Nonlinear distribution of cost for one activity

[nitsa]

In cases wher the cost of an activity is not uniform, is this possible, using
MS Project, to distribute cost in a nonlinear manner such as a bell shape or
a triangular curve distribution?

 

[Jan De Messemaeker]

Hi Dani,

As long as cost is calculated from resoruce's work, yes, because you can
contour the work of an assignment (double-click an assignment line in a
Usage view to get assignment information, you can select the contours)
HTH

 

[nitsa]

In other words you mean that if I have a task for which I entered three
values: start, finish and cost and no resource, then there is no way to
distribute the cost in a nonlinear way?

 

[Jan De Messemaeker]

Hi Nitsa,

First, apart from the cost issue, you may get unexpected results when
entering bith start and finish dates - normally you have both calculated by
Project (from durations and dependencies between tasks) oer at least only
enter satart a,d duration and have finish calculated.

This being said, no, Fixed cost (which is what you enter when you type in
cost without having resources) cannot be contoured.
I suggest you create a resource "cost" which you assign to the task to
create an assignment and then contour the work.

Hope this helps,

 

[davegb]

Hi Nitsa,

First, apart from the cost issue, you may get unexpected results when
entering bith start and finish dates - normally you have both calculated by
Project (from durations and dependencies between tasks) oer at least only
enter satart a,d duration and have finish calculated.

This being said, no, Fixed cost (which is what you enter when you type in
cost without having resources) cannot be contoured.
I suggest you create a resource "cost" which you assign to the task to
create an assignment and then contour the work.

Hope this helps,

I agree with the replies above. I'd also ask, what is the duration of
this variable cost rate task? If it's more than 2 weeks, that is a
problem. A good guideline for task durations is between 0.5 days and 10
days. Tasks shorter than this are micro-planning. Tasks longer than 10
days are very dangerous because you may not know how you're doing on
this task until it's 10 days late. That can be a problem in most any
schedule.

If the task's duration is longer than 10 days, I'd consider breaking it
down into multiple tasks. If you do that, you might use the "burn
rate", the rate the money is being spend, as a criteria for breaking it
down. Make a few tasks out of the one, each with a linear (or nearly
linear) burn rate. You've solved 2 problems at once.

I should also mention that if at some time you want to use Earned Value
Analysis, you will want every task to have a constant burn rate, or EVA
won't work.

Hope this helps in your world.

 

[nitsa]

Thank you.
Let me clarify my goal so you might be able to help me better:

I am trying to use MS Project in one or two levels above what it is normaly
used for. I have to monitor a project that is has a budget of $350K and a
duration of 3 years. The project is divided to roughly 60 tasks (ehch has a
duration and budget). I would like to build a plan for "cash flow" and then
to be able to track it using earned value method. Since I am providing the
funds, I would like to make sure that the project is on schedule and budget.

 

[Jan De Messemaeker]

Hi,

Good.
My advice should work for you.
HTH

 

[nitsa]

I tried what you suggested as a sample on one task:
I defined a resource called "cost" in the resource sheet and gave it a
standard rate of $1/hr.

I assigned this resource to the task of which duration was 10 and in the
"assignment information" I defined a bell shape work contour. The outcome
cost was $40.

If I wanted the cost to be a 100, the duration would increase to 25. And
since my planned duration and planned budgets are both given to me as an
input from "outside", all I could do is to change the rate. Am I working with
the right program?



I entered an activity, gave it a start date and a duration of 10 days. I

 

[davegb]

Thank you.
Let me clarify my goal so you might be able to help me better:

I am trying to use MS Project in one or two levels above what it is normaly
used for. I have to monitor a project that is has a budget of $350K and a
duration of 3 years. The project is divided to roughly 60 tasks (ehch has a
duration and budget). I would like to build a plan for "cash flow" and then
to be able to track it using earned value method. Since I am providing the
funds, I would like to make sure that the project is on schedule and budget.

As I mentioned above, EV doesn't work well if your tasks don't have
constant burn rates over the duration of the tasks. If you don't
believe this, do a simple example problem where you have tasks with
changing burn rates and see what happens. If you have lots of short
duration tasks, the changing burn rates might cancel out, if you're
lucky. That's one of several reasons for breaking a schedule down into
relatively small tasks.

