% Complete to Date

I'd like to easily report what percent of work is complete as expected to be
by date. Say I have tasks to be finished last week, this week, and next
week. Some of my tasks from last week are done, but not all, yet some of my
tasks to start next week have begun. Can project easily calculate some %
complete to Date? Or atleast, even if it doesn't give me credit for that
work I am ahead on, can it tell me if I'm up to date with what I said I would
be by now?

Jon S.

AndrewLavinsky
Sat May 03 07:49:00 CDT 2008

If you've got cost associated with your project, you could use Earned Value -
this is pretty much what it was designed to do.

For expected Expected % Complete, you would most likely be looking at a
custom formula like [BCWP] / [Baseline Work]. Then as you adjust the Status
Date in Project > Project Information, the Expected % Complete should change.

If you're not using cost, I am not sure how you would go about doing this.

-A

"Jon S." wrote:

> I'd like to easily report what percent of work is complete as expected to be
> by date. Say I have tasks to be finished last week, this week, and next
> week. Some of my tasks from last week are done, but not all, yet some of my
> tasks to start next week have begun. Can project easily calculate some %
> complete to Date? Or atleast, even if it doesn't give me credit for that
> work I am ahead on, can it tell me if I'm up to date with what I said I would
> be by now?
>
> Jon S.

AndrewLavinsky
Sat May 03 09:00:01 CDT 2008

Coffee wasn't kicking in yet, minor correction on the custom formula:

Expected % Complete = [BCWP]/[Baseline Cost]

"Andrew Lavinsky" wrote:

> If you've got cost associated with your project, you could use Earned Value -
> this is pretty much what it was designed to do.
>
> For expected Expected % Complete, you would most likely be looking at a
> custom formula like [BCWP] / [Baseline Work]. Then as you adjust the Status
> Date in Project > Project Information, the Expected % Complete should change.
>
> If you're not using cost, I am not sure how you would go about doing this.
>
> -A
>
> "Jon S." wrote:
>
> > I'd like to easily report what percent of work is complete as expected to be
> > by date. Say I have tasks to be finished last week, this week, and next
> > week. Some of my tasks from last week are done, but not all, yet some of my
> > tasks to start next week have begun. Can project easily calculate some %
> > complete to Date? Or atleast, even if it doesn't give me credit for that
> > work I am ahead on, can it tell me if I'm up to date with what I said I would
> > be by now?
> >
> > Jon S.

JimAksel
Sun May 04 09:22:00 CDT 2008

Pardon the bump in. Using Earned Value techniques, Micrsofto Project
automatically calculates Schedule Performance Index [SPI] and Cost
Performance Index [CPI]. Insert those columns.

SPI=Measure of Work Completed / Measure of Work Scheduled BCWP/BCWS. Above
1 is ahead of schedule, 1.0 is on schedule, less than 1.0 is behind schedule.
The value is usually displayed as a percent 100%=1.0

CPI=Measure of Work Completed/Actual Costs BCWP/ACWP. Above 100% is below
cost. CPI<100% means costs are exceeding plan... the work may be tougher
than you thought.

For our purposes BCWP=(%Complete)*(BAC).

Now we get crazy -- MS Project can calculate using two methods %COmplete and
Physical%Complete. %Complete deals only with duration and nothing else. If
you claim 25% Complete you are saying your project is 1/4 of the time it will
take to complete the task. Very few projects really work this way.

However, when most people say they are 25% complete, they mean they have
completed 25% of the work involved or 25% of the value (I have laid 25 of
100 bricks). Both of these measure are independent of duration. For
example, work may not produce value or cost in a linear fashion. Planting
corn has labor up front, while the corn grows there is less work, and
harvesting includes more work.

Generally, Physical%Complete is the most correct method to use in most
instances.
--
If this post was helpful, please consider rating it.

Jim

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



"Jon S." wrote:

> I'd like to easily report what percent of work is complete as expected to be
> by date. Say I have tasks to be finished last week, this week, and next
> week. Some of my tasks from last week are done, but not all, yet some of my
> tasks to start next week have begun. Can project easily calculate some %
> complete to Date? Or atleast, even if it doesn't give me credit for that
> work I am ahead on, can it tell me if I'm up to date with what I said I would
> be by now?
>
> Jon S.

Trevor
Sun May 04 10:34:22 CDT 2008

You really need to get more specific about what you are measuring % of.
And even if you can measure it, and can find a numerator and denominator to
get a % from, what does it mean and how can it be used for anything useful.
Gross numbers with a % sign after them don't really help you to run the
project.
EV can spit out a cost and and schedule variance and a cost and and schedule
performance index, and an EAC, but it doesn't tell you much about what the
problem is, or even if there is one, what caused it, or what to do about it.
That's all in the detail.
Variances, indexes and ratios don't tell you if you are doing the right
tasks. It is possible to get good EV numbers by doing the wrong tasks, ie
the noncritical instead of the critical tasks.

Say the Task is to lay 10000 bricks in 10 days, that's 80 person-hours for
one person at say $50/hour, ie $4000 plus $1/brick, $4000 + $10000 = $14000,
or $1400/day.
At the end of Day 6, 3000 bricks might be laid, the bricklayer may have
worked 7 person-hours per day (42 Hours), perhaps he only cost $40/hour,
bricks might have cost $1.20 each ($3600).

So you have bricks (30%), Duration (60%), Hours (42/80), Cost (((3000 x
$1.2) + (42 x 40))/84).
The denominator in each case here is the baseline total bricks, duration,
work and cost.
Another reasonable denominator would be the baseline (ie as planned) bricks,
duration, work and cost up to the status date.
ie 3000/6000 for bricks, 6/6 for Duration, 42/48 for Work, $8400 for Cost.

Trevor Rabey 0407213955 61 8 92727485 PERFECT PROJECT PLANNING
www.perfectproject.com.au
"Jon S." <JonS@discussions.microsoft.com> wrote in message
news:9D16A050-1BC0-4851-A005-C4B9CA58EA27@microsoft.com...
> I'd like to easily report what percent of work is complete as expected to
> be
> by date. Say I have tasks to be finished last week, this week, and next
> week. Some of my tasks from last week are done, but not all, yet some of
> my
> tasks to start next week have begun. Can project easily calculate some %
> complete to Date? Or atleast, even if it doesn't give me credit for that
> work I am ahead on, can it tell me if I'm up to date with what I said I
> would
> be by now?
>
> Jon S.

JimAksel
Sun May 04 13:14:01 CDT 2008

I wish to offer some respectful difference of opinion with Trevor. He and I
do this occassionaly, I think deep down we both agree but come to it from a
different angle. My personal opinion is that Physical%Complete can be
"morphed" to accomodate exactly what Trevor is getting at. If you want to
lay 1000 bricks and your are going to take credit "by the brick" then make
that your measure and key in the value to Physical%Complete. It is all how
you define the numerator and denominator.

If EV is taken properly, then indicies (ratios), variance, etc. are actually
meaningful. Trevor and I agree 100% that doing the "wrong" tasks will
artifically pad the indicies and variances.... especially at the higher
levels of summary. This is why I don't like a lot of the "trip wire" indices
touted here in the United States.

To properly do EV, you absolutely must look at the details, and at the
appropriate level of detail. The method must be applied consistently. Trevor
shows you how to get different answers using the various measures of bricks,
work, etc. Bottom line is that you have to have a specifically definable
measure of success... we document that in an Earned Value Managment Plan for
every work package. We plan work to the level that below the work package
level tasks must be completely binary: Lay 12 bricks. There were either
laid or they weren't. The %Complete is either 0 or 100%. This is
essentially a milestone technique. Doing things this way will allow a
consistent measure of Physical%Complete at the Work Package levels. We take
earned value weekly.

