Full download test bank at ebookmeta.

com

Practical Kanban From Team Focus to Creating Value

Leopold

For dowload this book click LINK or Button below

https://ebookmeta.com/product/practical-kanban-
from-team-focus-to-creating-value-leopold/

OR CLICK BUTTON

DOWLOAD EBOOK

Download More ebooks from https://ebookmeta.com

More products digital (pdf, epub, mobi) instant
download maybe you interests ...

DE IMPLEMENTATION CREATING THE SPACE TO FOCUS ON WHA

WORKS Peter M. Dewitt

https://ebookmeta.com/product/de-implementation-creating-the-
space-to-focus-on-what-works-peter-m-dewitt/

Value in Business A Holistic Systems based Approach to

Creating and Achieving Value 1st Edition Jeffrey Yi-Lin
Forrest

https://ebookmeta.com/product/value-in-business-a-holistic-
systems-based-approach-to-creating-and-achieving-value-1st-
edition-jeffrey-yi-lin-forrest/

Creating sounds from scratch : a practical guide to

music synthesis for producers and composers 1st Edition
Metcalfe

https://ebookmeta.com/product/creating-sounds-from-scratch-a-
practical-guide-to-music-synthesis-for-producers-and-
composers-1st-edition-metcalfe/

The Meaning of Shared Value: New Perspective on

Creating Shared Value 1st Edition Floriana Fusco

https://ebookmeta.com/product/the-meaning-of-shared-value-new-
perspective-on-creating-shared-value-1st-edition-floriana-fusco/
Social Entrepreneurship An Evidence Based Approach to
Creating Social Value 1st Edition Chao Guo Wolfgang
Bielefeld

https://ebookmeta.com/product/social-entrepreneurship-an-
evidence-based-approach-to-creating-social-value-1st-edition-
chao-guo-wolfgang-bielefeld/

Design Works A Guide to Creating and Sustaining Value

through Business Design Heather M. A. Fraser

https://ebookmeta.com/product/design-works-a-guide-to-creating-
and-sustaining-value-through-business-design-heather-m-a-fraser/

Value investing from Graham to Buffett and beyond

Second Edition Mark Cooper

https://ebookmeta.com/product/value-investing-from-graham-to-
buffett-and-beyond-second-edition-mark-cooper/

Operations Management Creating Value Along the Supply

Chain 2nd Edition Roberta S. Russell

https://ebookmeta.com/product/operations-management-creating-
value-along-the-supply-chain-2nd-edition-roberta-s-russell/

Worlds of Rankings 1st Edition Leopold Ringel

https://ebookmeta.com/product/worlds-of-rankings-1st-edition-
leopold-ringel/
Practical Kanban
From Team Focus to Creating Value

Klaus Leopold

This version was generated on 4 Jan 2018.

© 2017 - 2018 Klaus Leopold

ISBN for PRINT version: 978-3-903205-00-0

ISBN for EPUB version: 978-3-903205-05-5

ISBN for KINDLE version: 978-3-903205-04-8

Table of Contents
Why did I write this book?
1. Why do we use Kanban?
1.1 Building Ships in a Workflow
1.1.1 We Cannot Complete More Work, Even If We Work Faster
1.1.2 We Have Enough Time for the Work We Never Have Time For
1.1.3 When We Set Limits, We Become More Predictable
1.1.4 When Everything is Important, Then Nothing Is
1.1.5 The Later We Begin, The Better For the Customer
1.1.6 Local Optimization Brings Global Sub Optimization
1.2 Kanban: Long Live Evolution!
1.3 Generating Value: Thinking About Service
1.4 Layers of Design: Kanban Flight Levels
1.4.1 Flight Level 1: The Operational Level
1.4.2 Flight Level 2: Coordination
1.4.3 Flight Level 3: Strategic Portfolio Management
1.5 Summary

2. Using and Improving Kanban Systems

2.1 Visualization, WIP Limits and Work Flow
2.1.1 Working with WIP Limits
2.1.2 Value and Flow
2.1.3 Dealing with several Work Types
2.1.4 Changes to Work Types over Time
2.1.5 Unplanned Work
2.1.6 Definition of Done
2.1.7 Summary
2.2 Dealing with Blockades
2.2.1 Blocker Clustering
2.2.2 Dealing with Backflow and Defects
2.2.3 Prioritizing Solutions
2.2.4 Interview with Matthew Philip
2.2.5 Summary
2.3 Customer Validation
2.4 Knowledge Transfer
 2.4.1 Capacity Constrained Resource
2.4.2 Specialist Bottleneck (Non-Instant Availability)
2.4.3 Specialist vs. Generalist
2.4.4 Summary
2.5 Coordination
2.5.1 From Idea to Coordinated Input - Replenishment Meeting
2.5.2 From the Input Queue into the Kanban System - Regular Standup
Meeting
2.5.3 Getting Better - Retrospective
2.5.4 Summary

3. Large-Scale Kanban
3.1 A Practical Example: A Sales Platform with more than 200
Project Employees
3.2 Scaling up Kanban
3.2.1 Consolidating Services
3.2.2 Connecting Services
3.2.3 Shared Services
3.3 Large-Scale Kanban at Bosch Automotive Electronics

4. Forecasting
4.1 Forecasting Requirements
4.2 Forecast for a Work Unit
4.3 Forecast for several Work Items without Historical Data
4.3.1 Determining the Minimum and Maximum
4.3.2 Monte Carlo Simulation
4.3.3 Continuous Throughput Forecasting
4.3.4 Interview with Troy Magennis
4.4 Can You Trust the Forecast?
4.4.1 Relationship between Work in Progress, Cycle Time and Throughput
4.4.2 Measuring the Stability of a System
4.4.3 Interpreting Stability Patterns
4.4.4 Interview with Daniel Vacanti
4.5 Summary

5. From Prioritization to Risk Assessment

5.1 Managing Demand with Cost of Delay
5.2 Quantifying Cost of Delay
5.2.1 Step 1: Determine the Value
5.2.2 Step 2: Determine the Cost of Delay
 5.2.3 Step 3: Sequencing
5.2.4 Determining Cost of Delay in Practice
5.3 Additional Refill Factors: Risk Assessment
5.3.1 Types of Risk
5.3.2 Quantifying Risk
5.4 An Interview with Markus Andrezak
5.5 Summary

6. Kanban in the STUTE Logistics Company

Terms used in this book
About the Author
Acknowledgements
Notes
Why did I write this book?
“In theory, there is no difference between theory and practice. But,
in practice, there is.” 1 This book offers practical solutions for the
most significant problems, which could arise when using Kanban. For
Kanban practitioners, this book offers motivation to expand Kanban
within your business.

Why another book about Kanban when there is already “Kanban

Change Leadership”? Good question! In our first book, Kanban
Change Leadership, Sigi Kaltenecker and I presented a short
introduction of Kanban with an emphasis on establishing Kanban
within your company. Our priority was to discuss, in practical terms,
the essentials of change management and how to design and
implement a Kanban system. Implementing a Kanban system,
however, is only the beginning of a long journey towards continuous
improvement in your organization.

The title, Practical Kanban: From Team Focus to Creating Value, is a

good indication of what you can expect from this book. I am a
Kanban practitioner. I was not completely happy doing pure research
at the university despite the work being very satisfying, because the
results of the research were never implemented in the “real world”. I
often thought that I had great solutions without a corresponding
problem. Once I started working as a Kanban trainer and consultant,
this was no longer an issue. In 2015 alone, I was involved in over
100 Kanban System designs and implementations. Since 2008, when
I started my Kanban career, I have worked on several hundred
Kanban Systems. After the implementation of a Kanban System, I
often return to the company some time later, analyze the system
and hold improvement workshops. Experience has shown me there
are typical problems that continually occur within Kanban
implementations. When I understood the problems, I was able to
offer better solutions. I began to catalog the solutions for these
problems, and eventually had a library of solutions for all possible
issues. This book brings structure to my experience and solutions
and makes them available for everyone to use.

Which problems seem to continually pop up in Kanban

implementations? The following questions provide a good summary:

Are we using Kanban properly?

How can we improve our Kanban?
How can we scale our Kanban?
How can our work become more predictable?
What should we work on next?

Chapter 1 reviews the fundamentals of Kanban. Advanced users

may roll their eyes and think, “Yeah, we know this already.”
Regardless, take time to review the basics because lack of
improvement in your Kanban implementation could be rooted in a
misunderstanding of these fundamentals. The questions

Why are we using Kanban?

What is a workflow?
Why should we limit the work in the system?
Where can I use Kanban in my company?

are answered in this chapter.

Chapter 2, Using and Improving your Kanban, dives into the

practical side of Kanban. The road to poorly implemented Kanban is
paved with good intentions, as perhaps you have experienced within
your organization. Complaints are made about a project taking too
long, Kanban is implemented, WIP (Work in Progress) limits are
established and it is assumed the work will automatically be finished
sooner. When project completion time doesn’t improve, the culprit is
almost certainly based on the wrong work being limited. In this
chapter, I discuss properly setting WIP-limits and give you tips and
tricks about Kanban visualization, WIP-limits and workflow. I’ll also
show you, through personal experience, how to learn from
obstacles, how to deal with bugs and how to avoid bottlenecks. The
end of this chapter answers the question: Which Kanban meetings
are useful and how can we organize them effectively? This chapter
should bring ideas for improving your Kanban system and help you
determine if your Kanban system is implemented correctly.

Kanban is rarely implemented with a “big bang”, meaning there are

not, all of a sudden, hundreds or thousands of people in a company
working with Kanban. More often it begins with pilot teams who
gather experience using Kanban, and if they are not completely
inept in their undertaking, the first positive changes occur. The
thinking often goes; if a little Kanban is good, more must be better,
right? The Kanban consultant/trainer cries tears of joy at the
prospect of so many billable hours, because dozens (maybe even
hundreds) of teams will need to be converted to the Kanban system.
This is great for the consultant, less so for the company. In the best-
case scenario, the company’s performance levels remain the same,
although more likely the performance levels will decrease. In
Chapter 3, I’ll discuss why this is the case and present a typical
example of how to better implement Kanban on a large scale.

Chapter 4 is completely dedicated to forecasting. Customers and

employers want to know, and rightly so, when they can expect work
to be completed. This requires predicting the future, which is never
easy, because there can be so many different outcomes based on
factors which we cannot control. Fortunately, some outcomes are
more probable than others, and we can use this to our advantage
when preparing a forecast for work completion. The beauty of this
approach is no time is wasted on guessing. With a little data, you
can prepare a forecast which is far more accurate than most
estimations.

In principle, a good situation is one where there is a lot of work in

the idea pool, backlog or option pool of a Kanban board, meaning
demand for your products and services are higher than your capacity
to supply them. Hopefully there will always be more demand than
supply, because more supply than demand, i.e. employees without
work to do, will inevitably lead to personnel reductions. We need to
appreciate the circumstance of having more demand than the
capacity to meet it, which immediately leads to the question: What
should we work on first, and which work can wait? I answer this
question in Chapter 5. I do not claim, however, to present the only
proper method of prioritization, as there are many different
approaches, each with their own advantages. In this book, I use
Cost of Delay as my favorite approach. At the beginning of a Kanban
implementation, I recommend using the Cost of Delay prioritization
method because it shifts the focus to the value of the results rather
than the cost of development. Time receives a price tag, which
allows economically sound decisions to be made. When the idea of
economy of delay has been completely embraced by the entire
group, you can begin to expand your toolkit of prioritization methods
and risk evaluations. The bottom line: Prioritization is nothing more
than making a decision based on your risk assessment.

