Plaintext
Copyright © 2017, agile42 International GmbH
agile42 International GmbH
Grünberger Str. 54, 10245 Berlin, Germany
Tel: +49 30 200 539 58
Web: www.agile42.com
Email: info@agile42.com
Edition: first edition, 2017-07-31
ISBN 978-3-96243-000-9 (printed, A5)
ISBN 978-3-96243-005-4 (printed, 5.5” x 8.5”)
ISBN 978-3-96243-001-6 (ePub)
ISBN 978-3-96243-002-3 (Kindle)
ISBN 978-3-96243-003-0 (online PDF)
ISBN 978-3-96243-004-7 (promotional excerpt, 5.5” x 8.5”)
This book is distributed under the Creative Commons Attribution, Non-Commercial, No
Derivatives 4.0 license (CC BY-NC-ND 4.0) available at https://creativecommons.org/
licenses/by-nc-nd/4.0/.
The license allows you to download, copy and share this material on the condition that
you (a) give appropriate credits, (b) don’t change the material in any way, and (c) don’t
use it commercially. This summary is included for convenience only and does not re-
place the full license.
Contents
1 Introduction 1
2 What is agile coaching? 5
2.1 Other kinds of coaching . . . . . . . . . . . . . . . . . . . . 7
2.2 The agile coaching stances . . . . . . . . . . . . . . . . . . . 9
2.3 When not to coach . . . . . . . . . . . . . . . . . . . . . . . . 16
2.4 Try this . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3 Coaching Fundamentals 19
3.1 Code of conduct for a systemic coach . . . . . . . . . . . . 19
3.2 The mindset behind systemic coaching . . . . . . . . . . . 21
3.3 Curiosity and keywords . . . . . . . . . . . . . . . . . . . . . 22
3.4 Listening and responding . . . . . . . . . . . . . . . . . . . 24
3.5 Structure for a coaching conversation . . . . . . . . . . . . 34
3.6 Powerful questions . . . . . . . . . . . . . . . . . . . . . . . 42
4 Coaching conversations with a team 51
4.1 Bridging questions . . . . . . . . . . . . . . . . . . . . . . . . 51
4.2 The 5D model . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.3 Try this . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
5 Team dynamics 63
5.1 Synergy in teams . . . . . . . . . . . . . . . . . . . . . . . . . 64
5.2 Team development and performance models . . . . . . . . 68
5.3 Challenges in forming teams . . . . . . . . . . . . . . . . . . 74
5.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.5 Try this . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
6 Coaching your team 81
Contents i
6.1 What is built into your framework? . . . . . . . . . . . . . . 82
6.2 Daily standup . . . . . . . . . . . . . . . . . . . . . . . . . . 85
6.3 Refinement and planning . . . . . . . . . . . . . . . . . . . 87
6.4 Retrospectives . . . . . . . . . . . . . . . . . . . . . . . . . . 87
7 Structured coaching 95
7.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
7.2 Make observations . . . . . . . . . . . . . . . . . . . . . . . 97
7.3 Formulate a hypothesis . . . . . . . . . . . . . . . . . . . . . 100
7.4 Identify a goal . . . . . . . . . . . . . . . . . . . . . . . . . . 101
7.5 Define metrics . . . . . . . . . . . . . . . . . . . . . . . . . . 101
7.6 Choose the coaching tools . . . . . . . . . . . . . . . . . . . 103
7.7 Build a coaching structure . . . . . . . . . . . . . . . . . . . 104
7.8 Follow up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
7.9 Case studies for TCF . . . . . . . . . . . . . . . . . . . . . . . 106
8 Challenges 115
8.1 The lizard brain . . . . . . . . . . . . . . . . . . . . . . . . . 115
8.2 SCARF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
8.3 Influencing behavior . . . . . . . . . . . . . . . . . . . . . . 125
8.4 Building your case on metrics . . . . . . . . . . . . . . . . . 126
8.5 Safe-to-fail experiments . . . . . . . . . . . . . . . . . . . . 127
8.6 Try this . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
References 131
ii The Hitchhiker’s Guide to Agile Coaching
Chapter 1
Introduction
Agile coaching is at heart not such a complicated thing. It is the art of
observing, listening, forming an understanding and validating it, then
working with the coachee to come up with and enact solutions. By
coaching and observing others coach, you build up your own collection
of patterns and tools. Where a junior coach will have to rely on written
material and the support of others, an experienced coach can quickly
form and validate hypotheses and pick the right tools from the toolbox.
But starting out as a new agile coach is not so easy. Where do you go?
How do you get started? With this book we wanted to create an introduc-
tory reference for new agile coaches as well as experienced ScrumMas-
ters. It contains much of the theory that we teach our new colleagues
at agile42 and hopefully a lot of insights for new coaches. The intention
was to provide a secondary source for what we teach as a lasting mem-
ory to hand out to clients and students. We decided to publish it so that
even more people can enjoy the results.
What you are reading right now is the book we wish we had had when
we started out as agile coaches.
Despite the title this book is not about agility as such. We will assume
that the reader has a working knowledge of all things agile. We may
touch on the agile principles and lean thinking in this book, but we are
not going to explain e.g. why it makes sense to slice large work items into
smaller independent pieces, or how to choose between synchronized
and unsynchronized releases. There are plenty of good books around
for those purposes.
Introduction
We have tried to introduce the topics sequentially, so that later chapters
build on top of previous ones. The book therefore wanders between dif-
ferent topics, although there is a clear path from listening to one-on-one
coaching and onward to team coaching and how to empirically influ-
ence the behavior of a team. We do recommend that you at least skim
the preceding chapters before jumping into later chapters.
Throughout the book we will use the term agile coach, but don’t take
that too literally. It’s an umbrella term that covers everyone interested
in coaching agile teams and small organizations. We don’t expect that
every reader wants to become an agile coach. Rather we have designed
and written this book to be useful to a large number of people including
ScrumMasters, development managers and system thinkers. So when-
ever you see “agile coach”, please substitute it with whatever title you
feel is appropriate for you.
That said, some parts of this book are written for people who see a lot of
different teams, either inside one organization or across many compa-
nies. ScrumMasters, development managers and internal agile coaches
typically work with the same teams year after year. Agile coaches, on the
other hand, are expected to train and coach teams from scratch until
they become self-sufficient in Scrum, which usually takes 2–3 months,
before moving on to other teams. Over time a consulting agile coach will
find herself working with many different teams in many different con-
texts and with many different goals. We have therefore included some
material relevant to working with new teams. There are chapters and
sections in this book dedicated to team building models, or what you
can do to help your teams perform better.
At this point you may be wondering who these guys behind the book
really are. We are agile coaches working at what we believe is the most
awesome agile coaching company, agile42. Since this book was written
by a team, we have chosen not to stand out individually. Instead, we
have been writing this book as if it were software developed by a team.
The Hitchhiker’s Guide to Agile Coaching
This means that we started by formulating a strategic goal for why this
book should be brought to life. We drew up a product vision from the
user perspective; we named and described personas and created a story
map. By planning tasks and visualizing them on a kanban board, we
wrote the book through an iterative and incremental process. We fo-
cused on first developing a Minimal Viable Product (MVP) through a
number of Minimal Marketable Features (MMF). What you see here is
the MVP. For the software geeks among you, the book was entirely writ-
ten using markdown files in a git repository with a continuous integra-
tion build using pandoc to create various formats of the book.
Becoming an agile coach is a journey that has no end. And that’s why
the name of this book is “The Hitchhiker’s Guide to Agile Coaching.”
Introduction
Chapter 2
What is agile coaching?
Agile coaching is the art of helping people see reality using an agile and
lean perspective and change their paradigms, habits and roles accord-
ingly. Agile coaching has two very different sides — agility and coaching
— and you will need to know a bit of both. Being good or skilled in only
one of these two domains is not sufficient.
Agile is an umbrella term for all kinds of methods and practices that are
based on the values and principles defined in the Agile Manifesto (“Agile
Manifesto,” 2001). At present the most widely used agile methods are
Scrum, Kanban and XP. When we refer to “agile”, we usually mean any
and all of the agile methods, but we will sometimes specifically address
one of the agile methods.
In our experience, most agile coaches typically have a background as
ScrumMasters, project managers, general managers, process specialists
or quality managers. Working with different teams, they have noticed
that they can help the teams work better by introducing or strength-
ening the team’s agile thinking. At some point, they may start calling
themselves agile coaches, or they keep their ScrumMaster title. Through
failures and successes, they have slowly learned how to approach and
work with different teams. They have started out as new agile coaches
with reasonably good agile knowledge, but do not know enough about
coaching. If you belong to this group, this book is for you.
Agile thinking and agile practices are easier to learn than coaching skills.
There are agile books, Scrum training, workshops, conferences, com-
munities and a lot more available, however for many people the main
problem is, in fact, the unlearning of old thinking patterns.
What is agile coaching?
Coaching is also a skill that you can pick up and improve on, although
perhaps not as easily as agile knowledge. This is because coaching is a
way of acting, almost a personal habit, and habits are notoriously diffi-
cult to change.
In order to become a good coach you must learn to observe and lis-
ten, to keep your own biases in control, to not jump to premature con-
clusions and to communicate well. This naturally affects how you are
aware of yourself, how you react to other people and how others per-
ceive you. Many people are unwilling to even start this journey. For ex-
ample, people who enjoy the feeling of power that comes from bossing
people around, do not necessarily want to give it up.
Agile coaching is not magical handwaving. It is a soft but very very struc-
tured skill. We don’t want to encourage the perception that an agile
coach can just walk into an organization, talk to people, run some work-
shops, slap stickies on flipcharts and turn everyone super-agile. Instead,
we see agile coaching as structured and long-term work that stretches
over months, sometimes years.
The difference between a newly started agile coach and an expert coach
is simply that the expert knows more structures by heart and remembers
more tools; and has more experience to help her choose the right tool or
structure for the job. Where an experienced coach can pick and choose
from memory, new coaches may need to consult lists, books or mentors
and are perhaps more likely to occasionally choose the wrong tool for
the job. We hope that this book will give you the tools and structures
you need to become an effective agile coach.
We would encourage you to think of coaching as an additional set of
tools in your leadership toolbox. Coaching is something you can try out
and adopt gradually, tool by tool and technique by technique; and it is
very likely that you will find the skills and tools in this book useful in
many different situations. For example, the skill of listening will be very
The Hitchhiker’s Guide to Agile Coaching
convenient when interacting with customers and colleagues, as well as
your mother-in-law.
. Other kinds of coaching
In this chapter we would like to position agile coaching by comparing
it to two other well-known coaching domains: systemic coaching and
sports coaching.
If you haven’t heard about systemic coaching before, you can think of
it as the classical therapy or counseling approach. Imagine a patient
sitting in a comfortable lounger and a calm, neutrally dressed counselor
in a hardback chair with a notepad in her hand, asking “And how can I
help you today?”
A systemic coach works with the system. The information, interpre-
tations, goals and actions all come from the coachee and the systemic
coach merely facilitates discovery through discussion. She asks neutral,
open-ended questions related to keywords that pop up in the discus-
sion. For example “Tell me more about your time there?” or “How did
the others react to this piece of news?”
Agile coaching borrows a lot from systemic coaching. However, where
the systemic coach is non-directing, the agile coach usually has an
agenda of making the team more agile. Both systemic coaches and
agile coaches usually have a sponsor and have been hired to achieve a
goal. The difference is that systemic coaches try not to impose their
own thoughts and ideas in the discussions. For an agile coach, part of
the goal is to act as a role model, infusing the organization with agile
thinking and practices.
What is agile coaching?
Sports Coach Agile Coach Systemic Coach
Sets goals and targets Supports the team, Accepts any result
for the team provides
benchmarks
Defines the team May discuss roles Accepts the team as
with manager it is
Lays off Helps team members Helps team members
non-performing find ways to improve find ways to improve
team members
Tells the team what Outside the selected Helps the team
to do in a given method, helps the explore scenarios
situation team understand the and choose which
approach in a given one to pursue
situation
Has a somewhat Uses lean and agile Has a circular
linear understanding thinking to understanding of
of cause and effect understand cause cause and effect
and effects
Very transparent Moderate display of Aspires to be neutral
emotions emotions
An agile coach also works in quite similar ways to a sports coach.
Whereas professional athletes know a lot about their sports and their
main tool (the human body), many professional software developers
are lacking in theoretical knowledge as well as practical experience.
The barriers of entry into software development are very low and many
software developers never receive sufficient formal education in their
craft. At the same time, the pressure to release new features means that
best practices are forgotten and people collect the wrong kinds of
experience.
Exacerbating the issue, everyone wants to portray themselves as experts
The Hitchhiker’s Guide to Agile Coaching
in order to keep their jobs. The more senior developers who could per-
haps act as mentors and role models are all overworked, participating in
dozens of critical projects and workgroups simultaneously. This means
that agile coaches need to tread carefully when nurturing both basic
software practices, as well as agile project/product management and
team building practices.
Every coach should always have a clear goal. Unfortunately, agile
coaches are often called in to do something vague or implicit, such as
“we’ve been told to deploy Scrum” or “my team isn’t meeting their
sprint goals” or “could you just look at this team and say what’s
wrong?” Part of your job is therefore to observe or quickly assess the
organization, make a diagnosis or draw up some hypotheses and agree
on the goal with your sponsor. Without an understanding of what the
organization wants and needs, it is difficult to achieve results.
To do this, you will need to allocate “access time” with the organization.
In some cases, this might mean a concerted, structured assessment car-
ried out in a very short time. Other times, it may be more informal and
unstructured — to the extreme point of hanging around in the coffee
room trying to meet and interview as many different people as possible.
. The agile coaching stances
Interestingly enough, an agile coach needs to be able to do many differ-
ent things that are not coaching. One of the basic skills of agile coaching
is to understand when to coach and when not to coach and to know
what else you might do instead.
For example, you may notice that a team doesn’t understand why they
should split work into small independent tasks. If you were only allowed
to coach, you would need to ask the right questions in sequence, pa-
tiently guiding them forward step by step until they connect the dots,
understand the implications and come up with the right methods. This
What is agile coaching?
would require almost inhuman leaps of logic and induction from the
team, as well as almost inhuman coaching skills from you. It would be
more effective to just step out of the coaching role and spend a little mo-
ment as a teacher.
Coaching and teaching are two different “modes of operation” that we
call coaching stances. There are five stances in total (see fig. 2.1) and we
will explore them in this section. You should always be aware of which
coaching stance you are taking at any given moment. You will change
stances regularly as the situation dictates and you will need to take a
different approach in each stance.
Coaching — For the most part, you should operate as a coach and stay
in the base stance of coaching. This is where you listen and ob-
serve how the team works; challenge and question their assump-
tions and status quo; and facilitate and lead the activities.
The coaching stance is very close to systemic coaching as we de-
scribed in the beginning of this chapter (in sec. 2.1). It is the most
“neutral”, in that the coach tries to be unbiased and not push her
own ideas on the team. In order to avoid biasing the team that you
are coaching, you should try to step back into the coaching stance
as soon as the situation allows.
Teaching — The teaching stance that we mentioned in the beginning
of this section is one that you will find yourself using quite a lot.
In contrast to systemic coaches and to some extent also sports
coaches, agile coaches need to be able to educate and teach peo-
ple.
In this context, “teaching” means the ability to detect when the
group you are coaching is lacking knowledge on some topic,
explain the concepts in a clear way, answer questions related to
The Hitchhiker’s Guide to Agile Coaching
Figure 2.1: The coaching stances
those concepts, and verify that the group has indeed acquired the
new knowledge. It does not necessarily mean planning and
carrying out engaging, informative multi-day courses. Although
many agile coaches also work as agile trainers and vice versa, the
skillset required for giving multi-day courses is quite different.
As mentioned, you should find your way back from the teaching
stance to the coaching stance as soon as possible. Asking ques-
tions like “How do you think this new knowledge will affect your
future actions?” or “With the knowledge I just provided you, what
will in your perspective be the best thing to do now?” can help
you return to coaching. It also signals to the coachee that you are
giving the responsibility back to her and she needs to make wise
decisions with the new knowledge in mind.
What is agile coaching?
Mentoring — This is the process of transferring learnings from the
mentor to the mentored person, the mentee. The mentoring
stance is useful when the mentee has knowledge to some extent,
but lacks experience in the topic. The transfer of experience often
happens through discussions, stories, anecdotes and advice. It
can also happen through hands-on demonstrations or through
collaboration and pair work, in which case one could talk about
apprenticeship.1
Mentoring is often confused with coaching, possibly because a
good mentor must also be a good listener and a good coach. Many
mentoring programs contain aspects of career or life coaching.
Similarly, coaches must sometimes act as mentors. An agile coach
would most commonly mentor a junior agile coach or a Scrum-
Master.
Your mentee is likely to find it convenient to have a mentor and
she will encourage you to stay in that stance. Again, remember to
return to the coaching role as soon as possible. For that, you can
state something like “That was the story I wanted to tell you. What
kind of thoughts did it give you?” or “Now that we have done this
together, let me see you do it on your own.”
“I refuse to answer that question on the
grounds that I don’t know the answer.”
— Douglas Adams
Advising — Being an agile coach also includes the capability of giving
advice to a client when necessary. This includes giving options,
creating insights and evaluating their experiences. If you have
opinions on a topic, you will also need to justify and motivate.
1 See http://en.wikipedia.org/wiki/Mentorship and http://en.wikipedia.org/wiki/
Apprenticeship
The Hitchhiker’s Guide to Agile Coaching
You should be very careful when giving advice, as coaching and
advising are conflicting activities. As a coach you are supposed
to be neutral and unbiased, but as an adviser you are expected to
provide opinions and suggestions. You cannot do both at once.
This is one of the reasons why you are most likely better off adopt-
ing one of the other stances rather than advising.
Another reason is that people who ask for advice are not always
looking for an honest answer. It happens that people just want to
vent their frustration or would like to know whether they can use
you for support when driving their own agenda.
The third reason is that when you give advice or
recommendations, you must also take responsibility for that
advice. Sometimes you should not have any say in the matter or
you may have conflicting interests. Sometimes a simple question
requires a complex answer (or several complex answers) and
sometimes the client takes your answer and implements it
without understanding the implications.
We commonly find ourselves answering “it depends” and then
perhaps explaining the relevant theory (training), helping them
find their own answer (coaching) or providing examples that are
hopefully illuminating (mentoring). Only occasionally do we
actually give straight-out answers (advising) and it almost never
happens that we carry out the work for them (consulting).
Even when giving advice, you should always leave the responsi-
bility to the coachee. One trick to make that happen is to form
your suggestions as hypotheses: “Could it be a possibility for you
to. . . ?”, or provide several options: “Another approach that could
be worth exploring is . . . ” This sends a signal to the coachee that
it is her call to come up with the solution, since she is the one that
knows the context best. It also indicates that you are now stepping
back to the coaching stance. When you push your solution onto
What is agile coaching?
someone else, there is a risk that the other person will not be fully
committed to achieve the wanted result.
Role modeling — As agile coaches we constantly need to demonstrate
integrity. Some of the key pillars in your career as an agile coach
are transparency, honesty, commitment to help people grow and
maximizing the benefits for your clients. Role modeling is primar-
ily about living the agile and lean values.
An agile coach cannot simply say one thing and then do some-
thing else. For example, stating that meetings always start on time
and then arriving late yourself, or asking the ScrumMaster to be
prepared but showing up unprepared for your own engagements.
