Conversational Systems
Thinking
The power of group engagement
with the rigor of system
dynamics…for less!
Purpose
• To describe a set of skills and an approach
that can improve the quality of
thinking…about just about
anything…without requiring computer
simulation
Agenda
•
•
•
•
•
Overview
Seven Skills
Value Add
Leverage Points
Q&A
Acknowledgements
• I have had the honor to work with several
inspirational systems thinkers
 Barry Richmond
 First articulated Seven Thinking Skills
 First coined term Conversational Systems Thinking
 Dana Meadows
 Developer of Leverage Points concepts
 Steve Peterson
 Continues to push these ideas forward
 Paper Barry Richmond, System Dynamics and Public Policy
is a great example
Continuum from Archetype/CLD ST
to Computer-Facilitated SD
Archetype/CLD ST
• Easy to Learn
• Nearly anyone can write it
• Some systemic insights
but can be too generic
• Mostly right brain
• Sexy
Computer-Facilitated SD
• Possible to Learn
• Somewhat limited
• More rigorous and more unique
systemic insights
• Broad appeal - left/right brain
• Years to Learn
• limited to few
• Most rigorous and unique
systemic insights
• Heavy left brain
• Geeky
Barry Richmond’s
Value per Effort graph
Value
Derived
Complex model
Big honkin’ model!
Simple model
Simple stock/flow map
Conversational use of skills
Effort/Time
Required
• A lot of clarity can be derived with little time and effort, simply by
conversational application of (operational) systems thinking
Venn Diagram
Build Complex Models
2%
Apply Conversational
Systems Thinking Skills
95-100%
•
•
Build
Simple Maps
40-50%
Build
Simple
Models
15-20%
A majority of the population (e.g. in organization or policy discussion…dare I
say in the population!) can learn and apply Conversational Systems
Thinking
They would become better participants (consumers and contributors) to
strategy and policy processes
About the Current Age
"I wish it need not have happened in my time,"
said Frodo.
"So do I," said Gandalf, "and so do all who live
to see such times. But that is not for them to
decide. All we have to decide is what to do with
the time that is given us.”
-J.R.R. Tolkien
One crucially important thing we can do with
the time given us is to increase our systems
thinking capacity and of those around us.
Seven Thinking Skills
Overview
Seven Thinking Skills of a
Systems Thinker*
Syst ems Thinking skill
Dynami c Think ing – Focusing on patterns of
behavior (trends) over t ime
System-as-Caus e Think ing – Choosing to focus
on the system within the organiz ation’s control as
responsibl e for performance issues
Fores t Think ing – Taking the 30,000 foot view of
the system
Opera tional Think ing – Looking for causality (How
is this behavior gener ated?)
Closed-loop (Feedback ) Think ing – Understanding
the feedback and ongoing process responsible for
behavior
Quantitative Think ing – Under standing how to
represen t non-physical, immeasurabl e varia bles in
analysis
Scientific Thinking – Building the most useful,
entertainable theory of causality
Contrast s with…
Static Think ing – Focusing on specific eve nts
System-as-Effect Think ing – Choosing to focus on
forces outside the organiz ation’s control as
genera ting the performance issues (creating
“victimitis”)
Tree-by-tree Think ing – Focusing on the details,
often getting lost in spreadsheets!
Factors Think ing – Develop ing a list of factors
associated/correlated with the behavior
Straight-line Think ing – Believing causality is a
one-way, linear relationship
Qualitative Thinking – Including only those
variables believed measura ble
Proving Truth Think ing – Looking for “The Answer”
*Barry Richmond
7 ST Skills
Mindset &
Approach
Representing
Assumptions
Framing
Issues
Contrasting CST with Simulation SD
Skill
Conversational
Computer Supported
Dynamic
Agreeing on issue, over time, scale of time concerned
Ditto
30,000 Foot
(Forest)
Expanding boundary of focus, determining appropriate sectors
Ditto
System as
Cause
Focusing on how system generates its own performance, not
relying on exogenous causes for behavior
Ditto
Operational
Building a mental model that represents causality, using
stock/flow maps to represent, may or may not need generic
templates to get at essence of activities
Use computer simulation to test plumbing,
will use generic templates to specify flows
Closed-loop
Closing feedback loops, removing free lunches. Can do so with
connector arrows. In CST will link feedback loops through
stock/flow structure to see implications on accumulations.
Cannot determine loop dominance.
Use computer simulation to test loop
validity and determine dominance.
Quantitative
Include non-physical determinants of behavior. Ensures mental
model doesn’t exclude what’s unmeasurable but important.