In Chapter 6, I give a report about the eventful Kanban journey of

the STUTE Logistics Company. Their IT manager, Holger Reith,
transformed his department within 18 months into a high-performing
group. Before implementing Kanban, they were a “typical” IT
department: lots of open tickets and requests, long waiting times
and an improvement process in need of improvement. Today, tickets
arrive in a regulated workflow and the automated forecasting
informs the customer when the work will be completed. Employees
can bring ideas for improving the order of tasks to be completed
which ultimately determine the Cost of Delay. The improvement
process of the IT department at STUTE Logistics mirrors many of the
topics in this book and I am pleased to be able to share their story.
This practical example of a Kanban implementation, which shows the
ideas presented in this book being used on a daily basis, is the
perfect conclusion to my book.
By the way: You can find supporting materials to help you implement
the ideas presented in this book at www.practicalkanban.com.

Klaus Leopold

Vienna, April 2017

1. Why do we use Kanban?
A participant from our Applying Kanban training in Bangkok tweeted,
“Increased team performance can lead to decreased system
performance”. At such moments, I breathe a sigh of relief because the
participants truly understand the message: Kanban is not a method for
optimizing teams. Assuming Kanban is a team optimization tool already
limits the possibilities of using Kanban for your organization. This is,
without question, the biggest misunderstanding which continually
manifests itself when I’m working with companies or doing workshops
and trainings. Although the biggest, it is only one of many. I get called in
by companies to help them with their Kanban implementations and see
their Kanban boards. I observe how the employees build their boards
with gusto and, unfortunately, false supposition, and know it’s time to
take a step backwards. At such times, it’s necessary to gather the users,
review the principles and practices, and clarify exactly what it means to
use Kanban.

Like any new method, Kanban has become a trend, which in itself brings
risks. Whether brought on through their own actions, or through greater
economic crises and upheavals, companies search for the next magic
elixir which can solve all of their problems in economically unstable
times. The elixir being used at the moment is Agile. Companies hope to
find success by implementing Agile practices and, at some point, realize
just following the rules of a few practices is not the same as using them.
Behind every method stands an abundance of knowledge and insights
from principles and assessments which need to be understood in order
to use the method successfully.

This is exactly what I experience with Kanban; all hopes are put into one
method. Several times a month, I observe highly motivated people in
their various companies develop, within two days, the most amazing
Kanban systems. These works of art, composed of colorful Post-It®
notes, perfectly drawn columns and swim lanes, hang on the wall and
are proudly observed by the team. “I really hope Kanban can make us
faster” or “Hopefully Kanban will help us somehow” are statements
which commonly follow. While these statements slowly move to the back
of my mind, I imagine the daily routine in such a company. The team
sits in a semi-circle in front of the Kanban board as a member takes a
red Post-It® and writes “Block: Test Infrastructure not ready” and
proceeds to the board and places the Post-It® on a ticket in the Test
column, turns around and sits down again. His colleagues confusedly
look at him, often followed by an awkward silence, until someone finally
asks, “How does that help?”

Surprisingly, or perhaps not so surprisingly, it doesn’t help anything. Just

because you have visualized something with creativity and colorful
pieces of paper does not mean the weak points in your system magically
disappear. Have you ever been able to lose weight by just sitting next to
an exercise machine? Whether it’s a treadmill or a new method for
working more effectively, the effectiveness does not come from the
purchase. To be effective, it must be used. People are the driving force
behind every transformation. The other parts, whether it be the home
trainer or Kanban, are nothing more than tools. This is the point I wish
to convey about Kanban. It is a tool to help you see the weak points in
your system and generate better value for your customer. Sometimes
the Kanban board immediately reveals the problem, and often the
solution, at the same time and sometimes the sticking point reveals itself
more slowly. The Kanban board is only as useful, however, as the people
who are willing to see the problems and understand that they must
invest time and energy to bring about the change they seek. When they
have understood this, they can think about improvements, and make
precise and deliberate decisions about their system under two
conditions:

1. The participants develop their own process instead of blindly

implementing “Best Practices”. “Best Practice” is, in any case, a
dubious concept because it assumes to have explored every single
option and determined the best of all of them. Even if you’ve
attended many Kanban workshops or trainings to understand how
to improve your system, you still must do the work yourself.
 2. The participants understand their problem before they start thinking
about the solution.

Nevertheless, one of the greatest weaknesses in management today is

the haphazard installation of a solution without truly understanding the
underlying problem. The focus on solutions does not make the company
more adaptable to the uncertainties and emergencies of a complex
world. When Kanban doesn’t hold up to the expectations of
management, the problem is not with Kanban, but rather with lazy
management and their preconceived notions about improvement. If you
are searching for a To-Do list, or a few simple steps to remedy your
issues, don’t even bother trying Kanban and move on to the next trendy
method.

Toyota Production System: What is correct?

Many ideas of Kanban in knowledge work originated from the Toyota
Production System (Ohno & Bodek, 1988). Toyota willingly showed their
production system to anyone interested. However, the production system is
only a snapshot of the current solution to the given production
requirements, problems and challenges. Many visitors did not understand
this point.

Mike Rother, in his book “Toyota Kata”, tells the story of a friend who
traveled to Japan several times to visit a Toyota factory (Rother, 2010). On
his first visit, he observed how the assembly line workers took pieces off of
rolling shelves which were filled with parts from different car models. The
worker could pull the appropriate components from the shelf based on the
model of car which was rolling down the assembly line. Through the use of
these rolling shelves, several different models of cars could be assembled
on the same line, and many automakers copied this system.

Several years later, Mike Rother’s friend returned to the same factory, but
the rolling shelves had been removed. The component parts for each car
were arranged in a kit which rolled down the assembly line with the
appropriate car. When the car arrived at an assembly point, the worker
need only reach into the kit to retrieve the necessary part.

Mike Rother’s friend was indignant. He asked his Toyota host which
approach was the correct one, but the host didn’t understand the question.
The host answered, “When you were here the first time many years ago,
we were assembling four different car models on this line. Today we
 assemble eight different models on the same line - it was not possible to
hold the various components for each car model on one shelf. Not to
mention that we wanted to achieve a one-piece flow. Whenever you visit
us, you see the solution which we developed for a specific situation at a
given point in time.”

What problems are we trying to solve?

When I get a request for consulting, I try to steer the conversation so

the following questions are answered: Why does the company want to
implement changes? What problems are they trying to solve by
implementing these changes? As long as these questions remain
unanswered, every method - Kanban or not - will run into a brick wall.
Without answers to these questions, it is impossible to define the
parameters which measure improvements after implementing the
method. Look at Figure 1.1 and ask yourself which system you think is
better.

Figure 1.1: Which system is better?

In both charts, the x-axis is the cycle time in days and the y-axis show
the frequency of each cycle time. By calculating the average cycle time,
we see that System 1 makes a delivery every 20 days and System 2
makes a delivery every 35 days, so clearly System 1 is faster than
System 2. Faster is always better, correct? What if speed is important,
but something else was more important than speed? Let’s look at the
two systems again from a different angle.

Figure 1.2: Speed versus Predictability - what is better?

In Figure 1.2, the x-axis shows the deviation from the average cycle time
in days, giving us a different aspect of the systems being compared.
System 1 makes a delivery every 20 days on average, but with an
uncertainty of +/- 11 days. On the other hand, the owners of System 2
can deliver in 35 days, and the customer can be assured that they
receive their product within +/- 2 days of their delivery date. Thus
System 2 is more predictable. So, which system is better?

You probably have already noticed that neither system is better unless
we know in which context and in which industry these systems are used.
In the automobile industry, for instance, the production start of a new
model begins on a target date. Before this target date, all of the
components - which can be as many as 10,000 individual components -
for the new model must be developed. Once in production, the result of
this development is reproduced. Any additional delay to beginning
production would cost an arm and a leg. Predictability is, in this case,
the only criteria to evaluate the fitness of the development department
for the purpose of developing a new model. In contrast, speed of
delivery would be more important for a software start-up because the
customers want to see new features more quickly, even if quality is
perhaps still slightly lacking.

The real question is, what is important for the customer? Each company,
and each entity within a company, fulfills a specific purpose. For some
companies, the purpose is to make money as fast as possible. Other
companies prefer to be in the market long-term and believe in offering
the customer - whose purchases pay the bills - better service. The
internal processes of the company should be aligned to whichever
purpose they serve, thus giving the customer what they want and need.
In addition, a company must recognize what type of service a customer
values. By which fitness criteria (Anderson, 2013) does a customer
evaluate the type of service? The fitness criteria often are, but not
limited to:

Delivery time
Predictability/Adherence to delivery schedules
Quality
Security
Compliance with regulatory requirements.

Good fitness criteria, from the customer’s view, are those which are
definable and measurable. In reality however, such criteria - for
example, efficiency, agility or quality - are terminologies which each
person defines and understands differently. When I set up Kanban in a
company, I often hear the employees discuss their expectation of
increasing efficiency in the organization. Great, but at what point is the
company more efficient? Efficiency is a general term that needs to be
specifically defined since the idea of efficiency is different not only for
each company, but for each person. In such cases, I pose the following
question: What do we want to improve with Kanban and how can we
measure our improvement? By persistently repeating this question, the
empty umbrella terms of efficiency give way to tangible, quantifiable
metrics.

In the intensive process of determining objectives and their related

fitness criteria, it also becomes clear that you cannot have everything,
and priorities must be set. From an insider’s point of view, quick delivery
and high employee workload are rarely achievable, even if it appears
that way when compared to other companies. Even shorter delivery time
and predictability are polar opposites - flight traffic is the best example
of this. Pilots typically fly slower than what is technically possible
because flying faster brings no advantage. The airlines must hold to
their flight plans and adjust their flight times according to the air traffic
at the destination. Therefore, the airline does not use flight speed as a
measure for successfully completing a flight; rather, the pilot must fly at
the correct speed so he can land at the destination airport at the
planned time. In the automobile industry, a supplier would permanently
damage the customer relationship by delivering component parts a week
earlier than requested and needlessly using up precious storage space.
Just because something can be done doesn’t mean it makes sense to do
so.

Good customer service provided by a company is always linked with a

deliberate decision to define the focus points of that customer service
without sacrificing improvements to their development. This assumes
that you know your customer, first of all, and understand them, second
of all. Understanding the customer is usually where the company invests
the least amount of energy. I often observe this in companies where the
customer is never mentioned in meetings, and if they are, it is in the
context of being a nuisance. In classical change management, the
customer receives her 15 seconds of fame through the prominent
placement of her name on the first Power Point slide - by the Org Chart
slide, the role of the customer has been forgotten. Wanting to
understand the customer is the basis of every improvement and,
therefore, is also the point of the question as to why we want to work
with Kanban.

