Earned Value Analysis: 50/50 Technique

[Pete Smith]

Hi,

is it possible configure Project to use the 50/50 Technique for EVA?

Thanks in advance
Pete

 

[John]

Hi,

is it possible configure Project to use the 50/50 Technique for EVA?

Thanks in advance
Pete

Pete,
Sure, it's really a matter of how progress is reported. If value is
"taken" only as 50% and 100% (or other common values such as 25/75.
60/40, etc.), then the EV data will be what you want.

John
Project MVP

 

[davegb]

Pete,
Sure, it's really a matter of how progress is reported. If value is
"taken" only as 50% and 100% (or other common values such as 25/75.
60/40, etc.), then the EV data will be what you want.

John
Project MVP

I think it will work as long as your durations are relatively short. If
you have long duration tasks relative to the duration of your project,
then you will have unreported value being earned, but actual expenses
acruing, throwing off your EV.

 

[John]

I think it will work as long as your durations are relatively short. If
you have long duration tasks relative to the duration of your project,
then you will have unreported value being earned, but actual expenses
acruing, throwing off your EV.

Dave,
It depends on what you mean by "relatively short". In our case any tasks
that were greater than one month but less than three months (yeah that's
something like two isn't it), could have a 50/50, 25/75, or 40/60 earned
value method (we used monthly statusing/reporting). And yes, such
discrete EV methods do throw a skew into the overall data but it
balances out. I might add that this method is used and preferred on many
Government contracts.

John

 

[Trevor Rabey]

The more Tasks that you create, and the shorter their durations, the more
control and accuracy you obtain and the less you need to be concerned about
monitoring and reporting fractions of task completion.
You could settle on "no finer level of progress reporting than 50%
completion" but if you make enough tasks which are indivually short enough
you can happily adopt "no finer level of progress reporting than 100%
completion" without acquiring too much over/under reporting error in a given
period. What doesn't get picked up this month will get picked up next month
(or next week).
See the Tracking Toolbar. It is already set for 0, 25%, 50%, 75%, 100%.

 

[davegb]

The more Tasks that you create, and the shorter their durations, the more
control and accuracy you obtain and the less you need to be concerned about
monitoring and reporting fractions of task completion.
You could settle on "no finer level of progress reporting than 50%
completion" but if you make enough tasks which are indivually short enough
you can happily adopt "no finer level of progress reporting than 100%
completion" without acquiring too much over/under reporting error in a given
period. What doesn't get picked up this month will get picked up next month
(or next week).
See the Tracking Toolbar. It is already set for 0, 25%, 50%, 75%, 100%.

 

[davegb]

I agree with Trevor.
And I agree with John. It can balance out, it you have enough
overlapping tasks. Let's take an extreme example of what can happen
though.

Let's say you have one long duration (3 mo.) task that starts 1 month
into the project. While AV is acruing each month, we see no EV acrual
on this task until month 3 (2 months after it started and the first
month after it passed 50% completion). Such a task would work better if
it's a 40/60 or 25/75. Other lower cost tasks will somewhat dilute this
effect. But it could easity obscure some other problem EVA would
normally illuminate. On a 6 mo project, that one month of inaccurate
data could very easily be crucial to overall project success. This is
why granularity is so important in EVA. To some extent, this problem
can be minimized by doing EVA on weekly or bi-weekly basis on short
projects like this. But the issue should be addressed to make sure
you're getting what you need from doing your Earned Value.