Role modeling is different from contracting or consulting. A con-
tractor or consultant is hired to do a job on behalf of the client.
The difference is that role modeling is a form of on-the-job train-
ing for the organization’s own ScrumMasters. The role model also
only assumes responsibility on a limited level and for a limited
time e.g. ensuring that a retrospective is well facilitated or that the
team is paying attention to the resulting actions.
As a role model you could for example facilitate a couple of
sprints, gradually handing over to the organization’s own
ScrumMasters which falls nicely within the scope of almost any
coaching engagement. This is in sharp contrast to sports coaches
who typically do not go out on the field to score goals during a
match, and to systemic coaches who can’t change themselves on
behalf of their clients.
If your coaching engagement explicitly includes contracting
or consulting, we would talk about embedded coaching. An
embedded agile coach is expected to work full time as a team
member or ScrumMaster for a period of months or years and
simultaneously roll out agile practices or promote agile thinking
in the team and surrounding organization. While this may seem
The Hitchhiker’s Guide to Agile Coaching
like a great idea — two for the price of one, so to speak — we try
to avoid this mode of engagement as it has several pitfalls for
both the coach and the client organization.
As the coach is now inside the organization, some of her credibil-
ity and leverage will be lost within months and managers will start
bossing her around. Much of the coaching effort is spent on run-
ning the mechanics of the chosen agile method and the coaching
progress is slower than it could be. The organization also easily
becomes dependent on the coach, so that progress stagnates after
the coach eventually leaves the building.
Embedded coaching is also problematic for the coach herself.
Working with two different goals seldom leads to satisfactory
results, especially if the goals are conflicting. There is only so
much time. . . should I run a retrospective or write more
software? Furthermore, personal professional development can
come to a standstill. Compared to a dedicated agile coach, an
embedded coach has twice the number of professions to learn
about but even less time. The experience she collects will be
limited to this particular client and she will lose out on the wide,
generalized knowledge that comes from working with many
diverse clients.
As an embedded coach, it is important to make it explicit when
you are stepping from one role to another. Do not move con-
stantly back and forth or linger in between, because that will only
confuse people around you. You will also need to find somebody
to train as your replacement.
As an agile coach you should always be careful not to involve yourself
in content and product decisions when advising and role modeling. It
may be tempting to propose new and nifty features or nudge the user
interface just so. Remember that you have been engaged to improve the
organization and not the product.
What is agile coaching?
When you position yourself inside the system you are making it diffi-
cult to stay objective and unbiased. And unless you happen to be a
recognized domain expert, it is very likely that the organization you are
coaching understands their customers better than you do.
. When not to coach
“It can be very dangerous to see things
from somebody else’s point of view
without the proper training.”
— Douglas Adams
Coaching requires a certain objectivity and it is difficult to coach people
you deeply care for. Avoid coaching your family, friends or close busi-
ness partners unless they really ask for it. Even then you might want to
make clear that it’s a one-off thing. Get out of your home or office, go
to a coffee shop or take a walk while you discuss. And when the coach-
ing session ends, remember to step out of the coaching mode and be
yourself.
Sometimes you will meet individuals who mistake coaching for therapy
or are messed up beyond the help of a coach. Coaching is not therapy. In
these cases, we recommend that you gently back off and stay in coach-
ing territory. This is primarily because amateur therapists can do se-
rious damage, but also giving therapy is probably not what you were
engaged for.
Now, of course, there are agile coaches who are also trained and licensed
therapists. There might be situations where somebody engages such a
trained and licensed therapist in order to give therapy to individuals or
teams. If this is the situation, we are quite sure that you know it and can
handle it. In all other cases, we recommend that you step back and let
professionals with the right skills handle the issue.
The Hitchhiker’s Guide to Agile Coaching
How do you recognize that you are coaching a person who needs ther-
apy more than coaching? The best advice we have heard is from a pilot
in the context of landing an airplane: If you are in doubt, then you are not
in doubt! Meaning that if you suspect that you are on the wrong path,
then act as if you are on the wrong path.
. Try this
1) When working with a team, regularly remind yourself of your
coaching stance. Try to return to the base stance, coaching, as
soon as possible.
2) The next time someone is asking for your advice, try a different
stance than straight out advising. Teach the relevant underlying
theory (in a nutshell). Coach them to bring out the pros and cons
of different options. Mentor them by giving examples from other
teams or organizations you have worked with.
What is agile coaching?
Chapter 3
Coaching Fundamentals
“Attention is the currency of leadership.”
— Ronald Heifetz
In sec. 2.1, we discussed how agile coaching compares to and contrasts
with systemic coaching. In this chapter we will look more into the fun-
damentals of using systemic coaching approaches in your job as an agile
coach.
We will focus on how to coach individuals as a systemic coach.
Coaching teams is much easier once you know how to coach
individuals. You do not need to be an expert personal coach though —
even a little knowledge of personal coaching can help you a long way
when coaching teams. We will cover techniques for listening and
formulating questions and talk about how to structure discussions.
. Code of conduct for a systemic
coach
Let us first have a look at what a coaching relationship is. A coaching
relationship is primarily a powerless environment (Stelter et al., 2005).
This means the coachee and the coach are to be equal partners in the
conversations. You coach eye to eye — not from above and not from be-
low. The coachee should not have to be concerned about consequences
for her job or career from what thoughts and ideas she is sharing with
the coach. If she has, she will be holding back which will decrease the
effect of the coaching.
Coaching Fundamentals
Figure 3.1: The fundamentals of coaching
Secondly, a coaching relationship is a room with confidentiality. The
coachee should feel safe at all points in time, both during and after the
conversation. This means that you do not share details from the conver-
sation with others — at least, not without permission from the coachee.
In popular terms, one could say that the The Vegas Rule applies here:
What happens in the conversation, stays in the conversation. We will
return to the topic of confidentiality when we talk about establishing a
coaching conversation in sec. 3.5.
Finally, it is over when it is over. Meaning, you do not restart a coach-
ing conversation once you have completed it. There is a time and place
for the conversation. Within that you discuss challenges, you explore
possibilities, you define action plans and you agree on how and when
to follow up. Outside that the conversation is not going on. You do not
The Hitchhiker’s Guide to Agile Coaching
refer or add on to the conversation when meeting the coachee at the
coffee machine, in the canteen or anywhere else.
A coaching relationship can only exist within these restrictions. Always
keep them in mind and, when necessary, make them explicit to the per-
sons you are coaching.
. The mindset behind systemic
coaching
A coaching conversation is something that is co-created between the
coachee and the coach. It is like a dance to which both parties con-
tribute. In order to perform well, you must first agree about the pur-
pose, the style, the rhythm and so on. If in a real dancing situation one
tries to do a jitterbug while the other attempts a waltz, the mix-up will
become apparent within the first few steps. In a coaching situation how-
ever, misunderstandings of similar magnitude can go undetected for a
long time.
As coaching is an approach to help people grow, a positive mindset is
also required. People grow through appreciation, by being acknowl-
edged and respected for who they are, but also by being challenged by
new and more ambitious goals. Appreciation is not about praising peo-
ple for any random reason, but taking an approach like appreciative in-
quiry (Cooperrider and Whitney, 2005) as a starting point.
Appreciative inquiry is a research-backed approach where people in an
organization collectively imagine and describe a compelling future. Be-
cause this future state is desirable, people don’t have to be forced or in-
centivized into working their way there. Focusing on problems and how
to solve them is wearying in the long run and might even make the prob-
lems bigger. You will find that several of the models we present in this
Coaching Fundamentals
book, e.g. the 5D model in sec. 4.2, follow the appreciative inquiry ap-
proach.
A catalyst for this is to live out the Principle of Heliotropism, which states
that positive imagination leads to positive actions and results. By hav-
ing positive ideas about the future we are already laying the foundation
for this positive future. Positive thoughts create similar connections in
our brains as if we already were acting according to these ideas. This
frees up energy to see new possibilities and do constructive work to-
wards achieving our goal. Like a sunflower we are orienting ourselves
towards the sun and take advantage of the possibilities we are being
given. One of the most well-known uses of the Principle of Heliotropism
was President Obama’s victory speech in Chicago Grant Park, November
4th 2008. His primary messages were All things are possible, Changes are
coming and Yes we can!
With this in mind, let us dig into what tools and techniques you can use
for successful coaching.
. Curiosity and keywords
At the risk of stating the obvious, we would like to underline that a coach
should be curious and interested in people. This is a self-selective trait:
if you are not interested in people, you will find it very boring to work
as an agile coach. So don’t be afraid of asking questions. The only thing
worse than being clueless about something is trying to hide that you are
clueless. And frankly, who is the subject matter expert here — you or the
coachee?
That said, if you keep asking random questions or repeat the same pre-
pared questions, it sends a signal that you are not taking the conversa-
tion seriously. Make notes, review them and learn. People know that
names and in-house acronyms are difficult to remember, so those are
The Hitchhiker’s Guide to Agile Coaching
more excusable. And sometimes new tools or an organizational restruc-
turing may cause what you have learned to become obsolete. Even so,
the professional approach is to remember as much as possible or at least
give the impression that you care.
Keywords are words that unlock further ideas and concepts. They de-
pend almost totally on the coachee and the conversation and are not al-
ways easy to identify. We recommend that you keep notes, jotting down
what you think could be significant. If some word pops up or is men-
tioned repeatedly, ask about it or underline the words so that you can
return to them later. When there is a lull in the conversation, scan back
and look for interesting and relevant topics to return to. “You mentioned
. . . earlier. What does that mean?”
Asking about something may uncover more keywords. Jot them down
too and use the new keywords in combination with your agile exper-
tise to design more questions. There is a root cause analysis technique
called “five whys,” hinting at the fact that digging down five layers is usu-
ally enough. Sometimes there could be three layers and sometimes six,
but on average the causal branch you are exploring will bottom out after
five times of asking “and why does that happen?”
Be aware that it’s entirely possible to dig too far. You should look out for
circular answers and indications that the coachees are getting bored or
fidgety. In a coaching conversation, you need to walk a fine line between
going too wide and going too deep.
As a practicing agile coach you will stumble across patterns that repeat
inside and across companies. Don’t make the mistake of assuming that
these similar patterns have similar solutions. Keywords most often
point to symptoms, but the underlying network of causes and effects
can be very complicated and not all of the causes are necessarily visible.
For example, many new Scrum teams (and sadly many established
teams too) are struggling to meet their sprint goals. Clearly something
does not work as it should. . . but what? Perhaps the Product Owner is
Coaching Fundamentals
pushy, perhaps the managers don’t understand capacity planning,
perhaps they are dependent on another team delivering an unreliable,
low quality component? Perhaps all of the above. . . or something else?
Don’t make assumptions — go and see.
We will discuss this topic further in the chapter on structured coaching
(sec. 7). For now, remember to be curious and keep track of keywords.
. Listening and responding
We all think while we listen. We study the person whom we are in a con-
versation with, reflect on the words, tie back into our own experiences,
find sudden insights or interesting parables and connections. Often our
thoughts take us far away into our own spheres of interest and when we
come back to reality we realize that we have missed something. . . but
probably nothing important, right? Other times we find something re-
ally important to say and mull over the message, rehashing it in our own
minds while we wait for our turn to speak. . . and forget to listen. This is
how our brains are built and there is nothing wrong with that.
As coaches, however, these habits make it difficult for us to do our job.
They take focus away from the coachee, and while we sometimes get
good ideas, just as often we might jump to premature conclusions. As
coaches, we need to listen better. In this section, we will cover some
thinking models and effective techniques for listening.
Shifting or supporting
The waiting-to-speak syndrome occurs when the listener is not actu-
ally focusing on what the speaker is saying, but rather is waiting for a
break in the flow in order to infuse her own thoughts into the discus-
sion. When somebody does this to you it feels like the other person is
not actually listening to what you said. In response, people often resort
The Hitchhiker’s Guide to Agile Coaching
to repeating their words in a louder voice, rehashing their arguments in
different words or speaking on top of the other person.
This syndrome is the result of what we call shift response — when we
think more about how to shift the topic into whatever suits our purposes
than about what the other person is saying. The opposite, support re-
sponse, is when we are following along with the discussion in order to
keep it flowing. In general there is nothing wrong with either kind of
response and a good discussion needs a suitable combination of both.
When we are coaching, however, we need to be careful not to drive the
discussion. We need to be aware of which response we are using. Sup-
port response is generally the gentler and safer option. We already men-
tioned keywords in sec. 3.3 as a way to stay focused and in sec. 3.6 we will
discuss an effective method for creating questions that bring the conver-
sation forward without disrupting the topic.
Communication enablers
Every serious traveller has experienced situations in which they need
to communicate with someone but don’t know the language. It can be
something as simple as buying a meal or getting a taxi or something
as complex as trying to explain what you do for a living — presumably
some form of ScrumMastering or agile coaching. How do you explain
agility if you only have five words in common? How about coaching?
Without a common language communication becomes slow and error-
prone and is limited to simple and concrete things. If you can not point
to it or show a picture of it, it does not exist. Having sufficient command
of a common language is one of the basic communication enablers. As
the name suggests, the enablers help the flow of communication. If you
are missing one or more enablers, communication — and thereby also
coaching — becomes difficult. If many or all enablers are missing, com-
munication becomes impossible.
Coaching Fundamentals
A common culture is another enabler, similar to a common language.
People from different cultures look at the world differently and this can
cause misunderstandings during discussions. Cultural differences can
also cause prejudices which tend to prevent people from entering a con-
versation with open minds.
Listener biases and emotions likewise make it difficult to communicate.
As an agile coach, you will often meet people who are just doing things
wrong and refuse to understand it — in fact they may even try to con-
vince you that your own methods will never work here. It may help to
recognize that you are both biased in your own ways. Similarly, if a per-
son is upset or sad, she may find it difficult to focus on the discussion.
It could be a quarrel at home, some personal issues, perhaps bad news,
sometimes just too little sleep or a headache.
For communication to be effective you will also need enough time and
a distraction-free environment. For example, one of the authors of this
book gets caught by moving images and finds it difficult to hold a dis-
cussion if a television is on in the background. Even an information ra-
diator or a street window can be distracting. He solves this by picking
seats that face away from the TV or window.
Furthermore, the communication channel should be as wide as possi-
ble.1 Morse code or radio transmissions are at one extreme: these trans-
missions are very slow, unidirectional and require additional encoding
and decoding. Videoconferencing tools are somewhere in the middle,
although they have become much better over the last decade or so. Live
face-to-face discussion is at the other extreme. In a live conversation,
verbal communication is enhanced by postures, gestures, facial expres-
sions and body movements. Information flows in both directions be-
tween the participants.
1 “The most efficient and effective method of conveying information. . . is face-to-
face conversation.” (“Agile Manifesto,” 2001)
The Hitchhiker’s Guide to Agile Coaching
Active listening
Virginia Satir (1964) was the first to recognize that listening consists of
several stages. She divides reception into intake, meaning, significance
and response. As coaches, we like to use the simpler but still adequate
active listening model2 which contains the three phases of compre-
hending, retaining and responding.
In active listening, the first phase is comprehending. This includes hear-
ing and understanding the words, sentences and the context. As men-
tioned above, understanding the language and the cultural context is
very important and even critical for comprehension.
The second phase is retaining. The problem here is that speech is real-
time, wide-band communication. The words come at you fast and it
is not always possible to stop and consider a sentence from different
perspectives or even ask the speaker to repeat. Other problems include
the fact that short-term memory is lossy — previous sentences will be
drowned out by new ones.
Further, it sometimes happens that you underestimate the significance
of something when you first hear it. Later you get more information and
realize that the speaker made some important point previously, but you
can no longer remember exactly what it was. It is also easy to lose atten-
tion, especially in one-way presentations or in other situations where
you are not expected to contribute.
The third phase is responding, which includes both verbal and
non-verbal responses. Body language — a simple thing like raising a
finger — can be a powerful way of reacting without actually breaking
the flow of the speaker. Verbal responses come on different levels: we
can repeat what we just heard word by word, we can paraphrase the
message or we can add our own reflections to it. We might also bring
2 http://en.wikipedia.org/wiki/Active_listening
Coaching Fundamentals
the topic forward by replying in support or in contradiction, or perhaps
shift the topic to something else.
The good news is that listening and responding is a skill at which you
can become better. Let’s focus on a model that may help you understand
how listening works.
The levels of listening
There are different ways of listening, some of which are more suitable
and some less suitable for coaching. Kimsey-House et al. (2011) de-
scribes three levels of listening that approximately map to first person
(me), second person (you) and third person view (us). Let’s describe
these listening levels (see also fig. 3.2) and reflect on how they can be
useful for an agile coach.
Figure 3.2: The levels of listening
The Hitchhiker’s Guide to Agile Coaching
Level 1 – Internal listening (subjective) — In level 1 listening, the
narrator’s story is reflected through the listener’s own experiences
and memories. Keywords remind you of episodes or topics that
are of interest to you, but not necessarily to the speaker. It can
trigger a premature reply, an unexpected shift response that cuts
the flow and steers the discussion off topic.
For example, when hearing about somebody’s misfortunes on the
way to work, the listener may be prompted to tell about her own
commute. Or when hearing somebody complain about a prob-
lem, the listener may feel obliged to solve it. Or when the speaker
mentions a theory or method, the listener may be led to give a lec-
ture on the topic complete with “interesting” examples.
Sometimes people just feel the need to vent and do not want
much more than “tea and sympathy”, compassion and support.
Other times, they might be hoping for concrete advice related to
some issue and might not be interested in hearing tangential
stories. The speaker may be halfway through a story when she is
interrupted and not allowed to finish. In all of these cases, the
shift response may be quite rude to the original speaker. People
may get the perception that the responder is selfish, trying to
hijack the discussion or one-upping in an attempt to impress.
Of course, level 1 listening can also inject new and interesting in-
formation into a discussion. When mentoring, it can be useful to
share personal experiences. A drink and a funny anecdote told
at a party may trigger more drinks and even funnier anecdotes.
In coaching conversations, however, level 1 is mostly out of place
and should be used with care.
Level 2 – Focused listening (passive) — In level 2 listening, the listener
is fully focused on sucking in information and committing it to
memory. The story is heard without filters and the listener is al-
most sitting in the narrator’s chair.
Coaching Fundamentals
This can be a very effective way of learning about the other per-
son. We use the VIBE model to remember what to observe when
listening:
• Voice — the intonation, energy level, sighs and grunts
• Information — verbal (choice of words, what is said, what is
left unsaid) and non-verbal (shrugging, pointing)
• Body language — hand gestures, changes in posture, facial
expressions etc.
• Emotions — is the person coming across as happy, sad, dis-
tressed, calm, energetic, passive, neutral etc.
The main drawback of level 2 listening is that you sometimes for-
get to participate enough in the discussion and this can make the
speaker nervous or distraught. This is because people subcon-
sciously search for feedback while speaking. If the feedback is
missing or contradictory, the speaker gets confused and doesn’t
know how to continue. Most people just sputter to a stop, think-
ing that the listener is trying to be mean or sarcastic. As a coach,
this is not the message you want to send.
Level 3 – Global listening — Listening with empathy. Everything is in-
cluded in the listening: the speaker’s tone of voice, body language,
changes in energy level, etc. The listener is not only listening to
the voice and words of the speaker but also actively follows the
non-verbal cues. Global listening requires patience and curiosity
and the ability to put yourself in the shoes of the speaker.
