Discussion on:

Debunking four resource leveling myths

We'll take all you got!

Of the three leveling methods mentioned, what are people's preferences? Is it a good idea to do all three methods and compare the results to come up with a 'best fit'? I'm a little leery of using automated methods in Project from using it in the past. It may be that it has been improved.

Lots of good insight in the article.

Most project management concepts and tools, including resource leveling and its various implementations in MS-Project and similar tools, are based on construction projects where laborers are essentially interchangeable. For such projects, task dependencies are the only factor affecting schedule; and resource leveling is meaningful.

On the other hand, most projects that most of us are likely to be involved in have both task and resource dependencies. That is, it makes a difference who the individual is that performs a task. If that individual is already busy, the task that individual must perform cannot be started until that individual actually is available. When tools define the critical path, they address only task dependencies but fail to take such resource dependencies into account. Resource leveling also fails to recognize that who does certain work has additional ramifications beyond mere loading.

Together, these failures to recognize and adequately address resource dependencies lead to erroneous and deceptive critical path and schedule definitions that will be the undoing of even the most conscientious project manager.

More often than not, resources are not dedicated 100% to a single project. They may have business as usual duties or they may be shared across more than one project. In my opinion Resource Leveling at the Portfolio level is far more important than at the Project level and it is a management skill requiring judgement and compromise on a continual basis rather than an automated push-button solution.
Sure, MS-Project helps us to understand the problem and it's good at highlighting conflicts but it is just a tool and it makes the assumption that your estimates are perfect - in the real world they aren't so continual monitoring and adjustment is needed.

Automated resource leveling works very well when you have planned the project that included elapsed time and effort time. Not otherwise. I have rarely seen planning at this granularity.
And the second issue is project plans are padded arbitrarily (no metrics or policy for padding). Garbage In, Garbage out. This causes PMs to overload, make unrealistic projections, etc.
Project planning is a science. Project Management is an art.

One very common mistake made by project managers attempting their first resource leveling exercise is to precisely model reality. Today's tools, MS Project included, allow a scheduler or PM to set up Resources with various work schedules, vacations, sick leave, and more. This results in an enormous amount of time getting the schedule set up and as any PM or scheduler knows, nothing goes precisely according to schedule. Then it is a maintenance nightmare. It is better to allow for a certain degree of inaccuracy (because it is a prediction of the future that is going to happen anyow) and keep it simplified. How simple? That will vary by project and individual and budget -- but don't try to use every option available because it usually results in a less accurate schedule than a more accurate one.

One of the most common errors that I see is related to scope control and resource leveling. Those of you that work on major projects know what I am talking about. Scope adjustments happen all of the time and incorporating them back into your plan is a critical component of managing and adjusting. I often times see the same time-lines trying to be held even though the scope has demonstratively changed.

For a deeper understanding of this situation and more visit. http://www.upperedge.com/wp-content/uploads/2011/07/ue_whitepaper_gaincontrol_6_16_11.pdf