If a company wants to make themselves ready to use Kanban, they need

to find their own answers to the following questions:

1. How do we find out what the customer wants?

2. Once we know what our customer wants, how will we fulfill the
requirements? What must our workflow look like to generate value
for the customer?
The answers to the first question are addressed by methods such as
Discovery Kanban www.discovery-kanban.com or practices like Lean
Startup, Lean UX and Design Thinking. This is fascinating material and
highly recommended if you truly want to understand your customer. The
focus of this book, however, is to answer the second question. We
assume that a company has already answered the first question and
knows what the customer values. The next step is to build a work
system which can fulfill these customer values in the best way possible.

We have already established, that Kanban - like any other method -

cannot work wonders just by existing. No method or framework could
ever be suitable for every single company in the world. Why, then,
should Kanban be an appropriate tool for solving question 2?

1.1 Building Ships in a Workflow

Kanban is not always intuitive. Occasionally at the start of a Kanban
training, before the participants can settle into a comfortable state of
listening and learning, I will tell them… nothing. Instead I begin with an
exercise - difficult to describe in writing, but I’ll give it a try. For the next
few minutes you can feel like Aristoteles Onassis: time to build a ship!

A Workflow Experiment: Folding Paper Ships

What you will need:

A stack of paper
A stopwatch
At least six people, but no more than 12 people, in a single row
(designates the work system)
The first person in the row (work system) is responsible for
passing the paper into the system one piece at a time.
The last person in the row (work system) is responsible for
timing the work.

How to proceed:
 Figure 1.3: Ship Building Setup

The people in a work system sit side-by-side at a table (make sure the
table is long enough).
With the exception of the next-to-last person, each member in the
work system is responsible for a maximum two folds of the ship,
regardless of the total number of people in the system, and each fold
(activity) is distinct for each work system member. I define an
activity to be the individual steps or actions which are
executed in a work system that at the end produces tangible
value. An activity can have many different forms - in this experiment
it denotes the various folds needed to form a paper ship.
The next-to-last person must complete several complicated folds.
The last person time-stamps each completed ship.

The experiment is done in two rounds:

Round 1: There is no limit to the work which flows into the system.
The members work according to the push principle and, as soon as
their work (fold) is completed, move the ship to the next member.
Round 2: This round demonstrates the pull principle. Only one ship at
a time can be worked on in each individual activity (WIP limit of 1).
This means, as soon as a participant has completed her activity, she
raises her hand and leaves her ship on the table in front of her. She
can only take a new ship into her activity when her ship has been
pulled into the following activity and the ship in the previous activity is
finished and waiting to be pulled.

The following rules are used for both rounds:

Ships will be worked in a strict First In, First Out (FIFO) order; i.e, the
ships will be worked on in the order that they enter the system.
Team members are not allowed to help one another or improve the
process. This might seem strange, but it is important since we want to
measure the difference between the two rounds. If the process would
be improved in the second round, we could not compare the
measurements between the first and second round.
 Work needs to be done quickly because we want to maximize output.
In economic terms, if each ship produced can be sold for ten dollars,
then the goal is to produce as many ships as possible in each round.

Push it! Everyone knows this situation well. There is endless amount of
work to finish and the salespeople pull in one order after another.
Ursula, from today’s training, gets the roll of Sales Manager. Her job is to
fold the piece of paper in half and pass it into the work system. In the
first round, she does this without pause. At the beginning, everyone is
still having fun and Harald, the next one in line, only has to fold the
paper in half. He’s pretty swift and manages to pass the paper within
half a minute on to Anna, whose activity is a bit more complex because
she needs to fold the corners towards the middle. The next member,
Lisa, also completes her activity deftly, but slowly a backlog of work
starts to pile up. As the pile gets larger, the laughter dies down and a
concentrated silence takes over the room. Everyone hastily folds their
ships to reduce their pile as quickly as possible, but before a ship is even
completed, two more have landed on the pile. Tom, who sits at the end
of the row and has to complete three activities, has it the worst. I joke
with him, “Hopefully you don’t break your fingers,” as he continuously
folds the ships one after the other. For the first few seconds he keeps up
the pace fairly well, when all of a sudden two ships pile up, then three,
then five until he has a vast pile next to him: a textbook example of a
bottleneck. Too bad for Tom - everyone else is so busy doing their own
activity that nobody has time to help him. John, on the other hand, has
an easy job. He measures how many ships per minute are completed,
starts the stopwatch at the beginning of the round and each time a ship
leaves the final activity, he notes the time. I motivate the members to
work faster and faster because I can sell each completed ship for ten
dollars and I want to get rich!

After two minutes, I send a red piece of paper into our ship building
enterprise and give notice that it will be the last ship. Gradually each
member finishes her work, but Tom has a few minutes of work left.
While the members are waiting for Tom to finish his work, I give them
some food for thought. I tell them, “Don’t forget, in reality you would
need to continue working because there are always new customer
orders being placed.” Finally, Tom finishes the red ship and sends it to
John, at which point I say “Stop!” John measures the cycle time (see the
formula below) of the red ship:

Arrival time is 5 minutes 18 seconds, so the red ship was completed in

the first round with a cycle time of 3 minutes 18 seconds. Everyone gets
a short break (Tom can shake his hands out), and then we’re off to the
second round.

Pull it! The second round begins under different circumstances. In each
activity, only one ship can be worked on at a time, which denotes a WIP
limit of 1. Ursula, our Sales Manager, holds to the guidelines exactly and
sits calmly in front of her pile of paper. She folds the first piece, then
raises her hand. Harald pulls the piece of paper from Ursula, completes
his fold, then raises his hand. This is the signal for Anna to pull the piece
of paper from Harald, complete her folds, then raise her hand. It
reminds me a bit of a human “wave” moving through a football stadium
- a rhythm is established in the assembly line: complete the fold(s), raise
your hand, then pull the next ship. Tom, who is the process bottleneck
at the end of the row, has less stress in this round because he can
steadily fold his ships to completion without ships piling up in front of
him. However, since Tom must make four folds to complete the ship, his
tempo determines the speed of the entire production line. That means,
at some point during the exercise, everyone must wait for Tom to
complete his activity before the next ship gets pulled. When this
happens, I comment (again) how each ship can be sold for €10 and if
the workers in our “factory” are staring at the ceiling instead of working,
we will soon go bankrupt. It is interesting to see what the participants
do while they are waiting for the next ship. Some closely observe what is
happening in the production line. Others use the time to improve the
quality of their work by carefully going over the folds again. After two
minutes, I send a red piece of paper in to the system. At 2 minutes and
45 seconds, the completed ship arrives at John - the cycle time is 45
seconds. Stop.
1.1.1 We Cannot Complete More Work, Even If
We Work Faster
Let’s review shortly: In the first round using the push principle, everyone
worked as fast as they could. In the second round, several people had
time to relax and just look around or used the time to improve their
work. This can only mean the second round was less productive, or-? I
asked the participants how they felt after working with the pull principle
and I received the following answers:

It was more relaxed.

I was able to concentrate.
We could work on quality.

These observations are interesting. I immediately note (since I’m the

boss of the ship company) that I believe I made more money in the first
round. In the second round, I claim that I had to declare bankruptcy in
because everyone spent their time staring at the ceiling instead of
working. How can we find out the amount of money we made in each
round? We examine the throughput. Throughput tells us how much work
- in our case the number of ships - is completed within a given time.
When I take, for example, an average throughput of four ships per
minute, and I sell each completed ship for 10 dollars, I earn 40 dollars
per minute. Now the moment of suspense: we calculate the
measurements from John together. Can you guess how many ships were
completed per minute in each round? I draw a table on the flip chart.
We need to remove the first result from both rounds because the system
was not completely filled at that point. The final result is removed as
well, since it was a partial minute when the final ship was completed.
We are searching for the average throughput per minute, and the first
and last measurements would skew these results. We can now calculate
the average throughput of each round.
 Minute Push with no Limit Pull with WIP 1
0 3 3
1 5 6
2 6 5
3 6
4 5
5 2
Ø Throughput 5.5 ships per minute 5.5 ships per minute
Cycle time of
the red ship 3:18 minutes 0:45 minutes

I love this moment - the participant’s faces are filled with looks of
bewilderment. How are these results possible?

In both rounds, approximately five ships leave the work system

every minute.
The second round using the pull principle was completed more
quickly, since the red ship was delivered after 2 minutes 45 seconds
versus 5 minutes 18 seconds in the first round. The cycle time was
reduced by a factor of four, from 3 minutes 18 seconds to 45
seconds.
Stress in the first round, free-time in the second round.

The participants understand the results as soon as I look towards the

end of the row in our ship building enterprise: there was the bottleneck
in the process (mind you, only the process - people are not usually the
bottlenecks).

Whether working in a push system or in a WIP limited system, the

bottleneck determines the throughput, because the work is deadlocked
latest at this point in the system. It doesn’t matter how much work we
dump into the system or how fast the previous activities are finished: at the
end of the day the amount of work completed is the same.

For managers in manufacturing plants, this is rudimentary knowledge.

Even when folding our ships - our “manufacturing” plant - it was easy to
identify the bottleneck. In the area of knowledge work the concept of a
bottleneck is well-known, but it is often difficult to easily identify the
bottleneck. In knowledge work bottlenecks tend to occur temporarily in
one area and then move on to another area because each piece of work
done is almost never the same, i.e. a different “ship” is produced each
time. There are no two identical customer orders in knowledge work,
thus each requires different amounts of work in the various areas each
time value is created. For example, one work item may require a
majority of effort in the analysis phase and less effort in the
development phase, so you might assume the analysis phase is the
bottleneck. However, this can change with the next work item.

Often too much energy is invested trying to overcome perceived

bottlenecks, when the obvious solution is usually overlooked: The inflow
of work can be limited and the resulting available resources can be used
to ease the bottleneck. Bottlenecks are discussed in more detail in
section 2.4.

1.1.2 We Have Enough Time for the Work We

Never Have Time For
Let’s look at it from an economic perspective: When we can sell each
ship for 10 dollars, we will earn the same amount of 55 dollars per
minute (5.5 ships/minute x 10$ per ship) whether we use a push system
or a pull system. Or think about the usual situation in companies, where
although employees are already working to their limits, work continues
to be added. Such a heavy workload is, ironically, perceived positively
within many companies. Since the bottleneck determines the throughput
of the system, however, it makes absolutely no sense to continuously
increase the workload because the system will not become faster or
more productive; not to mention in a push system, a large amount of
capital is tied up in the backlog of work in progress. In knowledge work,
there is a lack of awareness of this tied-up capital because the backlog is
invisible - mostly it is stored data which is no longer worked on.

W. Edwards Deming would say, “It’s not about working harder, it’s about
working smarter” (Deming 2000). This is exactly what happens in a pull
system, because the work flows in only as fast as the bottleneck allows.
For one thing, workers in the system can budget their time accordingly
and second, they still have time left over (which is, ironically,
automatically interpreted as a bad sign). So I ask my training
participants, “What could we do with the free time available while
waiting for the bottleneck to clear?”