Now, the reason why I posted a second time.

The original post asked about status of the schedule.... items that should
be complete are not, items that should start "tomorrow" have already started.
This is a schedule maintenance issue. Here's the four rules we use:

Rule 0: Establish a status date (Project/Project Information/Status Date)
Rule 1: There is no such thing as an unstarted task to the left of the
status date. Establish a new date to the right of the status date.
Rule 2: There is no such thing as unfinished tasks to the left of the status
date. If the task is not complete, increase the duration so that it finishes
after the status date. You get that information from the task owner.
Rule 3: Nothing starts in the future. If a task has a %Complete there is no
chance it started to the right of the status date. Ask the task owner when
it started and enter that date in the "Actual Start" column, not "Start".
Then post your %Complete making sure the task owner has identified when they
believe they will finsih.
Rule 4: Nothing is finished after the status date. If the schedule says the
task finishes "next Tuesday" (to the right of the status date) and the task
owner says the work is complete, then it didn't finish next Tuesday did it?
It finished on or before the status date, so key that into the "Actual
Finish" column (not Finish) and claim 100%.

Something else to think about. The "brute force" logic in MS Project tells
us if two tasks are linked as Finish To Start, then if B follows A, then A
must be 100% before B can start. If B has started, then "officially" you
have a logic error in the relationship. Most often this means you do not
have enough detail in your schedule. An example may be software testing
cannot start until all the coding is complete. That is not entirely true...
If you are inventing "Microsoft Office" you can certainly start testing
"Word" while they finsih coding "PowerPoint".... true, you can't test the
integration of the two, but you can certainly see if Word will do some of its
functions.

We run into this type of logic all the time. As such, you should be
prepared to add additional detail to your schedule as you "learn" when the
project progresses.


And to keep rambling .....

Insert the "status" column. Although "status" works against a duration
driven version of %Complete (bad bad bad), it will tell you "late", "on
schedule" and "complete."

There have been so many posts on this forum of "I want to know what
%Complete I should be at the status date." The answer is use
Physical%Complete and compare it to SPI and that should be 1.00 for every
task. If not, look for tasks with SPI less than 1. There is a very deep
divide in this community over SPI and something called "Earned Schdule" ....
but that takes a book to explain.


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

Jim

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



"Trevor Rabey" wrote:

> You really need to get more specific about what you are measuring % of.
> And even if you can measure it, and can find a numerator and denominator to
> get a % from, what does it mean and how can it be used for anything useful.
> Gross numbers with a % sign after them don't really help you to run the
> project.
> EV can spit out a cost and and schedule variance and a cost and and schedule
> performance index, and an EAC, but it doesn't tell you much about what the
> problem is, or even if there is one, what caused it, or what to do about it.
> That's all in the detail.
> Variances, indexes and ratios don't tell you if you are doing the right
> tasks. It is possible to get good EV numbers by doing the wrong tasks, ie
> the noncritical instead of the critical tasks.
>
> Say the Task is to lay 10000 bricks in 10 days, that's 80 person-hours for
> one person at say $50/hour, ie $4000 plus $1/brick, $4000 + $10000 = $14000,
> or $1400/day.
> At the end of Day 6, 3000 bricks might be laid, the bricklayer may have
> worked 7 person-hours per day (42 Hours), perhaps he only cost $40/hour,
> bricks might have cost $1.20 each ($3600).
>
> So you have bricks (30%), Duration (60%), Hours (42/80), Cost (((3000 x
> $1.2) + (42 x 40))/84).
> The denominator in each case here is the baseline total bricks, duration,
> work and cost.
> Another reasonable denominator would be the baseline (ie as planned) bricks,
> duration, work and cost up to the status date.
> ie 3000/6000 for bricks, 6/6 for Duration, 42/48 for Work, $8400 for Cost.
>
> Trevor Rabey 0407213955 61 8 92727485 PERFECT PROJECT PLANNING
> www.perfectproject.com.au
> "Jon S." <JonS@discussions.microsoft.com> wrote in message
> news:9D16A050-1BC0-4851-A005-C4B9CA58EA27@microsoft.com...
> > I'd like to easily report what percent of work is complete as expected to
> > be
> > by date. Say I have tasks to be finished last week, this week, and next
> > week. Some of my tasks from last week are done, but not all, yet some of
> > my
> > tasks to start next week have begun. Can project easily calculate some %
> > complete to Date? Or atleast, even if it doesn't give me credit for that
> > work I am ahead on, can it tell me if I'm up to date with what I said I
> > would
> > be by now?
> >
> > Jon S.
>
>
>

Trevor
Sun May 04 23:14:58 CDT 2008

Jim,
You may as well write the book, since you've started it.

Jon,
We get a lot of questions here, sometimes almost the same question several
times on the same day, and a flurry recently, about how to measure progress.
Usually it is along the lines of "how are we doing compared to how we wanted
to be doing?".
I think, and I could be wrong, the problem is that everyone wants to boil a
complicated situation down to a description in one number.
Perhaps this is because everyone has senior management to report to but
senior management doesn't seem to have the necessary attention span to
understand anything other than one number, so you may as well cook up a
number which is palatable and immediately acceptable so you don't have to
explain it.
There is a lot more to how a game is going, and how it is likely to end up,
than just the score, eg:
How much time is left?
Who is injured and off the field?
Which team has the best reserves?

