Earned Value across different files

[Rose]

My company has archived a monthly copy of different project managers' project
plans, as they don't really trust that a baseline will be saved monthly by
the PMs.
Now they want an earned value report done for each project. I'm new to the
concept of earned value anyway, though I did read the help file on it. But
the help files talk about comparing baselines in the same file. How can I
compare 2 different baselines in 2 different files? Or will I somehow need to
import a baseline into a main plan to do earned value?

 

[Gérard Ducouret]

Hello Rose,
The Earned Value calculation will not compare different baselines, even in
the same file. It will compare actual data and *only one* baseline at a time
(at the Status Date). Project 2003 allows you to choose a baseline among
eleven as a reference:
Tools / Options... / Calculation / Earned Value... / Baseline for Earned
Value calculations...

Gérard Ducouret

 

[Steve House [MVP]]

Adding my .02 - why do they want the PM's to save a baseline every month?
IMHO, the PM who DON'T routinely save revised baselines are doing it
correctly. When you post actual performance into your plans, Project
updates the schedules to show that same dates as the actuals you've
obtained. That means that once you do work, you lose the original planned
data for those tasks. Saving a baseline copies the plan information into a
set of "bins" in the database that are not changed by posting progress.
This preserves the plan as you originally intended to work it so you can
compare actual performance back against your original intentions. But if
you save a fresh copy of the baseline every month, you overwrite the
original plan data with with whatever the schedule is showing when you save
the revised baseline, thus making it completely impossible to ever measure
whether you are running ahead of or falling behind your schedule or are
unning over or under budget. IMHO, once your project is planned and
approved, but before beginning work, you save a baseline for future
reference and comparison purposes. Once saved, that baseline is engraved in
granite and the only time you'd change it is if tasks are added or removed,
changing the plan so much as to turn it into a substantially different
project.

Here's the idea ... We had a task originally scheduled for 6 days starting
10 Jan and requiring 48 man-hours work @ $10/hr for a total cost of $480.
When you start the project and save your baseline, the baseline also shows
the same data. When you did the task, your original resource had been in an
accident and it took you a while to find a replacement. When you actually
did the work, the task actually started on the 17 Jan. But the replacement
resource turned out to be a better worker and he finished it up in 4 days
after 32 manhours work for a total cost of $320. After posting the actuals,
our plan shows the new figures but our baseline shows the old ones, allowing
us to see exactly where we stand. Started late but made up time and
finished not quite as late and we're $160 under budget as well - good job!
That information is what we'd lost if we saved a new baseline over the top
of the old one.

 

[Rose]

"Saving a baseline copies the plan information into a
set of "bins" in the database that are not changed by posting progress. "

That, I believe, is a possible use of the 11 baselines we can now use in
2002 - to be used as snapshots?
They would like not just an earned value report on right now vs the original
baseline, but also where things were a month ago vs the original baseline, 2
months ago vs the baseline, etc. I would never overwrite the original
baseline, as that would negate its whole purpose.
I could just run one now and only look at the date range I want to focus on.
But what about this scenario: 2 months ago the PM was keeping up with hir
project and everything was great. Last month either things slipped or the
project fell behind. But now the project looks back on track again. How would
I know *now* that there was a problem last month? Project doesn't track when
actual work was done - only that it got done by X date. If I only compare now
to the baseline, I could only see a general picture of slippage over the last
maybe 2-3 years...there's no way to be more specific. Unless I *do* have a
snapshot of the plan the way it was 1 month ago.
In this case I do because the plan's been archived. I was hoping to take
that snapshot and compare it to the current plan to look for slippage and
earned value that way, not as compared to the original baseline. Considering
how long our projects are (I work for the Department of Transportation),
comparing now to 3 years ago is pretty useless. Comparing to something more
recent is more meaningful.

 

[Steve House [MVP]]

Whew! You had me worried there a minute <grin>. As long as the original
baseline is safe you're ok. But as Gerrard said, unfortunately the Earned
Value metrics are always based on comparing the current plan (containing
actuals) with one of the baselines but can't compare one baseline against
another.

I suppose it is a matter of perspective. When I think of slippage, I
habitually think of how progress has gone versus how it was planned at
inception. If we lost ground in the first few months of our 3 year project
and after, say, the first 6 months we had slipped 30 days but had then
gotten our act together and since then things haven't gotten worse, today
we'd still show 30 days slippage. I'd only want that to get zero'ed out if
we had been able to recover that lost ground because to my perception,
gettin g back on track means that you have been able to correct the factors
that caused you to slip in the first place AND recover the time you'd lost
before you fixed it. OTOH, if we were still under-performing, over time
we'd see 30 days, then 40 days, then 50 days slippage, etc etc. If I have a
deadline to deliver our Project by the 1st of September and we're running
late, I want the slippage to show me how late I am against that deadline,
not how much later I've slipped in the last month.


--
Steve House [MVP]
MS Project Trainer & Consultant
Visit http://www.mvps.org/project/faqs.htm for the FAQs

 

[Rose]

In a normal world, where I created and maintained the project plan myself and
projects lasted only a year or so, that would apply. This isn't normal.
They've got a good reason to want to compare earned value for 6 months ago
to now, and not the original baseline vs now. Too bad that Project can't do
it, but thanks for confirming that it can't.