
Potential causes for little positive variation in activity duration or cost include the following.
◗ People work diligently to milestone dates and do not understand a desire to have the work completed early.
◗ Estimates are much less probable than were believed, leaving little potential for positive variations.
◗ The work expands to fill all available time and budget.
0 1 2 3 4 5 6
0 0.5 1 1.5 2 2.5
Actual activity time Scheduled time
Numberofresults
Figure 3.4 The distribution of actual task completion time differs from the estimate distribution and shows a remarkable percentage of completions right on the due date.
◗ Belief that the next activity would not be ready to use it anyway.
◗ In most organizations, there are significant penalties (threatened or real) for completing activities late. But what are the “rewards”
for completing activities early? Have you ever seen any of the following:
◗ Management gives the product a more detailed review than they would if you completed it near the due date and forces you to make unnecessary changes.
◗ The reduced budget for the performing organization leads to higher overhead rates and, in extreme cases, downsizing.
◗ Reduced credibility in the performer’s activity duration and cost estimates leads to increased pressure to reduce estimates.
Those factors add to the psychological reasons that cause projects to lose much of the potential positive variance. Project managers assign tasks and train people to respond to specific milestone dates. Thus, even if they are done early, they might hold on to the product until the due date.
Why not? Management usually does not take any advantage of early completion or reward the task performers if they do deliver early. If the resource performing the work is paid in accordance with the time they spend on the task, they are incentivized to use up all the resources author- ized. If your project uses a cost reimbursement contract with them (the usual practice for resources in the company and for certain types of exter- nal resources), they may even be incentivized to slow down the work to get overtime pay or more total revenue from the project.
If one resource gets its activity done early, what is the chance that the next critical resource down the line is ready to hop to and start working on its activity? If it is a critical resource, it is in demand and has limits on availability. It does not seem likely that the next resource will be able to work on the activity until the date they had planned for it. Therefore, the positive variance is lost and wait time is introduced. That means that the actual schedule time grows due to activity dependence.
All of that leads to the second conflict, illustrated in Figure 3.5. The upper path refers to the performing resource. To be a successful team member, I must contribute to early completion of the project. To contrib- ute to early completion of the project, I must turn work in early. On the
lower branch, to be a successful team member, I must have sufficient time in my task estimates to complete my commitments. To complete my com- mitments, I must turn in task early. The obvious answer is that I can always do extra checks and improve the quality of my project task result when it looks as if I might finish early. Even if I did finish early and turn it in to my manager to be checked prior to submitting it to the project, the manager (a very busy person) likely would not look at it until it is due anyway.
3.2.2.1 Student syndrome
Did you always study for exams weeks ahead, so you could go to bed early the night before? Did you always write your papers to get them done at least a week before the deadline, to avoid the gap in the library when all the books on the topic are out, and to get to the college computers before everyone else was on them all night?
Well, it is probably not news to you that most people have a tendency to wait until tasks get really urgent before they work on them. That is especially true for busy people in high demand, that is, all the most impor- tant people the project manager is counting on to get the critical path work done on time.
Figure 3.6 shows the typical work pattern of many people. They do less than a third of the work on an activity during the first two-thirds of the activity duration. They then do two-thirds of the work during the last third of the activity duration. Where are they more likely to find they have a problem to complete the activity in the remaining time, during the
A Successful team
member
B Contribute to early completion
D Turn in work
early
C Sufficient time in next project
D Do not turn in work early
′
Figure 3.5 The conflict underlying project schedule overruns.
first third of the effort or during the last third? If they are working above 100% capacity already to complete two-thirds of the work in one-third of the time, there is no chance to keep to the activity duration by a little extra effort. What is the chance they can recover from an unanticipated problem, like a computer crash?
Student syndrome behavior results in little chance of seeing the posi- tive side of activity duration variation. The effects described above make it unlikely we could take advantage of positive variation, even if we did see it. No wonder projects rarely complete early! The reality is that relative activity duration normally shows a skewed distribution, with a mean well above the average activity time. That is one reason why we often see overruns on activity time but rarely see underruns.
Most project management guidance recommends that project man- agers use an early start schedule. That means starting all noncritical path activities earlier than is necessary to meet the schedule date. People working on those activities know there is slack in their activity. How does that influence the urgency they feel in working on the activity?