Managing organisational design

Managing
organisational design

Managing organisational
design

Produced and published by:
Collinson Grant
33 St James’s Square
London
SW1Y 4JS
ISBN 978 0 9564337 5 6
Copyright © November 2012
Collinson Grant
Ryecroft Aviary Road Manchester M28 2WF
Telephone: +44 161 703 5600
E-mail: postmaster@collinsongrant.com
All rights reserved. No part of this book may be reproduced in any form or by any means without
permission in writing from Collinson Grant. The publisher makes no representation, express or
implied, about the accuracy of the information contained in this book. It cannot accept any legal
responsibility or liability for any error or omission that may be made.
Printed in the United Kingdom using paper produced from sustainable sources.
Design, typesetting and production – Centrix Q2 Limited www.cq2.co.uk

Managing organisational
design
Restructuring large businesses has been at the heart of Collinson Grant's
activities since the early 1980s. We have helped to get better returns from assets,
to increase operating results and to improve the competitiveness of companies
in the United Kingdom, in mainland Europe and in the United States. Our work
as management consultants focuses on costs, organisation and people. We
use this simple framework to manage complex assignments – often with an
international dimension – and to support managers on smaller, more focused
projects. Supporting clients in responding to changed circumstances, in seizing
opportunities and in strengthening their businesses is a constant feature of our
work. The notes at the back summarise what we do.
www.collinsongrant.com

Contents
Section Subject Page
1 Fundamental concepts 1
1.1 A belief in simplicity 1
1.2 The managerial structure 1
1.3 The corporate structure 2
1.4 Control and lean principles 3
1.5 Models of corporate devolution 4

2 Basic managerial structures 7
2.1 Information and data requirements 7
2.2 Managers 8
2.3 Layers and spans 10
2.4 Levels of work 17
2.5 Job design 20
2.6 Line and function 20
3 The business model 29
3.1 The starting point for corporate organisational design 29
3.2 Mission, policies and aims 30
3.3 Strategy 30
3.4 Value chain - process and job design 31
3.5 Configuration of structure 32
3.6 Accountability and measures of performance 33
3.7 Balance in all things 34
4 Corporate structures 39
4.1 Follow the business model 39
4.2 Integrated and devolved structures 39
4.3 Integrated organisations 40
4.4 Devolved structures 42
4.5 Profit centres 46
4.6 Support functions 48
5 The corporate centre 53
5.1 Corporate parenting 53
5.2 Structure and operations of the centre 54
5.3 A model for the function of the group centre 55

6 Matrix structures 59
6.1 Three degrees of complexity 59
6.2 Loose matrix structures 60
6.3 Strong matrix structures 64
6.4 The classic matrix 68
6.5 The matrix environment 71
7 Control and structure 77
7.1 The property of control 77
7.2 Accountability and responsibility 77
7.3 Business reporting 78
8 Shared services and outsourcing 83
8.1 Shared service centres 83
8.2 Do the right conditions exist? 84
8.3 Outsourcing 86
9 How lean structures keep down cost 91
9.1 Good design is naturally lean 91
9.2 Mapping the functional alignment of the population 92
9.3 Core and support 97
9.4 Duplication 99
10 The inter-trading organisational model 103
10.1 The inter-trading challenge 103
10.2 The inter-trading business model 104
10.3 Resource allocation and the value chain 105
10.4 Making inter-trading work 106
10.1 Transfer costing 108
10.2 International inter-trading 111
10.3 Terms of inter-trading 112
10.4 Strategy and structure 114
10.5 Reviewing the value chain 115
10.6 Contracting out 119

11 The transfer costing organisational model 123
11.1 Approach 123
11.2 A tough regime for managers 123
11.3 Organisational design 124
11.4 Accounting for transfer costing 125
11.5 Making transfer costing work 126
11.6 Transfer costs and transfer value 128
11.7 Accountability and control 129
11.8 A transfer costing universe 130
11.9 Role of the centre 130
11.10 Transfer costing – a case study 131
11.11 Managing cost reduction 133
12 Organisation of branch networks 137
12.1 Models 137
12.2 Authority 140
12.3 Authority to set prices 140
12.4 Authority over credit and customers 144
12.5 Authority over procurement 147
12.6 Authority over inventory policy 150
12.7 Control of resources, costs and operations 152
12.8 Authority to allocate and invest capital 155
13 A word in conclusion 159
14 Collinson Grant 163
14.1 Skills – the sort of work we do 163
15 Notes on quotations 167