If you have a task with a high to low burn rate, it will show poor cost
performance at first (lots of money being spend without appropriate EV)
and the reverse for tasks starting with a low burn rate which increases
over the life of the task. If you use Work % Complete, it helps if the
changing burn rate is due to increasing or decreasing labor usage. It
doesn't help if it's due to other costs which can't be accounted for by
using Work % Complete.

Hope this helps in your world.

 

[Jan De Messemaeker]

Hi,

First, if the durartion must be fixed, you better set the task type to fixed
duration.
Task Information, Avanced, Task Type
Assign the resoruce as you did
Select the contour as you did
Change Work for the assignment.
That's it.

You're not using the program for what it i designed to do so it is not
abnormal you have to enme more data is it?

 

[Steve House [Project MVP]]

IMHO, you're working backwards. Until you have explicitly defined the tasks
and the resources performing them, you don't know, CAN'T know, that the
project budget is $350k and will take 3 years. You have not set a budget or
timeline, you have set some parameters within which you hope you can
accomplish some business objective. The budget and timeline is an estimate
of what will really be required to achieve those objectives. So far, you
have set some goals based on the anticipated availability of assets, in
other words, a distribution of projected revenues. The actual budget may be
under or may exceed those desired parameters.

Don't define an arbitrary cost and then expect the work to expand or
contract to fit. A task creates a specific, concrete, deliverable. The
task will take how ever long it takes to complete exactly that deliverable,
nothing more and nothing less. You can't stop when the defined deliverable
is 90% complete and it's pointless to continue work beyond the point it is
100% complete. Thus each task has a specific, concrete, exactly measurable
amount of work associated with it. Work costs X dollars per unit - you have
to pay resources to do work or they don't show up and nothing happens. You
pay them a specifc rate or salary. So the cost of doing the work on a task
is defined by the amount of work required to create the deliverable which in
turn is defined by the nature of the deliverable itself. As a result, you
cannot define the budget until you have explicitly defined the tasks,
analysed how much work will be involved in each one, and identified the
human and material assets that will be doing that work and their associated
costs.

If the projected cost is greater than the assets on hand, you cannot
arbitrarily shrink durations to fit - instead you must redefine the
project's objectives to encompass reduced deliverables. As an example, if
my painter paints 10 square feet per hour and we have 1000 square feet of
wall to paint, it will take him 100 hours. If we can't afford to pay for
100 hours, we can't just pull out of thin air that he will do it in 75
hours. The only way we can really get the budget down to within the
resources we have available is erect fewer walls so he only has to paint 750
square feet.

HTH

 

[davegb]

IMHO, you're working backwards. Until you have explicitly defined the tasks
and the resources performing them, you don't know, CAN'T know, that the
project budget is $350k and will take 3 years. You have not set a budget or
timeline, you have set some parameters within which you hope you can
accomplish some business objective. The budget and timeline is an estimate
of what will really be required to achieve those objectives. So far, you
have set some goals based on the anticipated availability of assets, in
other words, a distribution of projected revenues. The actual budget may be
under or may exceed those desired parameters.

Don't define an arbitrary cost and then expect the work to expand or
contract to fit.

The cost assigned may not be arbitrary, Steve. I've used Comparitive
Sizing methods many times myself. If I've built 5 compressor stations
already, and I have, when I build the sixth I can make a pretty
accurate estimate of the cost if it is pretty similar. And I can factor
in some differences. Detailed estimating is not the only way to get a
reasonable project estimate. More accurate? Yes. But if there isn't
enough data yet to get one, I've started many projects on a Sizing
estimate based on experience or on published numbers. And they've
worked out quite well.


A task creates a specific, concrete, deliverable. The

task will take how ever long it takes to complete exactly that deliverable,
nothing more and nothing less. You can't stop when the defined deliverable
is 90% complete and it's pointless to continue work beyond the point it is
100% complete. Thus each task has a specific, concrete, exactly measurable
amount of work associated with it. Work costs X dollars per unit - you have
to pay resources to do work or they don't show up and nothing happens. You
pay them a specifc rate or salary. So the cost of doing the work on a task
is defined by the amount of work required to create the deliverable which in
turn is defined by the nature of the deliverable itself. As a result, you
cannot define the budget until you have explicitly defined the tasks,
analysed how much work will be involved in each one, and identified the
human and material assets that will be doing that work and their associated
costs.