Trevor Rabey 0407213955 61 8 92727485 PERFECT PROJECT PLANNING
www.perfectproject.com.au
"Jim Aksel" <JimAksel@discussions.microsoft.com> wrote in message
news:3D457A98-B2A9-453F-8B13-6A865D02273F@microsoft.com...
>I wish to offer some respectful difference of opinion with Trevor. He and
>I
> do this occassionaly, I think deep down we both agree but come to it from
> a
> different angle. My personal opinion is that Physical%Complete can be
> "morphed" to accomodate exactly what Trevor is getting at. If you want to
> lay 1000 bricks and your are going to take credit "by the brick" then make
> that your measure and key in the value to Physical%Complete. It is all how
> you define the numerator and denominator.
>
> If EV is taken properly, then indicies (ratios), variance, etc. are
> actually
> meaningful. Trevor and I agree 100% that doing the "wrong" tasks will
> artifically pad the indicies and variances.... especially at the higher
> levels of summary. This is why I don't like a lot of the "trip wire"
> indices
> touted here in the United States.
>
> To properly do EV, you absolutely must look at the details, and at the
> appropriate level of detail. The method must be applied consistently.
> Trevor
> shows you how to get different answers using the various measures of
> bricks,
> work, etc. Bottom line is that you have to have a specifically definable
> measure of success... we document that in an Earned Value Managment Plan
> for
> every work package. We plan work to the level that below the work package
> level tasks must be completely binary: Lay 12 bricks. There were either
> laid or they weren't. The %Complete is either 0 or 100%. This is
> essentially a milestone technique. Doing things this way will allow a
> consistent measure of Physical%Complete at the Work Package levels. We
> take
> earned value weekly.
>
> Now, the reason why I posted a second time.
>
> The original post asked about status of the schedule.... items that should
> be complete are not, items that should start "tomorrow" have already
> started.
> This is a schedule maintenance issue. Here's the four rules we use:
>
> Rule 0: Establish a status date (Project/Project Information/Status Date)
> Rule 1: There is no such thing as an unstarted task to the left of the
> status date. Establish a new date to the right of the status date.
> Rule 2: There is no such thing as unfinished tasks to the left of the
> status
> date. If the task is not complete, increase the duration so that it
> finishes
> after the status date. You get that information from the task owner.
> Rule 3: Nothing starts in the future. If a task has a %Complete there is
> no
> chance it started to the right of the status date. Ask the task owner
> when
> it started and enter that date in the "Actual Start" column, not "Start".
> Then post your %Complete making sure the task owner has identified when
> they
> believe they will finsih.
> Rule 4: Nothing is finished after the status date. If the schedule says
> the
> task finishes "next Tuesday" (to the right of the status date) and the
> task
> owner says the work is complete, then it didn't finish next Tuesday did
> it?
> It finished on or before the status date, so key that into the "Actual
> Finish" column (not Finish) and claim 100%.
>
> Something else to think about. The "brute force" logic in MS Project
> tells
> us if two tasks are linked as Finish To Start, then if B follows A, then A
> must be 100% before B can start. If B has started, then "officially" you
> have a logic error in the relationship. Most often this means you do not
> have enough detail in your schedule. An example may be software testing
> cannot start until all the coding is complete. That is not entirely
> true...
> If you are inventing "Microsoft Office" you can certainly start testing
> "Word" while they finsih coding "PowerPoint".... true, you can't test the
> integration of the two, but you can certainly see if Word will do some of
> its
> functions.
>
> We run into this type of logic all the time. As such, you should be
> prepared to add additional detail to your schedule as you "learn" when the
> project progresses.
>
>
> And to keep rambling .....
>
> Insert the "status" column. Although "status" works against a duration
> driven version of %Complete (bad bad bad), it will tell you "late", "on
> schedule" and "complete."
>
> There have been so many posts on this forum of "I want to know what
> %Complete I should be at the status date." The answer is use
> Physical%Complete and compare it to SPI and that should be 1.00 for every
> task. If not, look for tasks with SPI less than 1. There is a very deep
> divide in this community over SPI and something called "Earned Schdule"
> ....
> but that takes a book to explain.
>
>
> --
> If this post was helpful, please consider rating it.
>
> Jim
>
> Visit http://project.mvps.org/ for FAQs and more information
> about Microsoft Project
>
>
>
> "Trevor Rabey" wrote:
>
>> You really need to get more specific about what you are measuring % of.
>> And even if you can measure it, and can find a numerator and denominator
>> to
>> get a % from, what does it mean and how can it be used for anything
>> useful.
>> Gross numbers with a % sign after them don't really help you to run the
>> project.
>> EV can spit out a cost and and schedule variance and a cost and and
>> schedule
>> performance index, and an EAC, but it doesn't tell you much about what
>> the
>> problem is, or even if there is one, what caused it, or what to do about
>> it.
>> That's all in the detail.
>> Variances, indexes and ratios don't tell you if you are doing the right
>> tasks. It is possible to get good EV numbers by doing the wrong tasks, ie
>> the noncritical instead of the critical tasks.
>>
>> Say the Task is to lay 10000 bricks in 10 days, that's 80 person-hours
>> for
>> one person at say $50/hour, ie $4000 plus $1/brick, $4000 + $10000 =
>> $14000,
>> or $1400/day.
>> At the end of Day 6, 3000 bricks might be laid, the bricklayer may have
>> worked 7 person-hours per day (42 Hours), perhaps he only cost $40/hour,
>> bricks might have cost $1.20 each ($3600).
>>
>> So you have bricks (30%), Duration (60%), Hours (42/80), Cost (((3000 x
>> $1.2) + (42 x 40))/84).
>> The denominator in each case here is the baseline total bricks, duration,
>> work and cost.
>> Another reasonable denominator would be the baseline (ie as planned)
>> bricks,
>> duration, work and cost up to the status date.
>> ie 3000/6000 for bricks, 6/6 for Duration, 42/48 for Work, $8400 for
>> Cost.
>>
>> Trevor Rabey 0407213955 61 8 92727485 PERFECT PROJECT PLANNING
>> www.perfectproject.com.au
>> "Jon S." <JonS@discussions.microsoft.com> wrote in message
>> news:9D16A050-1BC0-4851-A005-C4B9CA58EA27@microsoft.com...
>> > I'd like to easily report what percent of work is complete as expected
>> > to
>> > be
>> > by date. Say I have tasks to be finished last week, this week, and
>> > next
>> > week. Some of my tasks from last week are done, but not all, yet some
>> > of
>> > my
>> > tasks to start next week have begun. Can project easily calculate some
>> > %
>> > complete to Date? Or atleast, even if it doesn't give me credit for
>> > that
>> > work I am ahead on, can it tell me if I'm up to date with what I said I
>> > would
>> > be by now?
>> >
>> > Jon S.
>>
>>
>>

SwapnaK
Mon May 05 08:37:04 CDT 2008


% COMPLETE CALCULATION

Hi, I am not understanding the % Complete Calculation.

I have reported status at the Summary Task level. Not understanding how
it's distributing status to it's subtasks. Is it done based on dates or
duration or work?

Please help.


--
SwapnaK
------------------------------------------------------------------------
SwapnaK's Profile: http://forums.techarena.in/member.php?userid=48275
View this thread: http://forums.techarena.in/showthread.php?t=960860

http://forums.techarena.in

JonS
Mon May 05 09:40:01 CDT 2008

All,

Thanks for your response. I'll try a few things, see where it gets me, and
post some comments later today or tomorrow.

Trevor,

You are exactly right. So I've got 5 metrics I'm tracking for management
reporting and scorecard. Two of which point to timing, and if we are on time
or not. I'm taking a straight count of deliverables completed vs. total due.
Then I wanted to do a % complete. Though in my project plan we are not
tracking dollars, or even work, I will have duration, which I can use as a
scale for work. If i do % complete i'll give myself a weighted completion
average, which will balance with # of tasks completed. I thought I'd take it
a step further, and look for a way to show % complete of what should be
complete.

Consider I have 3 tasks each with 4 sub-tasks but some sub tasks may take
longer than others. The 3 tasks are to begin and end at different times. At
any time during this project, I'd like to know of the work that should be
done, how much is? If I could attached a sample project to show my point, I
think it would help.

Second best would be to show % complete of work that should be completed,
but what would be best is to get credit for work that is completed ahead of
schedule.

"Trevor Rabey" wrote:

