duilio
Senior Member
Posts: 53
Registered: 10-20-2004
Member Is Offline
Mood: No Mood
|
![[*]](./images/VSUK/default_icon.gif) posted on 5-13-2010 at 07:02 |
|
|
Ideas on how to "sell" SD to managers
Hi Tom, Bob, JJ (et others)
I'd like to exchange ideas with you on the following topic. Let's imagine you need to introduce SD to a group of managers that did not have any
previous exposure to the subject. You have two hours.
Where do you start from? A theoretical lesson with some case studies? Or a game (like the beer game or similar ) to explain that their usual
decisional mechanism are not perfect? And after that some concept od SD ?
Any hints, opinion, ideas?
Thanks!!
Giovanni
|
|
|
LAUJJL
Senior Member
Posts: 696
Registered: 5-23-2003
Location: STRASBOURG FRANCE
Member Is Offline
Mood: good
|
| posted on 5-13-2010 at 10:47 |
|
|
selling SD
Hi Giovanni.
I think that SD is very difficult to adapt to management problems. Whatever they manage (business, war, country) they work in a context that needs
different qualities than those required by a modeler or anybody interested in the results of modeling. They deal with people and need to behave in a
very clear way with precise rules (sometimes only apparently) because they manage a lot of people (employees, shareholders, clients, suppliers,
bankers, etc..) who have an absolute need to understand their behavior. This needs people oriented towards action and not thinking, who have a
relative simple way of thinking, in which they believe, that permits to handle all problems probably less than perfectly and avoid making big
mistakes. This is enough. Wanting more than that will necessitate too much thinking with a lot of time lost, confusion and indecision which is the
worst that can happen for a manager who must look like being always sure of what he is doing.
I have myself managed small and bigger businesses (one of them was buying each year about 10000 small trucks, vans and cars, Eurli) and our family
business currently employed about 150 people.
We have recently sold most of our businesses for reasons of weak profitability and are looking for other opportunities.
To resume SD is too subtle for management problems.
I do not mean that managers are stupid people, they have learned from experience how to behave accordingly.
SD’ers do not understand this because they never had the occasion to manage anything important and build models instead of delivering solutions.
So expecting to explain SD in two hours, is at best misleading. I have never tried to explain SD to any of my manager friends and I know a lot of them
in Strasbourg. It would deserve their interest.
They will never have the time to learn this very difficult method by themselves. The method by itself needs to conform to an extremely strict way of
working that is most of the time never respected by SD’ers themselves and I do not see who could do any modeling for them, especially in France and
even elsewhere. The only time when it may be worse trying to work with an SD consultant is when he has an extensive practice of the type of the
problem the manager faces, an extensive experience of the business itself (is specialized in this type of business) and participates closely to the
decisions taken and their implementations. To resume is selling solutions instead of selling modeling or SD.
This does not look very positive, but I think that one must always expect the worse because it is often what happens and it is then better to be
prepared.
If you are a consultant and want still to work with SD, and want to deliver a true added value, the only way is to choose a growing field with a lot
of dynamical problems and become a specialist of it.
I am myself interested in SD, because I studied extensively mathematics when I was young, but not because I expect any added value by using it. I
expected once but no more. But I still enjoy making very simple models that may help me to think even is this thinking is not necessarily very
productive.
Regards.
JJ
[Edited on 13-5-2010 by LAUJJL]
|
|
|
duilio
Senior Member
Posts: 53
Registered: 10-20-2004
Member Is Offline
Mood: No Mood
|
| posted on 5-13-2010 at 12:17 |
|
|
SD for Managers
Hi JJ,
I agree with you that the timing requested by manager is often not exactly the same timing requested by SD. And I agree that managers have "simple
minds" as well.
An I used to manage people in the past so I know "real life".
What I was thinking is to use SD for figuring out possible outcomes from strategies. Maybe the strategy definition is the only moment in time where
managers sit down for a (short) while and think about future directions, new ideas...
Thanks again, your input to the forum are always of great vale.
Cheers
G
|
|
|
Barry
Member
Posts: 39
Registered: 10-1-2007
Member Is Offline
Mood: No Mood
|
| posted on 6-9-2010 at 20:05 |
|
|
Run a simulation that addresses a management challenge in an interesting way?
(I haven't had much luck, so speaking hypothetically)
Managers have PRESSING issues, and don't seem to have time for eventual issues.