Humans are social animals and we have evolved to use body
language and facial expressions as well as prosody — the tone of
voice, syllable stress and rhythm — when we communicate.
Non-verbal cues give additional information that can be very
useful in a coaching conversation. Often the cues support the
verbal message, but sometimes they can seem irrelevant or
The Hitchhiker’s Guide to Agile Coaching
even conflicting. Conflict in words and body language could for
example indicate that the other person is stressed or distracted.
Other times you can pick up non-verbal information that changes
or even inverts the meaning of the words. Sarcasm, for example,
relies on tuning the verbal message just right. While people can
express sarcasm just by choosing the appropriate (or inappropri-
ate) words, it is much easier if they can use non-verbal channels
as well.
Non-verbal communication is often more honest than verbal
communication. That is, somebody might say one thing yet
make it apparent from their stance and the tone they use
that they would rather say something else. There are also
microexpressions, visible emotions like disgust, anger and fear
that flicker over the face for a fraction of a second. While they are
difficult to catch, they are also almost impossible to suppress.
People who know what to look for might gain a lot of information,
such as whether the other person is being truthful or not.
That said, in the context of agile coaching, we are more interested
in setting up a safe and constructive environment than
constantly assessing the honesty of the coachee. Body language
is too large and fuzzy a topic to fit in here, so we will not go into
detail. Our message to you is to watch for gestures, postures,
emotions, inflections, pauses etc., that indicate whether the
coachee is alert and interested in the conversation. Combine this
information with the spoken words to gain a more complete
understanding of what is going on.
In addition to the three levels of listening, there are also several ways
of not listening. One might ignore the other person completely or
half-listen while doing other things. Obviously these are not good
approaches for an agile coach. If you find yourself distracted to the
point of listening with a half ear only or briefly ignoring the other
Coaching Fundamentals
person, the best thing you can do is apologize and either break the
discussion or continue with renewed focus.
There is nothing particularly good or bad in any of the three levels. In
fact, in a normal social conversation all three levels are needed in suit-
able amounts. In coaching conversations, however, level 1 and to some
extent also level 2 can disturb the balance of the discussion and under-
mine the trust between coachee and coach. This is why an agile coach
should continuously be aware of which level she is listening at and try
to move towards levels 2 and 3.
Try this
1) You can try out the different levels of listening on a test subject in
a short exercise of less than ten minutes. Explain the three lev-
els, then ask the other person to tell you a story e.g. about what
they did last weekend in three minutes. While they tell the story
you will switch several times between the three levels of listening.
After the story is finished, the storyteller should try to guess the se-
quence of listening levels that you used. You can also swap roles
and run the exercise again.
2) If you notice that you are falling into level 1 listening, feeling a
strong urge to tell of similar events or give solutions, you can try
some of the following “survival techniques”:
• Focus on identifying and writing down keywords as you hear
them, so that you can get them off your mind and onto paper
where you can pick them up later. If the same keywords are
repeated, underline them or draw boxes around them.
• If you run out of keywords, also note your own ideas and hy-
potheses. One technique is to draw a vertical line across the
paper and write keywords and observations on the left and
your own thoughts and hypotheses on the right.
The Hitchhiker’s Guide to Agile Coaching
• Ask neutral clarifying questions related to the keywords you
pick up: “Tell me more about . . . ”, “What do you mean when
you say . . . ?”, “What was the impact of . . . ?” or “Why do you
think they acted that way . . . ?” Don’t try to help the other
person or create any development in the discussion.
• In your own mind, quickly and silently repeat the words the
other person is saying. This technique is known as “rapid
repeating” and it helps with clearing out your own thoughts.
• Don’t talk, use only non-verbal signals (sounds, expressions,
grunts and body language) to keep the conversation going.
• Limit yourself to replies of one or two syllables (“Yes”, “OK”,
“Mhm?”, “And then?” . . . ) to keep the conversation going
without injecting topics.
• Admit that you got lost in the discussion: “Sorry, I was
distracted/had to write something down. Could you repeat
what you just said?”
• Repeat back what you heard to confirm your understand-
ing (active listening). You can repeat exactly what you heard
word by word (verbatim) or describe it in your own words
(paraphrasing).
• Adjust your own posture to change the dynamics. Lean
forward or back, cross/uncross your legs or just relax in the
chair. When the topic changes, you can take it as a cue to
change your posture.
• Mirror the other person’s body posture and voice level. This
technique is known as “pacing”.
• Apply some “mindfulness” in your listening. Without remov-
ing your focus from the speaker, try to relax and make your
awareness wider. Open up your senses to take in as much of
the environment you can.
3) The next time you are in a place with many others — a lounge,
classroom, concert, shopping mall, beach, park, restaurant, foot-
Coaching Fundamentals
ball match, train station etc. — take a moment to apply level 3
listening on the whole room. Observe what is happening. What
groups can you see? Who is talking to whom? Are people ener-
getic, expectant, stressed, happy or sad?
. Structure for a coaching
conversation
A coaching conversation is not just small talk about how things are go-
ing or not going. Coaching conversations are for committed individu-
als or teams that want to make a significant change or get wiser about
an important matter. To help you succeed as a coach, you can apply a
structure to the coaching conversation. This helps you and the coachee
(or team) to focus on the important topics and find specific actions to
carry out as results of the conversation.
It is important to remember that a coaching conversation is something
that you design together with the persons you are working with, at the
moment when the conversation is wanted. As a coach you never take
the coachee or team to places they do not want to go — this would be
strictly out of line. You should never guide the content of the discus-
sion, nor force somebody into a discussion. What you do instead is help
them to go where they want to go, while helping them reflect on what is
important.
As the coach, your task is to make sure that the environment is right. In
sec. 3.4, we mentioned the important communication enablers. Try to
ensure that all of them are in place. Allocating enough time and setting
up a distraction-free environment is easy, but if there are large cultural
differences or you lack a common language you may have to reconsider
whether the conversation is worth having at all.
The Hitchhiker’s Guide to Agile Coaching
We have found that many people do not know what coaching looks like.
Sometimes they just need to vent or want some advice and get upset
when you start by asking questions. In such situations, we have found
it useful to ask for permission to coach, for example, by saying: “Is it
ok if I ask you some questions to help you achieve the result you want to
achieve?”
The two levels of the coaching conversation
As shown below in fig. 3.3, a coaching conversation is conducted on
two levels: A) the conversation level and B) the meta level (Storch et
al., 2006). As the coach you are constantly acting on both levels. The
coachee or team is primarily on the conversation level, but will from
time to time be invited to the meta level by you.
Figure 3.3: The structure of a coaching conversation
Coaching Fundamentals
The conversation level is where the actual conversation happens. Here
you are using the levels of listening, showing curiosity and forming
questions based on keywords as we discussed previously. The coachee
or team are also on this level as they answer your questions.
The meta level is where you are observing, reflecting and designing the
conversation. Here you are deciding which powerful questions3 to ask,
which hypotheses to formulate and in which direction to take the con-
versation next. As a coach, imagine yourself as having a third eye ob-
serving the conversation from the meta level and your awareness about
the flow of the conversation. The answers you get will help you make
the right decisions.
Meet up at the meta level
As mentioned, you will from time to time invite the coachee or team to
join you at the meta level. The purpose for this is to have a conversation
about the conversation to collaborate on designing the conversation, re-
flect on the learnings so far and make decisions about where to go next.
There are normally at least three opportunities for meeting at the meta
level:
(1) When you are establishing a contract for the conversation
(2) When you are having a check-up during the conversation
(3) When you are concluding the conversation.
At the very start of the conversation, you should spend a little time on
making contact with the one(s) you are coaching. Establishing contact
helps people relax and feel confident in speaking freely. This is small talk
and chit chat, talking about the weather, the traffic, how’s-your-family,
etc. Keep in mind that you should start steering towards starting the
actual coaching conversation before too long.
3 The concept of powerful questions is explained up in sec. 3.6.
The Hitchhiker’s Guide to Agile Coaching
How you can establish the contract
When you are establishing the contract, you invite the other part to talk
about the conversation you are about to have. The contract consists of
an agreement that you will have a coaching conversation, the ground
rules for the conversation and the goal for the conversation.
First of all, it should be clear for both parties that you are not just hav-
ing a little chat. You are about to enter a serious conversation which
might include both sensitive information and personal decisions. This
expectation is sometimes implicit in the invitation, but it doesn’t hurt
to make it explicit. If you are in a discussion already and realize that it
might turn into a coaching conversation, you might ask “Is it OK if I ask
you some questions to help you solve this problem?” You can also set up a
later meeting in a similar way: “Should we sit down tomorrow and have
a coaching conversation about this topic?”
You will also need to establish the ground rules of the coaching conver-
sation. This normally includes a statement of confidentiality from your
side as well as the promise of an “emergency exit” for the coachee or in-
dividual team members. For confidentiality, we typically use the simple
and strict Vegas Rule that we described earlier, in sec. 3.1. However, if
the team you are coaching is large and the atmosphere in the organi-
zation is sufficiently constructive, forgiving and supporting, you could
also consider the Chatham House Rule4 . It states that all participants
are free to use any information, but are not allowed to reveal who made
any comment.
In recurring meetings the rules quickly become implicit and you don’t
need to repeat them, but when you are coaching someone for the first
time you will have to explicitly set down the rules. We have a kind of
boilerplate that we use that goes something like this:
4 https://www.chathamhouse.org/about/chatham-house-rule
Coaching Fundamentals
“Before we get started, I would like to make it clear that I will keep this
conversation confidential. I may bring up general issues with others, but
won’t talk about things that can be traced to you or to this discussion un-
less you give me permission. Also, you have the right to end the discussion
and walk out if you feel like it, no questions asked. If at any point and for
whatever reason you feel that you don’t want to be in this room anymore,
just say so and we will stop right there.”
We find it good practice to have the counterpart formulate the goal of
the conversation in one short sentence — and we usually also write it
down in a notepad or on a sticky note. We usually ask questions like:
• “What is the topic you want to elaborate and get insights on?”
• “How can I best serve you during this conversation?”
• “Are there questions you especially want me to ask or questions you
absolutely do not want me to ask?”
• “When this conversation is over, where do you expect to be and what
do you hope to have learned?”
Now the actual coaching conversation can begin. Start by picking out
keywords from the goals you just agreed on and formulate a question or
several questions from that.
Having check-ups
During the conversation, you should from time to time make a timeout
to check-up on the conversation. Look at the clock and the timebox,
summarize the learning so far and decide where to go next.
Check-ups help to co-design the conversation on the fly with the pur-
pose of bringing the most possible value into it. Think of it as a sort of
in-sprint inspect-and-adapt cycle on the meta level.
You can use check-ups when you feel the conversation is at a crossroads
in which you have to decide which path to take next. Be humble and
The Hitchhiker’s Guide to Agile Coaching
unbiased: do not take for granted that your personal decision will be
the best path. Instead ask the one(s) you are coaching what they think
and follow their choice. Remember: it is not about you! It is all about
them! You can also use check-ups to re-negotiate the goal if you and the
other person realize that another topic seems to be more important.
You can make as many check-ups as you feel necessary, asking ques-
tions like:
• “Let me summarize the discussion so far. We have been discussing
. . . and . . . , figuring out that . . . . Do you want to expand more on
this, or would you rather move on and look at other options? What
would that be?”
• “Let’s look at where we are in this conversation. As I see it, we can
either go in the direction of . . . or in the direction of . . . . You might
see a third direction. Where do you want to go from here?”
• “Let’s step out of the conversation and check how we are doing. In
the beginning of this conversation we agreed to speak about . . . but
it seems to me that we are now discussing . . . . What do you think,
should we return to the original topic or is the new topic more im-
portant to you?”
When the check-up is over you can continue the conversation taking
into account the decisions you have just made together.
As an aside, we have found it useful to have a clock unobtrusively avail-
able. The simplest way is to put your wristwatch flat on the table beside
your notepad. Smartwatches can be set to vibrate at certain points in
time. Some smartphones can be configured to always show the time
and there’s a plethora of timer apps. If all else fails, simply explain that
you want to see how much time is left before you pull out your phone or
watch.
Coaching Fundamentals
Define the next steps, starting with a clear conclusion
By the end of the conversation, it is time to come to a conclusion focus-
ing on the specific steps the coachee or team is going to do, in order to
initiate the desired change. Have him, her or them summarize the con-
clusion instead of you doing it. That fosters the sense of responsibility.
Remember: it is not your solution — it is their solution!
As a follow up to the conversation, it is great practice to ask about what
the next step will be, when this will be done and how you will know that
this has been achieved. To the last question the one(s) you coach will
usually answer something like: “I’ll send you an e-mail, letting you know
how it went”. In service to the other part, you could then reply: “And
if I do not receive this mail, will it be helpful for you if I ask you about
how it went?” This attitude sharpens the awareness about the coaching
conversation as something that serves a purpose, rather than just being
a chat about life, the universe and everything.
End by asking for feedback
Finally, being a coach that wants to improve your skills, you should also
ask for feedback about the coaching conversation. You can ask ques-
tions like: “How was this conversation for you?”, “What did I do that was
especially useful for you?” and “Which questions did I ask that were use-
less or disturbing for your understanding of the matter?”.
Receive the feedback with gratitude and ask clarifying questions if you
wish, but do not go into arguing about whether the feedback was right
or wrong. You are asking for opinions and everyone is entitled to their
own. The important matter is how the coachee or team experienced
your coaching — there is most likely a point behind the feedback re-
gardless of whether you liked it or not.
Just like you opened the conversation, it can also be good to close down
the conversation with chit chat and small talk as you collect your pens
The Hitchhiker’s Guide to Agile Coaching
and papers and pack your bag. If you are meeting the coachee or the
team again, you might also want to agree on the time and perhaps a
topic or even an agenda for your next meeting.
Try this
1) Coaching conversations are best practiced in a live setup. We
recommend that you start out by practicing with individuals and
once you are comfortable with this, you can move on to coaching
teams. This means finding someone who is willing to be coached
by you, even though you may still be a novice.
Ideally the initial setup should be three persons: one who is
coaching, one who is being coached (the coachee) and one who
observes the conversation. This works very well if all three of you
want to practice coaching. You can then take turns being the
coach, the coachee and the observer. You can do 20-minute
coaching conversations followed by five minutes of feedback
from the observer, then take a five-minute break, switch roles
and have a new coaching conversation. (Yes, as you probably
already have figured out, the Pomodoro technique5 can be used
for many purposes.)
2) If you can’t find an observer, then another option is to record the
conversation and learn from it. You will of course need the per-
mission of the coachee to record the conversation.
Our experience tells us that while audio recordings are good, video
recordings are much better for recall and understanding. Posi-
tion the camera so that your upper body is in the frame, not the
coachee. Make sure that the picture is wide enough for you to
move around, but still shows your postures and facial expressions
clearly.
5 See http://pomodorotechnique.com/
Coaching Fundamentals
Most modern pocket cameras or smartphones are more than ad-
equate for this purpose, but you may want to record a minute of
video beforehand to extrapolate whether there is enough memory
available. Downscale the video to a lower quality in order to save
space, as low-quality video is better than no video at all. A cheap
tripod, a fast charger and a long charging cable will also be useful.
3) Imagine a coaching conversation where you find it difficult to syn-
chronize with the coachee. You can either think back to a previous
coaching conversation that did not go as expected or think about
an upcoming conversation with a person you are uncertain about.
Now consider the communication enablers in sec. 3.4. Which
ones were (or will be) especially relevant for the success of the
conversation? Is there something related to the enablers that you
could have done (or could now do) differently to improve the
chances of success?
. Powerful questions
“It is folly to say you know what is
happening to other people. Only they
know, if they exist. They have their own
Universes of their own eyes and ears.”
— Douglas Adams
Asking the right questions is a challenging task, especially when you do
not want to impose your own opinion on to the person or team you
are talking to. Great coaching questions are open-ended, non-judging
and help foster new ideas and visions about possibilities. These kind of
questions we call powerful questions (see fig. 3.4).
The Hitchhiker’s Guide to Agile Coaching
Figure 3.4: Powerful Questions
There are several approaches to creating powerful questions. They de-
pend on which coaching school you are coming from. One approach
is to practice a deck of questions until you know them by heart which
helps you choose the right one in a given situation. Another approach is
to learn a strategy for designing the right questions as needed. One such
strategy that we have found works particularly well with both individu-
als and teams is based on a model developed by Canadian psychologist
Karl Tomm.
The past and the future — simple and complex assumptions
The Karl Tomm approach has its roots in systemic theory. It encapsu-
lates circularity and the understanding that each of us has a different
perspective of the facts in a given situation and is entitled to that view.
No one has monopoly on the truth; indeed different people can have
very different perspectives and hence different understandings of what
is happening and why.
The model in fig. 3.5 shows two dimensions. Time is shown on the hor-
izontal axis as the past and the future. On the vertical axis we can find
Coaching Fundamentals
Figure 3.5: Karl Tomm’s model for asking Powerful Questions
simple (obvious, evident, linear) and complex (non-evident, multi-path,
circular). While time should be easy to understand, the concept of sim-
plicity vs. complexity may require a bit of explanation.
In a simple and linear world, there is little room for doubt. Cause-and-
effect relations are easy to see and everybody can correctly predict the
result of any given action. In the complex world, however, you generally
can’t draw a straight line from cause to effect. There can be multiple
interacting causes, not all of them visible. There may also be circular
cause-and-effect chains of different lengths, for example vicious loops
and virtuous cycles. In a complex, circular world people may have quite
different interpretations of why something happened the way it did, and
sometimes even different opinions on what exactly happened.
Powerful questions are anchored in all four quadrants of Karl Tomm’s
The Hitchhiker’s Guide to Agile Coaching
model. The powerful questions can be aimed either at what has already
happened (past-oriented), or at what could happen (future-oriented).
The powerful questions can also be designed on the presumption that
there is one and only one truth (linear questions) or so that they ac-
knowledge the diversity in our understanding of the truth (circular ques-
tions).
When we ask questions that are past-oriented and linear (upper left in
fig. 3.5), we ask questions like a detective in an interrogation, for exam-
ple:
• “What happened then?”
• “What kind of velocity did you have before that?”
• “When did it happen?”
• “How did it work out?”
• “Was it the PO or the customer?”
As a detective, we are looking for facts that help us understand the issue.
Remember that facts are undisputable: something either happened or
it did not happen.6 All kinds of data, statistics, recordings, diaries, blog
posts, e-mails etc. are welcome.
But even when people are in the same place at the same time, they of-
ten end up with different observations. Memories have a tendency to
evaporate over time. People fill in with assumptions when they don’t
have all the facts. This means that people can have very different inter-
pretations of various events. Hidden assumptions are a major source of
conflict and unearthing them can help immensely.
When we ask questions that are past-oriented and circular (lower left
in the picture), we ask questions like an anthropologist doing research.
6 As far as we know, this is the case everywhere except in Washington DC, where it
might be helpful to explicitly say “a true fact”.
Coaching Fundamentals
We are looking for intentions and expanding our understanding of the
intentions. For example, we could ask:
• “What do you think was their motivation to do so?”
• “From which point of view could her action make sense?”
• “Could it be that they knew something that the rest of you did not?
What could it be?”