This book is about the basics of good organisational structure. It is based on
a belief that all businesses and institutions should default to the simplest
design possible.
All too often, extreme concepts of organisation are promoted to popularise
books on management. Recently published volumes have suggested that the
structures of companies should have no boundaries, that they should be
without a centre (bagel-shaped), or that there should be no top or bottom.
A well-known management guru, who surely should know better, wrote in
a respected journal that ‘management is a silly abstraction whose workings
cannot be illustrated by charts that pretend to show who reports to whom’.
He goes on to say that ‘humans behave like bees and have an innate sense of
what it is right to do, and if left alone will just do it.’
We choose to disagree with commentators who have their heads in the
clouds. Have any of these authors, we wonder, interviewed the managers
and the staff in an organisation where each colleague gives a different
version of how a process is meant to work, about who owns its performance,
and who in the department has authority to instruct whom? It is not unusual
to come face to face with inconsistences of these sorts. We have found that it
is all too common when the structural sinews of organisation are neglected
for long periods.
Collinson Grant’s work constantly demonstrates that clarity and robustness
in the design of organisational structures, and of the jobs within them, is a
source of strength where it exists, and a hazard to corporate health where
it does not. All things in nature have structure, and none is more highly
organised than the allocation of tasks in a hive. If only we were as good at it
as the bees!
That true sage of management practice, Peter Drucker, first began to describe
the rationale for certain generic forms of organisation back in the middle of
the 20th century. Following his lead, it has since been possible for managers
to think systematically and analytically about frameworks for the exercise
of authority, how people are led, processes managed, and managers held to
Introduction

account. Over time some general truths have emerged. Although common
mistakes have been recognised, they continue to be repeated because many
managers take so little interest in organisational design.
It is true that as organisations become increasingly complex it gets harder
to be an effective manager. But structural dysfunction is a substantial
contributor to most business catastrophes, usually with problems of
transparency and robustness of accountability at its heart. In the banking
and investment management industry in recent times, the formality of
line reporting was largely abandoned. Accountability had been jettisoned
in favour of individualised targets with rewards. Without much by way of
a management process, bankers who were supposed to behave like bees
managed to kick over the hive!
Those shooting star technology companies that keep outgrowing their
formal structures may need brief periods of creative anarchy in which their
worker bees do great things by just following their instincts. But is that really
the case for long? The reality is that leaders like Steve Jobs and Bill Gates
became successful visionaries partly because they had masterly authority
over their organisations, setting firm boundaries and expectations, often to
the point of being freakish about it. And we can be sure that finely tuned
accountability for company time and money, for objectives, and compliance
with organisational structure and process are not being ignored at the profit-
conscious wonders of the age such as Amazon and Google.
Finally, at the outset of this little book let us suggest that all who tread here
should heed two warnings. First, designing and changing organisational
structure is more politically charged than any other undertaking in the
world of business. It generates more resistance, runs up against more
blockers, and gives rise to more perceived threats to the comfort of a greater
number of people than any other type of change that can happen in an
organisation. This fact underlines the importance of studying structure to
get organisational design right.
Second, be aware that standards in organisational design are very low. Few
managers do it well. Many never make a serious attempt to be objective,
systematic or analytical. And any change is more often by drift rather than
intent. Must do better!