> Jim,
> You may as well write the book, since you've started it.
>
> Jon,
> We get a lot of questions here, sometimes almost the same question several
> times on the same day, and a flurry recently, about how to measure progress.
> Usually it is along the lines of "how are we doing compared to how we wanted
> to be doing?".
> I think, and I could be wrong, the problem is that everyone wants to boil a
> complicated situation down to a description in one number.
> Perhaps this is because everyone has senior management to report to but
> senior management doesn't seem to have the necessary attention span to
> understand anything other than one number, so you may as well cook up a
> number which is palatable and immediately acceptable so you don't have to
> explain it.
> There is a lot more to how a game is going, and how it is likely to end up,
> than just the score, eg:
> How much time is left?
> Who is injured and off the field?
> Which team has the best reserves?
>
> Trevor Rabey 0407213955 61 8 92727485 PERFECT PROJECT PLANNING
> www.perfectproject.com.au
> "Jim Aksel" <JimAksel@discussions.microsoft.com> wrote in message
> news:3D457A98-B2A9-453F-8B13-6A865D02273F@microsoft.com...
> >I wish to offer some respectful difference of opinion with Trevor. He and
> >I
> > do this occassionaly, I think deep down we both agree but come to it from
> > a
> > different angle. My personal opinion is that Physical%Complete can be
> > "morphed" to accomodate exactly what Trevor is getting at. If you want to
> > lay 1000 bricks and your are going to take credit "by the brick" then make
> > that your measure and key in the value to Physical%Complete. It is all how
> > you define the numerator and denominator.
> >
> > If EV is taken properly, then indicies (ratios), variance, etc. are
> > actually
> > meaningful. Trevor and I agree 100% that doing the "wrong" tasks will
> > artifically pad the indicies and variances.... especially at the higher
> > levels of summary. This is why I don't like a lot of the "trip wire"
> > indices
> > touted here in the United States.
> >
> > To properly do EV, you absolutely must look at the details, and at the
> > appropriate level of detail. The method must be applied consistently.
> > Trevor
> > shows you how to get different answers using the various measures of
> > bricks,
> > work, etc. Bottom line is that you have to have a specifically definable
> > measure of success... we document that in an Earned Value Managment Plan
> > for
> > every work package. We plan work to the level that below the work package
> > level tasks must be completely binary: Lay 12 bricks. There were either
> > laid or they weren't. The %Complete is either 0 or 100%. This is
> > essentially a milestone technique. Doing things this way will allow a
> > consistent measure of Physical%Complete at the Work Package levels. We
> > take
> > earned value weekly.
> >
> > Now, the reason why I posted a second time.
> >
> > The original post asked about status of the schedule.... items that should
> > be complete are not, items that should start "tomorrow" have already
> > started.
> > This is a schedule maintenance issue. Here's the four rules we use:
> >
> > Rule 0: Establish a status date (Project/Project Information/Status Date)
> > Rule 1: There is no such thing as an unstarted task to the left of the
> > status date. Establish a new date to the right of the status date.
> > Rule 2: There is no such thing as unfinished tasks to the left of the
> > status
> > date. If the task is not complete, increase the duration so that it
> > finishes
> > after the status date. You get that information from the task owner.
> > Rule 3: Nothing starts in the future. If a task has a %Complete there is
> > no
> > chance it started to the right of the status date. Ask the task owner
> > when
> > it started and enter that date in the "Actual Start" column, not "Start".
> > Then post your %Complete making sure the task owner has identified when
> > they
> > believe they will finsih.
> > Rule 4: Nothing is finished after the status date. If the schedule says
> > the
> > task finishes "next Tuesday" (to the right of the status date) and the
> > task
> > owner says the work is complete, then it didn't finish next Tuesday did
> > it?
> > It finished on or before the status date, so key that into the "Actual
> > Finish" column (not Finish) and claim 100%.
> >
> > Something else to think about. The "brute force" logic in MS Project
> > tells
> > us if two tasks are linked as Finish To Start, then if B follows A, then A
> > must be 100% before B can start. If B has started, then "officially" you
> > have a logic error in the relationship. Most often this means you do not
> > have enough detail in your schedule. An example may be software testing
> > cannot start until all the coding is complete. That is not entirely
> > true...
> > If you are inventing "Microsoft Office" you can certainly start testing
> > "Word" while they finsih coding "PowerPoint".... true, you can't test the
> > integration of the two, but you can certainly see if Word will do some of
> > its
> > functions.
> >
> > We run into this type of logic all the time. As such, you should be
> > prepared to add additional detail to your schedule as you "learn" when the
> > project progresses.
> >
> >
> > And to keep rambling .....
> >
> > Insert the "status" column. Although "status" works against a duration
> > driven version of %Complete (bad bad bad), it will tell you "late", "on
> > schedule" and "complete."
> >
> > There have been so many posts on this forum of "I want to know what
> > %Complete I should be at the status date." The answer is use
> > Physical%Complete and compare it to SPI and that should be 1.00 for every
> > task. If not, look for tasks with SPI less than 1. There is a very deep
> > divide in this community over SPI and something called "Earned Schdule"
> > ....
> > but that takes a book to explain.
> >
> >
> > --
> > If this post was helpful, please consider rating it.
> >
> > Jim
> >
> > Visit http://project.mvps.org/ for FAQs and more information
> > about Microsoft Project
> >
> >
> >
> > "Trevor Rabey" wrote:
> >
> >> You really need to get more specific about what you are measuring % of.
> >> And even if you can measure it, and can find a numerator and denominator
> >> to
> >> get a % from, what does it mean and how can it be used for anything
> >> useful.
> >> Gross numbers with a % sign after them don't really help you to run the
> >> project.
> >> EV can spit out a cost and and schedule variance and a cost and and
> >> schedule
> >> performance index, and an EAC, but it doesn't tell you much about what
> >> the
> >> problem is, or even if there is one, what caused it, or what to do about
> >> it.
> >> That's all in the detail.
> >> Variances, indexes and ratios don't tell you if you are doing the right
> >> tasks. It is possible to get good EV numbers by doing the wrong tasks, ie
> >> the noncritical instead of the critical tasks.
> >>
> >> Say the Task is to lay 10000 bricks in 10 days, that's 80 person-hours
> >> for
> >> one person at say $50/hour, ie $4000 plus $1/brick, $4000 + $10000 =
> >> $14000,
> >> or $1400/day.
> >> At the end of Day 6, 3000 bricks might be laid, the bricklayer may have
> >> worked 7 person-hours per day (42 Hours), perhaps he only cost $40/hour,
> >> bricks might have cost $1.20 each ($3600).
> >>
> >> So you have bricks (30%), Duration (60%), Hours (42/80), Cost (((3000 x
> >> $1.2) + (42 x 40))/84).
> >> The denominator in each case here is the baseline total bricks, duration,
> >> work and cost.
> >> Another reasonable denominator would be the baseline (ie as planned)
> >> bricks,
> >> duration, work and cost up to the status date.
> >> ie 3000/6000 for bricks, 6/6 for Duration, 42/48 for Work, $8400 for
> >> Cost.
> >>
> >> Trevor Rabey 0407213955 61 8 92727485 PERFECT PROJECT PLANNING
> >> www.perfectproject.com.au
> >> "Jon S." <JonS@discussions.microsoft.com> wrote in message
> >> news:9D16A050-1BC0-4851-A005-C4B9CA58EA27@microsoft.com...
> >> > I'd like to easily report what percent of work is complete as expected
> >> > to
> >> > be
> >> > by date. Say I have tasks to be finished last week, this week, and
> >> > next
> >> > week. Some of my tasks from last week are done, but not all, yet some
> >> > of
> >> > my
> >> > tasks to start next week have begun. Can project easily calculate some
> >> > %
> >> > complete to Date? Or atleast, even if it doesn't give me credit for
> >> > that
> >> > work I am ahead on, can it tell me if I'm up to date with what I said I
> >> > would
> >> > be by now?
> >> >
> >> > Jon S.
> >>
> >>
> >>
>
>
>

JonS
Mon May 05 10:19:01 CDT 2008

Jim,

SPI might work for what I need. Though I would need to know how it is
calculated. I set a status date, though all SPI shows 0 on a test project
plan I am experimenting on.

Any where you can point me?

Jon

"Jim Aksel" wrote:

> Pardon the bump in. Using Earned Value techniques, Micrsofto Project
> automatically calculates Schedule Performance Index [SPI] and Cost
> Performance Index [CPI]. Insert those columns.
>
> SPI=Measure of Work Completed / Measure of Work Scheduled BCWP/BCWS. Above
> 1 is ahead of schedule, 1.0 is on schedule, less than 1.0 is behind schedule.
> The value is usually displayed as a percent 100%=1.0
>
> CPI=Measure of Work Completed/Actual Costs BCWP/ACWP. Above 100% is below
> cost. CPI<100% means costs are exceeding plan... the work may be tougher
> than you thought.
>
> For our purposes BCWP=(%Complete)*(BAC).
>
> Now we get crazy -- MS Project can calculate using two methods %COmplete and
> Physical%Complete. %Complete deals only with duration and nothing else. If
> you claim 25% Complete you are saying your project is 1/4 of the time it will
> take to complete the task. Very few projects really work this way.
>
> However, when most people say they are 25% complete, they mean they have
> completed 25% of the work involved or 25% of the value (I have laid 25 of
> 100 bricks). Both of these measure are independent of duration. For
> example, work may not produce value or cost in a linear fashion. Planting
> corn has labor up front, while the corn grows there is less work, and
> harvesting includes more work.
>
> Generally, Physical%Complete is the most correct method to use in most
> instances.
> --
> If this post was helpful, please consider rating it.
>
> Jim
>
> Visit http://project.mvps.org/ for FAQs and more information
> about Microsoft Project
>
>
>
> "Jon S." wrote:
>
> > I'd like to easily report what percent of work is complete as expected to be
> > by date. Say I have tasks to be finished last week, this week, and next
> > week. Some of my tasks from last week are done, but not all, yet some of my
> > tasks to start next week have begun. Can project easily calculate some %
> > complete to Date? Or atleast, even if it doesn't give me credit for that
> > work I am ahead on, can it tell me if I'm up to date with what I said I would
> > be by now?
> >
> > Jon S.

