This document discusses classifying work by complexity using the Cynefin framework. It begins by explaining the four domains in the framework - simple, complicated, complex, and chaotic - based on the relationship between cause and effect. Typical problem solving approaches and behavioral loops are provided for each domain. The document then presents a model for classifying work based on technical and social complexity factors. Examples are given to demonstrate how specific work can be categorized. The document concludes by providing examples of how the classification approach could be used to label categories like project methodology or Kanban work item types.
The Role of Box Plots in Comparing Multiple Data Sets
Avoiding Analysis Paralysis
1. ‘Typing’ Work by Complexity
Avoiding Analysis Paralysis
Derek W.Wade
@derekwwade
dwade@kumido.com
gplus.to/derekwwade
www.derekwwade.net
www.kumido.com
www.kumido.com Attribution, Share-alike
First presented as “Stacey-Cynefin Mashup” at Agile Coach Camp U.S. — Indianapolis, IN — 2015
2. www.kumido.com Attribution, Share-alike
Snowden, D. and Boone, M. A leader’s framework for decision making. Harvard Business Review Nov 2007
https://hbr.org/2007/11/a-leaders-framework-for-decision-making
Complexity Determines Effective Action
Some Problems Don’t Yield Well to Analysis
Cause and effect is perceivable, repeatable, and
predictable.
Example:
Operating a hand-operated water pump
Approach: ‘Best Practices’ and ‘Standard
Operating Procedures’
SIMPLE
Domains
Relationship between cause and effect is
detectable, but separated over time and space.
Requires analysis, investigation, or expertise.
Example:
Operating an Airbus A380
Approach:‘Good Practices’/Systems Thinking
COMPLICATED
Domains
The relationships between causes and effects
are only perceivable in retrospect.The specific
cause and effect for one situation is unlikely to
apply again in the future.
Example:
Moving a hive of bees
Approach: Emergence (manage the whole)
COMPLEX
Domains
There is no detectable relationship between
cause and effect, even in retrospect.
Example:
Fighting a forest fire
Approach: Novel, rapid action to create stability
CHAOTIC
Domains
3. www.kumido.com Attribution, Share-alike
simple
complicated
complicated
complex
chaotic
increasing technical complexity
increasingsocialcomplexity
Common Sources of Complexity
People & Technology
Adapted from Stacey, R. Strategic Management and Organizational Dynamics. London: Pitman Publishing, 1996 and
Zimmerman,B. et al. Insights from complexity science for health care leaders Edgeware,VHA inc. Irving, Texas 1998
4. www.kumido.com Attribution, Share-alike
Domain Classification SOCIAL
COMPLEXITY
TECHNICAL COMPLEXITY
Factor Effect Factor Effect
Your work can be classified using:
1. The relationship between Cause and Effect
2. The best problem-solving Approach to use
3. Behavioral loops for this type of approach
4. An example for this type of domain
SIMPLE
Domains
1. Repeatable, perceivable, and predictable
2. ‘Best Practices’ and ‘Standard Operating
Procedures’
3. Sense ⟶ Categorize ⟶ Respond
4. Operating a hand-operated water pump
COMPLICATED
Domains
1. Detectable but separated over time/space;
requires analysis, investigation, experts
2. ‘Good Practices’/Systems Thinking
3. Sense ⟶ Analyze ⟶ Respond
4. Operating an Airbus A380
COMPLEX
Domains
1. Are only perceivable in retrospect, and do
not repeat
2. Emergence
3. Probe ⟶ Sense ⟶ Respond
4. Moving a hive of bees
CHAOTIC
Domains
1. No relationship detectable
2. Novel: rapid action to create stability
3. Act ⟶ Sense ⟶ Respond
4. Fighting a forest fire
Categorizing Complexity by Source
1. Familiarize your team with the categories of the framework
Adapted The Cynefin Framework by Cognitive Edge
5. www.kumido.com Attribution, Share-alike
Domain Classification SOCIAL
COMPLEXITY
TECHNICAL COMPLEXITY
Factor Effect Factor Effect
Your work can be classified using:
1. The relationship between Cause and Effect
2. The best problem-solving Approach to use
3. Behavioral loops for this type of approach
4. An example for this type of domain
# of biz processes affected
Biz process variability
# of process stakeholders
Stakeholder cohesion
+
+
+
-
# Systems touched
System technical debt
Frequency of releases
Type of development
+
+
-
(see below)
SIMPLE
Domains
1. Repeatable, perceivable, and predictable
2. ‘Best Practices’ and ‘Standard Operating
Procedures’
3. Sense ⟶ Categorize ⟶ Respond
4. Operating a hand-operated water pump
COMPLICATED
Domains
1. Detectable but separated over time/space;
requires analysis, investigation, experts
2. ‘Good Practices’/Systems Thinking
3. Sense ⟶ Analyze ⟶ Respond
4. Operating an Airbus A380
COMPLEX
Domains
1. Are only perceivable in retrospect, and do
not repeat
2. Emergence
3. Probe ⟶ Sense ⟶ Respond
4. Moving a hive of bees
CHAOTIC
Domains
1. No relationship detectable
2. Novel: rapid action to create stability
3. Act ⟶ Sense ⟶ Respond
4. Fighting a forest fire
Categorizing Complexity by Source
2. Team agrees on what factors drive complexity up (+) or down (-)
EXAMPLE ONLY!