• “Do you think she saw this as a problem at all?”
Future-oriented circular questions (lower right) are intended to explore
opportunities and different scenarios and expand possibilities. Here we
ask questions like a future researcher, such as:
• “What could you do to make things better?”
• “Is there something you could do to prevent this from happening in
the future?”
• “Can you see a possible solution to this problem?”
It can also be useful to apply appreciative inquiry and heliotropism to
explore a positive and compelling future rather than focus on the prob-
lems and issues. One trick is to ask a “miracle question”; something that
suggests that a miracle has happened overnight:
• “If you come to work tomorrow and the problem has unexpectedly
disappeared, how would you notice that?”
• “One day when you have solved this matter and look back on this
period in time, which decisions did you make that made a differ-
ence?”
Questions in this future-oriented, circular category might seem a bit
strange at first, but give it a try. The advantage of questions like this
The Hitchhiker’s Guide to Agile Coaching
is that they disconnect the person or team in front of you from the con-
strained situation they are currently in. They free up energy to see new
perspectives and decide on new courses of action based on those per-
spectives. They help the person or team to see which part of the miracle
or the desired future are already present today and how they can be used
as stepping stones towards the goal.
Finally, we can ask simple or linear future-oriented questions (upper
right). These are the kind of questions that a captain would ask. They are
more direct and practical, defining the next steps for the desired change:
• “What is the first thing you are going to do now?”
• “Who do you need to talk to in order to move this forward”
• “How can he or she help you?”
• “How will I know that you have done this?”
Like at the end of a retrospective, this is where specific tasks are defined
and prepared for action. As the coach your role is to ensure that the tasks
are clear and simple, but also that the coachee takes responsibility. You
can look back at sec. 3.5 for inspiration and more examples.
It is like driving a car
When coaching someone, we are mostly looking at the future and the
change we are going to make. However, from time to time we must look
back into the past to understand what has happened and why we are in
the current situation. It is like driving a car; we are mostly looking out
the windshield at the traffic in front of us, but from time to time we also
need to look in the rear-view mirrors to know what is behind us.
There is no specific route you must follow when using Karl Tomm’s dis-
cussion model. We have found that most of our coaching conversations
tend to start in the detective dimension, moving on to the perspective of
an anthropologist, then investigate as the future researcher and finally
Coaching Fundamentals
ending up as the captain. This is the order in which we introduced the
quadrants in the previous section. You can imagine drawing the letter
“U” over fig. 3.5, starting from the upper left. The path is not necessarily
linear, though. On your way, you can jump back and forth as intuition
tells you.
You have already learned about conversation flow, active listening and
identifying keywords. You are now adding powerful questions to your
toolbox of coaching skills. Asking powerful questions is a valuable
coaching skill and one that can be improved with practice.
Critics are saying
When first exposed to powerful questions, your coachee or team mem-
bers might think: “What the heck is this guy up to?” Especially if you are
asking miracle questions. You may also personally feel a bit uncomfort-
able using these kind of questions. We recommend that you review your
questions during and after the discussion, thinking back to which ques-
tions resulted in interesting and worthwhile answers and which ques-
tions seemed to break the flow or perhaps even offend the other person.
Especially questions from the anthropologist perspective are usually
grounded in the belief that behind any action, there are good
intentions.7 Not everybody believes in this postulate and we must
admit that we also regularly have to remind ourselves about it.
Nevertheless, taking this approach will let you help people reflect as
well as help solving conflicts.
You may also find it useful to remember that in any given situation, peo-
ple try to do the best they can with the information and tools they have.
We will return to the Prime directive of retrospectives by Norman Kerth
in sec. 6.4.
7 As stated by Humberto Maturana, one of the founders of the Systemic Theory.
The Hitchhiker’s Guide to Agile Coaching
While powerful questions are useful, they are not a silver bullet. For ex-
ample, we sometimes run across teams that feel powerless, suppressed,
oppressed or victimized. They may be stuck complaining about the past
and refuse to move on to imagining a future. “Yes, we tried something
like that four years ago and it didn’t work.” There are strategies for deal-
ing with such teams, such as arranging a quick win in some area just
to show that change can indeed happen or enlisting managers to help
(although often the managers are part of the problem).
Try this
1. Think of a possible coaching situation. It can either be with a team
or with a person one on one. It might be a coaching conversation
you already had with someone or you may be preparing for an
upcoming coaching situation.
Try walking around in the Karl Tomm model and formulate ques-
tions from the four different perspectives. Formulate at least three
questions in each dimension. You might have to imagine the an-
swers you will get in order to formulate new questions.
Next, practice asking powerful questions in live coaching situa-
tions. You can also use them in situations you would not normally
consider to be coaching situations, like in Daily Scrum for exam-
ple.
2. When you are in a suitable coaching conversation, spend some
extra effort at the beginning of the conversation separating facts
from interpretations. Write statements and keywords on stickies
and put them on a Karl Tomm model on a whiteboard. Ask ques-
tions that help you understand whether a statement is a fact or an
assumption and recategorize the stickies as necessary.
“The PO moved a difficult backlog item from their backlog to ours”
is a fact, while “the PO did it because he thinks the other team
doesn’t like difficult backlog items” would be an interpretation.
Coaching Fundamentals
Chapter 4
Coaching conversations
with a team
In the previous chapters, we covered the basics of systemic coaching,
such as how to listen and keep a conversation flowing without actually
steering it; how to open up a new coaching conversation and guide it to
closure; and how to formulate questions so that the coachee provides
the content while you provide the structure. As you start getting the feel
for how to run coaching conversations with individuals, you can move
on to having coaching conversations with teams.
Coaching teams is far more complex than coaching individuals. You
probably already have guessed that. When coaching a whole team, you
need to focus on more than one person; you need to be aware of the
perspectives and intentions of several people; and you need to master
both steering the conversation towards a goal, letting each of them have
airtime and help the team finding consensus. Tough job, isn’t it?
Well, Don’t Panic, even though the going gets tough. In this chapter,
we will introduce you to some tools and techniques that you can use in
team conversations.
. Bridging questions
While the powerful questions we discussed in the previous section are
useful for “opening the box” and creating progression throughout a dis-
Coaching conversations with a team
Figure 4.1: Bridging questions
cussion, the intention of bridging questions is to collect thoughts and
ideas and to build a common understanding in the team.
In team coaching conversations, bridging questions gather information
and conclusions and help the team find and expand on a common
agreement. They build, so to speak, a bridge between the different
team members’ attitudes, experiences, opinions and suggestions. They
bring out new information and help resolve conflicts.
Your ability to use bridging questions is the first step towards mastery of
team coaching conversations.
The Hitchhiker’s Guide to Agile Coaching
The idea behind Bridging Questions
One of the the major challenges in newly formed teams is making deci-
sions. Good decisions require the right amounts of exploration, dissent,
analysis and effort. Some teams can take a very long time to get this
right.
There are several ways in which the decision-making process can go
wrong. One common issue is that the team starts digging into the first
reasonable solution instead of exploring options, thereby limiting their
solution space. People can sometimes take criticism of an idea very per-
sonally and start defending that idea beyond what is reasonable. An-
other common problem occurs when two or more persons each have
individual agendas to drive. Having a strong opinion and being vocal
about it doesn’t mean that the opinion is the best.
As a great team coach, you must ensure that all members take part in
the conversation and contribute to the solution. Through the use of
bridging questions, you can create strong links between the team mem-
bers by linking the ideas generated by them. You do this by formulating
questions that nudge team members into relating their answers to what
was previously said by others, instead of just pursuing their own agenda.
Some examples:
• “What similarities do you see in what different people are saying
here?”
• “In what way do the things Paul is speaking about link to your in-
terest in the subject? What can you add to this?”
• “In what way are you inspired when you hear the others discuss the
subject?”
• “Is there something you feel that the others have forgotten in their
discussion?”
• “In which way can you see that your thoughts are aligned with
other people’s opinions? In which way do your thoughts differ
from the others’ opinions?”
Coaching conversations with a team
For example, we once encountered a team with a very respected, skilled
and resourceful developer who was also a strong open source advocate.
His pet feature was to maintain compatibility with a similar open source
project, even though this played no role in the company’s strategy. He
would inject backlog items and add tasks related to this project (e.g.
“Maintain compatibility with XML schema changes in OSS project”) and
while the team grumbled a little they did not speak out. The PO did not
have enough technical knowledge to see what was happening.
By the use of bridging questions, we made him aware that most people
on the team did not enjoy the extra work and in fact thought it a bit silly.
After a few rounds of this he came to the conclusion that it was time
to drop the topic. As a result the team became more focused and their
throughput increased markedly to the happy surprise of the PO.1
We have also seen teams where people don’t respond much to bridg-
ing questions, saying things like “I don’t know,” or “Nothing comes to
mind.” Sometimes they are just indifferent, sometimes afraid of rock-
ing the boat, and sometimes they are genuinely unable to come up with
anything. This could indicate that you are working with a workgroup
or pseudo-team that lacks synergy and heedfulness, something we will
discuss further in sec. 5. But there are also ways of creating peer pres-
sure within a discussion by drawing on people’s feelings and emotions.
In the next section we will explain how you can do that.
Reflecting participants
Some team members might not have a strong opinion about a certain
topic, so they might choose to be passive in a team discussion and let
two or three strong minded members drive their own arguments. This
has several problems. Since some of the expertise now lies latent, the
team risks reaching a premature conclusion and the passive team mem-
bers are not helping resolve the conflicts either.
1 We also like open source software but we think focus is more important.
The Hitchhiker’s Guide to Agile Coaching
A great team coach can invite the passive participants into the conver-
sation by asking them to reflect on what the strong-minded members
have said:
• “What do you especially like about the ideas of Peter? Paul? and
Mary?”
• “When Peter and Paul are arguing so strongly for each of their dif-
ferent opinions, which similarities do you see in their opinions that
they might not see?”
A reflecting participant can also provide a strong contribution for solv-
ing a conflict:
• “When you experience that Paul and Mary are having this conflict,
how does it make you feel? How do you think the others feel?”
This approach relies on the fact that people are social animals and as
such reluctant to upset others in the “tribe”. People who keep distressing
or angering others are generally going to be pushed towards the fringe of
the social group.2 In prehistoric times this would have a negative impact
on your chances of surviving and finding a mate and thus the asocial
tendencies would mostly be bred out of the population.
In modern times, people often belong to multiple “tribes”, either per-
manently or for a shorter time. This could include family, childhood
friends, old classmates, neighbours, hobbyist groups, various fan clubs
and, of course, teams at work. As coaches, we can exploit the social glue
in a workgroup to help the group come to a joint conclusion in the way
we just outlined above.
2 Watch out for bullies, narcissists and other sociopaths. They instead form the core
of such groupings and survive by exploiting this tendency in others.
Coaching conversations with a team
. The D model
Back in sec. 3.6, we looked at Karl Tomm’s approach as a way of gen-
erating powerful questions. We can also look at it as a complete path
through a team coaching conversation. We would apply the coaching
conversation structure from sec. 3.5, keep the discussion flowing using
active listening and use the powerful questions to create a coherent di-
rection in the discussion.
This approach works well enough for us to use it as the “default” in many
coaching conversations. It works especially well in situations of conflict,
because a large part of the discussion is spent on discovering facts to-
gether and creating a common understanding that includes different in-
terpretations. It also works well if the team is aware of some pain point,
but cannot formulate what it is.
However, while Karl Tomm’s model is good at digging into the past, it is
not as good at dreaming up ambitious goals. In this section we would
like to introduce another path through a coaching conversation that we
call the 5D Model. The 5D model helps the team choose or formulate the
problem to be addressed; share past experiences that might be worth
trying; explore what a great solution could look like; define the stepping
stones towards this solution; and decide who is doing what. It can be
especially useful if the team has found a problem that they want to solve
either in an ad hoc discussion or in a separate retrospective meeting.
The 5D Model, shown in fig. 4.2, is rooted in appreciative inquiry that we
mentioned briefly in sec. 3.2. It consists of five phases: Define, Discover,
Dream, Design and Deliver. You can run this as a guided discussion, or
you can plan exercises to draw out the best ideas.
Define
In this phase, we define the topic we would like to develop in this coach-
ing conversation. This phase is somewhat similar to establishing the
The Hitchhiker’s Guide to Agile Coaching
Figure 4.2: The 5D Coaching Model
contract in a classic coaching conversation. We describe the develop-
ment area through “transforming questions” — questions that have a
special power of change designed in. In contrast to problem-solving
questions, transforming questions are constructive and positive and fo-
cus on the goal rather than the problem, as illustrated by the following
example:
• Problem-solving question: “How do we prevent our meetings from
being so boring and unimportant?”
• Transforming question: “How do we create dedicated meetings
that help us realize the improvements we have set as goal?”
As you can see, the basic principle here is to base the questions on the
situation that you want to achieve rather than the situation you want to
Coaching conversations with a team
avoid. Problem-solving questions are not wrong as such, but they can
drain the energy from the team. In our experience, team coaching dis-
cussions benefit from “dreaming big”. Be ambitious and stretch this goal
by making it lofty, so that the questions will give constructive leverage
later in the conversation.
Discover
In this phase, you explore the past experiences of the team members
with a focus on successful events and incidents. The purpose is to de-
scribe and explain the contexts that illustrates the best practice of the
team members. It is important that the questions are as specific as pos-
sible.
Because the team is focusing on the things where they have succeeded,
we are already fostering the positive effect of the problem solving in this
phase. If the team has little or no prior positive experiences related to
the topic, ask the team members to think of other teams that have suc-
ceeded in similar situations. This can include teams that some of the
members have been a part of in the past or neighbouring teams in the
same company or competing companies or perhaps something from a
book or article.
As a coach, you can use multiple tools to facilitate and drive the work-
shop. If you run this as a discussion, remember to regularly sum up the
experiences the team is talking about. You can also facilitate a Brain-
storming exercise with stickies, or the Brainwriting variant that also ac-
tivates silent members. If the group is large (15+ people) you could run
a quick World Café.
You can also split the team and give the groups different targets, e.g. one
group to ideate and list mainstream ideas; one to work on the loftiest,
most inspiring ideas; one to come up with the weirdest possible ideas;
one on anti-ideas (ones that prevent you from achieving your goal); and
The Hitchhiker’s Guide to Agile Coaching
so on. Make presentations or a gallery walk at the end to spread the
knowledge around.
Dream
Having a common picture of the ideal situation for the team in the fore-
seeable future is a very powerful tool. It helps the team keep on track as
they are improving. Here the focus is on the future and the team’s desire
for ideal practices or an ideal situation. If “dreaming of the perfect fu-
ture” feels too fuzzy or idealistic, you can instead use terms like “vision”
or “ideas”.
At this stage, you should strive to create a playful and welcoming atmos-
phere — a “yes, and” mindset — so that team members gradually be-
come bolder as they experience the friendly welcome of their ideas and
imaginations. It is therefore important to avoid negative replies or re-
jection of a team member. Instead draw inspiration and build on top of
the ideas of others, providing more details and offshoots of each other’s
imaginations. All ideas are welcome.
With minor changes, many tools from the Discover phase also work well
in the Dream phase. If you facilitate a conversation, try using bridging
questions to build a common understanding. You can also have multi-
ple groups work on different perspectives on the same issue (for exam-
ple People/Process/Product) and merge their results. You can also try
the Park Bench or Fish Bowl exercises.
Design
Based on the wishes for the future and the best experiences of the past,
we now examine which of the possible future situations will be of par-
ticular relevance to the team. We help the team define what approach
or approaches they should focus on. Again, it is important to keep the
wording as concrete and action-oriented as possible.
Coaching conversations with a team
Deliver
In this final phase, the goal is to define the actions and plans of action
to achieve the objectives. The responsibility is defined, so it is clear to
everybody who is doing what and when. This clarification should be
explicit and detailed, so no one is in doubt about his or her role in the
near future. It should also be agreed about when to follow up, so the
implementation can be secured and supported. The concept of SMART
actions3 can be helpful.
Sometimes we find that teams are nervous about going forward with the
concepts they have Dreamed and Designed. The concepts look good
on paper, but reality strikes when they start drawing up concrete action
plans. Throughout the Design and Deliver phases, keep an eye open
for uncertainty. For example, we have found it useful to ask the team
members to individually judge how sure they are about the proposal. We
ask them to think of a number from 0 to 10, where ten means absolute
certainty and zero means that they really can’t say anything, then raise
that number of fingers.
If it turns out that the team feels generally uncertain about the proposal,
ask them what would increase the score. If there are different opinions
in the team, ask what makes people sure and what makes them unsure.
If they still feel uncertain, help them create time-limited safe-to-fail ex-
periments. The outcome of the actions will then be learning and knowl-
edge.
The 5D model is a powerful and effective communication structure be-
cause it breaks the way we usually think of improvements. By anchor-
ing the discussion in what has actually worked instead of what does not
work, we can increase motivation and enthusiasm and pave the way for
future learnings.
3 SMART = Specific, Measurable, Achievable, Relevant and Timely. See http://en.
wikipedia.org/wiki/SMART_criteria
The Hitchhiker’s Guide to Agile Coaching
The model does put a lot of responsibility on the facilitator, though. As
you will notice, guiding the team through the model takes some careful
balancing between guarded optimism and, well, unguarded optimism.
The “landing curve” from Dream into Delivery can be tricky, especially
if the team you are coaching has been conditioned over the years not to
rock the boat. Make sure to help them out: talk to the managers; drop
some words in the right places; bring the stakeholders’ attention to the
right issues.
. Try this
1) Reflect on the benefit of bridging questions — how can you see
yourself using them in practice? Formulate one or two bridging
questions that you can use in general. Memorize them and try
them out when you have the next opportunity to do so.
2) Think about an upcoming discussion where there might be an ar-
gument between two or more people. This might be a retrospec-
tive or perhaps a backlog refinement or sprint planning meeting,
where you know that people will promote different approaches.
Who of the other participants could you involve through a bridg-
ing question? (If there are none, you may want to invite a couple of
peers, for example.) Prepare a couple of bridging questions aimed
at different people. Write the questions down, then try them out
in the meeting.
3) The Karl Tomm and 5D models are best practiced when you have
an actual open conversation with a team, for example in a retro-
spective.
Prepare by drawing one of the models on a flip-chart (look at
fig. 3.5 or fig. 4.2 for the general layouts) and reflect on which
questions you could ask the team in each of the phases.
Coaching conversations with a team
Bring the flip-chart to the retrospective, give the team a short
introduction to the model and use it during your conversations.
Document your journey through the model by writing stickies
and putting them on the flip-chart. Point out to the team
whenever you change phase in the model.
The Hitchhiker’s Guide to Agile Coaching
Chapter 5
Team dynamics
Information systems today (and indeed the last 50+ years) are just too
large for one single software engineer to handle. Even though we have
excellent building blocks in the form of stable, low-cost software stacks
— operating systems, standard libraries, databases, etc. — it turns out
that we need a whole team and often several teams to get stuff done in
a reasonable time, before the customer grows tired or the market moves
on.
Having more people around brings a number of positive side effects.
More people means more ideas and this gives us a wider source of po-
tential solutions. Team members can become sparring partners, help-
ing each other improve in their craft and reviewing each others work to
improve the quality. The team can, as a whole, cover a larger experience
and expertise area than any one single developer and they also have bet-
ter opportunities to share and spread the workload.