Much more thought needs to go into how to sell Models. Much more thought. I hope that there are other contributors to this thread because LAUJJL
touched on the cultural gap between managers and modelers. It's like any other cultural gap. It just needs to be understood.
|
|
|
duilio
Senior Member
Posts: 53
Registered: 10-20-2004
Member Is Offline
Mood: No Mood
|
| posted on 6-10-2010 at 16:20 |
|
|
How to sell SD to managers
Maybe the key point is to convince them that they can count on a group of modelers that can help in solving real issues. The General Motors case
studies are an example of that, in the 70' they realized they need a team of "modelers" to help them solving real problems (See Sterman Bisiness
Dynamics for a description of one of this cases).
I see also potenutial in using SD to simulate a strategy.
But I agree with JJ, there's still a lot of step to be taken.
But I do not give up;)
Cheers
G
|
|
|
Barry
Member
Posts: 39
Registered: 10-1-2007
Member Is Offline
Mood: No Mood
|
| posted on 6-20-2010 at 18:43 |
|
|
Heck...searching jobs for simulations, models, system dynamics, systems thinking...etc...there are hardly none.
[Edited on 6-20-2010 by Barry]
[Edited on 6-29-2010 by Barry]
|
|
|
Barry
Member
Posts: 39
Registered: 10-1-2007
Member Is Offline
Mood: No Mood
|
| posted on 6-20-2010 at 20:28 |
|
|
I think the gov and education are the best customers. Most jobs I find want Modelers with advanced degrees in a bunch of other data stuff that don't
mesh.
[Edited on 6-22-2010 by Barry]
|
|
|
Barry
Member
Posts: 39
Registered: 10-1-2007
Member Is Offline
Mood: No Mood
|
| posted on 6-22-2010 at 16:29 |
|
|
The question is "How do you sell modeling in a formal systems thinking context". Managers HAVE plenty of "models" that are banging data together
and programming where the reality get lost in the analysis. Compared to all of the programming, a simple Sterman/Forrester model looks like kids
tinker toys. I'm trying to sell them as Graphic Design suppliments.
|
|
|
Barry
Member
Posts: 39
Registered: 10-1-2007
Member Is Offline
Mood: No Mood
|
| posted on 6-24-2010 at 16:26 |
|
|
I'm perplexed.
|
|
|
duilio
Senior Member
Posts: 53
Registered: 10-20-2004
Member Is Offline
Mood: No Mood
|
 posted on 6-28-2010 at 12:22 |
|
|
Hi Barry,
not sure your last answer is related to my comments, anyway we shouldn't necessarily agree on everything , right?
anyeay, the discussion was interesting for me
Cheers
G
|
|
|
LAUJJL
Senior Member
Posts: 696
Registered: 5-23-2003
Location: STRASBOURG FRANCE
Member Is Offline
Mood: good
|
| posted on 6-28-2010 at 13:21 |
|
|
Hi everybody
I have not understood what Barry really wants to say.
But the main thing to sell SD to managers is to verify that they need it.
It is easy to make a model, but one must be sure that it will really bring an added value to the client and not only understanding.
The added understanding must generate added profit.
And the probability that among your businessman there is one who really will benefit from an SD study is rather low.
The chance of a business to profit from an SD study has not been really studied by SD'ers. I never heard of a study about that subject and I study
each year the program of the SD conference. If anyone knows about the subject I would be very interested.
The reason of it is that SD'ers do not mind what the client will do with the model he buys.
And I can understand this.
I can rent a sport car to someone who is transporting his girl friend or his mother in law. I do not mind as long as he pays the fee.
And it will be much easier for me to convince someone to rent a sport car in the girl friend case than in the mother in law one.
It is the same for you. It will be extremely difficult to convince someone to buy SD if he does not need it. So the first thing to do is to select
business people who work in a very dynamic environment with many unsolved questions and where SD will show its real power and helpfulness.
But SD will not change the reality. If the reality is not favorable, nothing will do. So the first thing to do is to find the favorable environment.
This is known in French as “nager dans le sens du courant”.
Translation: swim with the draft and not against it.
I think too that if to proove the usefulness it is necessary to build a complex model, than it is probably not useful.
Regards.
