Re: scheduling multiplier

[Jack Shaw <jsh -at- SOFTWARE-AG -dot- DE>]

The ability to predict the doc./info. effort as a factor
of the time that the developers will need to do the work,
is bunk. There is no reliable factor/multiplier based on
expended dev. hours that can resolve to the doc. effort.

I doubt that any reasonable manager would require a reliable
and enforceable factor based on such naive input. More than
likely, this is an attempt to find an unknown (your doc.
effort needed) from the only predictable known (best-guess
estimate by the devl. group of their person-hours). But...
let's assume worst case unreasonableness...

So if you're forced to use such irrelevant bases for
developing a multiplier, you can do the following:

1. Require a definition of the number of changed/added
   user interface screens/menus from the devl. group.

2. Multiply that count by two for the number of total
   work days.

3. Add a day for each eight hours of estimated time for
   the devl. effort for a final "multiplier".

This of course _does not_ include review/beta test/print
and distribution time, which is pure non-dev.-related
and must be added on top of the devl.-derived "multiplier".
That's usually an experiential factor from past projects of
like size. And since this is all non-timeline related, such
things as holidays/vacations are not yet plugged in. It's
just raw work time, not a critical-path plan.

The reason that the QA people can play this game is, their
effort is more closely related to "lines of code/devel. hours"
for testing purposes. The doc. effort, though, is more related
to the effects of the devl. effort on the user interface. Neither
your nor QA's "multiplier" will hold much water when the
returns are in and you do a project post-mortem...after all,
QA has to test the doc. against the code, so how can they
predict their time when they don't know the extent of your
effort?! That's what makes this game-playing so frustrating...

So you'll be spending significantly more or less time on the
effort than the resulting "multiplier" predicts. But such
Kentucky windage can serve as a reference point for estimating
the next project--and that's usually what management really
wants, long term: a starting point for later guesstimating
and internal accounting--as unrealistic as it may seem.

IMHO and regards,

Jack Shaw