But teams can also suffer from communication and synchronization
problems. The more people and roles you have, the longer it takes for
information to disseminate and decisions to be made. Some people
find it difficult to work together. There are egos that need satisfaction
and people who drive their own agendas. Teamwork does not happen
by itself.
In our experience, successfully adopting the mechanics of Scrum
will generally give any workgroup a 15–30 % increase in productivity.
This comes from focusing on the content (delivering running tested
software) rather than the container (project plans, work breakdown
Team dynamics
structures, gantt charts, hourly reports, bug statistics). Basically, the
team starts doing the right things and doing them right.
From a traditional perspective, a 15 % increase in productivity is simply
incredible. A traditional manager would kill for that kind of improve-
ment! But those teams are still lacking synergy. By carefully nurtur-
ing your team past the initial formation stage, by letting them collabo-
rate and gain trust and start demanding mutual accountability, you can
grow a high-performance team. Even in strictly traditional organiza-
tions where Scrum and Kanban are seen as some kind of weird process
used only in R&D, we find teams that continue to ramp up productivity
by 20–50 % year after year.
Again, this does not happen by itself. It requires a dedicated line
manager, a skilled Product Owner, an effective ScrumMaster and, of
course, a team that is willing and allowed to rise to the challenge. In
this chapter we will explore the concept of groups, pseudo-teams and
high-performing teams and how you, as a coach, can help people move
them from one to the other.
. Synergy in teams
In the movie “Cast Away”, Tom Hanks plays the character of Chuck
Noland. As the title implies, Chuck is stranded on a desert island. But
he is not alone: he has a companion, Wilson, in the form of a volleyball!
Unfortunately, Wilson is of the silent type and does not offer much
help. Even though Chuck is very resourceful and innovative — he
extracts a bad tooth with a skate, among other things — he is never able
to achieve more than his own effort and ingenuity can provide. Chuck
Noland is solely one man, not a team.
What if we have several people? A group of people, for example, stand-
ing at a bus stop, still can not achieve more than the sum of the effort
The Hitchhiker’s Guide to Agile Coaching
from each person. In a working group, people typically have a profes-
sional relationship and exchange information related to their work, but
they do this only in order to achieve their individual goals. People at a
bus stop exchange information about when the next bus will arrive, but
will use this information to optimize their own journey and use of time,
sometimes to the point of competing for the limited space in the bus. A
working group lacks synergy.
At the other extreme, a Formula One team running a pit stop is a great
example of synergy. Their shared goal is to win the grand prix and in
order to do that they need to get a top position in the race today. Thus
their immediate target is to get the F1 car out of the pit stop as quickly as
humanly possible. Because of this shared goal, as well as having agreed
on roles and procedures, they are able to achieve more than if they were
working as individuals in a group. Everyone in the team collaborates to
get the car moving again. The team consists of experts, not specialists.
(A specialist only specializes in one thing, while an expert can have ex-
pertise in many areas.) If a team member is having problems, others
will step in and help. It doesn’t matter who does the job as long as it gets
done quickly and efficiently.
The difference between a team and a working group is synergy. Individ-
uals can’t achieve synergy for the obvious reasons. Groups don’t have
synergy by definition. Only when a group has worked together towards
a shared goal for long enough to jell together can we call them a high-
performing team. But what is this “jelling”? How do you recognize it?
And how do you make it happen?
We would like to refer to an important article by Weick and Roberts
(1993). As they studied teams that perform well in challenging envi-
ronments, they noted that the synergistic effects are born in interpreta-
tions and actions. Team members contribute to the goals by acting ac-
cording to their own interpretations of the situation and recent events.
Obviously if team members have different understandings of the situ-
ation and the goal, no synergy is possible. But if their interpretations
Team dynamics
Figure 5.1: Synergy in teams
are aligned so that they support and supplement each other, then the
contributing actions will also converge and give synergistic effects. The
whole collective activity system of habits, actions, signals and reactions
becomes greater than the sum of its parts.
For example, one team that we coached dug up an old Nabaztag1 and
hacked it so that for each commit, it would run different kinds of sta-
tic and dynamic analysis on the new code and then publicly announce
what it thought of it. If the rabbit gave negative feedback, other team
members would tease the hapless author but then generally help her
1 An internet-connected plastic rabbit with ears that could be rotated remotely. Yes,
you read that correctly. Look it up.
The Hitchhiker’s Guide to Agile Coaching
out. Soon people started to confer with others or work in pairs before
committing code, in order to get kudos from the “Killer Rabbit”. What
started as a nifty hack became a cornerstone in the team’s attempt to
strengthen the habit of writing good code.
Weick and Roberts note that this kind of social intelligence requires a
certain level of heed in the group members. Heedful people are atten-
tive, observant, careful, respectful and helpful towards others. The more
heedfulness in the system, the more developed the collective mind will
be. It can even transcend the individuals, meaning that the patterns and
habits remain when individuals come and go. When members start act-
ing out of habit or relying on luck, some of the collective intelligence is
lost and performance degrades.
From the perspective of agile coaching, the most important concept in
the whole article is that one can strengthen and renew the heedful inter-
pretation and contribution patterns that form the collective mind. We-
ick and Roberts note that heedful teams tend to:
• Make the patterns visible
• Model and discuss them
• Reward reinforcing actions
• Preserve anecdotes
Agile teams use working agreements, a Definition of Done, team boards,
team signals, information radiators, zero defect tolerance, build scripts
etc. etc. to make things visible. As an agile coach, you are in a unique
position to help them out in the daily standups and retrospectives. You
can teach them concepts like visual management and automation, and
facilitate meaningful conversations about what they see.
Of the surrounding roles, the ScrumMaster and the line manager are
best positioned to see reinforcing actions and give rewards. In the “Killer
Rabbit” case we mentioned above, the ScrumMaster was the one to drop
Team dynamics
the initial idea and the line manager was instrumental in helping the
team find a solution that was compliant with the IT security guidelines.
The same people can also help the team form a collective history. A
scratched snowboard carefully leaning against an old server or a dusty,
unopened magnum champagne bottle with an equally dusty and un-
opened congratulatory card are each telling their separate stories. If this
sounds like team identity and culture. . . it is. If it sounds like it will be
difficult in a distributed team. . . you’re right, again.
. Team development and
performance models
Team forming and performance does not happen by itself. You can-
not form a team by placing 5–9 persons in a room and shouting “SELF-
ORGANIZE!” as you close the door.
Given a moderately healthy environment and time to work together, a
group of people will generally start collaborating to achieve better per-
formance. This can take a long time, however, and not all groups will get
there in the end. There might be incompatible personalities or people
who simply do not get along. A skilled coach can guide groups past the
worst pitfalls and help them build a well-performing team much faster
than they could do it on their own.
In the 1960’s, Bruce Tuckman (1965) introduced a model for
understanding the development of teams. The model is interesting for
agile coaches, because if we can identify the stage of the team, then we
can determine what interactions with the team would be appropriate.
The model in its original form consists of four stages:
Forming — Group members are generally attentive and well-behaved,
but self-focused. The group of people is looking for answers on
The Hitchhiker’s Guide to Agile Coaching
questions like: Who is in the team? What are we expected to do?
How shall we do it? Who are we reporting to?
Here the agile coach should help the team members to get to know
each other and clarify the basic terms, objectives, vision and val-
ues. Assuming that these remain stable, the team should move
into Storming on their own.
Storming — The group members are starting to form opinions about
each other. They are trying to establish a common understanding
of the goals, roles and procedures, but have challenges coordinat-
ing and resolving difficulties.
Here the agile coach should work on diversity, dissent and conflict
resolution/dissolution. In other words, it is OK to have differing
opinions, but people need to learn how to constructively build on
top of the contributions of others.
Use bridging questions (see sec. 4.1) to help people understand
what others are thinking. Team building exercises can be very use-
ful at this stage and you could use exercises such as the Market of
Skills and Skills Matrix to help the team see what kind of people
the others are.
Norming — Members are starting to understand and accept the dif-
ferent working habits in the team. The team starts establishing
a common understanding, roles and procedures through self as-
sessment and agreements. The community will be established
and each individual will begin to accommodate herself.
Here the agile coach should encourage the development of the
team-specific common understandings, roles, routines, etc. Help
the team create working agreements, coding guidelines and pull
policies. Update the Definition of Done and streamline the task
board. Help the team build their identity: pick a team name;
encourage some fun and silliness; retain souvenirs; let them
Team dynamics
rearrange their corner of the office; help them set up their own
information radiators; etc.
Performing — A team in this stage can primarily concentrate on get-
ting the job done, rather than thinking about procedures, cooper-
ation and organizing. The cooperation is working well and there
is less talk about process and self assessment.
Here the ScrumMaster or agile coach should help encourage
work performance through a focus on excellence, further
potentials, new goals and targets, etc.
Note that this stage is very wide and doesn’t account for the pres-
ence (or lack) of synergy. It covers all working groups and teams
that have established their roles and procedures and are now fo-
cusing mainly on the actual work. The performance can be low or
high or anything in between.
Teams do not necessarily progress linearly through the model. Tuck-
man noted that many teams skip the Storming stage altogether, going
directly from Forming to Norming. Other teams can get stuck in Storm-
ing or Norming. And whenever the team setup changes (members leave
or join) or the context changes (the team relocates or is transferred to
another department), the team becomes a little less mature. It can, for
example, drop from Performing back to Norming or even all the way to
Forming. The agile coach should change her interactions to reflect this
new situation.
Research of team performance conducted by Jon Katzenbach and
Douglas Smith and described in their iconic book The Wisdom of Teams
(Katzenbach and Smith, 1993), shows how the performance of a team is
impacted when a group of people develops from being a working group
to a high-performance team. Katzenbach and Smith describe the
following stages, also shown in fig. 5.2:
The Hitchhiker’s Guide to Agile Coaching
Figure 5.2: Team performance
Working group — A working group is merely a group of people, for
which group performance is primarily dependent on individual
contributions. The working group is the sum of its parts but
nothing more and there is little potential for improvement.
Members do not take responsibility for results other than their
own, nor do they try to develop incremental performance
contributions requiring the combined, real work of two or more
group members. In other words, there is no synergy in the
working group.
Pseudo-team — This is a working group for which there could be a sig-
nificant, incremental performance need or opportunity, but the
Team dynamics
group either does not recognize this potential or is not interested
or capable of working towards it. The typical pseudo-team is not
focused on collective performance and is not trying to achieve it.
It can be very discouraging to be part of a pseudo-team where the
members are not interested in jelling. This can sometimes lead
to a vicious cycle of mediocrity and stress, an ongoing state of
low productivity that we call the Valley of Despair (see the end of
sec. 5.3 for more information).
When coaching a working group or a pseudo-team, you should
focus on fostering a team identity. Help them create a common
goal and also introduce opportunistic collaboration where it
makes sense. Show them that collaboration can be fun and
profitable individually (learning) as well as collectively (quality,
performance). Also help them visualize their work and
understand what is happening.
Potential team — This is a group with a significant, incremental perfor-
mance need that is interested in improving its performance and is
actively working to achieve this. Typically, the group would need
more clarity about purpose, goals or work products and more dis-
cipline in hammering out a common working approach. It has not
yet established collective accountability.
When coaching a potential team, show them how easy it is to try
out new techniques and working methods. Focus on improving
the retrospectives and ensure that the team is in control of its
working agreements, tools and processes. Ensure that they have
simple but good metrics with both leading and lagging indicators
so that they know when they are improving.
Real team — This is a small number of people with complementary
skills who are equally committed to a common purpose, goals and
working approach for which they hold themselves mutually ac-
countable.
The Hitchhiker’s Guide to Agile Coaching
In order to become a high-performing team, the team needs to
set up personal professional development plans together. They
will also need to know how to give and take feedback. As a coach,
you can help them achieve this.
High-performing team — This is a group that meets all the conditions
of a real team. In addition, the members are deeply committed
to one another’s personal growth and success and new members
will also join this commitment. The high-performing team signif-
icantly outperforms all other similar teams and also outperforms
all reasonable expectations given its membership. In the perform-
ing phase the team has reached synergy and are getting a lot more
out of their individual contributions.
We would like to mention that we especially like the Market of Skills
team exercise, as it contains many components that are important for
teams in the early Tuckman and Katzenbach-Smith stages. This is some-
thing we try to introduce in all teams that have worked together for less
than a year. As a follow-up, the Skills Matrix helps people see collabora-
tion opportunities, which is essential in pseudo-teams as well as poten-
tial teams.
The Tuckman and Katzenbach-Smith models look very similar and we
chose to draw them as parallel tracks in fig. 5.2. This is not always the
case. One could argue that the Katzenbach-Smith model continues
where the Tuckman model stops (with a bit of overlap). A forming
or storming team according to Tuckman can at most be a potential
team in the Katzenbach-Smith model and what Tuckman considers a
performing team may still be only a pseudo-team according to
Katzenbach-Smith.
Furthermore, the Tuckman model is team-internal only and team for-
mation can be strengthened regardless of the organization (assuming
of course that the budding team is not torn apart all the time). Im-
provements along the Katzenbach-Smith model are, however, strongly
Team dynamics
dependent on the surrounding organization and require understanding
and support from the ScrumMaster, PO, line manager and stakeholders.
For example, if the line manager keeps using new words but acts like
always before, the team will notice this.
Becoming a real team requires that each member is willing to take risks
and trust the others. Where they previously had individual actions and
separate work products, they will now have collective actions and joint
work products. They need to agree on a common purpose, set goals,
decide on the approach, and take mutual accountability. There will be
conflicts that need to be handled in constructive ways. People who call
themselves teams but take no such risks are at best pseudo-teams.
. Challenges in forming teams
As we mentioned earlier, teams do not “happen” by themselves. Even
assuming that the different personalities on the team can get along,
there can be multiple stumbling blocks along the way. According to the
book “Effective Teamwork” (West, 2012), there are several challenges in
forming teams, including:
Free-riding — Free-riding can happen when someone realizes that his
or her personal effort is no longer visible. With half a dozen other
people doing hard work, nobody will notice if one person just
“floats along”, right?
This was first described by Maximilien Ringelmann (1913)
through a simple experiment in a rope pulling contest. When the
test person is told that he has six strong firemen on his team, he
will pull less than if he believed he was one of the strongest.
Ringelmann also found that the individual involvement
decreases in larger groups.
The Hitchhiker’s Guide to Agile Coaching
Inefficient decision-making — Team decisions are hard to make.
Team decisions are usually better than the average individual
decisions made by the team members, but worse than a decision
made by the most brilliant team member. Team decisions are
highly influenced by the persons who speak first and loudest
within the team. These may be people who want individual
recognition or promotion or managers who want the team to
come to a specific decision. There are many tools to reduce this
influence, such as Roman Voting, Fist of Five and Planning Poker.
Hogging credit — In a well-working team, the team as a whole takes
responsibility for successes as well as failures. The road there can
be bumpy, however. Individual contributors can find it very chal-
lenging to trust other persons on the team if it can influence their
own personal standing in the company.
In this context, hogging credit means to make sure that your own
role in a success is emphasized or exaggerated (“Looks like the ar-
chitecture I drew up served us quite well”), while distancing your-
self from failure (“I said we should check it with the customer,
but they didn’t listen”). Credit-hogging can happen in private dis-
cussions as well as in public communications. It is often subtle
and difficult to disprove, except to those who see it first hand: the
team-mates.
The challenges listed above are all internal to the team. But there are
also external challenges. Many organizations are, in fact, set up in a way
that undermines or outright prevents the formation of high-performing
teams. Some factors include:
Deep specialization — A specialist has deep knowledge in one area,
which is usually quite narrow. Specialists are not rewarded for
helping others and can be reprimanded if they try. Therefore, they
Team dynamics
have a tendency to focus on their own narrow area and are not
used to collaborating.
Specialization is a good survival strategy in mechanistically de-
signed organizations and many managers are proud of the depth
of knowledge in their teams.
As a coach, you will need to make people understand that being a
generalist or expert (in many different areas) is an even better sur-
vival strategy. Both the organization and the individual will bene-
fit from having a wider profile of skills.
Belief in heroism — When a project is behind schedule, managers
often try to create more work hours per day. Engineers get to
spend long evenings and weekends at the office. Othertimes
strict processes and decision hierarchies force people to bypass
processes and do work under the radar in order to get things
done in a reasonable time.
To make up for it, managers reward these just-get-it-done people
with public recognition, gifts and bonuses, sending everyone the
message that heroism is highly valued. In the end, the whole or-
ganization believes that working overtime or bypassing processes
is normal.
A coach would need to promote strict timeboxes or WIP limits, as
well as sustainable development and team accountability, in order
to expose the danger with heroism.
Lack of transparency — In centrally directed organizations,
information often flows from the operative layers inwards or
upwards, while the outwards or downwards flow consists mainly
of decisions. Information may even be withheld from operative
personnel by the central layers.
Lacking broader information, people are unable to make good de-
cisions and will instead push their own opinions and agendas. In
The Hitchhiker’s Guide to Agile Coaching
those kind of systems, politicians who talk smoothly and know
what information to share with whom will prosper. Central sys-
tems are also slow to respond, although they are good at retaining
standardization.
Creating visibility and transparency is important. Presenting work
progress in a central obeya (war room) can be very useful, espe-
cially if managers start using that room as a base for making deci-
sions.
Individual incentives — Personal bonus plans make people work for
their own profit rather than for what’s best for the whole company.
For example, a developer may decline to help a teammate with a
project-related problem because that project is not on her incen-
tive plan. Another typical example is that a salesman sells a new
feature that turns out to be technically very challenging to imple-
ment.
We have found that replacing individual incentives with team in-
centives helps a lot. It is also a good idea to base incentives on ac-
tual outcomes and make them win-wins rather than guesswork-
based metrics: “Deliver as quickly and cheaply as possible” rather
than “Deliver by Oct 15 and at a cost of maximum 1.2 million.” We
have seen that relative targets work better than absolute metrics:
“Improve the lead time by 15 %” rather than “Reach a lead time of
8 days.”
Salary discrepancies — In many companies, salaries are set on an indi-
vidual basis. This means that people are in fact rewarded for ne-
gotiation skills rather than technical and teamworking skills and
for changing jobs rather than staying with their teams. Managers
may also be handcuffed by a limited increases budget, creating a
situation where employees have to compete with each other for
a raise. This is obviously a problem for managers as well as for
employees.
Team dynamics
Changing an existing wage system inside a large corporation is go-
ing to be very difficult. Managers may find opportunities when
setting up new teams from scratch. Many agile companies still
retain opaque wages. There are also companies with transparent
wage systems. For more insights you may, for example, want to
read “Joy, Inc.” by Richard Sheridan (2013).
Status rewards — Where wage discrepancies are often kept secret, sta-
tus rewards are more or less public. This can include such things
as a personal corner office, a standing authorization to travel first
class, a named parking place. . . or ownership of a red stapler.
They may start out as reasonable allowances or have some his-
toric background, but eventually take on a life of their own and
become a source of friction between people.
Agile companies take the status rewards off the table by making
the same rewards available to all employees. In cases where
supply is indeed limited, shared resource pools with usage
transparency usually work fine.
