@lkuceo Copyright Lean Kanban Inc. Presenter David J. Anderson TOCPA Utrecht Netherlands November 2013 Release 1.0.

[email protected] @lkuceo Copyright Lean Kanban Inc.

PresenterDavid J. Anderson

TOCPAUtrecht

NetherlandsNovember 2013

Release 1.0

Towards a framework for managing knowledge work

mailto:[email protected]


How I got into this…



Futility



My Evolutionary Epiphany





Motivation for the Kanban Method



Traditional Change is an A to B process

***either an internal process group or external consultants

CurrentProcess Future

Process

DefinedDesigned

transition

* Value stream mapping, ** Theory of Constraints Thinking Processes



Daniel Kahneman has given us a simple model for how we process information

Daniel Kahneman

System 1Sensory PerceptionPattern Matching

System 2Logical Inference

Engine

Learning byExperience

Learning from theory

FASTBut slow to learn

SLOWBut fast to learn



How we process change…

Daniel Kahneman

Silicon-basedlife form

Carbon-basedlife form

I logically evaluate change using System 2

I adapt quickly

I feel change emotionally using System 1

I adapt slowly



Adopting new processes challenges people psychologically & sociologically



TOC Observation #1



The Kanban Method



The Kanban Method…



The Kanban Method…

*also known as "kanban" in Chinese and in Japanese when written with Chinese characters



Water flows around the rock

“be like water”

the rock represents resistance



Kanban should be like water*

* http://joecampbell.wordpress.com/2009/05/13/be-like-water/



Principles behind the Kanban Method



The Kanban Lens



6 Practices Enable Process Evolution

The Kanban Method

VisualizeLimit Work-in-progressManage FlowMake Policies ExplicitImplement Feedback LoopsImprove Collaboratively, Evolve Experimentally

(using models & the scientific method)



(a) D-B-R

(b) CONWIP

(c) D-B-R + CONWIP (“CapWIP”)

(d) Kanban

Bottleneck

Any WIP Limited Pull System will work



TOC Observation #2



Flow Management



TestReady

Flow Efficiency

FH E

C A

I

GD

GYPB

DEMN

P1

AB

Lead TimeWaiting Waiting WaitingWorking

* Zsolt Fabok, Lean Agile Scotland, Sep 2012, Lean Kanban France, Oct 2012** Hakan Forss, Lean Kanban France, Oct 2013

Working

IdeasDev

Ready5

Ongoing

Development Testing

Done3 35

UATReleaseReady

∞ ∞Flow efficiency measures the

percentage of total lead time is spent actually adding value (or

knowledge) versus waiting

Waiting

LocalCycleTime

Flow efficiency = Work Time x 100%

Lead Time

Multitasking means time spent in working columns is often waiting

time

Flow efficiencies of 1-5% are commonly reported. *, **

> 40% is good!



Focus on Delay



Upstream Kanban Prepares OptionsReady

forEngin-eering

F

H

I

Comm-itted

D

4 Ongoing

Development

Done3

JK

12

Testing

Verification3

L

Commitment point

4 -

Requi-rementsAnalysis

2412 -

BizCaseDev

4824 -

Poolof

Ideas

∞

Min & Max limitsinsure sufficientoptions are alwaysavailable

Committed WorkOptions

Abandoned

OReject

P Q

$$$ spent acquiring options



Bottleneck should always be downstream of the commitment point

Poolof

Ideas

FH E

C A

I

Engin-eeringReady

GD

GYPB

DEMN

2

P1

AB

Pull

Abandoned

Ongoing

Analysis Testing

Done VerificationAcceptance3 3

Reject

Ongoing

DevelopmentDone3

Bottleneck should be here

Bottleneck workers should never be asked to work on something that is optional and may be

discarded. This includes any risk analysis (or estimation in legacy processes) that may be

required to assess viability of an option

Commitment point



Competing pressure of Last Responsible Moment versus Upstream Bottleneck

Poolof

Ideas

FH E

C A

Engin-eeringReady

GD

GYPB

DEMN

2

P1

AB

Ongoing

Analysis Testing

Done VerificationAcceptance3 3Ongoing

DevelopmentDone3

Commitment point

Last Responsible Moment Bottleneck early in workflow

Keeping the bottleneck early in the workflow means all downstream functions have slack

capacity. Reduces our need to manage WIP and reduces negative effects of variability in demand

There is a natural boundary between “the business” and the delivery/engineering function.