6. www.kumido.com Attribution, Share-alike
Domain Classification SOCIAL
COMPLEXITY
TECHNICAL COMPLEXITY
Factor Effect Factor Effect
Your work can be classified using:
1. The relationship between Cause and Effect
2. The best problem-solving Approach to use
3. Behavioral loops for this type of approach
4. An example for this type of domain
# of biz processes affected
Biz process variability
# of process stakeholders
Stakeholder cohesion
+
+
+
-
# Systems touched
System technical debt
Frequency of releases
Type of development
+
+
-
(see below)
SIMPLE
Domains
1. Repeatable, perceivable, and predictable
2. ‘Best Practices’ and ‘Standard Operating
Procedures’
3. Sense ⟶ Categorize ⟶ Respond
4. Operating a hand-operated water pump
0-1 processes affected
Invariant business processes
1 Stakeholder
High Stakeholder engagement
COMPLICATED
Domains
1. Detectable but separated over time/space;
requires analysis, investigation, experts
2. ‘Good Practices’/Systems Thinking
3. Sense ⟶ Analyze ⟶ Respond
4. Operating an Airbus A380
~2 processes affected
Stable processes with few exceptions
Joint stakeholders
High stakeholder agreement/
engagement
COMPLEX
Domains
1. Are only perceivable in retrospect, and do
not repeat
2. Emergence
3. Probe ⟶ Sense ⟶ Respond
4. Moving a hive of bees
Several processes affected
Unstable processes, many exceptions
Several stakeholders
Some stakeholder disengagement, or in
widely dispersed locations.
CHAOTIC
Domains
1. No relationship detectable
2. Novel: rapid action to create stability
3. Act ⟶ Sense ⟶ Respond
4. Fighting a forest fire
Many processes, or large-
scale business goals affected
No process; exceptions are the rule
Stakeholders unknown or changing
frequently
Categorizing Complexity by Source
3. Team determines specific parameters for those factors
EXAMPLE
7. www.kumido.com Attribution, Share-alike
Domain Classification SOCIAL
COMPLEXITY
TECHNICAL COMPLEXITY
Factor Effect Factor Effect
Your work can be classified using:
1. The relationship between Cause and Effect
2. The best problem-solving Approach to use
3. Behavioral loops for this type of approach
4. An example for this type of domain
# of biz processes affected
Biz process variability
# of process stakeholders
Stakeholder cohesion
+
+
+
-
# Systems touched
System technical debt
Frequency of releases
Type of development
+
+
-
(see below)
SIMPLE
Domains
1. Repeatable, perceivable, and predictable
2. ‘Best Practices’ and ‘Standard Operating
Procedures’
3. Sense ⟶ Categorize ⟶ Respond
4. Operating a hand-operated water pump
0-1 processes affected
Invariant business processes
1 Stakeholder
High Stakeholder engagement
~1-3 system touched
No technical debt
Daily/weekly releases
Updates to a legacy system
COMPLICATED
Domains
1. Detectable but separated over time/space;
requires analysis, investigation, experts
2. ‘Good Practices’/Systems Thinking
3. Sense ⟶ Analyze ⟶ Respond
4. Operating an Airbus A380
~2 processes affected
Stable processes with few exceptions
Joint stakeholders
High stakeholder agreement/
engagement
~3-5 systems touched
Some technical debt
Weekly/monthly releases
New features to a legacy system
COMPLEX
Domains
1. Are only perceivable in retrospect, and do
not repeat
2. Emergence
3. Probe ⟶ Sense ⟶ Respond
4. Moving a hive of bees
Several processes affected
Unstable processes, many exceptions
Several stakeholders
Some stakeholder disengagement, or in
widely dispersed locations.
~5-7 or more systems touched
Significant technical debt, brittle systems
Monthly or quarterly releases
Significant new functionality to existing
systems.
CHAOTIC
Domains
1. No relationship detectable
2. Novel: rapid action to create stability
3. Act ⟶ Sense ⟶ Respond
4. Fighting a forest fire
Many processes, or large-
scale business goals affected
No process; exceptions are the rule
Stakeholders unknown or changing
frequently
Many complex interacting systems, or don't
know which system(s) to touch!
Huge technical debt; "Code one fix =
get multiple bugs"
No regular release cycle
"Green field" development
Categorizing Complexity by Source
3. Team determines specific parameters for those factors
These now become
‘checklists’ that anyonecan use
12. www.kumido.com Attribution, Share-alike
Example: Kanban Method Work Item Type
20
3
5
5Project
Task Request
Note: Underlyingmodel must consider
other elements of demand,of course ;)
Exploration