Mandated corporate tools — While tools can be very useful, we find
that they can also cause unintended negative behavior. For ex-
ample, if your tool automatically assigns an owner to every back-
log item or task, it could cause team members to focus more on
their own tasks and thereby collaborate less. A tool that uses the
term “requirement” instead of “backlog item” could increase the
threshold for proposing changes. And a cumbersome tool with
strict access control might result in backlogs not being updated in
a timely manner.
To fix this you can try to find a reasonable compromise. For ex-
ample, many teams maintain control of their work by moving the
smallest level of detail (tasks that are measured in hours) out of
the tool and on to a physical task board. They move post-its on a
The Hitchhiker’s Guide to Agile Coaching
daily basis and update the larger-scale items in the corporate tool
when necessary.
Project resourcing — One particularly common problem is project re-
sourcing, meaning the way teams are routinely disbanded after a
project and reassembled with different people for the next project.
The reason is that a team needs to invest time and effort in becom-
ing a high-performing team. Both the Tuckman and Katzenbach-
Smith models assume that the group of people is stable for long
enough to make this investment.
On the way to high performance, the team must go through the
Valley of Despair (see fig. 5.2), where much of the effort is spent on
establishing the pecking order and coping with dissent and where
their productivity is even worse than it was at the start. Getting be-
yond this can take a long time. The normal timespan from start-
ing out as a working group to becoming a high-performing team
is between six to twelve months.
Whenever you disband and re-form teams, they lose everything
they have invested in bonding and jelling with their team mates.
Over time they learn that the process is not even worth starting.
There is instead a high risk that they take permanent residence in
the Valley of Despair. This is not a great place to be. Unfortunately,
many organizations are unaware of the magnitude of the prob-
lem as they have no stable and high-performing teams to compare
against.
. Summary
In this chapter, we discussed how teams differ from workgroups and we
covered several different models for forming and developing teams and
Team dynamics
for identifying and avoiding common pitfalls. Team synergy and build-
ing up trust are so important that an agile coaching engagement can be
considered an exercise in team building.
An agile coach continuously observes how people behave within the
system and tries to change the system so that people can change their
behavior for the better. In sec. 7 we will look at a structured method for
improving teams.
. Try this
1) Choose a team you are working with. Observe the team members
both as individuals and as a group and assess the team using the
Tuckman and Katzenbach-Smith models.
Next, choose an exercise, a work habit or a topic to raise that may
help the team move on to the next stage. Validate your chain of
reasoning with the ScrumMaster, the Product Owner, the devel-
opment manager, and other stakeholders surrounding the team.
Apply it and reflect on the results.
2) Observe a team you are currently working with. Which of the chal-
lenges listed in sec. 5.3 are most prominent in the team? Why do
they occur?
The Hitchhiker’s Guide to Agile Coaching
Chapter 6
Coaching your team
By now we have covered what agile coaching is, talked about how to
structure and facilitate coaching discussions with individuals and
teams, discussed how groups of individuals can develop into teams,
and how to form a new agile team. Next up, we will cover some of the
better-known agile and lean frameworks and think about how they
support agile coaching.
Figure 6.1: Inspect and adapt
You will be coaching the team every time you meet them, but there
Coaching your team
are subtle differences in what to focus on in the different meetings and
events. During backlog refinement and sprint planning meetings, you
should focus on facilitating — helping the team and stakeholders make
great product decisions. After the daily standups, however, you have a
great opportunity to grab five or ten minutes with the team following up
on how they are doing short term. In the retrospectives, you can drive
long term team questions.
. What is built into your framework?
The framework you have chosen supports coaching in certain ways.
Scrum conveniently provides a number of pre-scheduled meetings and
touch points with the team, so there is little to plan.
Scrum also has two specific roles besides the development team,
namely the ScrumMaster and the Product Owner. Both of these are
make and break roles. As the agile coach, your job will become difficult
and you will lose precious time if one of the designated persons is
unsuitable for the job or doesn’t have enough time to do it well.
Kanban, on the other hand, doesn’t define any meetings or roles. We
find that many teams pick up meetings from other agile methods and
these generally fall in one of three categories:
1. Just-in-time meetings: backlog replenishment, continuous
improvement meetings, quality circles
2. Asynchronous cadences: daily standups, backlog refinement, ret-
rospectives
3. Cadences synchronized with other teams: daily standups, release
demos
The Hitchhiker’s Guide to Agile Coaching
Just-in-time meetings are often scheduled on short notice — “hey, let’s
refine a couple of backlog items after lunch today” — which can be dif-
ficult if you are working with multiple teams or multiple clients. Ca-
denced meetings on the other hand can be scheduled months in ad-
vance. For this reason, we have found it best to set up cadenced meet-
ings during the coaching period also when working with Kanban teams.
Each of these touchpoints bring together the whole team or a subset of
the team and provide a great opportunity for you to observe or interact
with them. Don’t give them up too easily! New Scrum teams often think
that some of the built-in activities are obsolete or boring and want to
skip them. If that happens, we suggest that you ground the discussion
with the team in the values and principles in the Agile Manifesto (“Agile
Manifesto,” 2001). Together, figure out which values and which princi-
ples are behind each of the activities they want to skip. If the proposed
change does not undermine those values and principles, just go for it.
Otherwise a better option would be to figure out new and better ways to
meet the values and principles.
When observing your team, it’s useful to look at team dynamics, what
they talk about and the energy levels. We often use the following “cheat
sheet” of questions, adapted from Lyssa Adkins (2010):
• Is everyone who wants to getting the time to speak? Are there
dominant people in the room who need to listen more? Are there
quiet voices that want to be heard?
• Are the ideas of high quality or are people simply going with the
easiest solution?
• Is the team moving toward the simplest solution possible? Or are
they going future-proof and gold plating it too?
• Is the team getting tired? Do they need a break?
• Is the atmosphere getting tense? Do they need some comic relief?
• Is the team being audacious enough? Do they come up with great
ideas or break through barriers? Or are they avoiding taking a risk?
Coaching your team
• Are they taking on as much as they could or are they letting “ac-
cepted” barriers get in the way?
• Is the team considering things in terms of customer value? Or in
terms of their own effort?
• Are they stuck? Do they need a new perspective, one that brings
them more possibilities?
Figure 6.2: Making observations
This is intended to be a starter kit of observation points. Keep this list of
observation questions handy, somewhere that you can quickly access it
when a conversation just happens around you. Over time, you will come
up with your own observation questions arising from what you observe.
For example, if you work with several teams in the same company, you
The Hitchhiker’s Guide to Agile Coaching
may notice that they behave similarly. Add your questions to this list
and share them with fellow agile coaches.
. Daily standup
The daily standup is a short-term planning meeting. Although the meet-
ing is not exclusive to agile methods, it’s an integral part of many agile
methods including Scrum, Kanban and XP.
For an agile coach, the daily meeting forms a great opportunity to ob-
serve the team and ask some coaching questions. Let the meeting run
its course while you observe and listen, make notes and form hypothe-
ses. While every team is different, there are some items to keep in mind:
• Situation: Does the team report to itself or the ScrumMaster? Is
the situation presented honestly? Do they have facts and infor-
mation in front of them? Is the granularity right (tasks less than
one day in length)?
• Focus: Is the goal clear? Is the team focusing on getting the next
backlog item done, rather than ensuring everyone has work for
the day? Is there a lot of bureaucratic overhead?
• Speaking: Does everyone get the opportunity to speak? Who
speaks most, who is most silent? Do people listen intently or are
they just waiting for their own turn? Are people supporting each
other?
• Decision-making: Who makes decisions? Is it one person or the
whole team? Do they evaluate multiple options? Are decisions
based on facts? If they make assumptions, do they go on to vali-
date the decisions before investing time?
• Language: Does the team have their own “slang”? Does the body
language support the verbal message?
Coaching your team
• Trust: Are they showing respect for each other and for other
teams? Are they having fun together? Are they able to bring up
difficult topics? Are they showing courage?
Standups are also a good moment to evaluate team agreements, for
example retrospective action points, daily goals, checking burndown
charts, urgent tickets or any other things the team agreed to check daily.
If you have something to discuss with the team or just want to share your
observations, you can ask them to stay on for a few more minutes after
the meeting. It’s polite to make this request before the meeting starts.
Try this
• Focus on the work, not the worker. Avoid asking each person on
the team to give a status report; focus instead on the stories and
priorities.
• Pass a ball (or some token) around to indicate whose turn it is to
speak and to add dynamics to the daily standup.
• Team members should be speaking and making eye contact with
each other, not reporting to the ScrumMaster.
• If the team is not finding the meeting useful, find the root cause
and fix it, rather than abandoning the meeting. Often the granu-
larity of the tasks does not match the frequency of the meeting or
people do not see the need to collaborate.
• Use a “parking lot” for discussions that are too long or do not con-
cern the whole team. Keep the stand-up focused, finish it on time
and then anyone who needs to continue the parked discussions
can do so after the meeting is over. Anyone has the right to call
“time out”.
• Towards the end of the meeting, ask questions like: “Which story
are you going to finish next?” or “Do you have in front of you all
the information you need to make good decisions?”
The Hitchhiker’s Guide to Agile Coaching
. Refinement and planning
The Sprint planning and backlog refinement meetings are strongly fo-
cused on product questions. Your role as an agile coach is initially to
facilitate the meeting, helping the team and stakeholders have produc-
tive discussions and make good decisions about the product.
At the same time, you will show the ScrumMaster how to run a good
planning meeting. Over time, as the team learns what the planning
meeting is about, coaching the ScrumMaster will become your primary
focus point. Eventually you can step back and let the ScrumMaster take
the lead.
We would like to point out that an agile coach should not get involved in
product design questions. The problems inherent in advising and con-
sulting were discussed in sec. 2.2. You should help the team learn the
tools and methods they need (coaching and training) and help them
bring out information if necessary (facilitating). You can also provide
options if you receive a direct question, but it’s not your job to design
the product.
. Retrospectives
As we mentioned in the introductory chapter, this book is mostly about
coaching and not so much about the agile practices. However, good ret-
rospectives are so important that we have decided to make an exception
that proves the rule.
During a retrospective teams inspect and adapt both their methods and
the way they work together as a team. In an iterative development ap-
proach like Scrum, a team runs retrospectives at the end of each itera-
tion. In a flow based approach like Kanban, the cadence for feedback
meetings is defined by the policies agreed on by the team(s), like, for
instance, a bi-weekly Service Delivery Review.
Coaching your team
The primary reason for digging deeper into the topic of retrospectives
here is simply that they are a key characteristic of agile teams. An inef-
fective team that continuously improves will, in the end, beat an effec-
tive team that doesn’t. All truly agile organizations that we are aware of
take their retrospectives very seriously — sometimes almost to the point
of becoming religious about them. If you do only one thing as an agile
coach, make sure it’s getting the team to run effective retrospectives and
take responsibility for their own ways of working.
Secondly, retrospectives are key engagement point for agile coaches and
a great opportunity to use structured coaching conversations as dis-
cussed in sec. 3.5, sec. 3.6 and sec. 4.2. Thirdly, retrospectives are one of
the simpler ways of achieving continuous process improvement (CPI).
This in turn is explicitly mentioned in Agile Principle #12: “At regular in-
tervals, the team reflects on how to become more effective, then tunes
and adjusts its behavior accordingly.”
Safety first
Retrospectives are typically facilitated by a ScrumMaster or an agile
coach. The facilitator needs to establish a safe environment where
trust, transparency and openness allows each person to contribute
equally. While always remaining neutral in discussions, the facilitator
observes the room and the dynamics of the group, helping them reach
the goal.
It is important to understand that the retrospective is about learning and
looking ahead, not about blaming and looking back. Norman Kerth has
captured this intention perfectly in the the prime directive of retrospec-
tives:
The Hitchhiker’s Guide to Agile Coaching
“We understand and truly believe that
everyone did the best job they could,
given what they knew at the time, their
skills and abilities, the resources
available, and the situation at hand.”
— Norman Kerth
This quote emphasizes the fact that people learn through trial and er-
ror. There are complex situations that cannot be solved in any other way
than by actions or experiments. The outcome might be obvious in hind-
sight, where we always have 20/20 vision, but the prime directive helps
us remember that it wasn’t so simple at the time. The quote continues:
“At the end of a project everyone knows
so much more. Naturally we will discover
decisions and actions we wish we could
do over. This is wisdom to be celebrated,
not judgement used to embarrass.”
— Norman Kerth
We often start retrospectives by reminding the participants of the prime
directive.
Retrospective Formats
The format and structure of each retrospective may be different and
each team will typically have its own way of doing it. The key goal is
to make the meeting a team learning event, fostering change through
concrete actions.
One way to keep the retrospectives interesting is to have several different
kinds of retros: quick, standard and deep. Quick retrospectives are done
Coaching your team
ad hoc, in 5-10 minutes. The goal is to understand what is happening
and find a quick win. You can use powerful questions (see sec. 3.6) to
guide the discussion or the Cynefin model, Delta-Plus or Five Whys1 to
generate insights that you can later delve deeper into.
For standard retrospectives on a weekly or biweekly basis, you might use
the Starfish or Speedboat models2 to create suggestions, then dot vote
and form the top suggestion(s) into experiments, pilots or SMART Ac-
tions. As an alternative you could facilitate an open discussion using
either the 5D model (as explained in sec. 4.2), or the Karl Tomm model
(sec. 3.6), again making sure that you end up with actionable items.
It’s important to run deep retrospectives every now and then. They are of-
ten run by the book — “Agile Retrospectives” by Derby and Larsen (2006)
— and can take several hours. The Derby & Larsen model includes five
stages, which we will describe below. If the group is small enough to
hold a single coherent discussion, the 5D and Karl Tomm models can
again be very useful. In larger retrospectives, it is important to think
global and act local. In practice there is often a preparatory meeting or
team survey for collecting data. After that the data is collated centrally
and some basic analysis is carried out. The teams reconvene the follow-
ing day to generate insights and plan actions.
The five stages of a Retrospective
1. Set the stage
The aim here is to welcome people, make them feel comfortable, break
the ice and create a good, safe environment to get everyone involved in
the retrospective. It also helps to trigger people to start thinking about
the past sprint in a light way.
Some examples:
1 http://en.wikipedia.org/wiki/5_Whys
2 Look up Starfish or Speedboat on http://www.plans-for-retrospectives.com/.
The Hitchhiker’s Guide to Agile Coaching
• We can remind ourselves of our team values
• Remind everyone about the prime directive
• Ask powerful questions, e.g. if you had to represent the sprint as a
weather report, what would it be?
2. Gather data
Once we have people engaged, it’s time to take a step deeper into the de-
tails. You want to elicit as much information from the group as possible.
The facilitator can also bring data to the table for the team to discuss.
Examples:
• Look at some metrics, for example velocity, throughput, lead time
or quality
• Review the actions from the last retrospective
• 4L’s: what I Liked, Learned, Lacked, and Longed for
3. Generate insights
In this phase it about making sense of the collected details. We try to
distinguish symptoms from root causes and to discover patterns.
Possible approaches:
• 3H’s: what Helped, what Hindered, and what Hypotheses do we
have for improving?
• Timeline: plot the collected data on a timeline and identify pat-
terns
• Starfish: what should we Start, Stop, Continue, do More, do Less
of?
4. Decide what to do
Coaching your team
Once we have figured out what the problems are we need to find out
what to do about them so that we can prevent this from happening
again.
• Prioritize! Write SMART actions
• Goal - Action: Define a long term goal plus a concrete action for
next sprint
• Update agreements: Working Agreements, Definition of Done,
etc.
5. Closing
Since we may have had some heated discussions or shared some sensi-
tive personal details, we want to formally “close the box”. For the facili-
tator, it might also be good to ask for feedback from the team and thank
people for participating.
• Appreciations: ask people to share something they appreciate
about another team member.
• One word: state one word about the Sprint or the Retrospective.
• Return On Time Invested (ROTI): Ask people how much they
learned and improved in this retrospective.
Summary
In retrospectives, participants look at the process and discuss what
worked and what did not, trying to understand the reasons for each.
They then make hypotheses about how to address problems and define
actions that could solve them. To make this happen it is crucial to
clearly define the goal, who is responsible for achieving it, what is
the timeframe and which metrics will be used to verify the outcome.
The status of these action items should be checked at each following
retrospective.
The Hitchhiker’s Guide to Agile Coaching
It is important, from time to time, to leave room for new independent,
innovative ideas and experiments to improve the process.
Coaching your team
Chapter 7
Structured coaching
Some agile coaches have a fantastic memory. They remember exactly
where they left off and are able to pick up the trail where they left off.
Not all of us have this gift. In fact, we think that this gift is becoming
exceedingly rare. As you work on and off with lots of different teams,
it is very easy to lose your way or forget important details. Everyone,
including the memory geniuses and the multitaskers, can benefit from
some structure in their work.
The “Team Coaching Framework” (TCF) is our way of introducing that
structure to the work we do. For individual coaches, the coaching struc-
ture serves as a memory aid in intermittent engagements. In pairs or
small groups of coaches, the coaching structure forms a basis for collab-
oration. ScrumMasters, line managers and coaches can pool their ob-
servations, draw up hypotheses together, agree on the tools to be used
and then carry out actions designed to gently nudge the team towards
the same goal.
The structure also allows sharing experiences and ideas for improve-
ment with members of an organization or the larger professional com-
munity. In effect, you can take an anonymous coaching structure and
discuss the content with any other ScrumMaster or agile coach, regard-
less of whether they are involved with the same client or not. This allows
two or more coaches to learn from each other and enables more experi-
enced coaches to mentor a less experienced coach.
TCF is based on the Observe, Orient, Decide, Act (OODA) loop1
because it allows us to commit late. There are also goal-oriented
1 http://en.wikipedia.org/wiki/OODA_loop
Structured coaching
methods that follow the better known Plan, Do, Check, Act (PDCA)
cycle, but the PDCA cycle forces early commitment — we need to
plan up front, before we have enough facts. In the OODA loop we
can continuously add more observations as we go and change our
hypotheses and goals as necessary.
Figure 7.1: The OODA loop when coaching a team
This approach works especially well in complex environments. New ag-
ile teams tend to need help with their dysfunctions first. Only after they
have started working well enough together as a Scrum team does it make
sense to draw up goals and start working together towards them.
Note that there can be many different ways of introducing structure.
TCF is something that we have constructed over many years of successes
and failures. It’s not perfect, but it serves us well. Doubtless it will keep
evolving in the future.
The Hitchhiker’s Guide to Agile Coaching
. Overview
TCF relies on Coaching Tools. These are methods, techniques, work-
shop formats, etc. that an agile coach can introduce to change the be-
havior of a team. You can find a collection of coaching tools in the ag-
ile42 TCF Tool that we have developed2 .
The question arises when to use a specific tool, or, from another per-
spective, what tool to use in this specific context and situation? TCF
uses Coaching Cards to bind together a context and a couple of coach-
ing tools. The coaching card template, shown in fig. 7.2, is very simple.
It merely consists of four fields or headings. We sometimes use an A4
card for this, sometimes a brief text document. Several coaching cards
make up a Coaching Structure. The coaching structure can also con-
tain an introduction in the form of a description of the context and the
background situation.
We usually create new coaching cards when coaching teams, but since
there are situations that recur every so often, we have found it helpful to
create a small library of such cards that we can reuse.
The process of creating a coaching card or a coaching structure goes
generally from observations through hypotheses and goals to metrics
and finally, actions and coaching tools. Be prepared to jump back and
forth though, as you may unearth new information or come up with new
hypotheses. Next up, we will describe the general steps of creating a
coaching card and thereby a coaching structure.