Managing organisational design

1
1.1 A belief in simplicity
Making profit is the primary purpose of a business in the private sector.
Therefore,howandwhereprofitistobemanagedandmeasuredisaprincipal
factor in organisational design. This leads to the underlying concept of the
profit centre – a self-contained, relatively autonomous unit whose leader,
whether a Chief Executive or a Depot Manager, has to achieve a budgeted
net profit. Profit centres are fundamental building blocks of organisational
design, clarifying where profit is measured, who is accountable for it, and
who manages cost. They may encompass an entire enterprise or be one of
many small units into which a business is subdivided.
Profit centres give the person in charge of each trading entity the authority to
set prices and to direct marketing and selling activity. This simple approach
encourages entrepreneurial behaviour and allows maximum responsiveness
to local markets. It also positions decision making close to the customer and
makes it easier to link managerial incentives directly to profit.
But what if they lead to the creation of a patchwork of fiercely independent
territories which compete with each other to the detriment of the business as
awhole? Otherdownsiderisksinclude:ineffectiveuseofthecompany’stotal
resources; minimal exploitation of economies of scale; poor and sometimes
misguided direction; and minimal sharing of ideas and experience.
Deviations from the profit centre model are usually intended to mitigate
one or more of these risks. The division into customer-facing and support
functions, perhaps dubbed ‘Commercial’ and ‘Operational’, is only one
example. A hybrid organisation, based on a matrix structure, is another.
At Collinson Grant we try constantly to keep our feet on the ground. We
assume that managers and the staff in an established enterprise want to
understand how the business works best, and that such study merits more
than a passing thought.
We think successful businesses have managers who are clear about what
they have to do and about the responsibilities of others; they know to whom
everyone reports and who in turn reports to them; and these facts are widely
communicated.
1 Fundamental concepts

2
We feel comfortable with clarity about who has to make decisions and is
then accountable for outcomes. We favour measuring the performance of
individual managers and of the whole business and then connecting the two.
We firmly believe that the structure of an enterprise greatly influences what
it costs to run, and that the design of structure will affect everything that a
business or institution attempts to do.
For us an organisational structure is a framework with a boundary. It must
haveatopandabottom.Therewillbeacentralpointatthetopwhereultimate
authority lies, but it will delegate as scale and complexity require. We think
that to believe differently is to deny the responsibility of the organisation’s
leadership for setting a pattern in which managers manage with economy of
resources, efficiency of process and effectiveness of outcome.
It is not necessary to change. Survival is not mandatory.
W Edwards Deming
1.2 The managerial structure
Section 2 of this book is about the fundamentals: configuring structure in
the components of an organisation such as a department, using generic
guidelines on a small scale. The concern here is with individuals, groups or
functions in a single organisation. We refer to these as managerial structures
even though some of the principles can be readily scaled up and applied to
large, corporate models.
We emphasise the need to accommodate the staff and managers within a
sound structural framework according to how many people there are, how
they function and relate to each other to make processes work well, and how
accountability is put in place. Particular concerns are layers of hierarchy and
spans of control, of controlling costs by restricting the number of managers
to what is strictly necessary, and of the relationships between operational
and support staff.
The study and design of a managerial structure requires the application of
some immutable principles, for which there are simple tools and techniques
that help assess what is appropriate and shape the configuration accordingly.
“ ”

3
1.3 The corporate structure
In Sections 3, 4 and 5 we discuss the design of structures at the level of
the corporate enterprise. We begin by acknowledging the importance of
the business model in influencing the structure given to business units
and profit centres; and, in really large organisations, to divisions, sectors,
territorial entities and groups of companies.
Ultimately a corporate structure has to work properly in terms of how
it assigns responsibilities and accountabilities to the individual senior
managers running each part of it. So the principles for designing managerial
structures, discussed in Section 2, still apply. But our key considerations
here focus much more on the functions and purposes of business units and
less on individual job-holder accountability.
Design of the corporate structure is more crucial to the success of the
enterprise than design of the managerial structure. It has a more direct
impact on growth and how the business goes to market, organises supply
and so on. It is, therefore, strongly influenced by business policy and strategy.
Tools and techniques are needed to design managerial structures to
best practice standards, but designing corporate structure relies more
on the practitioner’s depth of business experience and well researched
understandings of markets, products, core operations such as production
and supply, and the generic organisational options that have, inevitably,
to be considered. We deal with business model principles in Section 3,
corporate structures and the centre in Sections 4 and 5, and more complex
matrix variations in Section 6.
1.4 Control and lean principles
Structuresneedtobedesignedformorethanclarityandlogic.Theyalsoneed
properties that put managers in full control of their fields of responsibility
and which, when certain design rules are applied, will tend to drive out
unnecessary cost.
In Section 7 we suggest some guidance on how certain generic structures
promote business controls. In Sections 8 and 9 we give examples of how
certain techniques can be employed to apply lean principles in organisations.