Need to specify range and build structure
to numerate non-physical. Often requires
lots of selling to stakeholders!
Scientific
Use a build confidence in analysis (map if done) approach. Not
proving truth.
Ditto, but building confidence in
simulations and results.
The Skills Part 1
Framing the Issue
Dynamic Thinking
•
•
•
Move from event to over-time
perspective
Power of group agreement on
“what’s happening”
I usually get individuals to
sketch out…then have small
groups discuss, integrate and
prioritize
Profit (in Millions of $)
“Much Longer Time Horizon” Lens
20
“Longer Time Horizon” Lens
“Short Time Horizon” Lens
0
-20
“Here & Now” Lens
Years
Dynamic Thinking
Don’t only look back!
• Very interesting discussions can occur when
asking to project how long it will take to achieve
some objective.
Revenue
Let’s double revenue!
How will that occur?
20
Different curves
represent different
mental models that can
be brought to the light of
day!
10
0
Years
Dynamic Thinking
Malnutrition (Peru)
Chronic Malnutrition in Peru 1975-2004
45
40
% malnourished, height (under age 5)
•
Here the behavior
of interest is…
Why have
improvements in
reducing
malnutrition
stopped!?
Understanding
why it worked for
a while and why
it’s stopped can
lead to more
systemic
conversations
1975
1984
35
1992
30
1996
25
2000
2004
20
19
75
19
76
19
77
19
78
19
79
19
80
19
81
19
82
19
83
19
84
19
85
19
86
19
87
19
88
19
89
19
90
19
91
19
92
19
93
19
94
19
95
19
96
19
97
19
98
19
99
20
00
20
01
20
02
20
03
20
04
•
More than one of every four children under the age of five suffer
from chronic malnutrition in Peru. After steady advances, progress
in addressing the problem essentially came to a halt – despite
roughly $300 million a year devoted to resolving the problem.
30,000 Foot (Forest) Thinking
•
•
•
Boundary expansion
One way to avoid
unintended consequences
or poor projections
Example: Oil Production


Top map projects oil
production solely on
production infrastructure
Second includes finite
resource, price and the impact
on demand (consumption
infrastructure)
pro ducing stuff
fro m raw m ateri als
Units
of Stuff
s
s
Production Infrastructure
Supply
rem oving
ol d
Prod uction
In frastructure
utilization of
capacity
pro ductivity o f
capacity
s
~
pro ducing stuff
fro m raw m ateri als
Potenti al
Raw
Mate rial
se llin g to
consu mers
Units
of Stuff
s
s
s
s
rem oving
ol d
Prod uction
In frastructure
s
utilization of
capacity
Production Infrastructure
Supply
s
in ve stin g
in new
Consumption Infrastructure
Demand
Cons umption
In frastructure
s
o
mo nths of
i nvento ry
rem aini ng
de sired uni ts
pe r co nsum er
o
capacity
ad ding
s
s
reven ue p er
capacity u nit
de sired
in ve stme nt
o
s
ad justing
$ per u nit
$ per
Unit of
Stu ff
Price
s
reven ue from
stuff
s
s
o
de sired
gro wth
rem oving
30,000 Foot (Forest) Thinking
Tips
• Resist the temptation to “dive into”
everything you know about a part of the
issue in any one sector until you’ve
captured the 30,000 foot view
• Use Sector Frames to help categorize
• Expand boundary by questioning clouds in
a stock/flow map
The Skills Part 2
Representing the Mental Model
Operational Thinking
• The stock/flow language is the operational language of
Conversational Systems Thinking
• It becomes a great framework for…
 Representing the main chain (core infrastructure)
 Determining the essence of how activities are generated (stock
vs. flow)
 Identifying levers through generic templates
Main Chains (or Core Infrastructures)
Examples of stock/flow maps
Treatment Capacity
Capacity in
Develop ment
beginning
devel opme nt
Financials
Treatment
Capacity
coming
on line
Cash in
Accou nt
removing
from use
revenu e
expen ses
Population Health
Healthy
Pop ulation
becoming
at risk
At Risk
Pop
bi rths
Workforce Development
Candidate
Pool
hiring
ine xp erien ced
Inexperienced
Staff
becoming
experi enced
Chronic
Pop
dyin g
devloping
chronic
condi tion s
losing risky
status
Experienced
Staff
Personal Health Habits
attrition of
experi enced
building good
habits through medical
& pu blic aware ness
losing good
hab its
Personal
Habits
Main Chains (or Core Infrastructures)
An example from the CDC
CDC was mandated by Congress with two measurable goals re: diabetes
policy…
1. Rate of diagnosing diabetes
2. Reducing Prevalence
3. Reducing deaths (dying)
What do you think of this strategy? Sufficient?
developing
but undiagnosed
Population
without
Disease
Undiagnosed
Prevalence
1
diagnosing
dying
Diagnosed
Prevalence
developing
and diagnosed
2
3
Turns out these goals are
incongruent!