. Make observations
A good coach always starts by observing the team or individual you are
coaching. Don’t be afraid to collaborate with others on collecting and
2 https://tcf.agile42.com/
Structured coaching
Figure 7.2: Coaching structure
sharing observations. Multiple people can throw in their observations
and, in doing so, build a better basis for the coaching structure.
An observation is some information that you have seen, remembered or
recorded, or received from a data system or from another person. Ob-
servations are facts: things either happened or they did not happen. As
such they are beyond dispute. People may have different observations
and different understandings of why things happened in certain ways,
but we will discuss that in the next section.
What to observe? Anything that seems significant to you, including the
elephant in the corner of the room. This would include:
• Team interactions — who is talking, who is not?
• Voice and intonation
The Hitchhiker’s Guide to Agile Coaching
• Postures and energy levels
• What do people talk about? What do they leave unsaid?
• What information do people look at? What do they ignore?
• Significant events originating inside or outside the team
• Who makes decisions? Is the team deferring to others? How are
decisions communicated?
When to observe? The best time is any time and all the time! In sec. 6, we
mentioned that different frameworks (Scrum, Kanban etc.) can give you
several different touchpoints with the team. Use all the opportunities
you get. Don’t forget to talk with stakeholders, specialists and neigh-
bouring teams.
How to collect information and observations is up to you. Many of
us write diaries or record our engagements and interactions some
other way. We use pen and paper, post-its, tablets, cellphone cameras,
pocket/system cameras, e-mail, etc. to collect data. Then we aggregate
the information into a reliable and secure collaboration environment
to share with colleagues and/or clients.
When observing, you should be aware of something called observation
bias, which means that the observer allows her own ideas to influence
what she observes or how she interprets the observations. In practice, it
means that people are more eager to take in facts that support their own
points of view. If you like knitting, you would subscribe to knitting mag-
azines, blogs and podcasts, and perhaps avoid those about car tuning or
technical physics.3
There are some tricks you can use to avoid observation bias. Try this:
• Avoid drawing any conclusions until you have a good amount of
observations
3 If you are a technical physicist who likes knitting and car tuning, let us know and
we will change this example.
Structured coaching
• Entertain multiple hypotheses and be prepared to change them
• Share and compare observations and conclusions with other
coaches
• Validate each and every hypothesis before you expend significant
energy on them
. Formulate a hypothesis
Now that you have some observations, you can start thinking about
what would explain the observed behaviors. Draw up an explanation
that is coherent and follows the observations. This is not always
straightforward and can require a lot of thinking, and support from
others. Many of the retrospective “Gather insights” methods work well
here, for example Five Whys, Fishbone or Force Field Analysis.
Always try to validate or invalidate your hypotheses before you commit
to it and start constructing a goal. Search for for additional information
that can strengthen or weaken it. The OODA loop allows you to dig up
more data and reformulate your hypotheses as needed. Make more ob-
servations and talk to people to get more information. The flip side is
that you shouldn’t get too attached to your hypotheses. Always be pre-
pared to throw them away and start from scratch.
Share observations and collaborate with other people as they can help
you come up with a wider range of hypotheses or help (in)validate a hy-
pothesis you have made. If you have access to a co-coach or mentor,
you can learn a lot by discussing your observations and hypotheses with
them. Again, remember that while facts are facts, different people may
have different interpretations and opinions of those facts.
Over time, you will learn to see patterns that are recurring within the
organization or across organizations. Experienced agile coaches have
seen many different situations and have learned what is likely to help in
each situation.
The Hitchhiker’s Guide to Agile Coaching
If you end up with several coherent hypotheses, you may want to create
a separate coaching card for each hypothesis and construct a coaching
structure from the cards. If they are aligned, you can just merge them
into the same document and call it your coaching structure.
. Identify a goal
From the hypothesis, it is usually quite easy to construct a goal. Given
the context and the hypothesis, what is the behavior we want to see in
the team? Try to phrase this as a transforming question or statement,
rather than a problem-solving one. (Please see sec. 4.2 for a description
of transforming questions.)
The goal can be lofty but should not be too abstract — rather than
“world peace and happiness,” you may want to aim for “ceasefire
agreement.” The reason for loftiness is that you may not actually reach
the goal before the coaching structure becomes irrelevant. As the team
improves along the chosen dimension, the Theory of Constraints by
Goldratt and Cox (2004) says that another dysfunction will eventually
pop up and become the most important bottleneck, which means that
you will need to create new coaching cards and structures as you go.
For the same reason, you should avoid making KPIs or incentives out of
these goals. The reason for having pragmatic goals is that abstract goals
can be very difficult to work towards — you may have to work along
several dimensions to get there and getting there can take too much
time.
. Define metrics
Metrics are integral to the idea of coaching cards and coaching struc-
tures. We define the metrics immediately after we have agreed on a
Structured coaching
goal, not after defining the actions, otherwise we run the risk of choos-
ing metrics that measure the actions we take, rather than the goals we
want to reach. We recommend that you use at least one lagging indica-
tor and a handful of leading indicators. And what are leading and lag-
ging indicators? Let’s approach this topic through an example.
The problem with most metrics is that they can be measured only after
we have an outcome and this may be too late. For example, many orga-
nizations measure the quality level of their software by making a release
candidate, then having a cadre of testers run regression tests. Because
manual testing is expensive, releases are done as seldom as possible and
the longer they go between releases, the longer the bugs go undetected.
The number of found and unfixed bugs in the software is a very common
lagging indicator. But by the time you get around to counting the bugs,
it is too late — the bugs are already in the software. You need to root
them out one by one and then test again to ensure they are gone and to
hunt for additional bugs hiding behind the first ones.
As you can imagine, it would be very useful to get some early indications
of what the quality level is likely to be. What should we look for? Here
are some ideas:
• The number of automated tests — two tests are more likely to find
a bug than one test
• Code coverage for automated tests — better code coverage means
that more bugs are likely to be found and fixed within the Sprints,
rather than in system testing
• Cyclomatic complexity — low CC indicates simpler code that is
easier to understand and, thereby, is less likely to have bugs
• Bug fixing rate — the more bugs we fix and the faster we fix them,
the fewer bugs we are likely to have in the product
These are what we call leading indicators. They are indirect measure-
ments of things that are associated with or correlated to what we want
The Hitchhiker’s Guide to Agile Coaching
to achieve. Leading indicators are also probabilistic — there is a chance
that each individual indicator is wrong, so we need to look at them in
aggregate. For example, even if the bug fixing rate looks good, the bug
introduction rate could be even higher.
We do not need to figure out the exact probabilities however. For our
purposes it is enough to use many different leading indicators and check
what they show as an aggregate. Leading indicators can sometimes also
help us understand alternative ways of achieving our goal.
You will find it comparatively easy to come up with lagging indicators
for the coaching card or coaching structure. However, leading indicators
often leave people stumped. To make matters worse, you need at least
three and preferably five or more of them. We would advise you to look
at what people do, what they discuss and the words they choose. We
sometimes use the “miracle question” to find ideas: assuming that the
problem was fixed overnight, how would you notice that things were
going the right way?
It is important to make all metrics very clear and actionable. Saying
“progress is updated regularly” is a good start, but not in itself suffi-
cient. Instead you could say “in the daily standup, the ScrumMaster
notes whether the Sprint board was up to date or not when the standup
started.”
. Choose the coaching tools
As mentioned in the overview, tools are methods, techniques, workshop
formats etc. that you can introduce to permanently change the behav-
ior of a team. Because of this very loose definition, the list of potential
coaching tools is very long — theoretically infinite.
Structured coaching
At agile42, we maintain a repository of 100+ central, well-defined tools4
that we expect all our coaches to learn and know. Several dozen of these
we use weekly if not daily and there are also many unlisted tools that we
use occasionally. We also sometimes combine components into ad-hoc,
hybrid tools or create new tools on the spur of the moment.
By observing and talking to other coaches, you may learn about new
tools that you can try out and make part of your toolbox. Other good
sources are blog posts and other books.
. Build a coaching structure
At this stage, you have all the information you need to make a coaching
card. Take a piece of paper or a text editor, write down the five headings
and fill in the rest. This is straight-forward manual labor that should
not pose any challenges. Don’t write a novel though, usually one or two
paragraphs will be more than sufficient.
If you have more than one coaching card, you can combine them into a
coaching structure. It is often sensible to merge the context descriptions
into a separate section at the start of the document:
1. Context
2. Coaching card A
1. Hypothesis
2. Goal
3. Metrics
4. Tools
3. Coaching card B
1. Hypothesis
4 https://tcf.agile42.com/
The Hitchhiker’s Guide to Agile Coaching
2. . . .
4. Coaching card C
1. . . .
. Follow up
If your hypothesis is wrong, this will usually become apparent at the lat-
est when you start applying the coaching tools. Be prepared to change
or even scrap the whole coaching card and make new observations to
refine your hypotheses.
It takes two weeks to build a habit. When the coaching tools are applied
consistently and regularly — on a daily or sprintly basis — you should
start seeing changes and improvements within the first sprint, and they
should be established after a couple of sprints.
You may need to swap out metrics as you go. Don’t be afraid of adding
or removing leading indicators. Because lagging indicators measure the
actual outcome, you may want to stick to them for longer periods of time
in order to build up long-term historical data.
As the team grows and the context changes, the coaching card even-
tually outlives its usefulness. As mentioned previously, the Theory of
Constraints by Goldratt and Cox (2004) states that every system has one
major bottleneck that constrains the whole system. As you remove or
reduce this bottleneck, another takes its place and working further on
the original bottleneck will not improve the system. There will come a
time when you must archive the current coaching structure and create
a new one.
Is it OK to share a coaching structure with the team? It depends. There’s
nothing that prevents you from sharing the structure, but you might get
unanticipated side effects. For example, it might affect the team’s be-
havior adversely or they might start gaming the metrics. On the upside,
Structured coaching
they could also take the desired change to heart and make it part of their
team goal. In general, however, we would advise keeping the coaching
structure as a tool for the eyes of the ScrumMaster and related stake-
holders only. This is because the team already has a mechanism for
improvement (the retrospective meeting) but the ScrumMaster doesn’t
really have anything.
If you really want to take something up with the team, you can kick off
a retrospective by telling them what you have observed. This is usually
a safe approach that is appreciated by the team. Before you start, re-
member to ask if they want your feedback, and respect their decision.
Also make sure that they understand that you are merely stating your
observations and not making any judgements. We find that most teams
appreciate the information, and are able to hold good and constructive
conversations about it.
. Case studies for TCF
In this section, we will present a number of case studies from which you
might be inspired. Each of the case studies is described just as we would
prepare a coaching card. This includes the observation (context), the
hypothesis, the goal, the metrics and the tools involved.
Coaching card: The hurried tests
Context:
At the review meeting the team demonstrates a product that has not
been fully tested. The tests are never ready. Often the tests are run on the
tester’s machine or on the developer’s laptop. They never describe the
test environment nor do they reflect the normal production situation.
There is no report on the tests run to verify the product. The Product
Owner seems to be okay with this and the ScrumMaster says that the
time is so short that “it is a miracle that they have time to test anything”.
The Hitchhiker’s Guide to Agile Coaching
Hypothesis:
We think the team does not fully understand the principles behind the
XP practices. Also it seems that there is considerable technical debt,
which is preventing the team from investing enough time in test au-
tomation and checking the quality of the delivered User Stores. We be-
lieve the lack of automated tests results in a fear of making mistakes.
Goal:
The Scrum Team is delivering with higher quality and faster than
before. Thanks to a Continuous Integration infrastructure, they have
developed a sense of safety which improves their focus on value and
performance in implementing new features. Automation is perceived
as having a great value by all the teams working on the software
platform, even those that are not yet using agile methods. The number
of defects after deployment have decreased dramatically and teams
have more time to deliver value.
Metrics:
Leading indicators:
• The Development Team operates a continuous integration (CI)
server on which the latest software is automatically deployed and
uses this for the Sprint Review demonstrations.
• There are automated build and test scripts within the CI server.
The team receives feedback on the quality of the code within min-
utes of check-in.
• For every User Story, the Development Team creates at least one
automated functional test.
• For every change in the code, the Development Team creates sev-
eral new unit tests.
Lagging indicators:
Structured coaching
• The Development Team is delivering consistently with higher
quality.
• The number of interruptions during the Sprint due to defects and
problems in production has decreased significantly if not disap-
peared completely.
• The code coverage level is high and increasing.
• The team is of the opinion that the Continuous Integration infra-
structure helps them develop faster and with higher quality.
Coaching Tools:
• TDD Workshop: Introduce TDD practices in a safe-to-fail envi-
ronment, so that the team not only understands the principles,
but can also apply them.
• Slow Lane: Pick one lower priority User Story for which the team
commits to “dot the I’s and cross the T’s”. . . which includes writ-
ing proper automated tests.
• Reproduce bugs: The team agrees to write tests first to reproduce
the most severe defects, and then fix the defects so that they can
immediately prove that the fix works. The tests are added to the
test suite in order to prevent regression.
Coaching card: Team reports to ScrumMaster
Context:
Team members report to the ScrumMaster in the daily standup. Except-
ing one or two senior engineers, the others do not ask or volunteer infor-
mation. The ScrumMaster takes a very controlling role in the meeting,
walking everyone through the three questions. When needed, he tells
the team what the managers are saying, doing and thinking about some
issues (“Manager N.N. called the customer yesterday and they will de-
cide. . . so don’t do anything before you get an e-mail.”). He is quick to
The Hitchhiker’s Guide to Agile Coaching
delegate work and ends the meeting by asking if everyone has “enough
work for today”. The ScrumMaster is a former Project Manager and took
his CSM around 8 months ago.
Hypotheses:
1) We think that the ScrumMaster has not internalized the mecha-
nisms that he is expected to use. He may have difficulties un-
derstanding the role and the dynamics between ScrumMaster and
Dev Team.
2) The ScrumMaster is used to being the hub of information between
“managers” and “coders” and is afraid of losing that position.
3) The team may be afraid of taking responsibility, and is happy to
hand it to the ScrumMaster.
4) There are indications that the managers see the dev teams as a
cost sink rather than a value source. That is, the managers are
more concerned about how the coders spend their time rather
than what they produce.
Goal:
The goal is to make the team members talk to each other rather than the
ScrumMaster and to plan their Sprints and workdays as a team.
Metrics:
Leading indicators:
• What percentage of questions in the standup are posed by the
ScrumMaster compared to team members?
• How often do the team members say “us” and “we” rather than
“me” and “I”?
Structured coaching
• How often does the ScrumMaster NOT ask if everyone has work
for the day?
• Ask the ScrumMaster to skip a daily standup. Silently observe the
meeting when the ScrumMaster is not present. Do the interac-
tions change?
Lagging indicators:
• All work items are pulled by team members, not pushed by the
ScrumMaster
• The ScrumMaster listens more than he speaks
• Managers interact with the team directly, not through the Scrum-
Master
Coaching Tools:
• Teach the ScrumMaster in the role.
• Role-model for the ScrumMaster. Offer to run the daily standups
a couple of times, then explain to him what you did and why.
• Gently move the ScrumMaster outside the circle in the daily
standup. If team members still turn towards the ScrumMaster,
move him all the way behind the speaking team member.
• Talk to the ScrumMaster to find out what stakeholders he is re-
porting to. Then systematically meet all stakeholders and work
with them to replace the reports with transparency.
• Help the team create a pull policy.
• In a safe setting, discuss with the ScrumMaster how he feels about
the changed responsibilities, then help the ScrumMaster choose
the right path forward.
• Educate managers on value-focus vs. cost-focus and the implica-
tions of each.
The Hitchhiker’s Guide to Agile Coaching
Coaching card: Weak collaboration in the team
Context:
The team opens up multiple Sprint Backlog items in parallel and have
problems finishing them by the end of the Sprint. There is no collec-
tive feeling of responsibility and people hold back on asking for help
because the other team members are all busy with their own tasks and
would not have time to help anyway.
This attitude, combined with the size of the stories, generates the behav-
iour that individuals feel overwhelmed and focus primarily on complet-
ing their own tasks. Those few who are done early polish their own work
or open new stories before helping others. Most of the time, it is too late
to get a story done by the end of the Sprint. Furthermore, as there are no
consequences from not delivering a story, the stories get dragged along
from Sprint to Sprint.
Hypothesis:
The team is more interested in being efficient as individuals, as opposed
to being effective as a team. This leads to many dysfunctions.
The team does not understand the meaning of commitment and does
not understand their collective responsibility. Therefore, they fail to
manage the risk and to control the process properly.
Goal:
The goal is to learn to appreciate team effectiveness. Helping each other
to move collectively forward is the first priority of the team.
Metrics:
Leading indicators
• Team members are actively seeking someone to pair with on tasks
(active collaboration), rather than waiting to ask for help until the
Structured coaching
Sprint is almost over and the PO and stakeholders are starting to
ask questions (passive).
• No more than three stories are in progress simultaneously.
• When you enter the room, you will regularly see team members
working in pairs or small groups on the same computer.
Lagging indicators
• The hit rate of the Sprint (number of stories delivered vs. planned)
stabilizes at 90% or above
• The variance of the velocity goes down. In other words, the veloc-
ity curve becomes smoother and exhibits less zig-zagging than it
does now.
• The velocity keeps increasing over time. The trend can be slow but
is positive.
Coaching Tools:
• Create a Skills Matrix and use it to promote collaboration for im-
proved knowledge transfer
• Use the Working Agreement to limit the number of stories started
during a Sprint. In case of exceptions, note them down and dis-
cuss them at the Retrospective. This will encourage collaboration
on a single story (including testing very early).
• Add pair work on tasks to the Working Agreement for better qual-
ity (building quality in).
• Agree that each task or User Story needs to be the responsibility of
two people rather than one and make sure this can be visualized
on the board.
• Highlight tasks dependencies at the planning meeting (use
sequential numbering and flow sequence).
The Hitchhiker’s Guide to Agile Coaching
• Write down the Sprint Goal as a marketing motto and hang it in
the team room. Challenge the team to pull stories and evaluate
multiple options to achieve the Sprint Goal.
Structured coaching
Chapter 8
Challenges
As agile coaches, we often face challenges — one could argue that that
is what the job is about. But even assuming that you yourself feel com-
fortable in your role as an agile coach and teacher and have the tools
and structures you need to do your job, you will often encounter other
people who are wary, scared or doubtful or simply have a different un-
derstanding of what has happened and what needs to be done next.
This chapter explores a couple of models that have helped us under-
stand why people react as they do to change and specifically to agile
adoptions and transformations. We will also look at some techniques
for lowering the resistance to change: constructing change proposals,
building consensus, anchoring ideas, running safe-to-fail experiments
and getting buy-in.
We will open this chapter by discussing two thinking models that are
useful for understanding why people resist change: the Lizard Brain and
SCARF.
. The lizard brain
All humans have a hard-wired notion that change is dangerous.
Over-reacting pays off in the jungle or on the veldt — those who scram
away without thinking have a good chance of seeing their offspring
grow up. And so we have evolved a mechanism for getting ourselves
out of dangerous situations quickly and with a minimum of thinking.
In fact, thinking is high on the list of things not to do. Those who start
analyzing whether the waving grass really does conceal a man-eating
Challenges
Figure 8.1: Coaching challenges
saber-tooth tiger are more likely to end up analyzing the tiger from the
inside.