If the projected cost is greater than the assets on hand, you cannot
arbitrarily shrink durations to fit - instead you must redefine the
project's objectives to encompass reduced deliverables. As an example, if
my painter paints 10 square feet per hour and we have 1000 square feet of
wall to paint, it will take him 100 hours. If we can't afford to pay for
100 hours, we can't just pull out of thin air that he will do it in 75
hours. The only way we can really get the budget down to within the
resources we have available is erect fewer walls so he only has to paint 750
square feet.

HTH

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

I tried what you suggested as a sample on one task:
I defined a resource called "cost" in the resource sheet and gave it a
standard rate of $1/hr.

I assigned this resource to the task of which duration was 10 and in the
"assignment information" I defined a bell shape work contour. The outcome
cost was $40.

If I wanted the cost to be a 100, the duration would increase to 25. And
since my planned duration and planned budgets are both given to me as an
input from "outside", all I could do is to change the rate. Am I working
with
the right program?



I entered an activity, gave it a start date and a duration of 10 days. I

 

[Steve House [Project MVP]]

Of course, parametric estimating such as you describe for your compressor
stations is an excellent method of doing a rough approximation when you're
in the preliminary feasibility study and planning stages, don't
misunderstand me. The problem comes when we try to go the next step and say
"The compressor station 'should' cost XXX dollars, now design a project plan
to spend exactly that amount, no more and no less." That 'budget' isn't
taking into account the way this instant of that type of project may differ
from all the other instances that have come before, changes in labour rates,
changes in material costs, etc etc, all the things that will go into the
actual cost of doing the work. I'll reiterate - IMHO a 'budget' is not what
you're allowed to spend. It is instead a detailed estimate of what you'll
need to spend to get the job done. I think of the budget as the operational
budget, a projection of actual costs in other words, the budget that is
actually used in the day-to-day management of the project and to track
progress against. Hopefully the operational budget is less than the
allocated assets - if not, update your resume.

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

Steve House [Project MVP] wrote:

Don't define an arbitrary cost and then expect the work to expand or
contract to fit.

The cost assigned may not be arbitrary, Steve. I've used Comparitive
Sizing methods many times myself. If I've built 5 compressor stations
already, and I have, when I build the sixth I can make a pretty
accurate estimate of the cost if it is pretty similar. And I can factor
in some differences. Detailed estimating is not the only way to get a
reasonable project estimate. More accurate? Yes. But if there isn't
enough data yet to get one, I've started many projects on a Sizing
estimate based on experience or on published numbers. And they've
worked out quite well.

<snip>

 

[davegb]

Of course, parametric estimating such as you describe for your compressor
stations is an excellent method of doing a rough approximation when you're
in the preliminary feasibility study and planning stages, don't
misunderstand me.

Actually, I phrased it wrong. I did compressor stations using Sizing
estimates without ever doing a Detailed estimate, just factoring in the
differences. No problem. Plenty accurate.

The problem comes when we try to go the next step and say

"The compressor station 'should' cost XXX dollars, now design a project plan
to spend exactly that amount, no more and no less."

Forturnately, there were no morons present, so no one said anything
like that. We had a reasonable contingency, and on all projects, a 10%
overrun was no problem.

That 'budget' isn't

taking into account the way this instant of that type of project may differ
from all the other instances that have come before, changes in labour rates,
changes in material costs, etc etc, all the things that will go into the
actual cost of doing the work. I'll reiterate - IMHO a 'budget' is not what
you're allowed to spend. It is instead a detailed estimate of what you'll
need to spend to get the job done. I think of the budget as the operational
budget, a projection of actual costs in other words, the budget that is
actually used in the day-to-day management of the project and to track
progress against. Hopefully the operational budget is less than the
allocated assets - if not, update your resume.

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

Steve House [Project MVP] wrote:

Don't define an arbitrary cost and then expect the work to expand or
contract to fit.

The cost assigned may not be arbitrary, Steve. I've used Comparitive
Sizing methods many times myself. If I've built 5 compressor stations
already, and I have, when I build the sixth I can make a pretty
accurate estimate of the cost if it is pretty similar. And I can factor
in some differences. Detailed estimating is not the only way to get a
reasonable project estimate. More accurate? Yes. But if there isn't
enough data yet to get one, I've started many projects on a Sizing
estimate based on experience or on published numbers. And they've
worked out quite well.

<snip>