4
1.5 Models of corporate devolution
In Sections 10, 11 and 12, we have selected three models to illustrate generic
approaches to devolution in large organisations. These all support the
principle of forcing decision making to the lowest competent level while
allowing the organisation’s centre to have transparency of operations and
keep ultimate control.

5
2. Basic managerial structures

7
2.1 Information and data requirements
The common form of managerial structure is the chart we know as the
‘family tree’. It illustrates how the reporting structure is configured so we
can see who reports to whom. From it we should get a clear account about
the intended hierarchy of managers and the staff, and about the layers and
spans in the organisation – the sinews of its structure.
But a chart is not the whole deal. It is the tip of an iceberg of other information
and data that are needed to construct and validate the organisational chart,
and to which access is needed if the chart is to be properly understood. Some
of this information is needed to build a chart in the first place, and there are
a lot of data with which charts ought to be annotated if the whole workings
and profile of the organisation are to be fully described.
Some of the information and data needed for a chart to fully explain
organisation include:
n members of staff by name and unique identifier (such as employee
number)
n population headcounts and full-time equivalent (FTE) numbers related for
each location, job and managerial node on the chart
n job titles
n job descriptions (preferably in Collinson Grant’s one-page format, an
example of which is at Appendix 1)
n job grades (with a cross reference to the overall grade structure in force)
n lengths of service by individual, cross-referred to some analysis showing
breakdowns and trends
n ages/dates of birth by individual, cross-referred to a similar analysis as
that for lengths of service
n pay – with all its components, rates, earnings, including overtime and
incentive structure, and the fully loaded cost by individual and with
managerial node totals, cross-referenced to the pay structure in force and
some analysis of breakdowns and trends
n equivalent hours of work with resource planning assumptions
n the employee status of individuals and groups by managerial node –
full-time, part-time, term contract, et cetera.
2 Basic managerial structures

8
Anyone drawing a chart from scratch should have researched much of this
information. It can be used to populate the chart, colouring in the bare facts
about numbers managed, and for other purposes such as assessing whether
the resulting structure and population is fit for purpose to undertake the
volume and type of work to be put through departmental processes.
A complete picture of the organisation will also include some details of:
n Key performance indicators (KPIs) for each managerial job
n layers and spans statistics – average spans and number of layers in each
part of the organisation and overall, actual and target
n staff turnover and absenteeism
n history of changes (staff and managers’ population growth, decline, last
change made, et cetera).
The fitness for purpose of the organisational segment can finally be assessed.
2.2 Managers
What is a manager?
The job of manager is the single most important building block in any
organisational structure. At its simplest a manager is any job into which at
least one other person (a subordinate) reports. The concept is essentially
about hierarchy, and how hierarchical authority is exercised over
subordinates in the workplace.
This idea may sit a trifle uncomfortably with readers for whom modern
employee empowerment and the ‘worker bee’ theory are attractive. Much is
said these days about the time being past when people need to be managed
in an old-fashioned sense. It is said that, instead, people merely need to
know in which team they play and what skills they ought to contribute, with
the consequence that there can be leadership and drive without managers.
This is wrong. We believe that most organisations need to make it known to
whom authority has been given and therefore where accountability lies, and
how the personal performance of managers and that of the process they are
charged with are measured.
The managerial node – the position of a manager in the structure chart, is
the point at which responsibility for their subordinates, and all their activity,