Change and fear are connected on quite a deep level in our brains. As
shown in fig. 8.2, when subjected to a threat of some kind — a threat
to their habits, territory, or very existence — people go into panic mode
and will either fight, flee or freeze. Like all models, this model is incom-
plete: it’s a simplification of reality and full of holes and errors. However,
we have found it very useful. It explains some of the reasons why people
fear change, and also helps make sense of why people react as they do
in the face of change.
The name lizard brain comes from fact that panic reactions are handled
(mostly) by a part of the brain called the amygdala, which can be found
just above the brainstem in all vertebrates. We share the amygdala with
The Hitchhiker’s Guide to Agile Coaching
Figure 8.2: The lizard brain — primal reactions
all mammals, birds and — as the name implies — with lizards.
When a vertebrate animal is subjected to a strong threat, the amygdala
causes an immediate primal reaction. This is not a simple reflexive ac-
tion: reflexes typically only yank a couple of muscles. But the primal
reaction is still well below rational thought. The amygdala in fact pre-
vents rational thought by subverting and disconnecting the upper brain
functions. People suffering from primal fear are not only unreceptive to
rational arguments but actually incapable of rational thought.
Interestingly enough for an agile coach, similar patterns can be found
in prolonged situations of uncertainty, fear or anxiety, as is the case in
e.g. agile deployments. Case in point: “We will pilot Scrum in the orga-
nization and your team is the best candidate!” In other words: “We say
you Scrum.” Ouch! How do people interpret this?
Challenges
Habits — Most people are initially vary of agile methods as they need to
change their own working habits. Harsh corporate environments
cause people to form their own patterns and strategies for survival
and they may be very suspicious of “team work” and “collabora-
tion”.
For example, we regularly meet engineers who have learned not to
ever give honest estimates. They have formed habits of hiding the
actual state of their work, giving vague reports like “90% done but
we have some technical problems” or having a perpetual “coding”
task that they move back and forth on the task board.1 Others feel
that the daily standups are stupid —– it takes weeks to implement
this feature anyway so what need is there to report every day?
Territory — Scrum can also be a threat to people’s territory. Former
project managers that are now relabeled ScrumMasters or Prod-
uct Owners suddenly find that they have lost a large portion of
their former decision power. Deep specialists are forced to open
up their work to the scrutiny of others. Software architects find
that teams are doing unauthorized changes directly into the code
base.
Existence — Some people see agile methods as a threat to their exis-
tence in the organization. These include line managers whose
teams are suddenly self-organizing; testers who learn that all test-
ing will soon be automated; or specialists who learn that they are
in fact not so special.
To the vast majority that is not directly involved in driving the agile tran-
sition, Scrum is a cause of anxiety or outright fear. How do they respond?
1 When we see a task named “coding”, this gives us the signal that the team is tracking
how they spend time rather than the outcome they produce.
The Hitchhiker’s Guide to Agile Coaching
Fight — Direct and indirect attacks. Badmouthing. Spreading fear, un-
certainty and doubt. Passive-aggressive behavior in the Scrum
meetings. “Here, let me help you with that. . . oops, so sorry.”
Appearing to play along for a while, then a sudden explosion of
this-will-never-work. And so on —– the examples are countless.
Flee — Some people grab the opportunity and move to other teams or
escape the company altogether. The decision to flee is often made
early, sometimes as soon as the news has broken, but it can take
months for the fallout to appear. By the time people start under-
standing and accepting the new methods, they have already made
new commitments and signed new contracts, and the decision is
no longer reversible.
Freeze — A surprisingly large number of people seem to lose their drive
and become unable to produce anything of value for months on
end. They are tapping on the keyboard and moving the mouse but
nothing real comes out.
In almost every team we work with, we find examples of fighting, freez-
ing and sometimes, but luckily very seldom, also of fleeing. This model
has helped us make sense of why people behave irrationally during an
agile deployment. As an agile coach, it is your job to navigate people
past this stage of anxiety. The main problem is how to do that when
people are not thinking rationally.
You will need to start by identifying and addressing their concerns,
showing that the change is happening on their own terms. We have
noticed that it is good to have a brief meet and greet with the team
before the coaching starts. This could happen in conjunction with
some team training or as a stand-alone meeting between 20 and 45
minutes in length. During the meet and greet we go through the
upcoming coaching schedule, point out that we are here to support
Challenges
and guide the team, and underline that things will happen on their own
terms.
With especially afflicted persons, we have also had good luck with in-
formal one-on-one coffee discussions, alleviating their fears and giving
ideas for how to move forward. Sometimes the best approach is to just
listen and make it clear that you have heard their concerns. In the end,
agile adoption is about teamification and while you may be able to help
people find their new positions in the team, there are always going to be
people who need to make some hard decisions for themselves.
. SCARF
Because agile methods introduce new patterns of communication and
decision power, this will obviously have an impact on the existing social
networks and patterns. SCARF is a model for understanding the drivers
of human social behavior. In this context it can help explain some of the
fears and dislikes people show when subjected to an “agile adoption” or
“agile deployment”. Although the model itself is well described by David
Rock (2008), we will recount the key points here.
Similar to the lizard brain model that we covered previously, SCARF de-
scribes low-level threat or reward scenarios in the amygdala: people un-
consciously try to minimize threats and maximize rewards. We do this
by tagging stimuli as either “good” (associated with rewards) or “bad”
(associated with threats) and try to respectively approach or avoid such
stimuli. The model, shown in fig. 8.3, contains five different social inter-
action domains: Status, Certainty, Autonomy, Relatedness and Fairness.
We will describe them next.
Status — Your importance relative to other people, pecking order, se-
niority. High status is linked to lower stress, better health and a
longer life. Being subjected to pecking and mobbing activates the
The Hitchhiker’s Guide to Agile Coaching
Figure 8.3: SCARF
same areas as physical pain, and even the perception of a reduc-
tion in status can cause a strong threat reaction. People may de-
fend irrational ideas in order to avoid the threat and pain associ-
ated with a loss of status.
For example, we often see organizations where projects are
strongly linked to the Project Manager or PO, to the point
that people talk about “Pete’s project” and “Linda’s project”.
Reprioritizing Linda’s project over Pete’s may cause a perceived
increase or loss in status, triggering gloating from one and
complaints from the other. Fighting to get a position on the most
strategically important project becomes more important than
collaborating on getting all projects done.
Challenges
Status is decreased when people receive unwanted instructions or
technical advice, or are subjected to public shaming or criticism.
Public acknowledgement, recognition, promotions etc. increase
status. Working on a well-regarded team can also increase status.
Certainty — Your perceived ability to predict the future. This is impor-
tant because the brain works as a continuous pattern-matching
machine. Your senses continuously give feedback that your brain
filters and matches with known patterns, and adjusts your actions
accordingly. If something feels out of kilter — e.g. your foot slides
sideways when walking — your brain immediately refocuses on
the urgent issue. This immediate need takes attention away from
your long-term goal. Continuous uncertainty, such as not know-
ing your role in this new agile process, can disrupt people’s ability
to produce results.
Because the mere perception of predictability is what counts, peo-
ple are happy to believe in project plans even if they rationally
know that the software project failure rate is immense. People are
also happy to engage in depth-first problem solving even if width-
first problem solving (e.g. the wisdom of crowds) would result in
better solutions.
Conflicting visions, overly large goals, time-limited job contracts
and organizational changes can increase the uncertainty and
make people feel concerned. Clear objectives, reachable goals,
credible plans and solid careers increase certainty.
Autonomy — Your sense of being in control of events. This is some-
what similar to Certainty, but reflects your own ability to control
the future rather than your ability to predict it.
Micro-management reduces people’s autonomy, which can
be very debilitating. Allowing people to arrange their working
The Hitchhiker’s Guide to Agile Coaching
environment and choose their tools, pick their work tasks,
choose their working hours etc. can have a positive effect.
Working in a team also reduces the autonomy because the indi-
viduals have less say. However, there are increases in status, cer-
tainty, relatedness and fairness that makes the sum total positive.
Relatedness — Your sense of being safe with others, of being
surrounded by “friends” rather than “foes”. This likely stems from
humanoids having lived in tribes and flocks for hundreds of
thousands of years, where strangers are potential competitors
and likely to cause trouble.
The need for relatedness is a good mechanism for team-building.
Having a strong team identity helps, as does informal meetings,
water-cooler discussions as well as mentor-style relationships.
In companies with tall hierarchies, long budgeting cycles and
complicated incentive systems people can start perceiving other
groups as foes. This leads to withholding of information, internal
politics and eventually to siloing. Specialists who work alone or
are often shifted from project to project can also suffer from low
relatedness.
Sharing food, alcohol or other mild stimulants help people relate,
thereby reducing this threat. For example, in many cultures it is
traditional to serve vodka, coffee or tea during important meet-
ings because this helps build a kind of temporary relatedness.
Fairness — Your perception of exchanges being equitable. Unfairness
is linked to intense negative emotions such as disgust, which
makes this a surprisingly strong dimension of human social
interaction. It can drive people to e.g. spend significant spare
time on building open-source software, take unpaid leave to
volunteer for the Red Cross or throw themselves in front of
advancing tanks.
Challenges
There are many things that can ruin the sense of fairness, includ-
ing internal politics and unclear strategic decisions. We also see
problems related to pay discrepancies, individual incentives and
fake rewards i.e. “Employee of the Month” -style awards. For ex-
ample, a certain company moved their headquarters to the coun-
tryside far outside a major metropolitan area. Thousands of em-
ployees were forced to take long commutes and were not amused
when they noticed that most of the executives lived in an upscale
residential area nearby.
Fairness can be increased by improving transparency and com-
munication, by establishing clear expectations, and by having the
same rules apply to everyone including top managers.
In an agile transition, the SCARF model can help us understand why
different people react in different ways. Traditional command and con-
trol organizations are often “business monarchies” or “technical monar-
chies” which means that managers or technical leads are accustomed to
receive the rewards: they enjoy higher status, they provide directions
and action plans, they decide who does what, they control the budget,
they decide who is in the team and who is not and they can choose to
share or withhold information.
The team, on the other hand, is under threat. Their status comes from
how well they can serve the manager and they depend on the manager
to provide plans for the future. They have little autonomy, receiving de-
signs and action requests from elsewhere. Furthermore, they are often
working on individual goals, and things like stack ranking and large dis-
crepancies in pay undermine their sense of relatedness and fairness.
In an agile organization, the most important thing is to ensure that the
team can work as effectively as possible. The team receives the rewards,
and the managers are there to facilitate and support. This change is both
rapid and disruptive, and it may come as a shock to many managers and
The Hitchhiker’s Guide to Agile Coaching
technical leaders. Often, they try to retain their rewards and avoid the
threats by making agile go away.
. Influencing behavior
We have found the two models mentioned above very useful in under-
standing why people resent the new and scary agile methods. Often the
best approach is to sit down over lunch or a cup of coffee and make it
clear that while roles are going to change, nobody is getting fired. The
people involved will also be in control of the process (after an initial
boot-strapping period) and you will be there to help people become
proficient.
As an aside, it can help to take a so-called “yes, and” approach and
build on top of suggestions posed by the organization. Using Microsoft
Project for backlog management? Yes — and let’s see if it has a team
board view. Full traceability from requirements to test cases? Yes — and
this means that we should ensure that the requirements are testable
(training, workshops, hands-on coaching) and that the traceability
report can be generated automatically.
The second half of this chapter covers a couple of more structured ap-
proaches. The first approach is to collect data. If the data supports your
hypothesis, you have a case for pushing a potential solution. Otherwise,
drop it and do something else. The second approach is to design and
carry out safe-to-fail experiments, in a way that is transparent and in-
volves as many people as possible in the process. This lowers the friction
and resistance.
Challenges
. Building your case on metrics
“All opinions are not equal. Some are a
very great deal more robust, sophisticated
and well supported in logic and argument
than others.”
— Douglas Adams
Many problems are complex, and the more people they involve, the
more complex they tend to be. Formally, a complex problem has causal
chains that are partially obscured, multi-linked or cyclical. More prag-
matically, a complex situation can have several likely causes and the so-
lutions may have unintended side-effects. In a complex situation people
can argue for a very long time about the nature of the problem. People
may even disagree about whether the problem actually exists or not.
The great quote by Douglas Adams that we chose to open this chapter
with states that opinions should be founded in logic and argument. We
would like to add to the list and underline the importance of metrics.
Metrics are the foundation of empirical process control. While opinions
can be hashed and rehashed until the cows come home, it is difficult to
argue against hard data.
One approach for resolving complex situations is to collect data and
quantify the problems. Make a tick on a piece of paper every time some-
thing occurs or count the bright red post-its on the board at the end
of the sprint. After you have collected enough data, analyze it to see
whether the issue is large enough to make into a business case. How
many times does this happen per sprint? How much time do we lose
and how much does that cost us? How long does it take for our proposed
solution to pay itself back?
The Hitchhiker’s Guide to Agile Coaching
All things can’t be expressed as money, though. One team member was
complaining about people entering the big open office area and ask-
ing him for directions to other teams and individuals in the area. We
asked this team member to note on a piece of paper whenever he felt
disturbed. At the end of the sprint, he had to his own surprise barely
collected three such incidents. This put the problem into perspective,
and he agreed that the issue was smaller than he thought. Even so, the
team ideated and executed a couple of solutions, including pasting an
area map and a “do not disturb” sign on the wall beside his desk.
. Safe-to-fail experiments
There are several different mechanisms you can rely on in order to lower
the threshold and reduce the risk for trying out new things. One such
mechanism is to create change proposals collaboratively, so that you get
more ideas for proposals and the chosen proposal gets a wide buy-in.
We touched on this approach in the chapter on “Structured Coaching”
(sec. 7) and it is applicable to all kinds of proposals.
Another mechanism is the approach known as “Which Arm?” The name
comes from the question you are asked when donating blood: which
arm should we draw the blood from? The end result will be the same
in any case — you will lose half a liter of blood — but it gives you the
perception of being in control.
How does “Which Arm?” relate to experiments? Well, many compa-
nies have a static perspective on organization. They believe that the
current setup is stable and that any change requires energy. The more
agile approach is to assume that change is inevitable: the question is
in which direction to guide it. In the context of organizational change,
avoid asking “Shall we try this particular proposal?” and instead try ask-
ing “Which of these proposals are we going to try next?” This obviously
requires more than one proposal and a method for comparing them.
Challenges
A third mechanism is to frame the proposal as limited time pilot or ex-
periment. We try it out for a sprint or two and assess whether it is head-
ing towards success or failure. If it looks like it’s going to fail, we cut it
short and revert to our previous process. If it looks like it’s succeeding,
we let it run until we are sure, then embed the new method into our ways
of working.
It turns out that for a proposal to be a safe-to-fail experiment it is es-
sential to have a way of dampening the side effects, a kind of rollback
plan. The rollback can be time consuming and expensive but it must
be doable. You will also need to list some early indicators of impending
failure.
For many years now, we have used the Cynefin complexity model for
sensemaking by Snowden and Boone (2007). The model has many
implications, for example, that organizations are complex and that
planning and executing a change program is futile. Organizational
change can however be directed. Dave Snowden has created a compact
template for defining safe-to-fail experiments aligned with the Cynefin
model. It helps ensure that the proposals are coherent and motivated
but also that you are watching for early signs of failure and have a solid,
clear roll-back plan.
Snowden advocates running several experiments in parallel with differ-
ent parameters and ensuring that at least one is oblique and another is
naïve. An oblique experiment does not attack the root cause but rather
focuses on one or more of the symptoms. This can lead to valuable
learning and can be a good approach for problems that look intractable.
Naïve experiments are set up without much analysis or even thought.
Usually this would involve throwing money (or consultants. . . but we
are repeating ourselves!) at the problem to make it go away.
Have you considered hiring a hex doctor to exorcise the bugs from your
code lately? No? It may be a stupid idea, but it is also a valid experiment
that is both oblique and naïve. There is no logical analysis behind it and
The Hitchhiker’s Guide to Agile Coaching
it focuses directly on the issue (bugs) rather than underlying root causes
(you are not preventing bugs effectively enough). Experience tells us
that this particular experiment is unlikely to work as planned. . . but in
some organizations it could be a useful learning experience with bene-
ficial side effects.
Using all three mechanisms mentioned above, and working on a Kan-
ban board, it is possible to evolve an organization by probing, sensing
and adapting.
. Try this
1) At the end of sec. 8.2 we described a traditional command-and-
control organization using the SCARF model. What would an agile
organization look like? Take a moment with a like-minded col-
league/friend and try to build a SCARF case for a hypothetical
hyper-agile organization. Then compare the two SCARF cases.
What do you see? What could you do to alleviate the issues?
2) Take a moment to refresh your memory on the concepts of coach-
ing cards and coaching structures in sec. 7. How do they differ
from the safe-to-fail experiments outlined in this chapter?
Challenges
References
“So long, and thanks for all the fish.”
— The Dolphins in The Hitchhiker’s Guide
to the Galaxy (and the agile coaches)
Adkins, L., 2010. Coaching agile teams: A companion for scrum masters,
agile coaches and project managers in transition. Addison-Wesley.
Agile Manifesto, 2001.
Cooperrider, D.L., Whitney, D., 2005. Appreciative inquiry - a positive
revolution in change. Berrett-Koehler Publishers.
Derby, E., Larsen, D., 2006. Agile retrospectives: Making good teams
great. The Pragmatic Programmers.
Goldratt, E.M., Cox, J., 2004. The goal: A process of ongoing improve-
ment, 20th anniversary. ed. North River Press.
Katzenbach, J., Smith, D., 1993. The wisdom of teams. Harvard Business
Review Press.
Kimsey-House, H., Kimsey-House, K., Sandahl, P., Whitworth, L., 2011.
Co-active coaching. Nicholas Brealey Publishing.
Ringelmann, M., 1913. Recherches sur les moteurs animés: Travail de
Challenges
l’homme. Annales de l’Institut National Agronomique, 2me série 12, 1–
40.
Rock, D., 2008. SCARF: A brain-based model for collaborating with and
influencing others. NeuroLeadership Journal 1, 44–52.
Satir, V., 1964. Conjoint family therapy. Science; Behavior Books, Palo
Alto, CA.
Sheridan, R., 2013. Joy, inc.: how we built a workplace people love. Pen-
guin.
Snowden, D.J., Boone, M.E., 2007. A leader’s framework for decision
making. Harvard Business Review.
Stelter, R., Hansen, S.E., Møller, L., Holmgren, A., Rosenkvist, G.,
Hansen-Skovmoes, P., 2005. Coaching - læring og udvikling. Dansk
Psykologisk Forlag.
Storch, J., Søholm, T.M., Juhl, A., Dahl, K., Molly, A., 2006. Ledelses-
baseret coaching. Børsens Forlag.
Tuckman, B.W., 1965. Developmental sequence in small groups. Psy-
chological Bulletin 63, 384–399.
Weick, K.E., Roberts, K.H., 1993. Collective mind in organizations:
Heedful interrelating on flight decks. Administrative Science Quarterly
38, 357–381.
West, M.A., 2012. Effective teamwork: Practical lessons from organiza-
tional research. John Wiley & Sons.
The Hitchhiker’s Guide to Agile Coaching
9 783962 430030