JJ
[Edited on 28-6-2010 by LAUJJL]
|
|
|
Barry
Member
Posts: 39
Registered: 10-1-2007
Member Is Offline
Mood: No Mood
|
| posted on 6-29-2010 at 22:57 |
|
|
Thanks. This is what I mean. There are plenty of SD modelers in the engineering field. People model steering and tracking systems, automatic systems
etc. Most of what is discussed here is Sterman, Forrester's use...and that is a specific use of SD directed towards increasing understanding. It has
to be read and understood. Forrester and Sterman have written a LOT about how to use and interpret the models for managing everything from traffic to
cities. Hard systems modeling doesn't have that understanding element. They are components of automatic systems, like a computer system on a car.
Other writers have written much more on the subject, and most of those writers focus on Dialogue first, and modeling second, as a tool for Dialogue.
The most popular is Senge. My interpretation is that Dialogue helps groups to think about issues in more systemic ways (instead of only focusing on
here and now....if we do this now, we will get this...but later we will get that)...and that models are tool for use in those Dialogues. When the
whole group can acquire habits of thinking more systemically, then people stop bumping heads. We can TALK about global warming because we can talk
about something that would normally be too complicated to talk about. The model didn't generate the insight, it was just a tool that let us talk
about it and move past our fixed assumptions. I'm interested in seeing groups solve issues like global warming, but I'm not a huge enthusiast of
models. Even though Sterman, Forrester and others discuss modeling in the context of management...hard systems modeling seems to outnumber management
modeling 10 billion to one.
[Edited on 6-29-2010 by Barry]
|
|
|
SDLEE
Junior Member
Posts: 2
Registered: 7-12-2010
Member Is Offline
|
| posted on 7-12-2010 at 14:29 |
|
|
I'm working at one of the automotive companies as an engineer and I had an advanced degree on SD. Based on my experience, selling SD to low-level or
mid-level managers are not easy because they do not work on dynamic issues where SD is beneficial. It is easier for high-level managers to accept SD
than low- or mid-level managers.
I've been working with engineers whose work involve in modeling feedback system such as automobiles. Unfortuntely, my expereince is that they hardly
apply the feedback thinking to dynamic issues other than hard systems modeling. I wonder why... One fact I can think of is that there are three
distint yet conneted factors: (1) issues we want to understand thus solve, (2) modeling the issue and (3) then analysis of the mathmatical model using
MATLAB. Most engineering books on dynamic problems focus on the last part, "analysis" of mathmathical model. SD is an activity of "modeling"
rather than "analysis", which requires understading the issues first.
Sangdon Lee
|
|
|
LAUJJL
Senior Member
Posts: 696
Registered: 5-23-2003
Location: STRASBOURG FRANCE
Member Is Offline
Mood: good
|
| posted on 7-12-2010 at 14:48 |
|
|
how to sell SD to managers
Hi Langdon
Engineer technical problems are easier to define than social problems. This is why the work of technical issues focuses on the analysis part.
Social problems are much more difficult to define and can be differently defined. One should then take a lot of time with the problem definition,
which is generally not done, people rushing to model the problem having not tried to determine the utility of doing so and how to do it.
But the analysis of social models is still difficult, mainly because of the lack of precision of the definition of the problems, which is one of their
characteristics.
Regards.
JJ
|
|
|
SDLEE
Junior Member
Posts: 2
Registered: 7-12-2010
Member Is Offline
|
| posted on 7-12-2010 at 20:08 |
|
|
I noticed in this and other SD forums that people ask about "advanced" SD books or materials. I think looking for advanced material for SD might be
futile because the difficulty of developing SD is not that people lacks the skills/knowledge/theory of SD (stock and flow, feedback, etc), rather
people lack knowledge on the issue to tackle. Once I studied several important books on SD, I think it is beneficial to read books on a specific
issue or collect information.
Sangdon Lee
|
|
|
Barry
Member
Posts: 39
Registered: 10-1-2007
Member Is Offline
Mood: No Mood
|
| posted on 7-17-2010 at 21:36 |
|
|
>they hardly apply the feedback thinking to dynamic issues other than hard systems modeling. I wonder why...
