Projects in Less Time
|
Projects in Less Time
|
|
|
Chapter 13 — Why Projects Take Too Long©2005 Mark Woeppel Back at the Executive MBA program class for project management, the students were to have completed their assignments documenting why employees at various levels in their companies build safety time into their projects. There is, however, only one paper turned in and it appears that nobody has any real answers because the students all had trouble getting anyone at their companies to admit that safety time was actually being added to estimates. Professor Rick probes. The class finally realizes that they really do have the answer after all. It is revealed that the vast majority of employees expect they have at least an 80% chance of finishing on time as long as they are 1) not delayed by others or 2) loaded up with too much else to do. They also discover that the boss adds his own safety time cushion on top of theirs. When numerous layers of management are involved, each layer adds more safety time. Then, top management makes a global cut in the projected time it will take for the entire project to finish — a cut that managers had already padded their estimates to absorb. Professor Rick concludes that there are at least three reasons excessive safety is added to time estimates: 1. Estimates are influenced by previous negative experiences. 2. The more layers of management there are, the bigger the final estimate will be. 3. Global cuts are anticipated, so time estimates are increased to absorb them. The professor also concludes that safety time makes up the lion’s share of the estimated time it will take to complete a project. The class then determines that delays that occur during one step get passed on to the next step, but the converse is not true — extra progress made during a step is rarely passed along. So, delays accumulate but advances do not. Professor Rick equates guarding the performance of each step to “cost world” philosophy while he equates acknowledging the performance of the project as a whole to “throughput world” philosophy.. One class member attributes “wasted safety” to “students’ syndrome.” Another says it is “bad multitasking” that impacts negatively by inflating lead times and wasting set up time. Self-fulfilling prophecy is mentioned as a negative contributor, too, because if one thinks a project will take longer, the project will expand to fill the time allotted. Concern that there will be times when people are not working steadily and efficiencies may drop (especially when dependencies between steps are involved) also plays a role. For the next class session, Rick asks his students to be ready with an example for each of the three devices that people use to increase safety and three for the devices that waste it. Commentary — Chapter 13In spite of the fact that task durations are often conservatively estimated to begin with, the presence of certain behaviors can cause them to increase even more. Three important behaviors make project durations longer than necessary. They are: Deliberate Padding Once the people doing the work have conservatively estimated their tasks, the estimates are then passed through several layers of management where they are increased even more. Because managers feel they must protect their own performance, in many organizations task estimates are not treated as “estimates,” they are treated as “commitments.” People don’t want to be late on commitments, thus, they “pad” their estimates of how long a given task will take. Student Syndrome
This “Murphy” is really two things: common cause process variation and special cause process variation. The two types of variation are not differentiated in the text, but in the implementation, must be treated differently. Common cause variation can be predicted and managed using the CCPM approach. Special cause variation must be treated separately in a risk analysis process. Multitasking occurs when an individual is working on more than one task at the same time. There are two kinds of multitasking: good and bad. Good multitasking is moving two or more tasks along together smoothly, such as catching up on customer calls while heading to a meeting. On the other hand, bad multitasking is anything but smooth. It’s the dropping of work on one task before it is finished in order to start another, only to stop and begin yet another task or go back to a previous task. All too often, people aren’t able to complete a task without getting pulled off onto something else, so “task time” grows each time a change is necessary. Goldratt wants you to see that the majority of task completion time is not used for the actual work, but is waiting or queue time. Tasks ready to be worked on cannot be worked on because there is no available resource. If the estimates are too long, during execution the actual time will grow even longer! No wonder projects consistently finish late and over budget. Parkinson’s Law These two behaviors, student syndrome and multi-tasking, have the same root cause — the lack of clear priorities. Student syndrome occurs when you believe the real due date is distant relative to the amount of time needed to complete the task, while bad multitasking is caused by not recognizing the real priority of tasks until they become late relative to the “need by” date. Why Do Projects Take So Long? 1. We add too much time to the original plan. We allow x amount of time, so it takes x amount of time. 2. Our resources multi-task, adding unnecessary work (additional setups) to the project 3. The Student Syndrome causes us to waste whatever buffer we did have, adding more time to our already generous estimates. 4. Parkinson’s Law blocks us from taking advantage of any favorable variation (tasks finishing early) the project experiences. Implications for Management Critical chain seeks to reduce / eliminate these behaviors, and therefore they are not planned for in the project. We can overcome deliberate padding, student syndrome, bad multitasking and Parkinson’s Law through better management and communication. If we can eliminate these behaviors from our projects, the time to finish the project is reduced. For example, looking at our project with the critical path (A, C2, & D), you see the length of the path is 15 days.
After removing the safety time from these tasks, the critical path is shortened significantly, to 8 days. This is one of the reasons that critical chain projects consistently finish in less time than projects that do not use this approach.
Glossary — Chapter 13Bad Multitasking — The dropping of work on one task before it is finished in order to start another. Common Cause Variation — a source of variation caused by unknown factors that result in a steady but random distribution of output around the average or mean of the data Special Cause Variation — variation caused by known factors that result in a non-random distribution of output; also referred to as "exceptional" or "assignable" variation. Parkinson’s Law — The satirical statement that work expands to fill the time allotted to it expressed as a law by a British economist in the early 1900s. Students’ Syndrome — The type of procrastination students are prone to when they are assigned a project or are facing a test.
|
|
|
“Student
syndrome” is a term that pertains to the psychology of procrastinating,
something students are particularly prone to do. The analogy is to students who
are going to take a test. When do they study for it? The night before! Why?
Because they have much more important things to do! Often in projects, people
start too late, using their safety time to work on other things, thinking they
still have enough time to complete the task on time. After they begin the task,
they run into problems, causing it to take even longer than the original padded
estimate. The student syndrome causes longer durations because some of the time
needed to complete a task is lost when it’s started too late or even when it’s
started “just in time.” Then, Murphy causes the task to take even longer.
