Primavera P6 Resource Leveling - Too Complicated? No Way!

In reading a P6 setup recommendation document for one of our IT program clients I almost fell off of my chair when I saw the recommendation for resource leveling. It said not to use resource leveling as it introduces too many variables into the scheduling process. For an IT group not to be using resource leveling in a program is crazy!

Believe it or not, we see this advice suggested a LOT.

Trust Issues?

The typical arguments range from “I have no idea what P6 is doing” or “I don’t trust P6” to “I want to decide for myself where the tasks should go”. Just last week in a P6 training class in Toronto I heard “resource leveling does not work, so we do not use it”.

These points are all interesting, but all ill-founded. To use P6 Resource Leveling you are going to have to get over some trust issues.

What to do With That Expensive Recommendation Report?

Now I have to say, after 27 years of doing P3 and P6 consulting, this is one of the most unpleasant parts of the role of “Expert”. How do you politely explain to your client that they paid good money for bad advice? On this point there is no grey area; AVOIDING RESOURCE LEVELING IS TERRIBLE ADVICE.

I can maybe understand not being sure how to proceed to get resource leveling working for you, but to try and do this manually or not at all in a resource intensive project like IT is a recipe for disaster. How can anyone hope to manually manage such complex and changing resource requirements without resource leveling?

Well, they can't. So instead of working at developing a resource leveling criteria that works, what tends to happen is clients put a whole lot of hard logic where it does not belong. This makes it even harder to get a realistic schedule without a lot of out-of-sequence tasks which drives them into the very scary world of progress override.

How to Get Resource Leveling Going

• Input your resources into P6 with some realistic ideas of resource availability
• Create some activity codes that will help you get some prioritization that makes sense
• Ask the site supervisor or foreman how they decide what they are going to work on next. If they can explain it in words, you can usually get it into a code.
• Do some test scenarios and come up criteria that works and can be explained. Even if you use the out-of-the-box (OOTB) leveling priority, which is virtually manual, get started.
• Stick with it. The worst thing you can do is change your criteria every time you run it. This will make it impossible to explain or defend and no one will trust what you are giving them.

In the below example, we needed to deliver several barracks buildings, but the contract had turnover dates, buildings, and floors we had to adhere to. So these were naturally part of the leveling prioritization. Then we used total float and remaining duration to get work done that could not slip and to get tasks that were longer to get started sooner so they would not delay us down the road. It worked really well and the schedulers stopped adding crazy logic which saved themselves a lot of man-hours and late nights at the job site.

A few Tips

I like to make a layout with the fields I am going to use as my prioritization criteria. That way I can make sure I can explain what is happening. I also store the unleveled dates and unleveled total float using global change if you do not have the P6-Calculator. That way I can see what is going on in the Gantt chart.

With Emerald's P6-Calculator, we have modeled very complicated prioritization into priority codes. For example:

• Is it emergency work?
• Is it a safety system that cannot be down?
• Is it a regulatory requirement?
• If so, what is the penalty cost?
• Is a a priority system?
• Is it a priority to get your facility to capacity?

All these factors were weighted and scored to come up with a specialty leveling criteria.

If you cannot get it to do what you need, you can even go further with PSOP.

Be brave. Give it a try. Do several tests and iterations.

YOU WILL NOT REGRET IT.