Earned Schedule

[Wilson]

I'm trying to identify metrics that the PMO can use to control this project
and I'm having difficulty because the planning team decided to not input cost
or resource information. They have simply assigned responsibilities via a
text field and manage the cost aspect somewhere else. The PMO is looking for
"EVM" type metrics but without cost and resource info I'm sort of at a loss.
Is it possible and or does it even make sense to jimmy the EVM metrics to use
only duration? I.e. Instead of using BCWP - BCWS, use # of Activities
completed - # of Activities scheduled to be completed.... Another example I
considered was a modified Earned Schedule (*we started the project in Feb):

(# of months complete) + {[(# of tasks completed)-(# of tasks scheduled
through April)] / [(# of tasks scheduled in May)-((# of tasks scheduled
through April)]}

Using the data from the above formula, I would then be able to arrive at
SPI(t) & SV(t)

QUESTION - Does this make any sense (considering the limited amount of data
I have to work with) or am I crazy?

 

[salgud]

I'm trying to identify metrics that the PMO can use to control this project
and I'm having difficulty because the planning team decided to not input cost
or resource information. They have simply assigned responsibilities via a
text field and manage the cost aspect somewhere else. The PMO is looking for
"EVM" type metrics but without cost and resource info I'm sort of at a loss.
Is it possible and or does it even make sense to jimmy the EVM metrics to use
only duration? I.e. Instead of using BCWP - BCWS, use # of Activities
completed - # of Activities scheduled to be completed.... Another example I
considered was a modified Earned Schedule (*we started the project in Feb):

I think you mean Earned Value. And no, it won't work at the activity level.
And activity could involve anything between 8 hrs (usually shouldn't be
less than a day) and thousands of hours (though it should have a duration
of less that 10 days, which could be thousands of hours if enough resources
were working on it). So there is too much variation between activities size
to make them a useful metric.

(# of months complete) + {[(# of tasks completed)-(# of tasks scheduled
through April)] / [(# of tasks scheduled in May)-((# of tasks scheduled
through April)]}

Using the data from the above formula, I would then be able to arrive at
SPI(t) & SV(t) But not meaningful ones.

QUESTION - Does this make any sense (considering the limited amount of data
I have to work with) or am I crazy?

It sounds to me that you don't have anything to hang your hat on. You're
team has chosen to not monitor anything that could be used to track
progress.
In some cases, like construction, consumables used as metrics, though not
for EV. This would be things like cubic yards of concrete installed, tons
of rebar installed, tons of pipe installed, yards of conduit installed,
etc. But these measure only a specific kind of progress, not overall
progress. But maybe there's some way you can find a similar metric for your
kind of projects.
Hope this helps in your world.

 

[Wilson]

I appreciate your help. This is a large software implementation (upwards of
70MM and approx 6000 tasks through assembly test phase). The planning team
did input an 'effort' value into one of the text fields but I have little
faith the values have remained accurate since planning finished. I agree
that it makes sense that the variances between activity efforts would nix
using Earned Value at the activity level.

Thanks for your help though.

 

[Jim Aksel]

There may be some saving tatics, but this will get sticky.
You may be better of with some of the trip wire indices such as Basline
Executability Index (you alluded to it). You can also compare actual
duration to baseline duration to see if that might get you anywhere.

You really need to check my profile to see if you see anything useful.
There are a lot of tools to help you get this done, but those tools are not
"free ware"

First, memorialize your %Complete someplace. Stick it in spare field. Also
save all the current start/finish dates as well. You may want to use
something like Baseline1 for this data. Change all the task types to fixed
duration.

Now, clear out your %Completes. Insert the Baseline1 Start/Finish columns
and compare your current start/finish to those dates. You will need to adjust
schedule logic so your dates mesh back up.

Now you should be back to "go." Decide what you want to track for your
baseline. If resource costs are about the same, you can reasonably fake this
with one resource named "Worker1" costed at $1/hr. At that point,
physical%Complete and %WorkComplete will be the same. You can peanut butter
the worker1 on all descrete (not summary) tasks and this will give you some
measure of "value" that is (duration) weighted.

How do you weight each task as per the "baseline" the PM is keeping
someplace. It can be done, you have to agree with the PM on the value of each
work package.

Once there is a properly costed/resource loaded schedule in place, you can
set a baseline, repopulate your %Completes and see where you are. If you
properly status the schedule, you should have a pretty good idea of your EV.
In your case, "properly costed resource loaded" can be worker1 $1/hr.

I could write for pages.... this should get you started.



--
If this post was helpful, please consider rating it.

Jim

Visit http://project.mvps.org/ for FAQs and more information
about Microsoft Project

I'm trying to identify metrics that the PMO can use to control this project
and I'm having difficulty because the planning team decided to not input cost
or resource information. They have simply assigned responsibilities via a
text field and manage the cost aspect somewhere else. The PMO is looking for
"EVM" type metrics but without cost and resource info I'm sort of at a loss.
Is it possible and or does it even make sense to jimmy the EVM metrics to use
only duration? I.e. Instead of using BCWP - BCWS, use # of Activities
completed - # of Activities scheduled to be completed.... Another example I
considered was a modified Earned Schedule (*we started the project in Feb):

(# of months complete) + {[(# of tasks completed)-(# of tasks scheduled
through April)] / [(# of tasks scheduled in May)-((# of tasks scheduled
through April)]}

Using the data from the above formula, I would then be able to arrive at
SPI(t) & SV(t)

QUESTION - Does this make any sense (considering the limited amount of data
I have to work with) or am I crazy?