9
resides. Work done by subordinates is an extension of the manager’s work.
Poor work by subordinates is also poor work by managers, who need to
recognise, mitigate and correct it. People want, need and must be led by
managers with properly designed managerial remits which also define the
remits of their subordinates.
No other system of working is sustainable, despite claims from some quarters
that a hierarchical chain of command ought to be consigned to history and
replaced by management structures where people do not sit at the top or the
bottom but are said to inhabit inner and outer zones!
Man cannot conceive of an organisation
that some are not capable of subverting
Russell Ackoff
Supervisors and team leaders
A supervisor is someone with all the powers and accountabilities of a
manager but in a more junior capacity. Supervisors are often referred to as
first line managers and, apart from their less senior standing, they have all
the characteristics of any manager and a chart should treat them as such.
Team leader is a more problematic description since any sort of team can
have a leader. The Chairman may be the leader of the Board, and the Chief
Executive of the company at large. Nevertheless, it can be a conveniently
specific term for someone who is not a manager or supervisor because they
mainly do the same work as the people around them, but are also relied on
to help organise the work within the relevant process. To us, a team leader
is someone who does not have personal authority over the conduct and
performance of individuals in their work group.
Typicalteamleaderresponsibilitiesarecoaching,assisting,training,guiding
or informing others in a team, setting an example for them to follow, or for
acting as a communication or administrative channel linking the team and
the first line manager. It is a position of responsibility but not of managerial
accountability or power.
“ ”

10
For example, in times past, a manufacturing operative on whom the
foreman relied for support in organising work was known as a charge-hand.
This person was responsible for the work but responsibility for the people
remained with the foreman and was never delegated. A job-evaluated
premium was usually paid but otherwise the terms and conditions were the
same as those for the rest of the workforce.
Specialists
It is also best to distinguish managers – who always manage people, from
those non-managerial staff (often quite senior) who do specialised work,
including ‘managing’ things but not directing people. These are usually
specialists or experts and are on the limb of a reporting line, outside the
vertical chain of command.
Organisations have an unfortunate habit of appointing specialists to be line
managers on the basis of their technical skills or length of service even if
they have no skills in managing people. Worse, an extra layer may be created
for a deputy, reporting one-on-one, who is instructed to manage the people
in order to leave the specialist unencumbered by the manager’s principal
task of managing people!
It is best to resist this at all costs. Respected specialists with no leadership
or people skills should be given a neutral title like ‘executive’ which is both
accurate and carries no ‘baggage.’
...if any particular organisation isn’t screwed
up now, it used to be, or soon will be!
That is the reality of organisational life.
Geoffrey M Bellman
2.3 Layers and spans
Concept
The concepts of layer and span are fundamental in analysing any structure
that has the form of a managerial tree. Quantifying structure by counting
its depth (layers) and width (spans) is a valuable exercise in understanding
“
”

B
E F G
C D
A
H I
the resources it is absorbing and some of the problems that may challenge
it. Even a brief study of any chart may throw up some obvious shortcomings
in the configuration.
In the example below there are three layers of managers. A is at layer 1 being
the head of the organisation. At layer 2 are B and C – A’s direct reports –
while F, who reports to B, is at layer three.
D, E, G, H and I are not managers because no one reports to them. In counting
down vertically we should be clear whether our reference is to layers of
managers or of all employees.
Manager A has a span of control of three; managers B and F have spans of
control of two while C has just one. The five non-managers have no spans to
manage. Each of the four managerial positions is a ‘managerial node’ within
the whole structure.
Where the number of layers is small, structures are referred to as flat, and
where they are great, they are deep. Where the number of nodes within a
span is great, it is wide, or conversely narrow. Inevitably, these terms are
relative.
Counting layers
The number of layers to be counted depends at which node the count starts
– that is to say which manager is at layer 1.
11