Prevalence & deaths
(measurable)
must increase if
diagnosing increases!
Main Chains (or Core Infrastructures)
Homework for Session 3: Map the expenditure
An Example from working with GA Legislators
a.
Map current
or desired
initiatives
onto main
chain
Proposed/
Current
Expenditure
becoming
at risk
Healthy
& Safe
Population
Afflicted &
Chronic
Population
At Risk
Population
c.
d.
becoming
afflicted
dying from
chronic
complications
b.
returning to
healthy & safe
Impacts
flow (a-d)
Why?
How long till
impact seen? Why?
Potential Unintended
Consequences
School fitness
programs
a&b
Develops healthy
behaviors in students;
prevents the onset of
obesity
Most visible after 20-40
years. Takes till that long
until students reach
“complications” ages
Funding from other
programs
….
….
….
…
…
Main Chains
Tips
gaining advocates
from resistant
deciding
for
deciding
against
Aware of
& May
Use DL
Resistant
to DL
becoming
aware
Advocates
for DL
losing
advocates
• Although you can map non-physical, it’s likely best to
make core infrastructure physical
• A structure mapping the building Support for a program
at Boeing becomes a chain of Advocates and Resistants
• Non-physical variables can then drive these more
physical flows
Activity (Flow) Generation
Healthy
& Safe
Population
Afflicted &
Chronic
Population
At Risk
Population
becoming
at risk
dying from
chronic
complications
becoming
afflicted
returning to
healthy & safe
treatments by
specialists
improving
degrading
Living
Environment
• Understanding whether a flow is generated primarily from a stock or
flow can lead to insights in where focus has been and what might be
done
• Have conversations with stakeholders regarding the essence of how
something is generated
Activity (Flow) Generation
Generic Templates
Templates
• Stock-based




Resource
Compounding
Draining
Gap-adjustment
• Flow-based
Resource
Gap-adjustment
pumping
changing
Barrels
of Oil
Oil Wells
Perceived
Quality
adjustment fraction
or
adjustment time
well productivity
Compounding
Draining
becoming
infected
discharging
Infected
Population
 Co-flow
perception
gap
Charge on
Capacitor
discharge
rate
Co-flow
infection
rate
criticizing
Using generic
templates helps
identifies levers
Criticism
Delivered
Self
Esteem
losing
self esteem
per criticism
Current
Quality
Example: Auto manufacturer
Strategy to reduce auto-related fatalities
Features by
automakers to
reduce fatalities
can reduce:
• Accident
probability
• Fatality
probability
Co-flows
annual
fatalities
Cumulative
Fatalities
accidents
per year
fatalities per
accident
miles driven
per year
accidents
per mile
Vehicles
•
•
miles per
vehicle
per year
External resource
But can also
reduce:
• Autos (UIO)
• Miles driven
(VMT)
Manufacturer classified initiatives as active vs. passive
Turns out using generic templates made it easier to classify
Closed-Loop Thinking
• Showing feedback loops
that drive changes in
dynamics
 Reinforcing
 Counteracting
(Balancing)
• Can do so just as well in
stock/flow maps…
• And you don’t lose the
main chain nor levers
identified
add 10% more!
producing stuff
from raw materials
Available
Raw
Material
Limit to Growth
Balancing Loop
selling to
consumers
Units of
Stuff
s
s
o
Supply Balancing
Loop
Demand
Balancing Loop
o
s
s
Price
units produced per
Production
unit of production
Infrastructure
infrastructure
and its Utilization
s
Demand
Can do with/without
generic templates
Healthy and Chronic Population discussion
CST Exercise with GA Legislators
spending on
afflicted & chronic
Cume
Spending
$ per afflicted
& chronic
demand for spending
on afflicted & chronic
Reinforcing
Loop
Healthy
& Safe
Population
becoming
at risk
At Risk
Population
mortality %
becoming
afflicted
Afflicted &
Chronic
Population
dying from
chronic
complications
returning to
healthy & safe
•
Current treatment strategies might create a vicious cycle (Reinforcing Loop)

•
More demand & spending for treatments on the Chronic Population means less spending on the At
Risk Population means more Chronic Population needing treatments)
What might be the most effective way to limit this loop’s impact?