1. We could assist our colleagues at the bottleneck. In practice,

one of the most common objections to this idea is “We are not
specialists for the bottleneck work.” The good news is, that is beside
the point of helping a colleague! It’s going about easing the burden
of work and this can be done in many different ways. For example,
you could take over some administrative duties or go as a substitute
to meetings - those wanting to help don’t necessarily need to
perform specialized work. More important is that the colleagues in
the bottleneck can concentrate on completing their work. What are
the consequences? Even when a non-specialist assists at the
bottleneck, the throughput increases (see also section 2.4). It may
be slower than a specialist (who is presently unavailable) stepping
in to help, but it still improves.
2. Is this even an optimal process? When everyone in a
department is working at full speed and, nevertheless, work is being
completed slowly, the fundamental question about balance needs to
be asked. It is possible that the manner in which the work is
distributed and performed is not ideal. In our experiment the
imbalance was readily seen. While every other process step only
had one task to complete, at the bottleneck there were three tasks
to complete. Distributing the work more evenly would resolve the
bottleneck and the throughput increases.

More throughput means more product can be delivered in the same

amount of time. As we saw in the ship experiment, the free time can
also be used to improve the quality of the work. One person would be
extremely happy when more product of better quality can be delivered:
the customer. The beauty of it is, quality improvements stemming from
employees reflections on the process during the available free time, cost
the company nothing. There is no need to engage expensive process
engineers or consultants (such as myself) to take over contemplating
what can be improved.
 Slack offers an opportunity, with the same amount of output, to identify
and evaluate the weak points in a system, as well as working on quality and
finding ways for improvement. In a non-limited system, these opportunities
do not exist due to chronic overload.

1.1.3 When We Set Limits, We Become More

Predictable
The stability and hence the reliability of a system are directly associated
with one of the least favorite topics in the business world: estimation. In
the classical overloaded push system with - you guessed it - employees
working to capacity, someone is given the thankless job of making a
project plan. So, the project planner goes to the individuals, teams or
departments and to each one in turn says, “I have a question. When will
you be finished?” Each one picks a number then doubles it as a
precaution. The experienced project manager adds the partial
estimations together and multiplies it by three, just to be safe, and
finally arrives at the delivery date on which the product or project will
most certainly not be finished. Consequently, because the delivery date
is so often missed, the project manager preoccupies his colleagues,
while in a system where no work seems to get completed, by constantly
asking, “How much do you have left to do?” This question manifests
itself in supervisory measures, such as status reports, which require
more administrative work and prevents the employee from getting to the
real work, thus exacerbating the problem further. On top of this, the
employees are sent to an estimation workshop so they can finally learn
how to do estimation correctly, although the employees and their
estimating capability are not really the problem. The cause of
incorrect estimations is the amount of work in the system is
constantly changing. A reliable prediction method, regardless
which one, can only function on the basis of a stable system.

Nobody can give a valid estimation for when work will be completed in
an unlimited system where new work is continuously added that needs
to be worked on immediately. In addition, people are conditioned to
estimate effort. They ask themselves the question “How long do I
need?” instead of focusing on the cycle time by asking “When will it be
finished?”

Think once more about our flow experiment: What would have
happened if I had placed the red piece of paper into the system after 15
minutes instead of two minutes?

In the push system, a massive backlog would have formed and we

wouldn’t be able to say when the red ship would be completed. We
would only know that it would take much longer - but how much
exactly? In other words, this system is unstable. The customer
receives no reliable statement as to when he will see his ship in the
water.
In the pull system, the answer is simple: We know how much work
is in the system. This system is more stable. Therefore, it is
possible, through the use of estimation or forecasting for example,
to say when the work will be completed. Utilizing a pull system for
folding ships does not change the cycle time. On account of this
stability, the ship building enterprise can themselves, after 15
minutes, give their customers reliable information about the
completion time of the ship. In knowledge work it’s not quite that
easy because the scope of the work varies, rather than the same
object being built over and over again. Nonetheless, in knowledge
work a limited system is a prerequisite for being able to say when a
job will be completed.

One of the greatest advantages of WIP limits for both the worker and the
customer is predictability. On-time delivery is only possible when the
amount of work which gets started is limited, which results in a stable
system.

1.1.4 When Everything is Important, Then

Nothing Is
Let’s assume the customer puts the red ship into our unstable push
system at the 15-minute mark. As a result, the cycle time increases
dramatically and after an hour the ship is only approximately 80 percent
complete. The following exchange is what typically follows in real world
companies everyday:

Customer: “I could fold the ship together myself in two minutes and you
are telling me that after one hour it still isn’t finished? Hurry it up!”

Manager: “As you know, we place our focus on the customer.”

Which means the red ship gets top priority and is sent past all other
ships in the assembly line. Bad luck for the customers waiting for white
ships; they now have to wait longer. Thank goodness, the red ship will
be finished before the others - at least the impatient customer is
somewhat satisfied.

However, a company usually has more than one customer. The possibility
increases within an unlimited push system that there are many unhappy,
waiting customers who want their orders pushed forward. At some point
the system is flooded with red ships. What happened?

The advantage of priority at some point turns into a disadvantage.

If more high priority red ships need to be built, the normal priority
ships will need even longer to complete, although this is already the
case due to the individual process steps. Not only will the white
ships take longer, production of the red ships will also gradually slow
down.
When the system is eventually filled with nothing but high priority
red ships, priority no longer exists. If everything has priority, then
nothing has priority. The red ships now have the same cycle time
which the white ships had before, and the white ships will most
likely never be finished - at least until someone makes it top priority.
The entire system becomes slower, again.

You would not believe the bizarre excesses which can occur from an
uncontrolled prioritization in the real world. In many companies, I have
seen not only priority 1, but also priority 1+ and 1++. When a company
reaches this stage, a task force is established, which itself gets
prioritized - a fascinating endless circle.
 If priority should be guaranteed to a customer, it is necessary to limit not
only the system, but also the priorities.

1.1.5 The Later We Begin, The Better For the

Customer
Let’s go back to our special customer who gave the order for a red ship
at the 15-minute mark. Since the ship building enterprise hasn’t
delivered anything and the world has continued turning in the meantime,
the customer wants something new after the fourth activity. Now he
wants a green ship. Remember our experiment: in the first round the
red ship entered the system after 2 minutes and had a cycle time of 3
minutes 18 seconds. If the ship had entered the system after 15
minutes, the cycle time would have been approximately one hour. This
means the longer you work in an unstable system, the longer the cycle
time becomes. On the other hand, in the second round the cycle time
remained the same because the ships already in the system were
completed before something new was started.

In a push system, the following would occur: the customer’s red ship
would be taken out of the process and instead a green ship would be
placed at the beginning of a system which is already overloaded. Not to
mention the work already performed on the red ship is wasted. That
hurts the bottom line.

What would happen in a WIP limited pull system? Nothing. The work on
the red ship would have not started when the customer changed their
mind, thus there would be no economic loss.

One of the fundamental principles of a limited system is to keep the

work outside of the system as long as possible (Late Commitment). The
simple reason is once a job is started, it should be finished as fast as
possible. Work that is being planned will not be stuck in the system,
because this would increase the cycle time. Instead, the planned work is
put on hold before entering the system. In Kanban, this is known as the
Option Pool. For the work system, the effects can be advantageous:
 In the case where the customer changes their mind, the likelihood
of wasted work due to the Late Commitment is reduced.
The lower the WIP limit of the system is set, the shorter the cycle
time. Shorter cycle times reduce the probability that the customer
wants or needs to change something in the job while it is being
worked on.

With a limited system, we strive to keep the cycle time short and at the
same time start the undertaking of a job as late as possible (Late
Commitment). Thereby we reduce the risk that job changes requested at a
later time would make the work already performed useless.

1.1.6 Local Optimization Brings Global Sub

Optimization
Imagine the individual activities of our unlimited push system for
building paper ships were performed by a team versus just one person.
With everything else remaining the same, management decides to
perform a team optimization. The first two teams are chosen for the
optimization to use a new process to deliver faster than before. Each
week the selected teams improve their performance, their throughput
increases and the cycle time decreases. Management happily pumps
more jobs into the system - a reason to be satisfied with their decision.
Why, however, does the cycle time of the entire system suddenly
increase?

The local optimization of the teams at the beginning of the process has
the effect that more jobs can be pushed into the system. As more work
is shoved into the system, the amount of work in progress (WIP)
increases. More work in progress means longer cycle time. If a
bottleneck is present in the system somewhere (and there is always a
bottleneck), it determines the total throughput of the work system. The
local improvement of individual teams results in the entire system
slowing down.
It is a firmly held belief in many companies that local optimization has a
positive effect on the entire system. In the Agile way of thinking, the
team is the focal point of optimization because the value stream consists
of building a team with representatives of each required discipline. This
leads to a false conclusion, because rarely can a single team cover the
entire value chain. In the current theme of Agile scaling is this
particularly clear: many companies wanting to become agile believe they
merely need to educate their teams on agile methods to become an
agile business. By no means is this true. I am a fan of the organization
theorist Russell L. Ackoff from whom this timelessly valid statement
stems:

“And therefore, the performance of the whole is never the sum of the parts
taken separately, but it’s the product of their interactions.” (Ackoff &
Gharajedaghi 1984).

Unless the newest miracle work improvement weapon is currently being

tested (such as Scrum or Kanban), the false belief of local optimization
manifests itself often in the form of objectives for a single person or
teams tied-in with bonuses or other perks. With this indirect attempt at
optimization, the pressure on the work system is merely increased. In
the best case, this has no effect on the overall performance of the
system. Normally, however, a decrease in performance over the entire
system will occur. What would the motivation be to help a colleague (in
a different team or department) at the bottleneck when the team itself is
measured exclusively on accomplishing their local objectives? The fact
that a team is only one element along the path of producing value for a
customer goes completely unnoticed. Local optimization is for me a clear
case of failed vision: the understanding of the system as a whole is
missing. Let’s look at local optimization from the customer’s point of
view. For them, it is completely irrelevant if Team 1 or Team 2 works
faster. Referring to the example of ship building, this form of
performance improvements brings no value to the customer because a)
they still have to wait for the entire ship regardless of how quickly the
individual activities are completed and b) they have to wait longer for
the ship to be finished because Team 1 and 2 slow down the entire
system with their performance. Local optimization leads to global sub-
optimization.

Perhaps you understand now why I get ruffled when Kanban is

described as a team method. Using Kanban solely for individual teams or
departments is, in many cases, shooting yourself in the foot. Kanban
best reveals its potential when initiatives for change are removed from
the team level and instead the optimization considerations are applied at
the highest overall level possible (more on the Kanban Flight Level
theme is discussed in section 1.4). Visualizing is a central practice in
Kanban for good reason. Above all, at the beginning of a change
initiative, visualizing helps to see as much as possible from the system in
order to answer two basic questions:

1. How do we generate value for our customers?

2. Where do the problems lie in our system?

Had the management of our imaginary ship enterprise looked at the

overall picture, one thing would have become clear: The bottleneck is
the point in the process where optimizations should be implemented. In
the ship experiment, the idea was proposed that a single individual was
the bottleneck. In reality, the same thinking occurs. When people work
in a system, then the optimizing must occur at the people level. Thus,
trainings and workshops are given with the goal to make each person
work faster, not the system as a whole.

