Eli Goldratt and Restructuring your Project Schedules in 5 Minutes

Eli Goldratt wrote a lot about project management and its not realistic to try and cover everything in five minutes in this article, but I thought I will present one of his best ideas here for you.

Goldratt said that when a project manager asks a project worker for an estimate of how long a task will take, the answer will invariably be pessimistic. This is because people do not want to be responsible for causing delays to projects and there is always a multitude of things that can go wrong with an activity for it to not finish in its best or even normal time. The answer provided is therefore normally a good estimate of when the activity will definitely be finished by. Since tasks can normally be completed faster than this, it means that a project schedule normally has a large amount of invisible buffer built in.

He also argued that there are then several reasons why despite this large invisible buffer, projects do not finish much sooner than expected, among other things because they have been planned with more time than needed for each task.

How Can Projects be Structured to Take Advantage of this?

The obvious question, if you believe everything Goldratt has said, is then how should projects be scheduled in order to make this invisible buffer visible again, and once you do so, how that can be leveraged to help the project finish on time?

Here are three steps to modify a project schedule to reflect Goldratt’s ideas:

• The project manager should not add their own buffer to the project, since there is already buffer in the project (but it is not visible).
• To make the buffer visible, the existing tasks should be split into “core” and “buffer” parts, typically with 25-50% being buffer, and labeled as such (the total duration remains unchanged).
• All the buffer should be collected in a common buffer placed at the end of a project and this should be used to measure a project’s progress. If a task finishes ahead of schedule, the total project buffer grows and conversely when a task finishes later, the project buffer shrinks.

If the project team agrees to these changes, the result should be a project that is completed more quickly but these changes will also affect the way the project is run in other ways. In return for agreeing to shorten the task durations, the project manager can no longer demand that the project team finishes all their activities on time. In fact, it is now to be expected that some tasks in the project will take longer than planned. On the other hand, because the project’s progress is measured by how the project buffer is changing, if individual activities finish on time also becomes less important.