I think the uses are entirely different. That is clear in almost every book that I've read, but it's not spelled out as being an important
distinction. One is a "communication tool" (As in "soft systems modeling", and when writers say that "the model building process is important",
isn't it clear that they mean "talking about the issues under question and bringing misguided assumptions to light") while hard systems modeling is
for prediction, analysis, and tracking. Many dialogue and democracy promonents like Donella Meadows promote the use of models to facilitate
discussions, while a "modeling approach" that treated people as hard systems would be anything but Democratic. Does anyone get what I am trying to
say here? I can't remember if it was Forrester or Sterman who said that Models are a last resort when all else fails (ie communication about products
and outcomes). I think that the "systems thinking" movement actually intends, and does not articulate, that they are using tracking models as a
METAPHOR. People are always confusing metaphorical writings with literal interpretations, in all sorts of writings.
|
|
|
Barry
Member
Posts: 39
Registered: 10-1-2007
Member Is Offline
Mood: No Mood
|
| posted on 7-17-2010 at 21:41 |
|
|
if "hard systems modeling" is more popular, it's probably because "hard systems management" ie "industrial management theories" are more
popular than decentralized theories of management (where any discussions, much less discussions around models would be seen as valuable.) In other
words, "control" is more popular than the "let's talk about it" approach. Senge included Bohm and other Dialogue writers (who are not at all
about automatic control), but it's only natural that "hard systems" will appeal more to Western Management.
|
|
|
RogerD
Junior Member
Posts: 3
Registered: 7-29-2010
Member Is Offline
|
| posted on 7-30-2010 at 00:11 |
|
|
"selling" modelling
"selling" modeling is a very dangerous concept.
Look how the Climate Change debate has been oversold, become an epistemological debate and derailed - almost fatally.
Modeling can provide insights rather than predictions.
The insights are only useful to the extent that they can help guide action.
Start with the familiar: every manager is already modeling; usually using a spreadsheet and preparing annual budgets for their area of
responsibility.
Start here - SD is a much better tool than excel for providing the insights needed to manage.
Analyze a budget spreadsheet which they have prepared and present it back to them as an SD model.
Roger
|
|
|
Barry
Member
Posts: 39
Registered: 10-1-2007
Member Is Offline
Mood: No Mood
|
| posted on 8-12-2010 at 18:22 |
|
|
Hey, that's a good idea. A spreadsheet substitute.
It's funny that no books really address selling.
|
|
|
LAUJJL
Senior Member
Posts: 696
Registered: 5-23-2003
Location: STRASBOURG FRANCE
Member Is Offline
Mood: good
|
| posted on 8-13-2010 at 13:16 |
|
|
It does not make sense to sell SD as a substitute to speadsheet budgeting. Budgets are made on a relatively short time horizon, generally one year and
with the purpose of ressources allocation, forecasting the profit and limiting the risk of cash shortage.
It is made for the managers, the shareholders and bankers who will easily modify some hypotheses with a tool they master.
There is no feed back within such a short horizon and using a tool like Vensim just to make budgets or even to test different budgets scenarios does
not add a sufficient added value to justify using such a complex tool compared with speadsheets.
It is much better for someone who sells SD to be really competent and to find clients who really will profit both from SD and from his knowledge and
experience and will buy his sevice the correct price that is needed by the amount of work that is necessary to provide a useful service.
Regards.
JJ
[Edited on 14-8-2010 by LAUJJL]
|
|
|
RogerD
Junior Member
Posts: 3
Registered: 7-29-2010
Member Is Offline
|
| posted on 8-23-2010 at 00:50 |
|
|
I agree SD is NOT a substitute for spreadsheet modeling.
My suggestion to "sell SD" was to "START with the familiar".
Its about building a bridge to the SD concept.
One will quickly discover the limitations to the spreadsheet and can move to SD where appropriate.
I find the opposite problem: the spreadsheet paradigm is so pervasive many fall back to spread-sheeting, using spreadsheets where inappropriate, and
often placing too much confidence in an illusion of precision.
The time horizon problem is something you need to stretch. Again, start with the familiar: shorter and less uncertain, to longer and more
speculative.
Yes, one year is typical for profit and cash flows.
But many managers are familiar with dealing with a longer time horizon for capital budgeting or strategic planning, and so on.
Build on what clients already know.
Roger
[Edited on 23-808-1010 by RogerD]
|
|
|
LAUJJL
Senior Member
Posts: 696
Registered: 5-23-2003
Location: STRASBOURG FRANCE
Member Is Offline
Mood: good
|
| posted on 8-23-2010 at 10:04 |
|
|
How to sell SD to managers
Hi Roger
From your point of view, selling SD this way may be useful, beacause as you say, start with something people master, and you can then extend the time
horizon.