Trevor
Mon May 05 20:34:50 CDT 2008

Jon, you can send me a small file if you like, but I can imagine it anyway.
You start off saying that your tasks don't have Work (or Costs) and then go
on to discuss "Work" anyway.
If you only have duration, that's all you can track.
Your phrase "% complete of work" indicates an worrying degree of confusion.
% Complete in MSP is about Duration only (I would like $1 for every time I
have had to say that).
Perhaps instead of "Work" you mean the Task itself.

I think you are having some trouble defining what "achievement", "progress"
and "success" are.
Of course, you are tempted to define whatever you "achieve" as "progress"
and "success".
This is just a distraction and could easily lead to a delusional situation
in which all of your scores and KPIs say that everything is rosy while your
project is crumbling around you.
When they interview the losing captain after the game and he says something
like "we lost but we played better because we had the ball more often and
ran faster" we all know he is just kidding himself.
Think Enron. Think various recent wars.

Not all tasks are equal.
There is no credit for doing 50 tasks before they were scheduled if the
really important one that should have been done last week is forgooten or
bogged down.
Doing tasks early could just be the result of ignoring or sabotaging the
plan, or it could just indicate the plan was very wrong in the first place,
especially if it turns out that successors are being done before their
indicated predecessors.

Trevor Rabey 0407213955 61 8 92727485 PERFECT PROJECT PLANNING
www.perfectproject.com.au
"Jon S." <JonS@discussions.microsoft.com> wrote in message
news:B144F4F4-6378-486C-8E52-92C795D6F150@microsoft.com...
> All,
>
> Thanks for your response. I'll try a few things, see where it gets me,
> and
> post some comments later today or tomorrow.
>
> Trevor,
>
> You are exactly right. So I've got 5 metrics I'm tracking for management
> reporting and scorecard. Two of which point to timing, and if we are on
> time
> or not. I'm taking a straight count of deliverables completed vs. total
> due.
> Then I wanted to do a % complete. Though in my project plan we are not
> tracking dollars, or even work, I will have duration, which I can use as a
> scale for work. If i do % complete i'll give myself a weighted completion
> average, which will balance with # of tasks completed. I thought I'd take
> it
> a step further, and look for a way to show % complete of what should be
> complete.
>
> Consider I have 3 tasks each with 4 sub-tasks but some sub tasks may take
> longer than others. The 3 tasks are to begin and end at different times.
> At
> any time during this project, I'd like to know of the work that should be
> done, how much is? If I could attached a sample project to show my point,
> I
> think it would help.
>
> Second best would be to show % complete of work that should be completed,
> but what would be best is to get credit for work that is completed ahead
> of
> schedule.
>
> "Trevor Rabey" wrote:
>
>> Jim,
>> You may as well write the book, since you've started it.
>>
>> Jon,
>> We get a lot of questions here, sometimes almost the same question
>> several
>> times on the same day, and a flurry recently, about how to measure
>> progress.
>> Usually it is along the lines of "how are we doing compared to how we
>> wanted
>> to be doing?".
>> I think, and I could be wrong, the problem is that everyone wants to boil
>> a
>> complicated situation down to a description in one number.
>> Perhaps this is because everyone has senior management to report to but
>> senior management doesn't seem to have the necessary attention span to
>> understand anything other than one number, so you may as well cook up a
>> number which is palatable and immediately acceptable so you don't have to
>> explain it.
>> There is a lot more to how a game is going, and how it is likely to end
>> up,
>> than just the score, eg:
>> How much time is left?
>> Who is injured and off the field?
>> Which team has the best reserves?
>>
>> Trevor Rabey 0407213955 61 8 92727485 PERFECT PROJECT PLANNING
>> www.perfectproject.com.au
>> "Jim Aksel" <JimAksel@discussions.microsoft.com> wrote in message
>> news:3D457A98-B2A9-453F-8B13-6A865D02273F@microsoft.com...
>> >I wish to offer some respectful difference of opinion with Trevor. He
>> >and
>> >I
>> > do this occassionaly, I think deep down we both agree but come to it
>> > from
>> > a
>> > different angle. My personal opinion is that Physical%Complete can be
>> > "morphed" to accomodate exactly what Trevor is getting at. If you want
>> > to
>> > lay 1000 bricks and your are going to take credit "by the brick" then
>> > make
>> > that your measure and key in the value to Physical%Complete. It is all
>> > how
>> > you define the numerator and denominator.
>> >
>> > If EV is taken properly, then indicies (ratios), variance, etc. are
>> > actually
>> > meaningful. Trevor and I agree 100% that doing the "wrong" tasks will
>> > artifically pad the indicies and variances.... especially at the higher
>> > levels of summary. This is why I don't like a lot of the "trip wire"
>> > indices
>> > touted here in the United States.
>> >
>> > To properly do EV, you absolutely must look at the details, and at the
>> > appropriate level of detail. The method must be applied consistently.
>> > Trevor
>> > shows you how to get different answers using the various measures of
>> > bricks,
>> > work, etc. Bottom line is that you have to have a specifically
>> > definable
>> > measure of success... we document that in an Earned Value Managment
>> > Plan
>> > for
>> > every work package. We plan work to the level that below the work
>> > package
>> > level tasks must be completely binary: Lay 12 bricks. There were
>> > either
>> > laid or they weren't. The %Complete is either 0 or 100%. This is
>> > essentially a milestone technique. Doing things this way will allow a
>> > consistent measure of Physical%Complete at the Work Package levels. We
>> > take
>> > earned value weekly.
>> >
>> > Now, the reason why I posted a second time.
>> >
>> > The original post asked about status of the schedule.... items that
>> > should
>> > be complete are not, items that should start "tomorrow" have already
>> > started.
>> > This is a schedule maintenance issue. Here's the four rules we use:
>> >
>> > Rule 0: Establish a status date (Project/Project Information/Status
>> > Date)
>> > Rule 1: There is no such thing as an unstarted task to the left of the
>> > status date. Establish a new date to the right of the status date.
>> > Rule 2: There is no such thing as unfinished tasks to the left of the
>> > status
>> > date. If the task is not complete, increase the duration so that it
>> > finishes
>> > after the status date. You get that information from the task owner.
>> > Rule 3: Nothing starts in the future. If a task has a %Complete there
>> > is
>> > no
>> > chance it started to the right of the status date. Ask the task owner
>> > when
>> > it started and enter that date in the "Actual Start" column, not
>> > "Start".
>> > Then post your %Complete making sure the task owner has identified when
>> > they
>> > believe they will finsih.
>> > Rule 4: Nothing is finished after the status date. If the schedule
>> > says
>> > the
>> > task finishes "next Tuesday" (to the right of the status date) and the
>> > task
>> > owner says the work is complete, then it didn't finish next Tuesday did
>> > it?
>> > It finished on or before the status date, so key that into the "Actual
>> > Finish" column (not Finish) and claim 100%.
>> >
>> > Something else to think about. The "brute force" logic in MS Project
>> > tells
>> > us if two tasks are linked as Finish To Start, then if B follows A,
>> > then A
>> > must be 100% before B can start. If B has started, then "officially"
>> > you
>> > have a logic error in the relationship. Most often this means you do
>> > not
>> > have enough detail in your schedule. An example may be software
>> > testing
>> > cannot start until all the coding is complete. That is not entirely
>> > true...
>> > If you are inventing "Microsoft Office" you can certainly start testing
>> > "Word" while they finsih coding "PowerPoint".... true, you can't test
>> > the
>> > integration of the two, but you can certainly see if Word will do some
>> > of
>> > its
>> > functions.
>> >
>> > We run into this type of logic all the time. As such, you should be
>> > prepared to add additional detail to your schedule as you "learn" when
>> > the
>> > project progresses.
>> >
>> >
>> > And to keep rambling .....
>> >
>> > Insert the "status" column. Although "status" works against a duration
>> > driven version of %Complete (bad bad bad), it will tell you "late", "on
>> > schedule" and "complete."
>> >
>> > There have been so many posts on this forum of "I want to know what
>> > %Complete I should be at the status date." The answer is use
>> > Physical%Complete and compare it to SPI and that should be 1.00 for
>> > every
>> > task. If not, look for tasks with SPI less than 1. There is a very
>> > deep
>> > divide in this community over SPI and something called "Earned Schdule"
>> > ....
>> > but that takes a book to explain.
>> >
>> >
>> > --
>> > If this post was helpful, please consider rating it.
>> >
>> > Jim
>> >
>> > Visit http://project.mvps.org/ for FAQs and more information
>> > about Microsoft Project
>> >
>> >
>> >
>> > "Trevor Rabey" wrote:
>> >
>> >> You really need to get more specific about what you are measuring %
>> >> of.
>> >> And even if you can measure it, and can find a numerator and
>> >> denominator
>> >> to
>> >> get a % from, what does it mean and how can it be used for anything
>> >> useful.
>> >> Gross numbers with a % sign after them don't really help you to run
>> >> the
>> >> project.
>> >> EV can spit out a cost and and schedule variance and a cost and and
>> >> schedule
>> >> performance index, and an EAC, but it doesn't tell you much about what
>> >> the
>> >> problem is, or even if there is one, what caused it, or what to do
>> >> about
>> >> it.
>> >> That's all in the detail.
>> >> Variances, indexes and ratios don't tell you if you are doing the
>> >> right
>> >> tasks. It is possible to get good EV numbers by doing the wrong tasks,
>> >> ie
>> >> the noncritical instead of the critical tasks.
>> >>
>> >> Say the Task is to lay 10000 bricks in 10 days, that's 80 person-hours
>> >> for
>> >> one person at say $50/hour, ie $4000 plus $1/brick, $4000 + $10000 =
>> >> $14000,
>> >> or $1400/day.
>> >> At the end of Day 6, 3000 bricks might be laid, the bricklayer may
>> >> have
>> >> worked 7 person-hours per day (42 Hours), perhaps he only cost
>> >> $40/hour,
>> >> bricks might have cost $1.20 each ($3600).
>> >>
>> >> So you have bricks (30%), Duration (60%), Hours (42/80), Cost (((3000
>> >> x
>> >> $1.2) + (42 x 40))/84).
>> >> The denominator in each case here is the baseline total bricks,
>> >> duration,
>> >> work and cost.
>> >> Another reasonable denominator would be the baseline (ie as planned)
>> >> bricks,
>> >> duration, work and cost up to the status date.
>> >> ie 3000/6000 for bricks, 6/6 for Duration, 42/48 for Work, $8400 for
>> >> Cost.
>> >>
>> >> Trevor Rabey 0407213955 61 8 92727485 PERFECT PROJECT PLANNING
>> >> www.perfectproject.com.au
>> >> "Jon S." <JonS@discussions.microsoft.com> wrote in message
>> >> news:9D16A050-1BC0-4851-A005-C4B9CA58EA27@microsoft.com...
>> >> > I'd like to easily report what percent of work is complete as
>> >> > expected
>> >> > to
>> >> > be
>> >> > by date. Say I have tasks to be finished last week, this week, and
>> >> > next
>> >> > week. Some of my tasks from last week are done, but not all, yet
>> >> > some
>> >> > of
>> >> > my
>> >> > tasks to start next week have begun. Can project easily calculate
>> >> > some
>> >> > %
>> >> > complete to Date? Or atleast, even if it doesn't give me credit for
>> >> > that
>> >> > work I am ahead on, can it tell me if I'm up to date with what I
>> >> > said I
>> >> > would
>> >> > be by now?
>> >> >
>> >> > Jon S.
>> >>
>> >>
>> >>
>>
>>
>>

