Best Practice Using Buffers for Risk and Non-Productive Time

[Gary Furash]

I was hoping to get some ideas on best practices for
documenting risk buffers and non-productive time in MS
Project.

For example, let's say task "X" is estimated to take 3
days, if the resource could focus on it exclusively and
things went the way we expected. However:

1. given the fact that they probably can't focus it on it
exclusively (emergencies, meetings, etc.), they can
really only work at 70% of capacity;

2. if one of the accepted risks becomes realized, it
could take longer (say an additional week).

You could make the task longer, reduce the resource
availability, etc.

Given that we do not really track people to the hour here
(we don't care about that level of detail), what are some
best practices.

 

[Jack D.]

I was hoping to get some ideas on best practices for
documenting risk buffers and non-productive time in MS
Project.

For example, let's say task "X" is estimated to take 3
days, if the resource could focus on it exclusively and
things went the way we expected. However:

1. given the fact that they probably can't focus it on it
exclusively (emergencies, meetings, etc.), they can
really only work at 70% of capacity;

2. if one of the accepted risks becomes realized, it
could take longer (say an additional week).

You could make the task longer, reduce the resource
availability, etc.

Given that we do not really track people to the hour here
(we don't care about that level of detail), what are some
best practices.

As for risk, there are a number of approaches.
Project has PERT capability built in. You could use that.
There are vendors who sell monte carlo simulation tools for risk analysis in
project. You can use them.
There is Theory of Constraints and tools which facilitate it. You can use
them.
You can use contingency budgets and tasks.
You could try and improve your estimates and have plans to mitigate your
risks.
There are many different ways to do it. All of which could be considered
best practices.

As for people with non-productive time.
If it is a problem and you want to measure it so you can fix it then model
it in the project.
If it is a constant, then have people make their estimates with it in mind.
Either of these could be a best practice.


--
Please try to keep replies in this group. I do check e-mail, but only
infrequently. For Macros and other things check http://masamiki.com/project

-Jack Dahlgren, Project MVP


+++++++++++++++++++