One generates options, the other converts options / delivers

commitments

Different governance is required for each



TOC Observation #3



Common cause variability



Variability in Delivery Rate (velocity)



Velocity Variation

South African Team from 2011plotted per Sprint (2 weekly)Mean 29, UCL (+1 sigma) 43 (+1.5x), LCL (-1 sigma) 15 (- 2x)



Mattias Skarin client based in Paris in 2009/2010, plotted weeklyMean 42, +1 sigma = 55, -1 sigma = 29 (+/- 1.4x)

0

10

20

30

40

50

60

70

80

90

DBA Team Velocity

Total VelocitySmall support tasks

(not includedin total velocity)Trend

Week of Christmas

Trend



Investment Bank, London, Extreme ProgrammingWeekly Mean 10, Max = 16, Min = 6 Spread (+/- 1.6x)



TOC Observation #4



Evolutionary Management



Evolutionary change has no defined end point

EvolvingProcess

Rollforward

Rollback

InitialProcess

Future process is emergent

EvaluateFitness

EvaluateFitness

EvaluateFitness

EvaluateFitness

EvaluateFitness

We don’t know the end-point but we do know our emergent

process is fitter!



Institutionalize feedback systems to enable evolutionary change

OperationsReview

SystemCapability

Review

StandupMeetingmanager to subordinate(s)

(both 1-1 and 1-team)



Fitness criteria are metrics that evaluate capabilities external stakeholders care about



If we order a pizza we know what we care about…



Disintermediate!Risks, fitness criteria & classes of service

should be explicit & transparent

OperationsReview

SystemCapability

Review

StandupMeeting

manager to subordinate(s) (both 1-1 and 1-team)

Expose risk, classes of service & fitness criteria

at all 3 levels of feedback

Lead timeQualityPredictability





Kanban viewed through a Cynefin* Lens

Simple

ComplicatedComplex

Chaotic

Best Practice

Good PracticesEmergent Practices

Novel PracticesSimpleKanbanSystem

DeepKanbanSystem

KanbanMethod

*http://en.wikipedia.org/wiki/Cynefin

Unordereddomains

Kanban systems alone aren’t enough in the unordered domain

Ordereddomains

EnlightenmentReductionistPhilosophy

Deterministic

NewtonianPhysics

PragmaticHolistic

Philosophy

Probabilitic

QuantumMechanics

Multiple work typesMultiple classes of service

disorder

Single work typeSingle class of service

KanbanMethod

NotApplicable



TOC Observation #5



Other TOC concepts I’ve tried and discarded



Critical Chain



Throughput accounting



TOC Observation #6



Conclusion



Differentiators of the Kanban Method



A future for TOC in knowledge worker industries



Thank you!



About

David Anderson is a thought leader in managing effective software teams. He leads a training, consulting, publishing and event planning business dedicated to developing, promoting and implementing sustainable evolutionary…

He has 30 years experience in the high technology industry starting with computer games in the early 1980’s. He has led software teams delivering superior productivity and quality using innovative agile methods at large companies such as Sprint and Motorola.David is the pioneer of the Kanban Method an agile and evolutionary approach to change. His latest book, published in June 2012, is, Lessons in Agile Management – On the Road to Kanban.David is a founder of the Lean Kanban Inc., a business dedicated to assuring quality of training in Lean and Kanban for knowledge workers throughout the world.




@lkuceo Copyright Lean Kanban Inc. Presenter David J. Anderson TOCPA Utrecht Netherlands November 2013 Release 1.0.

Documents