Naturally a person’s performance and how quickly she completes her

work is based on her abilities. I have done this experiment hundreds of
times on different continents with different people. Regardless of where
or with whom the experiment is done (with the exception of the
occasional extreme case, good or bad), the throughput is always
between five and six ships per minute - much more or much less is
simply not possible. Fascinating, isn’t it? The result that the ship goes
slower through the system does not lie in the folding skills of the
participant with the final task, rather the participant has more work to
do than the others, thus our work system is the problem. Not only with
the ship experiment, but also in companies with which I work, I see the
abilities of individual people have an extremely small effect on the entire
output of an organization, unless these skills are part of a particular
specialization. Looking more closely, the bigger issues are always with
the work process, not with the people doing the work. The goal of
Kanban is not the local optimization of individual pieces of a system,
rather the improvement of the value stream from the time of
commitment (Work Agreement) up through the delivery to the customer.

Local optimization, in most cases, results in a decrease of performance

over the entire system. To improve the performance of a system, the focus
needs to be on the interactions between the individual parts.

At the end of the ship experiment, before delving deeper into the
principles and practices of Kanban, I always ask participants what they
can take away from this experience and what conclusions they reached.
Following is a sample of the answers I receive:

One should understand the processes used.

Throughput is determined by the bottleneck.
What is really important to the customer?
Implement WIP limits and observe the effect. Base decision-making
on the most suitable WIP.
Prioritization should occur before work enters the system, not when
the work is already in the system.
Encourage and demand active management.
Without WIP limits there is no time for improvement.
With WIP limits, there is reduced stress with equal
output/throughput.
With WIP limits it is possible to adjust to new requirements even at
a later time.
WIP limits results in more predictable and faster cycle times.
Even priorities need to be limited.
The focus should be on the entire system, not just pieces of it.
1.2 Kanban: Long Live Evolution!
The goal of the ship experiment is to show the difference between the
two work systems. In one system, a consistent and continuous workflow
is never achieved because the rhythm is determined by external
pressure. The other system achieves equal, if not better, results because
it regulates the inflow of work into the system through which a
continuous workflow is then established. It is an effective system,
although there are many things in a flow-based system that seems
counter-intuitive. Some of these we observed in the ship experiment
(itself actually a flow experiment):

We are faster, even if we are not working at full speed.

We have enough time for work for which we normally never have
time.
We become more predictable when we set limits for ourselves.
When everything has priority, then nothing has priority.
The later we start the work, the better it is for the customer.
Local optimization results in global sub-optimization.

I would like to come back, at this point, to the objection which some
participants bring up during the discussion about the flow experiment.
When we are talking about the use of Kanban, we are not talking about
industrial manufacturing, rather about knowledge work with
unforeseeable complexities. The term “Kanban” was made famous by
the Taiichi Ohno designed pull system at Toyota Production System and
simply means “signal card”. During the assembly of an automobile, these
signal cards showed that a production step had been completed and the
next delivery of parts could be received. Kanban systems, however, are
not an invention of the 20th Century. At the imperial garden in Tokyo,
the number of visitors allowed into the garden was controlled by the use
of tokens which were handed out at the entrance and collected again at
the exit (Anderson, 2010). This way, the number of visitors was limited
to prevent overcrowding in the garden, thus allowing each visitor the
chance for enjoyment and relaxation.

Knowledge work has very little to do with tangible objects: knowledge

doesn’t lie around in half-finished pieces and we cannot measure the
workload based on the number of people hanging out in a room.
Obviously, the practices from Toyota cannot be transferred directly into
the area of knowledge work, but there are system independent
principles behind the practices which are applicable to every work
system. Every system has limited capacity, thus limiting the work which
gets started makes sense regardless if the system is new software
development, manual labor activities, or industrial production.
Bottlenecks, as well, exist in every type of system in the world. Whether
it’s the imperial garden in Tokyo, the production plant at Toyota or
Kanban in knowledge work, the crucial point in each system is:

a. creating value for the customer, and

b. the desire to keep the system flexible and look for ways to
continuously improve.

Kanban, in knowledge work, offers no rigid schema for process

improvement, since a rigid schema is geared towards a particular
delivery or output. If this were required, nobody would be successful at
implementing Kanban, because no two companies or systems are exactly
the same. As you remember, each work process should fulfill the
customer’s wishes - making each process unique - and cookie-cutter
procedures do not work for unique processes. Kanban is not a pre-
defined or standardized process. In knowledge work, Kanban is custom
tailored to each system, allowing the flexibility to define how it is
implemented, without dictating what the end product should look like. It
is built upon evolutionary thinking; start with the work you do at the
moment, and improve it based on the practices and principles which
Kanban offers. In time, a process which is specifically tailored to the
needs of the individual company, is formed.

Of course, this is completely opposite of what typically happens in a

company. In an artificially initiated process improvement, someone, who
often has nothing to do with the daily work processes, attempts to
invent an optimal work process to keep the company competitive. As
soon as the new policies are installed, process improvement is
considered finished.

Principles and Practices

Kanban, however, is a bit more modest. If everything within a company
was truly terrible, the company would have already fallen out of the
market. As such, Kanban takes what already exists, analyses the current
situation, and builds from there, without assuming to know how the
employees should be working. Kanban helps discover, in increments,
how work processes can be improved. When a company implements
Kanban, it should be combined with perpetual reflection: what was
impacted by the last improvement? Did it improve anything? If it was an
improvement, let’s continue with it - if not, let’s try something else. Best
practices, which are given as one-size-fits-all in many other methods,
only occur in Kanban, if at all, in the context of an individual company.
The idea of evolutionary improvement in Kanban is based on three
principles:

1. Start with what you are doing now.

2. Follow evolutionary improvement.
3. Promote leadership at all levels of the organization.

In discussions, I often notice that for many, the term “evolutionary”

sounds like a tactic for taking small steps. This is not necessarily the
case. Evolutionary doesn’t define how large or small the improvements
need to be. Evolutionary simply means to analyze the current situation in
order to understand what is happening in the system and adjust the
desired degree of improvement accordingly. An evolutionary
improvement can certainly involve major changes: it is possible, after
the first draft of a Kanban system, that a completely unnecessary
process step is uncovered and can be immediately eliminated. From the
Kanban view, this is an evolutionary improvement because observing the
system in its current state provided the insight for the improvement.
Nevertheless, this evolutionary improvement is very profound. The
introduction of visualization and WIP limits is no small undertaking. It is
a significant change which must be clarified with the stakeholders (i.e.
those who are directly involved or affected by the process) of the
system.

Such evolving and massive improvements need to be differentiated from

those which are nothing more than a repackaging of old habits under a
new name. Kanban is perceived to be the new method to solve all
problems. As such, it also is put on the drawing board and
conceptualized for a “Big Bang” implementation, just as it was done for
other change initiatives which have failed. Management works out a
complete Kanban system and declares everything else used till now as
unnecessary. They throw the new system in their employees’ faces and
then demand improvement with this new method. This may be done
with good intention, but this is not what is meant by evolutionary
improvement. Even well-intended changes, especially within rigid and
hierarchical organizations, seem like control and surveillance
mechanisms to the employees. Most likely, the initiators of the change
are met with passive resistance. The introduction of Kanban requires the
agreement of all those affected by its implementation, and is necessary
for the diligent preparation of a change initiative. Further details can be
found in my first book, written with Siegfried Kaltenecker, Kanban
Change Leadership: Creating a Culture of Continuous Improvement
(Leopold & Kaltenecker 2015). Designing a Kanban system with the
cooperation of all participants and their delegates is an essential element
of a successful Kanban initiative. Nonetheless, even these systems are
not meant to be forever, and are meant only for the current state of
affairs in an organization.

If the decision is made to follow evolutionary improvement, the third

principle of Kanban is unavoidable: Promote leadership at all levels of
the organization. I would like to share my perspectives on this principle:

1. As already mentioned, Kanban can only be successful when those

directly affected are included in the initiative, and not simply
dictated to. The people who are meant to work with the Kanban
system need to be brought on board. This is the point we discuss in
Kanban Change Leadership: How do you get those affected involved
in order to create broad and stable support for the start of the
Kanban initiative?
2. Evolutionary improvement only works with those people who are
directly involved with the daily activities. These people know best
what does and doesn’t work. In some companies, employees may
not feel free to speak about their observations and ideas. When
Kanban is introduced, the participants need to be encouraged to
make recommendations for improvement, and this functions best
 when there is an honest interest in their views. Otherwise Kanban
quickly falls into the old paradigm of a small group of people telling
a larger group of people what needs to be done.

Kanban is not a top-down approach in which management knows the

best way to make improvements and directs the pertinent measures into
the company. Kanban is also not a bottom-up approach. This would be
no different than a top-down approach, just in the opposite direction,
and this approach often fails because the superiors may tolerate the
initiative, but are not necessarily convinced of its usefulness. A bottom-
up approach would result in local optimization and nothing more. All
levels in the hierarchy of a company should be involved in order to
develop the Kanban system to its full potential. As you remember, the
point of using Kanban is to bring value to the customer. Preferably, the
goal is to have a consistent, flow-based system which generates value.
The system should be able to show what can be improved, for the
customer, during the process of completing a project, product or service.
There are six practices in Kanban which facilitate the accomplishment
of this goal:

1. Visualizing
2. Limiting work in progress
3. Managing flow
4. Making policies explicit
5. Implementing fast feedback loops
6. Driving improvements based on methods and models

Even if it sounds like a mantra, my experience has shown that the

following must be continually emphasized: Kanban is nothing more than
the implementation of these three principles and six practices. It is not
dictated how they should be implemented - it is not even dictated that
these principles and practices need to be implemented. The beauty, and
perhaps for many the tricky part, of Kanban is like they say at Burger
King: you can have it your way! It puts the system under a magnifying
glass and clarifies what is and isn’t in the best interest of the company’s
health. Kanban neither improves nor worsens the situation in a
company; rather it, at most, triggers a reaction to something that is
suddenly apparent. Whoever hopes for a never-changing and ready-to-
use paradigm from Kanban should perhaps look for a different solution.
In Kanban, thinking for yourself is emphatically allowed!

Let’s take a closer look at the practices that can bring a flow-based
approach to knowledge work.

Visualizing

In our flow experiment, when the unfinished ships piled up and the poor
guy at the end of the assembly line nearly broke his fingers folding
paper, it was quite obvious where the bottleneck in the system was.
Problems in the manufacture of physical products manifest themselves in
ways that are readily apparent, such as overfilled warehouses, poor
quality, or warning signals from the machines. In knowledge work, the
object of improvement is concealed in the head of the employee. The
usual scenario in knowledge work is someone sitting in front of a
computer and typing. What is not obvious, however, is what they are
working on, why they are working on it, how long have they been
working on it, how the work is going, and whether or not everything is
okay. We cannot see where, in our work process, we should apply
improvements, and hence, we cannot see if our work system is fulfilling
the expectations of the customer. By visualizing the work, relationships
in the process can be revealed allowing for better decision making.

