John M. said:
Regarding the 80% utilization - which I agree is a generally accepted
planning assumption... I'm curious what the most common means is for
incorporating this into a plan. I'm sure much is driven by personal
preference, but if somebody didn't have much of an opinion, which would be
recommended.
I can think of the following:
1. Set the availability for each resource to 80% and assign resources to
tasks at 80%. This would spread the non-project time evenly across the plan
by stretching durations by 25%. If you plan based on durations (e.g. 2 hour
meeting from 9-11), you would need to knowingly add 25% as hopefully people
don't only spend 1.6 hours of work in the 2 hour meeting.
2. Leave availability at 100% and account for the remaining 20% by changing
the working times. Instead of having 8 hours of working time per day,
change it to be 6hrs 24min (I always wanted to go to work at 8 and leave at
3:24).
3. Leave availability at 100% and create separate task(s) that span the
planning duration to account for the 20% of non-project time. Similar to #
1, except that the non-project tasks are included in the plan. Assumes the
project is responsible for the cost of the admin/vacation/sick time.
4. Leave availability at 100% and inflate the duration/effort for each task
by 25%...essentially spreading the 80% utilization throughout the plan.
To further complicate this, some projects have resources that are only
available part of their time (e.g. 50% available). The reality is that many
times these resources are available 100% of the time for 20 hours in a week
vs. available 50% of the time for 40 hours. This seems like semantics, but
if this resource is in a 2 hour meeting with a full-time resource, s/he
should be planned to participate for 2 hours and not just 1 hour. In these
cases, I generally push for defining time periods that the resource will be
dedicated to the project (e.g. Monday - Noon Wednesday) vs. trying to plan
based on a floating 20 hour availability. And then layer ontop of this one
of the 4+ options listed above.
John M.
John,
You certainly have laid out several approaches for "dealing" with the
80% rule of thumb but I never got into that level of refinement. I guess
your item 4 comes the closest to a simple approach we use. In my opinion
trying to refine the plan with resource efficiency factors is best left
to those who have a lot of time on their hands and can use it
effectively on detail analysis. For me, the world, particularly when
dealing with the human element, just isn't that easy to capture or
define.
For every project plan I have developed, my preferred approach is to
base effort on historical data, whether that data is used directly or
whether it is used as a basis for the current project. For example, say
we are developing a plan that includes development of a new power supply
(or it could be framing a house). I would first look for similar
projects that were performed in the past and review the data for
applicability. If a past project also developed a power supply, I ask
myself if the power supply for the current project is less complex,
about the same, or more complex than the previous design. Based on that
knowledge, I apply a complexity factor when estimating the work content
for the current plan. Obviously I need to have enough knowledge
(technical in this case) to make a valid assessment. If I'm an engineer
I can most likely make the assessment myself. If I'm a business person
doing the estimating for the current plan, I will need to consult with
appropriate experts. This is why I believe the best people to estimate a
plan are those who will be performing it.
So what do I do in the absence of historical data? Like everybody else I
use educated guestimation. Again, this is why I feel it is so important
when estimating to either have the people who will be performing the
work to do the estimate, or make sure their input is factored in. Plans
developed my management alone or in estimating think tanks are doomed to
failure or serious overruns.
Now, I'm sure you are thinking that approach 4 is open to estimate
"padding" by whomever, and you are right. There will always be those who
"game the system". The best guard against bogus estimates is an open
dialogue between the customer, management, and those who will perform
the actual work. Of course honesty, integrity and a commitment to be a
part of the "team" don't hurt.
John
Project MVP