The Skills Part 3
Framing the Issue
Scientific Thinking Process
Observe/Identify
Issues
•
In applying CST use same confidence building
process you would with a model
 Building confidence not proving truth
Build/Revise
Causal Theory
•
•
•
Develop/Test
Strategies
•
•
Communicate &
Disseminate Solutions
& Insights
Iterate through applying skills
Revisit reference behavior
When building maps start small - core
infrastructure
Best if core infrastructure is physical
Add a piece of structure, discuss implications,
revise and add from there
Value Add in Practice
Pulls out concept of
systemic orchestration
• Operational strategy
maps identify where
systemic orchestration
will determine strategic
effectiveness
• Knowing timing and
magnitude is just as
important as knowing
levers!
Cumulative
Sales
Large Account Sales
SF In
Training
becoming
fully
productive
6 month
delay
productivity
hiring new
sf
Large Account
Sales Force
3 year
delay
beginning
Info System
in Development
coming on
Info System
line
Identify Unintended Consequences
Rigorous approach to improving mental models
Policy Question
How can we reduce the number of people dying with AIDs?
be com ing
in fected
In fected
Pop ulation
Heal thy
Pop ulation
dyin g
Expanding
boundaries
leads to…
be com ing infe cte d %
avera ge time
in fected befo re
dyin g
be ginn ing
ris ky be havior
Risky
Beh avior
Pop ulation
be com ing
in fected
In fected
Pop ulation
dyin g
bi rths
be com ing infe cte d %
avera ge time
in fected befo re
dyin g
be ginn ing risky
be havior %
If our primary strategy is to develop interventions that
increase the lifespan of the Infected Population, what will
happen to the rate of “dying” over time?
Annual deaths
Years
Able to Identify Leverage Points
• Conversational Systems Thinking – because it applies
stocks and flows – allows better analysis using Donella
Meadow’s “places to intervene in a system” framework
• Her list of leverage points (next page) works from
constants (e.g. productivity terms) down to rules and
mindsets that determine the system
• CST allows you to examine the gamut of her list
Identifying Policies
Places to Intervene in a system*
In order of increasing effectiveness
12. Constants, parameters, numbers (such as
subsidies, taxes, standards)
11. The sizes of buffers and other stabilizing
stocks, relative to their flows
10. The structure of material stocks and flows
(such as transport networks, population age
structures)
9. The length of delays, relative to the rate of
system change
8. The strength of negative feedback loops,
relative to the impacts they are trying to
correct against
7.
6.
5.
4.
3.
2.
1.
The gain around driving positive feedback
loops
The structure of information flows (who does
and does not have access to what kinds of
information)
The rules of the system (such as incentives,
punishments, constraints)
The power to add, change, evolve, or selforganize system structure
The goals of the system
The mindset or paradigm out of which the
system - its goals, structure, rules, delays,
parameters - arises
The power to transcend paradigms
*Donella Meadows, Sustainability Institute and author of The Global Citizen
http://www.sustainer.org
By using CST, you can more easily explore implications of
policies with regard to leverage
Summary
Dana Meadows said:
We have just enough time…
starting now!
A Brief Bibliography
Resource
Author/Publisher
Notes
Seven Thinking Skills
Barry Richmond, Pegasus
Communications (2000)
Introduced concept of sevent thinking skills. Excellent
resource guide.
Introduction to Systems
Thinking
Barry Richmond, iseesystems (2002)
Extremely accessible work produced by one of the leading
lights of the field.
Strategic Modelling and
Business Dynamics
John Morecroft, Wiley Press (2007)
One of the most accessible modeling books. Perhaps too
business-y for the layperson.
The Strategic Forum
Chris Soderquist, Berrett-Koehler
(2000)
Description of using modeling for strategic planning
The Systems Thinker
Newsletter
Pegasus Communications
www.pegasuscom.com
Newsletter format. Broad range of topics covered. Rigor of
treatment varies.
Pegasus website is a clearinghouse for a host of dynamic
modeling/system dynamics/systems thinking resources.
Operational Strategy
Mapping
Soderquist, C. and Shimada, M.,
Pegasus Communications
Feature article on using strategy mapping at Boeing
The Fifth Discipline
Peter Senge, Doubleday (1994)
Qualitative treatment of broad array of dynamic modeling
skills and topics. Popularized “systems thinking” approach.
Many follow-on “field books” published over the years.
Industrial Dynamics
Jay Forrester, Pegasus
Communications, (1961)
Classic book on system dynamics, written by the founder of
the field.