Limiting the Work in Progress

The flow experiment vividly shows how a work system can calmly
operate when the amount of work started is limited. In the flow
experiment, we used a simple method to achieve a WIP limit: each
activity was allowed a maximum of one item at a time. This, however, is
not the only possible approach and, in many cases, a WIP limit of one
may not be feasible. What is the consequence when we work in a WIP
limited system? In the case of the Imperial Garden in Tokyo, the visitors
were forced to wait at the entrance if all of the entry tokens were in use
- the system had reached its WIP limit. A WIP limited work system
functions in much the same way: eventually, the point is reached in
which no new work can be started until work within the system is
completed. Hence, the motto of Kanban is “Stop starting, start finishing!”
The simple economic consideration is that one piece of work which is
100% completed brings more value than ten pieces of work which are
only 10% completed.

Managing Flow

WIP limits are the key to establishing and controlling workflow in a

system. Managing a workflow, however, starts with visualizing the
activities that a product (or its parts - such as features, user stories,
etc.) or service passes through when generating value. The goal of a
workflow is to allow the work items to pass smoothly from one activity
to the next. The instruments of flow management include not only
setting up WIP limits, but also learning and interpreting the indications
of a system suggested by its visualization.

It is often not readily apparent, that the workflow should even be

optimized. Many times, working resources to full-capacity, or more
specifically working the people in the system to full-capacity, is the
objective of an optimization. If we are working under the premise of
wanting to generate value for the customer, high workload is not
necessarily the most sensible strategy. When all of the employees are
always working, work will be completed more slowly. Do you like driving
on a highway filled at 100% of its capacity? Or does your computer work
best when it is running at 100% capacity? It is usually more important
to the customer to receive her product as soon as possible - regardless
of what percentage of capacity the employees of a company are
working. Running at capacity removes the opportunity for a company to
correct the weaknesses in the system and incorporate sustainable
improvement. It is exactly these improvements, however, which are
needed to generate value for the customer.