JonS
Mon May 05 22:28:00 CDT 2008

Trevor,

Not sure what I said to set you off, but clearly your tone went from trying
to help, to wanting to lecture.

I might have bounced around a bit with some terms, but thought my point was
clear enough to guide suggested solutions.

Obviously there is "work" and "cost" on this project, though I won't need to
track them in MSP for this specific project. I understand that % Complete is
a calculation of duration, and understand each of the pitfalls that could
arise. Beyond counting 100 tasks complete out of 200 total tasks as 50%
done, I thought I could improve. So using duration as a SCALE (relative
magnitude) of work I could give a better picture of project status. I
thought to take it a step further, I could use a percent complete and an on
task metric all in one. It could be calculated, so I thought MSP had an easy
option. Clearly there isn't. And although you claim that there is no merit
for "doing 50 tasks before they are scheduled" I would say there are
situations where it is worth acknowledging. If you are loading a moving van
with things from your house, and your plan outlines that you are to do each
room at a time, starting from the front and working towards the back,
wouldn't you want to take into account the few pieces that are packed ahead
of schedule? It happens in all projects, and no amount of project planning
can prevent situations like this.

Thanks for your help and consideration on my topic, but if you don't want to
help me figure this out, you can save your preaching for another newbie.
Otherwise, I look forward to your ideas.

"Trevor Rabey" wrote:

