How to model tasks that will be used at an unknown point during the project

[Kryn13]

I am budgeting a process improvement project. I have a number of
sub-projects to evaluate various processes. I have a single budget amount
for time to be spent documenting and fixing problems that are discovered.
Those problems could be in any of the processes and the "fix" budget could
be spent any time during the 6 month project timeframe.

What is the best way to model this? The budget for "fixing things" is
several weeks but I don't know where they will be used the project and the
time will definately not be spent in one consecutive period.

Thanks,

Kryn

 

[Steve House [MS Project MVP]]

You are describing a classic example of a top-down budget but that's not the
basis of a project plan nor what Project is best at. Creating a project
plan is NOT a process of taking a pre-defined timeframe, work allowance, or
dollar amount and distributing them out over the various tasks you've
identified for the project. MS Project is based on the notion of bottom-up
budgeting, a totally different approach to estimating for both the schedule
and the costs - in other words, things will take as long and cost as much as
they're going to require to complete the project's deliverables and it
doesn't matter (at least at this stage) how much you're being allowed by
management. You use Project to estimate the time and budget that be
required to complete the project successfully and compare the result to your
allowances - if you're under budget you're in gravy but if you're over
budget you have to either find more efficiencies in the plan, renegotiate
the budget, or reduce the expectations for the deliverables. Think of it as
a reality-check for the budget you're being given. How do they/you KNOW it
will take several weeks to "fix things?" "Several" implies to me something
like 2 or 3 weeks - what happens if it gets really fouled up and there's a
total of 6 weeks of fixup to do? Do you just bailout and drop the project
half-done or do you bite the bullet and fix it anyway even if it puts you
over budget?

The best way to handle the contingencies is to forget for the moment the 6
week fixup budget. Look at the various sub-projects or key deliverables and
try to estimate what the most likely "fixup" will be for each one, based on
the nature of the process and prior experience with similar projects in the
past. Insert the fixup tasks at each point they might occur with an
estimate of how much will be required if they do occur. This will give you
a plan based on "worst case" budgeting. As you work the plan, if you DON'T
need to do any fixup in sub-project X as you though you might, drop the
fixup task from the schedule or reduce its duration. THis, BTW is one
element of a planning process called PERT - where you come up with a best
case, likely case, and worst case plan.

Hope this helps