I'd agree that SPI is not really about money but that doesn't mean it's
bogus. I see the monetary aspect of it as merely establishing a common,
convenient unit of measure - regardless of whatever units you choose to
track, the index is a ratio and whatever unit you choose is canceled out in
the calculation. It's simply a way of tracking the ratio of man-hours
scheduled with man-hours performed. If you used hours for schedule
performance calculations and currency for cost performance, you'd double the
number of data fields you'd need to worry about in the baseline. Since the
resource rate establishes an identity relationship between scheduled
man-hours and scheduled cost, and actual man-hours and actual cost, one can
use currency as a common measure and simplify the reporting process. It
also simplifys extrapolating the "S" curves to determine Estimated
Completion Date and Estimated Cost at Completion.
Tracking duration doesn't cut it because it doesn't take into account the
effort required versus the effort expended. Time is only coincidental to
progress - the real measure is how much useful outpout is required versus
how much has output been achieved. True, output is generated at a rate that
associates it with the passage of time but the reason for doing the project
is not to pass the time, it is to generate the output. Measuring time
doesn't seem like its measuring anything of signifigance in terms of whether
you've gotten any benefits for your efforts.
Not that you said otherwise, but I'd interject the least reliable method of
tracking progress is % Physical Complete because it's such a loosy-goosy
notion. If one is designing an engine, just exactly what objective measure
constitutes "50%"? Half the drawings done? The engine block but still
working on the cylinder head? Fuels system done but still working on the
cooling and electricals? Front half done but still working on the back
half? I just can't wrap my head around why so many people want to use it
since except in very rare circumstances it's something virtually impossible
to measure.
Steve,
In my way of thinking schedule and effort are separate entities.
Certainly one can argue that the time to do a task (duration) is
worthless without a resource to accomplish the task (effort) - perhaps.
I certainly agree that SPI in and of itself is not bogus. After all, SPI
is simply a dimensionless ratio. Where I depart is buying into the
process of using work/cost to measure schedule performance. I contend
that duration is relevant to schedule, effort is relevant to efficiency
and cost is relevant to budget. For example, based on history or
educated knowledge I know it takes a certain amount of time to do a
particular task. Resources are applied to that task to actually perform
it and money is allocated (budget) to pay for its completion. In my
mind, a unbiased assessment of completion compared to an end goal
requirement should determine how well the schedule is going. On the
other hand, the amount of effort expended compared to the original
estimate relates to how efficient the task is being worked. Finally, the
money spent relates to how effectively money is being spent. The first
metric should be SPI, the second metric should be EFI (a new term
representing resource efficiency) and the third metric is the standard
CPI. Off the wall thinking - perhaps - but then the original poster did
mention that he wanted some type of SPI measurement that didn't rely on
resource effort.
I agree that the passage of time is not a measure of accomplishing
anything, except in the case of level-of-effort type tasks. Rather I
believe schedule performance is best measured by comparing percent
complete to Total Slack against an end goal (e.g. final project
milestone). Thus the simple passage of time doesn't buy anything (except
a very bad SPI value) because tasks are not being completed in time to
meet the end goal. It doesn't matter how much effort is applied or when
the task was baselined, if the Total Slack is maintained (or increased),
the project is getting performed on time (i.e. favorable SPI). The
efficiency may be lousy and the cost prohibitive, but schedule wise, it
is in good shape.
I don't really have any feelings about % Physical Complete because I've
never studied it or used it. However, from what I have read it seems to
be a positive step in better defining "completion" in some cases. You
mention that % Physical Complete is "loosy-goosy" and virtually
impossible to measure. Measurable metrics are fantastic if they
represent valid data and can be developed at all. However in my
experience meaningful measurable metrics are very hard to come by. More
than once I have seen measurable metrics that weren't worth the paper to
print them on but somehow they were "sold" to the powers that be and are
being used to measure something. Sometimes the best and only valid
assessment is provided by a seasoned individual with tons of experience
and no "axe to grind". Granted, those people are hard to find.
No, I don't have all the answers. It's just that with regard to
calculating SPI, I think there is a better method.
John
Project MVP