Regardless of what you are optimizing, you are sending a signal to the
people with whom you work. Giving people space to focus on their work
communicates a sense of appreciation and trust. Ordering people to
work harder and micromanaging their work gives the subliminal
message that those people are not working hard enough for their
money. In Kanban, the workload is also considered, but generating value
is the primary concern. The focus lies in optimizing the workflow and not
Another random document with
no related content on Scribd:
 “Sure, people buying things for school. Grouchy Greenway was in,
he bought a lot of homework paper—pity the fellers in the third
grade. Ruth Binney’s scared of that ladder that rolls along—oh
bimbo, that’s my middle name. I can take a running jump and ride it
all the way to the back of the store.” He did not mention that he
played the harmonica for the girls to dance; he was a good sport and
did not tell tales out of school.
“I think Ruth and Annie Terris will miss you when you go to
Montana,” said Mr. Walton playfully.
“Such nonsense,” said Mrs. Walton. “Don’t put those ideas back
into his head.”
“I may go sooner than you think,” said Hervey.
He stood in the doorway to the dining room, pausing before
making his late evening attack on the apple barrel. A blithe, carefree
figure he seemed, his eyes full of a kind of gay madness. One
rebellious lock of hair sprawled over his forehead as he suddenly
pulled off his outlandish hat in deference to his stepmother. He never
remembered to do this as a regular duty; he remembered each time
separately, and then with lightning inspiration. He could not for the
life of him adapt his manners or phraseology to his elders.
“You know me, Al,” he said.
“Are you going to wash your face when you go in the kitchen?”
Mrs. Walton inquired.
“Sure, is there any pie?” he asked.
They heard him fumbling in the kitchen, then trudging up the
stairs.
“I think it would be just as well not to harp on Montana,” said Mrs.
Walton. “It’s odd how he hit on Montana.”
“One place is as good as another,” said Mr. Walton. “I’m glad it’s
Montana, it costs so much to get there. If he had Harlem in mind, or
Coney Island, I might worry.”
“He talks of them both,” said Mrs. Walton. “Yes, but I think his
heart is in the big open spaces, where the fare is about a hundred
dollars. If it were the Fiji Islands I’d be content.”
“Do you think he’d like to go to Europe with us next summer?”
Mrs. Walton asked. “I can’t bear to leave him alone.”
 “No, I’m afraid he’d want to dive from the Rock of Gibraltar,” said
Mr. Walton. “He’ll be safe at Temple Camp.”
“He seems to have just no balance-wheel,” Mrs. Walton mused.
“When I look in his eyes it seems to me as if they saw joys, but never
consequences.”
“Sort of near-sighted in a way, eh?”
“I do wish he had stayed in the Scouts, don’t you?”
“No, I don’t,” said Mr. Walton in a matter-of-fact way. “He didn’t
see it. Some day he’ll see it, but it won’t be because anybody tells
him. The only way Hervey can learn that a tree is high is for him to
fall out of it. That’s what I mean by his being near-sighted in a way.”
“Do you think those railroad workers are a good set?”
“Oh, they’re a good lot; good, strong men.”
“Well, I don’t care for that Hinkey, do you?”
Mr. Walton did not go into raptures over anybody from New York.
He was a good New Englander. Nor had he been carried off his feet
by the “million dollar theatre.” But being a true New Englander he
was fair in judgment and of few words, especially in the field of
criticism. His answer to this last question was to resume reading his
book.
 CHAPTER XXIV
IN THE SILENT NIGHT
In his own room Hervey opened the satchel which circumstances
had caused him to carry home. He thought that since kind fate had
brought the opportunity, he would like to give one exceedingly low
blast on a real musical instrument. He was astonished to find that
there was no musical instrument in the satchel, but a tin box
containing a small account book, a number of bills with a rubber
band around them, and an envelope containing some loose change.
He contemplated this treasure aghast. Counting the bills he found
them to be in amount a trifle over a hundred dollars. Never before
had he handled so much money. He was a little afraid of it. He shook
the sealed envelope which was fat with coins; that alone seemed to
contain a fortune. He glanced at the book and found it to full of
figures, entries of receipts and expenditures. On the flyleaf was
written:
Farrelton Merry Medley Serenaders,
Horton Manners, Treasurer.
He was greatly excited by this revelation. Here was a serious
business, a very grave consequence of a mischievous act. To be
sure, the bringing home of the satchel that did not belong to him
would have been the same in any case regardless of its contents.
But just the same the sight of so much money come into his
possession in such a way, frightened him. He had not thought of
such a thing as this. You see Hervey never thought at all—ever.
But he thought now. He had “colloped” (whatever that meant) the
treasury funds of this musical organization and he felt uneasy that he
should have to be the custodian of such a princely sum over night.
Money that did not belong to him! Would his wanton act be
construed as just harmless mischief? He had always wanted to have
a hundred dollars, but now he was almost afraid to touch it. He
replaced the box in the satchel and put the satchel under his bed.
Then he pulled it out again and put it in his dresser. Then he closed
and locked the window. When he was half undressed, he took the
satchel out of his dresser and stood holding it not knowing where to
put it. Then he put it back in the dresser.
He thought of going downstairs and telling his stepfather and
getting this awful fortune off his hands. But then he would have to tell
how he had come by it. Well, was that so very bad? Tripping a fellow
up? But would any one understand? He was very angry at the
deserter Hinkey. And he was equally angry that this dextrous little
tripping stunt should bear such consequences. It seemed to him that
even poor Horton Manners had taken a mean advantage.
He resolved that he would hunt up the musical treasurer in the
morning and return the satchel to him. He would hang on to it pretty
carefully going down the street, too. He did not know Horton
Manners, but he could find him. Of course, he would have to tell the
man that he was sorry he had tripped him up. And his explanation of
why he had carried the satchel home might sound rather queer. He
was not too considerate of the tripping treasurer. He was doomed to
a sleepless night on account of that “bimbo.” It was odd, more than it
was significant, that Hervey, who was afraid of no peril, was in panic
fear of this hundred and some odd dollars. He was just afraid of it.
Several times during that long night, he arose and groped his way
to the dresser to make sure that the satchel was safe. In the wee
hours of the night he was sorry that he had not hunted up Horton
Manners immediately after his escapade. But then he might have got
home too late. On every hand he seemed confronted with the high
cost of mischief.
He wondered if the tripping treasurer was searching for the culprit
with the aid of the police. He felt sure that no one dreamed he was
the culprit. Would they, might they not already, have traced Hinkey?
And what would Hinkey say? He had a reassuring feeling that
Hinkey could not be identified as one of the culprits. He certainly
would not tell on Hinkey. And he hoped that Hinkey would not be
incriminated and tell on him before he had a chance to return the
satchel. But surely Mr. Horton Manners had not gone home and to
bed, doing nothing about the theft of more than a hundred dollars. To
the young treasurer the affair was a plain robbery. Of course, Hervey
could not sleep when his imagination pictured the whole police and
detective force of the town aroused by a bold hold-up.
In the hour just before dawn Hervey, in his troubled half-sleep,
heard a knocking sound. Trembling all over, he pulled on his shirt
and trousers, crept stealthily downstairs and with a shaking hand
and pounding heart opened the front door.
 CHAPTER XXV
LIFE, LIBERTY⸺
No one was there. Hervey looked out upon the dissolving night;
already the familiar scene was emerging in the gray drawn—the
white rail fence, the gravel walk with its bordering whitewashed
stones, the big whitewashed tub that caught the rain-water from the
roof trough. He smelled the mist. There was no one anywhere about;
no sound but the slow dripping into the tub. Drop, drop, drop; it was
from the rain of two or three days ago. How audible it was in the
stillness! He crept upstairs again and went to bed. But he did not
sleep. He wished that dreadful satchel were off his hands. Over a
hundred dollars!
He arose in the morning before the household was astir and stole
out with his guilty burden. He knew that Kipp’s Railroad Lunch was
open all night and that it had a telephone. He would look in the
telephone book for Manners. That way he would find the address.
He thought of leaving the satchel at the Manners’ door, ringing the
bell, and running away. The recovery of the money would end the
trouble. But suppose the satchel should be stolen again—not again;
but suppose it should be stolen? Of course, it had not been stolen
before.... Just the same he was desperate to get it off his hands.
Things looked strange about the station so early in the morning;
there were so few people to be seen, and no shops open. Somehow
the very atmosphere imparted a guilty feeling to Hervey. He felt a
little like a fugitive.
He could not find the name of Manners in the ’phone book and
thus baffled, he felt nervous. For while he was losing time, the victim
and the authorities were probably not wasting any time. He thought
he would wait in the station a little while and try to decide what to do.
He knew that the family of Denny Crothers, a scout, was identified
with the big white church. There was an idea! Denny would know
where Horton Manners lived, or could soon find out. Perhaps he
might even take Denny into his confidence. It is worth considering
that in his extremity he was willing not only to use, but to trust, this
scout whose troop he had repudiated.
Well, he would sit in the station a little while (it was still very early)
and if he could not think of any other plan, he would go to Denny’s
house. It would seem strange to the Crothers, seeing him there so
early. And it would seem stranger still to Denny to be approached by
an arch enemy. But Hervey’s troubled thoughts could not formulate
any better plan.
The station was not yet open and he strolled back and forth on the
platform where a very few people were waiting for the early train—a
workman wearing a reefer jacket and carrying a dinner-pail, a little
group of girls who worked in the paper mill at Brierly, and a couple of
youngish men near the end of the platform. These two were chatting
and one of them gave a quick glance at Hervey. It seemed to him
that the talk which followed had reference to himself. He wished that
the station would open, for it was a raw fall morning; there was a
penetrating chill in the air. He wanted to sit down; he was tired of
holding that dreadful satchel, yet he would not set it down for so
much as a moment.
Suddenly, a rattling old car drove up and a brisk young man in an
overcoat got out and dragged two huge oilcloth grips to the platform.
He looked as if he might be a salesman who had completed his
assault on Farrelton. He stopped and lighted a cigarette, and while
he was doing this the two men strolled over and spoke to him. He
seemed annoyed, then laughed as he took out some papers which
the two men examined. Hervey overheard the word hardware. And
he overheard one of the men say, “K.O., Buddy.” They handed back
the papers, nodded sociably, and moved away. It seemed by the
most casual impulse that they approached Hervey. But he trembled
all over.
“You’re out early, kiddo,” said one of them. “Waiting for the train?”
Why, oh why, did he flush and stammer and answer without
thinking? “No—y-yes—I guess it’s late, hey?”
 “Guess not,” said the man with a kind of leisurely pleasantry.
“What you got in the bag, kiddo?”
“Bimbo, do I have to tell you?” Hervey demanded with the air of
one whose rights are outraged.
“Might be just as well,” said the man. “What’s your name
anyway?”
“My name is Hervey Willetts and you let go of that!” Hervey
shouted, tugging at the satchel. “You let go of that, do you hear!” He
not only pulled, but he kicked. “You let go of that or you’ll get in
trouble, you big⸺”
He was the center of a little group now; it was astonishing what a
number of persons were presently on the scene considering the few
early morning stragglers. The men put a quick end to Hervey’s ill-
considered struggle by taking the satchel while one held him firmly
by the collar. There is not a decent person in the world but rebels
against this collar grip which seems the very essence of effrontery.
Few boys so held will fail to use that potent weapon, the foot, and
Hervey, squirming, administered a kick upon his captor’s shin which
made the burly fellow wince and swear.
But it was all to no avail. They opened the satchel and noted its
contents. Hervey’s sense of indignity now quite obliterated every
other feeling. His struggles subsided into a wrathful sullenness; he
could not, or he would not, explain. He knew only that he was being
held and that fact alone aroused the demon in him. Of course, if
Walton could not manage him, and the Scouts could not win and
hold him, it was hardly to be expected that these low-bred detectives
could get closer to him than to hold him by the collar. A dog would
have understood him better. He was not the kind of boy to grab by
the collar.
These two detectives, apprised of the “robbery,” had taken their
stand at the station to note if any suspicious looking strangers were
leaving town on the first train. The boy had almost escaped, because
of his youth.
And escape was the one thought in his mind now. Twice he might
have explained; first to his good stepfather, and again to these
minions of the law. But they had the grabbing instinct and (oh, the
pity of it) had diverted his thoughts from honest restitution to a
maniac desire to beat them and baffle them, to steal indeed his
liberty if nothing else, and let the satchel with its fortune go hang! He
would steal; yes, he would forget all else now in this crazy mixup! He
would steal what was the very breath of life to him—his freedom. He
forgot the whole sorry business in this dominant thought—Horton
Manners, the satchel, everything. They had grabbed him by the
collar and he could feel the tightness in his neck.
As long as the squirrel has teeth to bite, he will bite. You cannot
tame a squirrel. The fact that he is caught stealing in your tree is
quite a secondary matter. Hervey Willetts never thought of stealing
anything in his life—but just the one thing.
Freedom!
So he did a stunt. With both hands he tore open his shirt in front,
and as he felt the loosening grip in back he sprang forward only to
feel a vice-like hand catch hold of his arm. And that hand he bit with
all his vicious might and main. Like lightning he dodged both men
and was off like a deer while the circle of onlookers stood aghast.
Around the end of the freight platform he sped and those who
hurried there beheld no sign of him—only a milk-can lying on its side
which he had probably knocked over.
Off bounded one of the detectives; the other lingered, sucking the
cut in his hand. He didn’t know much about wild life, poor man. This
was a kind of stealing he had never seen before—the only kind that
interested Hervey Willetts. The only thing that interested him—
freedom. As long as the squirrel has teeth to bite, he will bite.
You cannot tame a squirrel.
 CHAPTER XXVI
OUT OF THE FRYING PAN
But they caught him, and caged him. They found him in the camp
of railroad workers near Clover Valley where he had spent a week or
so of happy days. And they left nothing undone. They investigated
the histories of that rough and ready crew, for they were after the
man higher up, the “master mind” in back of the robbery.
They unearthed the fact that one of them, Nebraska Ned, had
been a sailor and had deserted his ship to assist in a revolution in
South America. It was then that Hervey made a most momentous
decision. He abandoned Montana quite suddenly and chose South
America as the future theatre of his adventurous career.
No master mind was discovered, not even the true master mind,
Harlem Hinkey. He was not implicated and he neglected to uphold
the chivalrous honor of Harlem by coming forward as the originator
of the prank which had such a grave sequel. In the hearing in court,
Hervey never mentioned his name. And there you have Hervey
Willetts. You may take your choice between the “million dollar
theatre” and South America.
There was a pathos about the quiet resignation, the poise and
fairness in face of all, which Mr. Walton presented in that memorable
scene at the hearing. I like Mr. Walton, good man that he was. He
sat, a tall, gaunt figure, one lanky limb across the other, and listened
without any outward show of humiliation. His tired gray eyes, edged
by crow’s-foot wrinkles singularly deep, rested tolerantly on the prim
young man, Horton Manners, who was having his day in court with a
vengeance.
And Hervey, too, looked upon the young treasurer musician with
interest, with dismay indeed, for he recognized in him the very same
young man into whose lap he had stumbled on the train coming
home after his triumphal season at helpless Temple Camp. Horton
Manners looked down from his throne on the witness box, gazing