> Jon, you can send me a small file if you like, but I can imagine it anyway.
> You start off saying that your tasks don't have Work (or Costs) and then go
> on to discuss "Work" anyway.
> If you only have duration, that's all you can track.
> Your phrase "% complete of work" indicates an worrying degree of confusion.
> % Complete in MSP is about Duration only (I would like $1 for every time I
> have had to say that).
> Perhaps instead of "Work" you mean the Task itself.
>
> I think you are having some trouble defining what "achievement", "progress"
> and "success" are.
> Of course, you are tempted to define whatever you "achieve" as "progress"
> and "success".
> This is just a distraction and could easily lead to a delusional situation
> in which all of your scores and KPIs say that everything is rosy while your
> project is crumbling around you.
> When they interview the losing captain after the game and he says something
> like "we lost but we played better because we had the ball more often and
> ran faster" we all know he is just kidding himself.
> Think Enron. Think various recent wars.
>
> Not all tasks are equal.
> There is no credit for doing 50 tasks before they were scheduled if the
> really important one that should have been done last week is forgooten or
> bogged down.
> Doing tasks early could just be the result of ignoring or sabotaging the
> plan, or it could just indicate the plan was very wrong in the first place,
> especially if it turns out that successors are being done before their
> indicated predecessors.
>
> Trevor Rabey 0407213955 61 8 92727485 PERFECT PROJECT PLANNING
> www.perfectproject.com.au
> "Jon S." <JonS@discussions.microsoft.com> wrote in message
> news:B144F4F4-6378-486C-8E52-92C795D6F150@microsoft.com...
> > All,
> >
> > Thanks for your response. I'll try a few things, see where it gets me,
> > and
> > post some comments later today or tomorrow.
> >
> > Trevor,
> >
> > You are exactly right. So I've got 5 metrics I'm tracking for management
> > reporting and scorecard. Two of which point to timing, and if we are on
> > time
> > or not. I'm taking a straight count of deliverables completed vs. total
> > due.
> > Then I wanted to do a % complete. Though in my project plan we are not
> > tracking dollars, or even work, I will have duration, which I can use as a
> > scale for work. If i do % complete i'll give myself a weighted completion
> > average, which will balance with # of tasks completed. I thought I'd take
> > it
> > a step further, and look for a way to show % complete of what should be
> > complete.
> >
> > Consider I have 3 tasks each with 4 sub-tasks but some sub tasks may take
> > longer than others. The 3 tasks are to begin and end at different times.
> > At
> > any time during this project, I'd like to know of the work that should be
> > done, how much is? If I could attached a sample project to show my point,
> > I
> > think it would help.
> >
> > Second best would be to show % complete of work that should be completed,
> > but what would be best is to get credit for work that is completed ahead
> > of
> > schedule.
> >
> > "Trevor Rabey" wrote:
> >
> >> Jim,
> >> You may as well write the book, since you've started it.
> >>
> >> Jon,
> >> We get a lot of questions here, sometimes almost the same question
> >> several
> >> times on the same day, and a flurry recently, about how to measure
> >> progress.
> >> Usually it is along the lines of "how are we doing compared to how we
> >> wanted
> >> to be doing?".
> >> I think, and I could be wrong, the problem is that everyone wants to boil
> >> a
> >> complicated situation down to a description in one number.
> >> Perhaps this is because everyone has senior management to report to but
> >> senior management doesn't seem to have the necessary attention span to
> >> understand anything other than one number, so you may as well cook up a
> >> number which is palatable and immediately acceptable so you don't have to
> >> explain it.
> >> There is a lot more to how a game is going, and how it is likely to end
> >> up,
> >> than just the score, eg:
> >> How much time is left?
> >> Who is injured and off the field?
> >> Which team has the best reserves?
> >>
> >> Trevor Rabey 0407213955 61 8 92727485 PERFECT PROJECT PLANNING
> >> www.perfectproject.com.au
> >> "Jim Aksel" <JimAksel@discussions.microsoft.com> wrote in message
> >> news:3D457A98-B2A9-453F-8B13-6A865D02273F@microsoft.com...
> >> >I wish to offer some respectful difference of opinion with Trevor. He
> >> >and
> >> >I
> >> > do this occassionaly, I think deep down we both agree but come to it
> >> > from
> >> > a
> >> > different angle. My personal opinion is that Physical%Complete can be
> >> > "morphed" to accomodate exactly what Trevor is getting at. If you want
> >> > to
> >> > lay 1000 bricks and your are going to take credit "by the brick" then
> >> > make
> >> > that your measure and key in the value to Physical%Complete. It is all
> >> > how
> >> > you define the numerator and denominator.
> >> >
> >> > If EV is taken properly, then indicies (ratios), variance, etc. are
> >> > actually
> >> > meaningful. Trevor and I agree 100% that doing the "wrong" tasks will
> >> > artifically pad the indicies and variances.... especially at the higher
> >> > levels of summary. This is why I don't like a lot of the "trip wire"
> >> > indices
> >> > touted here in the United States.
> >> >
> >> > To properly do EV, you absolutely must look at the details, and at the
> >> > appropriate level of detail. The method must be applied consistently.
> >> > Trevor
> >> > shows you how to get different answers using the various measures of
> >> > bricks,
> >> > work, etc. Bottom line is that you have to have a specifically
> >> > definable
> >> > measure of success... we document that in an Earned Value Managment
> >> > Plan
> >> > for
> >> > every work package. We plan work to the level that below the work
> >> > package
> >> > level tasks must be completely binary: Lay 12 bricks. There were
> >> > either
> >> > laid or they weren't. The %Complete is either 0 or 100%. This is
> >> > essentially a milestone technique. Doing things this way will allow a
> >> > consistent measure of Physical%Complete at the Work Package levels. We
> >> > take
> >> > earned value weekly.
> >> >
> >> > Now, the reason why I posted a second time.
> >> >
> >> > The original post asked about status of the schedule.... items that
> >> > should
> >> > be complete are not, items that should start "tomorrow" have already
> >> > started.
> >> > This is a schedule maintenance issue. Here's the four rules we use:
> >> >
> >> > Rule 0: Establish a status date (Project/Project Information/Status
> >> > Date)
> >> > Rule 1: There is no such thing as an unstarted task to the left of the
> >> > status date. Establish a new date to the right of the status date.
> >> > Rule 2: There is no such thing as unfinished tasks to the left of the
> >> > status
> >> > date. If the task is not complete, increase the duration so that it
> >> > finishes
> >> > after the status date. You get that information from the task owner.
> >> > Rule 3: Nothing starts in the future. If a task has a %Complete there
> >> > is
> >> > no
> >> > chance it started to the right of the status date. Ask the task owner
> >> > when
> >> > it started and enter that date in the "Actual Start" column, not
> >> > "Start".
> >> > Then post your %Complete making sure the task owner has identified when
> >> > they
> >> > believe they will finsih.
> >> > Rule 4: Nothing is finished after the status date. If the schedule
> >> > says
> >> > the
> >> > task finishes "next Tuesday" (to the right of the status date) and the
> >> > task
> >> > owner says the work is complete, then it didn't finish next Tuesday did
> >> > it?
> >> > It finished on or before the status date, so key that into the "Actual
> >> > Finish" column (not Finish) and claim 100%.
> >> >
> >> > Something else to think about. The "brute force" logic in MS Project
> >> > tells
> >> > us if two tasks are linked as Finish To Start, then if B follows A,
> >> > then A
> >> > must be 100% before B can start. If B has started, then "officially"
> >> > you
> >> > have a logic error in the relationship. Most often this means you do
> >> > not
> >> > have enough detail in your schedule. An example may be software
> >> > testing
> >> > cannot start until all the coding is complete. That is not entirely
> >> > true...
> >> > If you are inventing "Microsoft Office" you can certainly start testing
> >> > "Word" while they finsih coding "PowerPoint".... true, you can't test
> >> > the
> >> > integration of the two, but you can certainly see if Word will do some
> >> > of
> >> > its
> >> > functions.
> >> >
> >> > We run into this type of logic all the time. As such, you should be
> >> > prepared to add additional detail to your schedule as you "learn" when
> >> > the
> >> > project progresses.
> >> >
> >> >
> >> > And to keep rambling .....
> >> >
> >> > Insert the "status" column. Although "status" works against a duration
> >> > driven version of %Complete (bad bad bad), it will tell you "late", "on
> >> > schedule" and "complete."
> >> >
> >> > There have been so many posts on this forum of "I want to know what
> >> > %Complete I should be at the status date." The answer is use
> >> > Physical%Complete and compare it to SPI and that should be 1.00 for
> >> > every
> >> > task. If not, look for tasks with SPI less than 1. There is a very
> >> > deep
> >> > divide in this community over SPI and something called "Earned Schdule"
> >> > ....
> >> > but that takes a book to explain.
> >> >
> >> >
> >> > --
> >> > If this post was helpful, please consider rating it.
> >> >
> >> > Jim
> >> >
> >> > Visit http://project.mvps.org/ for FAQs and more information
> >> > about Microsoft Project
> >> >
> >> >
> >> >
> >> > "Trevor Rabey" wrote:
> >> >
> >> >> You really need to get more specific about what you are measuring %
> >> >> of.
> >> >> And even if you can measure it, and can find a numerator and
> >> >> denominator
> >> >> to
> >> >> get a % from, what does it mean and how can it be used for anything
> >> >> useful.
> >> >> Gross numbers with a % sign after them don't really help you to run
> >> >> the
> >> >> project.
> >> >> EV can spit out a cost and and schedule variance and a cost and and
> >> >> schedule
> >> >> performance index, and an EAC, but it doesn't tell you much about what
> >> >> the
> >> >> problem is, or even if there is one, what caused it, or what to do
> >> >> about
> >> >> it.
> >> >> That's all in the detail.
> >> >> Variances, indexes and ratios don't tell you if you are doing the
> >> >> right
> >> >> tasks. It is possible to get good EV numbers by doing the wrong tasks,
> >> >> ie
> >> >> the noncritical instead of the critical tasks.
> >> >>
> >> >> Say the Task is to lay 10000 bricks in 10 days, that's 80 person-hours
> >> >> for
> >> >> one person at say $50/hour, ie $4000 plus $1/brick, $4000 + $10000 =
> >> >> $14000,
> >> >> or $1400/day.
> >> >> At the end of Day 6, 3000 bricks might be laid, the bricklayer may
> >> >> have
> >> >> worked 7 person-hours per day (42 Hours), perhaps he only cost
> >> >> $40/hour,
> >> >> bricks might have cost $1.20 each ($3600).
> >> >>
> >> >> So you have bricks (30%), Duration (60%), Hours (42/80), Cost (((3000
> >> >> x
> >> >> $1.2) + (42 x 40))/84).
> >> >> The denominator in each case here is the baseline total bricks,
> >> >> duration,
> >> >> work and cost.
> >> >> Another reasonable denominator would be the baseline (ie as planned)
> >> >> bricks,
> >> >> duration, work and cost up to the status date.
> >> >> ie 3000/6000 for bricks, 6/6 for Duration, 42/48 for Work, $8400 for
> >> >> Cost.
> >> >>
> >> >> Trevor Rabey 0407213955 61 8 92727485 PERFECT PROJECT PLANNING
> >> >> www.perfectproject.com.au
> >> >> "Jon S." <JonS@discussions.microsoft.com> wrote in message
> >> >> news:9D16A050-1BC0-4851-A005-C4B9CA58EA27@microsoft.com...
> >> >> > I'd like to easily report what percent of work is complete as
> >> >> > expected
> >> >> > to
> >> >> > be
> >> >> > by date. Say I have tasks to be finished last week, this week, and
> >> >> > next
> >> >> > week. Some of my tasks from last week are done, but not all, yet
> >> >> > some
> >> >> > of
> >> >> > my
> >> >> > tasks to start next week have begun. Can project easily calculate
> >> >> > some
> >> >> > %
> >> >> > complete to Date? Or atleast, even if it doesn't give me credit for

