The terms SPI and schedule variance in EVMS is

							The terms "SPI" and "schedule variance" in EVMS is a misnomer. They do not measure
efficiency variance in the program schedule. What they do measure is a "spend rate"-
they measure the rate at which a program is spending money on a program to accomplish
work against the rate it was planned to spend money at, both using dollars as a basis. This
relates to schedule in that it assumes the baseline schedule is resource loaded and so there
is some causal correlation between the rate at which the program is spending and
progress in the program, which is generally true, that's not a one-for-one correlation.
Although there's ways to manipulate the numbers and approximate a variance in terms of
time, you cannot take the "SPI" and "schedule variance" and calculate accurately how
much the program has slipped as you can when you look at an IMS and/or SRA.

And this is where I think this measure falls short. Although spend rate analysis is
important, it is not something that your average program manger understands, uses or
cares about. What he does care about (and has to brief) is whether his program is on
target in terms of cost and schedule and redefining the schedule portion of EVMS is a
way to do that.

The traditional calculations and metrics used in EVMS would still apply, except that the
schedule pieces would be calculated in terms of days ( or weeks or months) instead of
dollars. For example, the SPI (Budgeted cost of Work Performed/Budgeted cost of Work
Scheduled) would become Budgeted Days of Work Performed/Budgeted Days of Work
Scheduled) which is another way of saying "how efficient has the Contractor been at
meeting his schedule?" The same sort of thing applies to Percent Complete, EAC and
VAC calculations and so would show the user how a program is doing in terms of dollars
AND time.

Making the switch from dollars to days for schedule items supports critical chain
scheduling as the program would have both a cost and budget account which would relate
to each WBS. This would allow the WBS elements to be scheduled at the 50%
probability level and the risk of the schedule to be included in a program level schedule
buffer, which would then function like management reserve does for the cost risk in the
traditional EVMS system.

With respect to the programs critical path, EVMS does not track it nor incorporate it in
any of the traditional metrics, principally because EVMS has always focused on dollars
and scheduling and critical path analysis is focused on time. The twain have never met,
but updating the EVMS concepts to include days would allow them to. Because the days
used in the schedule pieces of EVMS would be the same as those used in the program
IMS, EVMS would highlight the program's schedule performance and problem areas and
traditional schedule analysis and critical path analysis would provide the tools to analyze
the problems, delays and slippage of the program.

So basically, making changes to the schedule portion of EVMS would allow an
integration of EVMS and scheduling data, provide better and more useful status to the
users and more opportunities for in depth analysis.
Jim