The problem is that to be able to extend the time horizon you must go to the system thinking paradigm that works completely differently.
But SD has his own limits as spreadheets and the limits are not on SD, nor on modelers but on clients.
SD is poorly used, because most clients have not capitalized enough experience with the use of SD and are not able to choose a competent modeler or to
build by themselves models (the best solution for most business problems) and to interpret and use the preferably simple model again for most
business problems, knowing the limits of the method.
The most difficult thing when using SD for a manager is the ability to mix his own knowledge, experience and intuition with the insights given by a
model and make them work together. Doing this requires years of experience from the client. Not being able to do this when he first deals with SD, he
will fall into two alternate reactions. the first is to reject SD, the second is to accept blindly its results. Both solutions are bad. The solution
is to make the correct blending of practical experience and intuition with a rationale tool like SD.
Blending intuition and rationality is difficult and requires a lot of time and effort too, but it is absolutely necessary in the business world which
is only half rationale.
The worst thing to sell SD is to present it as a miracle solution that will at last solve the prospect's problems.
The prospect may become a client, but will be disappointed later on, and this will be counter productive. This will generate a negative word of mouth
loop.
Regards.
JJ
[Edited on 23-8-2010 by LAUJJL]
[Edited on 23-8-2010 by LAUJJL]
|
|
|
RogerD
Junior Member
Posts: 3
Registered: 7-29-2010
Member Is Offline
|
| posted on 8-24-2010 at 00:27 |
|
|
We're on the same wavelength here!
I have been amazed at what people try to do with a spreadsheet! (and usually when more appropriate tools are available.)
Some of them can make the paradigm shift and take up a new investment in learning SD. Most get lost and persist with their spreadsheets.
I find a helpful step is to introduce them to influence diagrams. This they find very helpful as a presentation or thinking tool.
The adaption of excel formulas to SD is a hurdle that many find a leap too far. At least they have some kind of a handle on what SD is about.
(btw forio have an interesting on-line tool that will read a simple excel spreadsheet and convert it to a systems dynamics model - I have found this a
useful "icebreaker" for some clients.
Expose was another attempt to bridge to the spreadsheet.)
Roger
[Edited on 24-808-1010 by RogerD]
|
|
|
LAUJJL
Senior Member
Posts: 696
Registered: 5-23-2003
Location: STRASBOURG FRANCE
Member Is Offline
Mood: good
|
| posted on 8-24-2010 at 11:13 |
|
|
How to sell SD ...
Hi Roger
Good luck in your endeavour to introduce business people to SD!
Being a business man myself and not selling SD, I have never tried to explain SD to any of my business friends. I think that their business are not
big enough to justify such an action, the biggest I knew and where i was a shareholder was dealing with a turnover of about 150 millions Euros, by
aggregating all the turnover of all the shareholders of this central organisation that was buying cars for them and other thinkgs like insurances,
credit etc.. and being a thinking tank too . i think that taken into account that they were most of the time not sharing the same ideas, Sd would have
been inappropriate.
I know several other firmas but they will never get to more than 100 million euros and I think that at this size SD is not the right tool to use.
Maybe influence diagrams but not quantitative SD.
SD books are generally to big and to complex.
Kim Warren's book is rather close to business people problems but is big too and could to my opinion say the same thing being three times thinner
without insisting and repeating things that business people know already. BD is too big and complex. One good book is Coyle's 'System dynamics a
practical approach'. It is now unfortunately a bit updated, deals with a language no more used and has not a very appealing presentation.
Cavanna's last book is maybe good, but I have not bought it.
Regards.
JJ
|
|
|
duilio
Senior Member
Posts: 53
Registered: 10-20-2004
Member Is Offline
Mood: No Mood
|
| posted on 9-1-2010 at 09:11 |
|
|
How to sell SD to manager
Hi all,
I'm really glad to see that my initial question generated such an interesting thread.
Maybe the time are now mature to introduce a bit of "quantitative" or statistical/or "analytical (as we are hearing as the new trendy expression)
techniques to manager.
I'll try to do it in the next months and keep you posted, so let's keep this forum alive!
Cheers
Giovanni aka Duilio
|
|
|
