Organization design using SD
Posted: Sat Jan 22, 2000 10:36 am
rahmanra@ind.sharif.ac.ir wrote:
> Involved in the process of designing a research organization, we want to base
> the process initially on SD to investigate and design growth and learning
> feedbacks for the system, before planning the physical organization.
> I am looking for resources discussing this approach in the design of a system.
Hazhir
Whats worked for me is to define my goals for the organization, lay
out a high level strategic model that can achieve such a goal, and use
the information paths that were necessary for the models success as
templates for the information paths in the real organization. When I
have had to deviate due to real-world constraints, Ive on occasion
gone back to the model to verify that the changes still worked in the
model context.
In a bit more detail, Ive treated the organization as the "plant" and
the information systems (manual or automated) as the feedback in a
control theory sense.
Of course, you start with the organizations objective and, if you are
correcting an existing problem, the picture of the existing problem.
Lay out the flow of work product as a stock and flow. If you are
fixing an existing problem, do all the usual stuff of verifying the
model can exhibit the problem behavior. If you are starting from
scratch, you may not yet be able to execute more than a stable flow
through the process, as the feedback is largely missing.
Then look for what seems like sensible information feedback paths.
Add them to the model and see if you get the behavior you want. If
you do, see if the model is realizable. If not, iterate on the
feedback or the "plant" until you succeed.
In my experience, a key benefit is a better linkage of the information
fed back to the goals of the organization. The informal "Its obvious
we need to measure X, Y, and Z" is replaced with "The system doesnt
care whether we measure X or Y or not. It works well enough if we
only measure and make decisions based on Z, but we have to do that at
least monthly [or whatever] or we get instability."
Hope this helps.
Bill
From: Bill Harris <bill_harris@facilitatedsystems.com>
--
Bill Harris 3217 102nd Place SE
Facilitated Systems Everett, WA 98208 USA
http://facilitatedsystems.com phone: +1 425 338-0512
> Involved in the process of designing a research organization, we want to base
> the process initially on SD to investigate and design growth and learning
> feedbacks for the system, before planning the physical organization.
> I am looking for resources discussing this approach in the design of a system.
Hazhir
Whats worked for me is to define my goals for the organization, lay
out a high level strategic model that can achieve such a goal, and use
the information paths that were necessary for the models success as
templates for the information paths in the real organization. When I
have had to deviate due to real-world constraints, Ive on occasion
gone back to the model to verify that the changes still worked in the
model context.
In a bit more detail, Ive treated the organization as the "plant" and
the information systems (manual or automated) as the feedback in a
control theory sense.
Of course, you start with the organizations objective and, if you are
correcting an existing problem, the picture of the existing problem.
Lay out the flow of work product as a stock and flow. If you are
fixing an existing problem, do all the usual stuff of verifying the
model can exhibit the problem behavior. If you are starting from
scratch, you may not yet be able to execute more than a stable flow
through the process, as the feedback is largely missing.
Then look for what seems like sensible information feedback paths.
Add them to the model and see if you get the behavior you want. If
you do, see if the model is realizable. If not, iterate on the
feedback or the "plant" until you succeed.
In my experience, a key benefit is a better linkage of the information
fed back to the goals of the organization. The informal "Its obvious
we need to measure X, Y, and Z" is replaced with "The system doesnt
care whether we measure X or Y or not. It works well enough if we
only measure and make decisions based on Z, but we have to do that at
least monthly [or whatever] or we get instability."
Hope this helps.
Bill
From: Bill Harris <bill_harris@facilitatedsystems.com>
--
Bill Harris 3217 102nd Place SE
Facilitated Systems Everett, WA 98208 USA
http://facilitatedsystems.com phone: +1 425 338-0512