Finishing tasks early, before baseline start date

[David]

So, say that I originally scheduled a task to start on 12/1 and complete
12/5, so my baseline start and complete dates are those dates.

What if I complete this task early and the actual start is 11/1 and actual
complete 11/5.

If I set my status date to 11/9, my SPI value would be incorrect, saying it
was 0.
SPI = BCWP/BCWS. So even if I am ahead of schedule, the spi value is
showing me as behind schedule? It should really be greater than 1.

Is this a known problem of Project Pro? Or is there a solution to this?

David

 

[JackD]

What is 1 divided by 0?
This is a problem inherent to any implementation of Earned Value.
Switching to a different tool will not solve it.
At the early stages of a project you will always find problems with the
ratios.
You have to wait until you have more data before you can draw a meaningful
trend.
If you have a few tasks which are scheduled to start, then BCWS will no
longer be zero and you will start to have a valid ratio.
At the early stage look at cost/schedule variances.
If you are a month ahead, why worry?

 

[David]

Jack,

For a large schedule, you'll have to track a lot of tasks that do start
early. Also, if have to report valid SPI and CPI values on a periodic basis
to higher ups, you need accurate data. You can not just wait until the tasks
start to begin.

David

 

[JackD]

If the project hasn't started then why have the tasks started? If you are
still before the project is planned to start you can simply report that you
have started early and therefore there is no valid SPI and that you will
report progress against the plan as soon as you can. It is completely
accurate to state that. I think most people who are expecting EV reports
should have enough intelligence to understand that a SPI for a project which
isn't even planned to start yet is meaningless. They should know the
limitations of the method. If they don't then you have a management problem.

Another possibility is that you have an incorrect baseline or a bad plan.
You should probably try to fix these before you bother reporting SPI.

Finally if indeed all your tasks finish before ANY were planned to start,
then you have a planning problem.

None of these problems can be solved with a scheduling tool.

The reality is that in a large project there are tasks which start early and
tasks which start late. EV is most useful/accurate as an aggregate measure
(the project SPI = sum of BCWP / sum of BCWP right?) so for any large
project which HAS started you WILL have some BCWS in the denominator and
this will not be an issue at all.

 

[Steve House]

If I can jump in - a project plan is not supposed to be a passive document
simply recording when people do work. IMHO, it is a PROACTIVE document -
your role as project manager is similar to the conductor in an orchestra and
the project plan is the symphonic score. You plan so as to accomplish the
project's goals on time and under budget and then you advise the resources
where they need to go with what tools and when they need to be there to work
on the tasks at the times you have determined they are supposed to take
place. You may have the occasional task that you find can start earlier
than you originally thought but that should be a relatively rare event. If
it COULD start earlier, that's where your plan should have showed it
scheduled in the first place. In the planning process you should have
determined the earliest that conditions would permit it to start and
directed the resource to work on it at that time.

Steve House [MVP]