through Hervey rather than at him, and adjusted his horn spectacles
in a way that no one should do who is under fifty years old. He held
one lapel of his coat and this simple posture, so common with his
elders, gave him somehow the absurd look of an experienced
business man of about twenty-two years.
He was not in the least embarrassed. He testified that he was
treasurer of the Farrelton Band and confessed that he played a small
harp. If he had said that he played a drum nobody would have
believed him. He said that he had lived in Farrelton but a short while
and made his home with his married sister. Then, on invitation of the
likely looking young man representing the prosecutor, he told how
Hervey had mentioned on the train that he was going to Montana
and that he was going to “collop” the money to get there.
“And when did you next see him?”
“Not till this very day; in fact—here in court.”
“When he spoke of Montana, did he ask you how much it would
cost to get there?”
“He did, and I informed him that it would cost at least a hundred
dollars. I advised him against going.” There was a slight titter of the
spectators at this.
“I think that’s all, your Honor,” said the interrogator. “Since the boy
admits he took the satchel, we need not prove that.”
“Just one moment,” drawled Mr. Walton, drawing himself slowly to
his feet. He had employed no lawyer, and would not, unless his
stepson were held for trial on the serious charge of robbery.
“You say you live with your married sister?” he drawled
ruminatively.
“Mrs. Winton C. DeGraw, yes.”
“Then your name would not be in the ’phone book?”
“Presumably not.”
“Hmph.”
“I don’t see any significance in that,” said the young prosecutor.
“I simply want to find out if my boy has told me the truth,” said Mr.
Walton. “This isn’t a trial, of course. When I have satisfied myself
about certain matters I will ask the court to hear me. One more
question, Mr. Horton—I mean Mr. Manners. Do you know the
meaning of the word collop?”
“I never investigated it.”
“Well, I have investigated it,” said Mr. Walton, with the faintest
twinkle in his eye. Hervey looked rather surprisedly at his stepfather.
“It does not mean to steal. It means to earn or to get by the
performance of a foolhardy act—what boys call a stunt. Do you know
what a stunt is?”
“I suppose when I was knocked down⸺”
“You mean tripped.”
“Well, tripped. I suppose that was a stunt.”
“Exactly,” said Mr. Walton. “That’s all it was and nothing more. I
have talked with boys and I find that if a boy jumps from a high fence
to get another boy’s jack-knife, he collops it. It’s a long time since
you and I were boys, Mr. Horton Manners,” Mr. Walton added with a
smile. “Do you really want to charge this youngster with a felony?” he
continued in a tone of quiet kindness. “Isn’t the case hard enough
without that? Did you never perform a stunt?”
Oh, Hervey Willetts, if you had no thrill in that moment for the
patient, kindly, harassed man—your friend and counselor; then
indeed was there no hope for you! But he had a thrill. For the first
time in all his life his eyes filled and brimmed over as he looked at
the man who wanted only to make sure of him, to know that he was
not dishonest; who could stand for anything save that.
“I think, your Honor,” said Mr. Walton quietly, “that this affair
simmers down to a piece of mischief with an unintendedly serious
consequence. I know, of course, about the recent affair of the fire.
My boy gave himself up because he would not be despicable. He
does not lie, much less steal. I believe the story he told me; that he
thought the satchel contained a musical instrument and that he
intended to blow it and cause panic to those gathered in the church.
He saw the police officer, thought he was watched, and carried out
the part of innocence by bringing the satchel home. It proved an
elephant on his hands, a guilty burden to one really innocent. He told
me he could not find this young man’s name in the ’phone book and
it develops that the name is not there. I have here two men who saw
him looking in the ’phone book in a lunch room near the station⸺”
The judge interrupted and surprised him. “I think we need not
prolong this,” said he. “I think the boy had no intention of committing
a serious crime, or any crime at all. I believe the story he told when
arrested. I’d like to think the consequence will prove a lesson to him.
But do you think it will?”
“I’m afraid it will not,” said Mr. Walton. “And I may say now that it is
my intention to send him somewhere where he will be under rigid
discipline. I think I may be left to deal with him.”
“Well, the charge of robbery is dismissed,” said the judge. Then he
appeared to ruminate. “But the boy is still with us and there’s the
problem. This is the second time he has been brought into court. He
kicked up quite a rumpus and bit an officer. Where is this kind of
thing going to end?” He seemed kindly and spoke rather sociably
and not as an official. “Why don’t you put him in the Boy Scouts?” he
added.
“The Boy Scouts haven’t given him a knockout blow yet,” smiled
Mr. Walton. “I’m always hoping they’ll reach him. But I suppose
they’ll have to do a stunt that pleases him. Meanwhile, I’m going to
send him to a military school. It seems like a confession of defeat,
but I’m afraid it’s the only thing to do.”
The judge turned to Hervey. “You’d better go home with your
father,” said he. “And you take my advice and get into the Boy
Scouts while there is time, or the first thing you know you’ll land in a
reformatory. So you want to go to Montana, eh?”
“Sure, they have train robbers out there?” said Hervey.
“And how do you like having a hundred dollars that doesn’t belong
to you?”
“Nix on that stuff,” Hervey said gayly.
“Yet you like train robbers.”
“Bimbo, that’s different.”
Mr. Horton Manners, still sitting like an owl on the witness stand,
gazed at Hervey with a look of utter bewilderment.
“But in South America they have rebellions,” said Hervey.
“Well, let us have no more rebellion here,” smiled the judge.
And he winked at Mr. Walton.
 CHAPTER XXVII
AT LAST
Of course, Hervey was never in any danger of being sent to prison
for robbery. As soon as he was arrested and made to tell his story,
Mr. Walton annoyed, but unruffled, saw the thing in its true light. He
went to the all night lunch room near the station and made sure that
Hervey had gone there; then he verified the boy’s statement that the
name of Manners was not in the ’phone book.
Quietly he even inquired among boys the meaning of collop. And
he learned on the highest juvenile authority that it did not signify
stealing nor an intent to steal. But Horton Manners had made the
charge of robbery and so the whole business had to be aired in
court. Mr. Walton was a man of few words; it would be interesting to
know what he really thought of Horton Manners.
As for Hervey, he quite forgot the affair within an hour of the time it
was over. He had been appalled to find himself the custodian of a
hundred and more dollars, but now that he had got it off his hands,
he went upon his way rejoicing. He never looked either backward or
forward; the present was good enough for him. It is significant that
he bore no malice toward Horton Manners. Once or twice he referred
to him as Arabella; then he forgot all about him. He could not be
bothered hating anybody; nor caring a great deal about anybody
either.
A few prominent townspeople financed the Firemen’s Carnival
and it was held after all. Shows and acts were engaged, the merry-
go-round revolved to the accompaniment of its outlandish music, the
peanut and lemonade men held form; you could see the five-legged
calf for “a dime ten cents,” and Biddle’s field presented a gala scene.
The boys of Farrelton went round and round trying to stab the brass
ring, they drank red lemonade and time after time gazed spellbound
at the five-legged calf.
Hervey did not care about seeing the five-legged calf unless he
could sneak in under the canvas fence, and he could not manage
that because of the man who kept shouting and slapping the canvas
with his stick. In common with all the other boys he was thrilled at the
sight of Diving Denniver who ascended a ladder to a dizzy height
and dived from it into a small tank directly below. Diving Denniver did
this thing twice a day, and his night performance was the more
thrilling because it was in the glare of a searchlight whose long beam
followed him in his slow ascent of the frail looking ladder and showed
him in a circle of light when he paused for one thrilling moment at the
top. He earned his living in this way, going around exhibiting at
carnivals and amusement parks, and he was the big feature of the
Farrelton carnival.
Hervey was not content simply to behold this daredevil exploit. He
saw it twice in the daytime and once at night, and he could not stand
the strain of being restricted to the enjoyment afforded a gaping
audience. That is where he differed from other boys. It was this
something in his nature that prevented him from reading boys’
books; he could not intrude into the hair-raising adventures and so
he had no use for them. The most thrilling stories were utterly dead
stuff to Hervey.
But here he could intrude. It was after he saw the night
performance that he felt the urge to penetrate to the hallowed spot
whence that enchanted daredevil emerged in his theatrically
cautious ascent of the ladder. The nature of the spectacular feat
required that it be performed at a distance from the body of the
carnival. As soon as the band started playing Up in the air mid the
stars, the long column of light was directed on the ladder which
appeared as if by magic a hundred yards or so from the thronged
area of the carnival. Every eye was then fixed with expectancy as a
white figure arose into view, moving up, up, up, to a little
surmounting platform. Then the sensational dive, after which the
pleasure seekers ate, drank and were merry again.
But Hervey could not go back to any merry-go-round after that,
and red lemonade had no solace for him. He wandered off from
those festoons of electric lights, away from the festive groups, into
the darkness. Before him, down near the edge of Biddle’s field, was
a tiny light. Soon he came to a rope fence which cut off the end of
the field from the public. Beyond this were wagons and huge cases
standing in the darkness, the packing and transporting paraphernalia
of the motley shows. In a monstrous truck that stood there the multi-
colored prancing horses of the merry-go-round would be loaded and
have a ride themselves.
On an upright of this rope fence was a sign which read
POSITIVELY NO ADMITTANCE. Hervey entered just where the sign
was placed. A hundred or so paces brought him to the holy of holies,
a little tent at the foot of the towering, slender ladder. In the darkness
its wire braces, extending away on each side to their anchorages in
the earth, could not be seen. Almost at the foot of the ladder was a
tank perhaps fifteen or eighteen feet square. Close by the tent was a
Ford sedan, and Hervey crept reverently up to it and read the words
on the spare tire cover DIVING DENNIVER. On the lower part of the
circumference was printed THREE HUNDRED FOOT DIVE. Diving
Denniver believed in advertising. In that tent lived the enchanted
mortal.
Hervey lingered in awe as a pilgrim might linger at a shrine before
entering. Then he walked rather hesitatingly to the open flap of the
tent. On a mattress which lay atop a huge red chest reclined Diving
Denniver in a bath robe. The chest had DIVING DENNIVER printed
on it, as also did a large leather grip, which bore the additional
information WONDER OF TWO CONTINENTS. If the world could
not see Diving Denniver on his dizzy perch, it at least could read
about him. Besides the makeshift divan the tent contained a rough
table formed by a red board laid on two saw horses.
On this was a greasy oil-stove and one or two plates and cups. In
his illicit wanderings, Hervey had at last trespassed through the
golden gates into heaven.
“I was walking around,” said he, rather unconvincingly.
Diving Denniver, a slim young man of about thirty, was smoking a
cigarette and looking over a magazine. It seemed incredible that he
should be thus engaged so soon after his spectacular descent.
 “Bimbo, that was some pippin of a dive,” said Hervey. Then, as
Diving Denniver made no attempt to kill him, he ventured to add, “Oh
bambino, that’s one thing I’m crazy about—diving.”
“Didn’t the cop see you?” the marvel asked.
“Leave it to me,” said Hervey. “There isn’t any cop there anyway.
Cops, that’s one thing I have no use for—nix.”
“Yere?” queried Diving Denniver, aroused to slight amusement.
“Do you—do you feel funny?” Hervey ventured as he gazed upon
the wonder of two continents.
“Where did yer git that hat?” asked the god of the temple. “What’s
all them buttons you got on it?”
“I climbed way down a cellar shaft to get one of those buttons,”
said Hervey, anxious to establish a common ground of professional
sympathy with this celebrity. “That’s the one,” he indicated, as he
handed Denniver his hat; “the one that says VOTE FOR TINNEY. He
didn’t get elected and I’m glad, because his chauffeur’s a big fool; he
chased me, but he couldn’t catch me. Some of those holes I cut out
with a real cartridge shell, like you cut cookies. I bet you feel funny,
hey?”
“Yere?” said Diving Denniver, examining the hat. “Well, do you
think yer could go back up there where the big noise is and then
come back here again—without gettin’ stopped?”
“You mean you dare me to?”
Diving Denniver roused himself sufficiently to reach over to a box
and grope in the pocket of a pair of ordinary trousers, the kind that
mortals wear. Then he tossed a quarter to Hervey. “Chase yourself
back there and get a frankfurter,” he said; “get a couple of ’em. And
don’t leave the cop see yer.”
So the wonder of two continents ate frankfurters—and scorned
cops. More than that, he and Hervey were going to eat a couple of
frankfurters together. At last Hervey felt that he had not lived in vain.
 CHAPTER XXVIII
THE LAW AGAIN
Hervey felt that he and Diving Denniver were pretty much alike
after all. The wonder of two continents beat all the boy scouts put
together. And he had now a fine precedent for his repudiation of
authority. Diving Denniver cared naught for cops and signs. Hervey
would have been glad to go into any court and cite this high
authority, confounding the powers with this frankfurter episode. He
was sorry he had not told Diving Denniver of his swimming across
the lake at Temple Camp (during rest period which was against the
rules). Instead of an honor he had received a reprimand for that. He
was a little afraid that some of the other boys would visit the wonder
in his tent, but in fact there wasn’t much danger of that. The wonder
was too much off the beaten track for most boys. Their thoughts did
not carry behind the scenes.
Hervey was now in much perplexity whether to witness the thrilling
exploit from the audience the next night or to view it from the
sanctum of the hero. In either case he intended to visit the remote
scene of enchantment with two frankfurters. He decided that he
would not demean himself by gazing at his hero with the idle throng.
He even negotiated an extra hour out from Mr. Walton in anticipation
of his second visit to the hermit of the ladder.
He could not possibly reach the place in the daytime, and besides,
he had to take up some bulbs for his stepmother the next day. For
this and other services he was to receive fifty cents. Twenty-five of
this would pay his admission to the carnival. With the other twenty-
five he intended to furnish forth a banquet of frankfurters for his hero
and brother daredevil. He could not afford to go twice in the day. He
had some thought of effecting an entrance over the high fence into
the field and having his entire fifty cents for the post-exploit feast. But
reckless as he was, he was cautious in this matter of reaching the
tent—there was so much at stake! So he decided to go respectably
in through the entrance and then cross the rope fence where the
“Positively No Admittance” sign was placed. It was not often that he
showed such a conservative spirit.
At half past eight, he found Diving Denniver strolling around in his
bathrobe outside the tent. Within, the odor of fried bacon and coffee
still lingered.
“You back again?”
“Sure, I want to see you from right here, and afterward I’m going
to go and get some more frankfurters. After you’re finished will you
let me go about ten or fifteen steps up the ladder and try it?”
Diving Denniver did not trouble himself to answer, but he ruffled
Hervey’s hair good-humoredly as he ambled about smoking his
cigarette. “Much of a crowd over there?” he asked.
“Oh bimbo, they’re all waiting. They stop dancing even when you
go up,” Hervey said.
“You’re a pretty slippery kid, all right, ain’t yer?” Denniver said.
“Ain’t there no guy up there at the rope?”
The words were scarcely out of his mouth when both he and
Hervey became aware of a policeman who had just come around the
side of the tent. But Hervey, though astonished, was not perturbed,
for he believed that the wonder of two continents would protect him.
One word from Diving Denniver and he would be safe. He even
ventured a defense himself.
“I’m going to do an errand for him,” he said.
“You can ask him yourself. So I’ve got a right to be here.”
But it appeared that it was Diving Denniver with whom the officer
had business. “Are you Charles McDennison?” he asked.
“Yere, what’s the dope?” the wonder asked, with a kind of
weariness in his voice.
Hervey was astonished, not to say shocked, that Diving Denniver
acknowledged the name of Charles McDennison.
“Let’s look at your permit,” said the officer.
Mr. McDennison entered the tent, presently emerging with a
paper.