Trevor
Tue May 06 00:35:44 CDT 2008

Relax.
If it seems pedantic to be careful about the vocabulary it is only because
many people (not you, obviously) get themselves more and more confused just
because they are sloppy with the terms and the definitions, and it makes it
harder to answer the questions.
Here is what I would be concerned about if this was my problem: is it
do-able, and, if it is, is it worth doing, ie what's the up-front effort
compared to the pay-off?
It is hard to see how, no matter how you slice and dice it, you can read
very much more into duration than just plain duration, actual and remaining.
If you do tasks earlier than the plan says they can be done (it shouldn't be
possible, but it happens) then this will show up as an actual start date and
an actual finish date earlier than the planned earliest start date and the
planned earliest finish date and then, depending on what's the successors
are and what's critical or not, other tasks will be able to be done earlier
and maybe the project will be able to finish earlier.
So I think the metric that matters is how many days have been shaved off the
original planned project earliest finish date, and % of anything doesn't
come into it, besides perhaps being able to say that if the project was
planned to take 100 days but now looks like it will take only 75 days that's
a 25% reduction in overall duration.

Reporting "100 tasks complete out of 200 total tasks as 50% done" is
uninformative, ambiguous or even perhaps deliberately misleading.
The remaining 100 tasks might take 90% of the total duration.
If noncritical tasks are done early it's nice to have them in the bank but
it doesn't necessarily help the overall outcome.

We can calculate anything that we have the input data for but I am still not
sure of your goal.
I hope this is in some way helpful even if it is not a direct solution, and
I'm happy to keep trying to help but may need a bit more information.

Trevor Rabey 0407213955 61 8 92727485 PERFECT PROJECT PLANNING
www.perfectproject.com.au
"Jon S." <JonS@discussions.microsoft.com> wrote in message
news:9FB4ED64-4293-41F5-A8BA-7A6AAD27FE12@microsoft.com...
> Trevor,
>
> Not sure what I said to set you off, but clearly your tone went from
> trying
> to help, to wanting to lecture.
>
> I might have bounced around a bit with some terms, but thought my point
> was
> clear enough to guide suggested solutions.
>
> Obviously there is "work" and "cost" on this project, though I won't need
> to
> track them in MSP for this specific project. I understand that % Complete
> is
> a calculation of duration, and understand each of the pitfalls that could
> arise. Beyond counting 100 tasks complete out of 200 total tasks as 50%
> done, I thought I could improve. So using duration as a SCALE (relative
> magnitude) of work I could give a better picture of project status. I
> thought to take it a step further, I could use a percent complete and an
> on
> task metric all in one. It could be calculated, so I thought MSP had an
> easy
> option. Clearly there isn't. And although you claim that there is no
> merit
> for "doing 50 tasks before they are scheduled" I would say there are
> situations where it is worth acknowledging. If you are loading a moving
> van
> with things from your house, and your plan outlines that you are to do
> each
> room at a time, starting from the front and working towards the back,
> wouldn't you want to take into account the few pieces that are packed
> ahead
> of schedule? It happens in all projects, and no amount of project
> planning
> can prevent situations like this.
>
> Thanks for your help and consideration on my topic, but if you don't want
> to
> help me figure this out, you can save your preaching for another newbie.
> Otherwise, I look forward to your ideas.
>
> "Trevor Rabey" wrote:
>
>> Jon, you can send me a small file if you like, but I can imagine it
>> anyway.
>> You start off saying that your tasks don't have Work (or Costs) and then
>> go
>> on to discuss "Work" anyway.
>> If you only have duration, that's all you can track.
>> Your phrase "% complete of work" indicates an worrying degree of
>> confusion.
>> % Complete in MSP is about Duration only (I would like $1 for every time
>> I
>> have had to say that).
>> Perhaps instead of "Work" you mean the Task itself.
>>
>> I think you are having some trouble defining what "achievement",
>> "progress"
>> and "success" are.
>> Of course, you are tempted to define whatever you "achieve" as "progress"
>> and "success".
>> This is just a distraction and could easily lead to a delusional
>> situation
>> in which all of your scores and KPIs say that everything is rosy while
>> your
>> project is crumbling around you.
>> When they interview the losing captain after the game and he says
>> something
>> like "we lost but we played better because we had the ball more often and
>> ran faster" we all know he is just kidding himself.
>> Think Enron. Think various recent wars.
>>
>> Not all tasks are equal.
>> There is no credit for doing 50 tasks before they were scheduled if the
>> really important one that should have been done last week is forgooten or
>> bogged down.
>> Doing tasks early could just be the result of ignoring or sabotaging the
>> plan, or it could just indicate the plan was very wrong in the first
>> place,
>> especially if it turns out that successors are being done before their
>> indicated predecessors.
>>
>> Trevor Rabey 0407213955 61 8 92727485 PERFECT PROJECT PLANNING
>> www.perfectproject.com.au
>> "Jon S." <JonS@discussions.microsoft.com> wrote in message
>> news:B144F4F4-6378-486C-8E52-92C795D6F150@microsoft.com...
>> > All,
>> >
>> > Thanks for your response. I'll try a few things, see where it gets me,
>> > and
>> > post some comments later today or tomorrow.