12
For example the top node might be the group’s Chief Executive Officer (CEO)
(the most senior executive in a larger organisation), or a CEO of just one of
its divisions, or the Chief Operating Officer (COO). In each case the scope
shrinks – from the group to a division, to (merely) the COO’s operating
organisation.
Optimising layers and spans – the ‘8 x 5 test’
It is a useful to test managerial structures against an ‘ideal’ benchmark. We
usually set this at a maximum span of eight with an ‘ideal’ of not more than
five layers – usually referred to as the ‘8 x 5 test’.
The ideal number of layers, five, is thus relative to the context – the size and
scope of the organisational structure under review. There are no strict rules,
and judgement is required. In this regard we may want to refer to the concept
of levels discussed later on. Layers and levels are different but the latter is a
useful tool in making judgements about the former.
Structural shape – the dangers
The importance of structural shape is often lost on company executives who
follow their own, untutored logic in designing the structures they manage.
A common mistake is to assign one manager to each separate task, type of
task, or activity irrespective of the number of managerial jobs (nodes) this
generates. Too many nodes will result in too deep a structure and spans that
are unnecessarily narrow. For example, an accounts department of twenty
people might give rise to eight managers configured in four layers whilst
2/1
3
4
5
2 3 4 5 6 7 8
1

a correctly shaped structure would be more likely to have four managers
and two managerial layers. One common feature of bad shape is one-to-one
reporting of the type shown here.
Worse, having unnecessary managers creates fertile conditions for the
uncontrolled increase in the whole staff population. Managers tend to
‘attract’ members of staff to bolster the standing of their own jobs, sometimes
building in more ‘slack’ for their own comfort or for ‘emergencies.’ The
employed population can balloon needlessly and expensively.
So it is particularly important that spans should not be allowed to be too
narrow or levels too numerous because the proliferation that follows drives
up the size and cost of the payroll with no guarantee of adding any value.
Not that this is the only danger – for example, the job at B is almost certainly
ill defined, and the accountability of A’s job may be diluted as a result. It
also encourages other bad habits as B can be an effective barrier to proper
oversight, communications and the flow of information up and down the
chain of command, and may block A’s vision of the work being done, or not
being done, below B.
13
C D
A Three Layers
B
C D
A
B
C D
A
Two Layers
☺
B
Two Layers
☺

B
E F G
C D
A
H I
14
Calculating the average span
To counter the balloon effect, it is a useful test to quantify the average span.
This is calculated from a ratio of the number of managerial nodes and the
total population. In the example there are nine staff in total and four of those
are managers (A, B, C and F), thus 9/4 computes as an average span of 2.25.
The 8 x 5 test suggests that eight is optimal so a figure as low as 2.25 must be
sub-optimal by quite some distance.
Effect of diversity of jobs on width of span
In practice a span may need to differ according to how diverse the jobs are
of the subordinates being managed. This usually varies at different depths
in the structure. For example, the number of people reporting into a senior
manager, such as a chief executive, is often greater because they are all well
qualified in their different fields and require less managing from above than
those lower in the structure.
But middle managers in charge of more junior managers and members
of staff will find that the interventions they need to make in supervising
subordinates limits the time they have available for each one, restricting
further the number that they can manage.
Continuing down the organisation, where there are large numbers of non-
managers performing jobs to the same or similar design (less diversity within
the managed group) junior managers’ spans can often be very wide indeed.

15
For example, a call centre supervisor can manage more than 20 subordinates,
because managing simpler, repetitive work takes proportionately less
managerial time per head.
Reducing the number of layers and managers
Flat structures are best because they reduce the number of layers. Generally,
the wider the average span the fewer the total number of managers and
members of staff will be needed for any given workload. The narrower the
span, the more likely it is that layers will multiply.
Flatter structures are associated with better managed and motivated staff,
better managerial performances, and senior managers who are less likely to
be out of touch with the quality of the staff’s work, and the processes they
operate.
Flatter structures can also be an antidote to managerial remoteness.
Common sense indicates that the quality of communication throughout the
organisation will deteriorate as its structure becomes deeper. Goals are more
easily shared when structures are flat.
A manager will take six months to get to know his staff,
but they will take only six days to get to know him.
John Adair
Our research suggests that managers should expect to spend at least
two-thirds of their time interacting with those directly reporting to them.
Managers should have enough subordinates reporting to them and a broad
enough span to justify this norm.
So, first and foremost, managers should be skilled at managing people
and spend most of their time doing it. If more than half of a manager’s
time is spent on work that is distinctly different work from that done by
subordinate staff, that manager may be using the skills of a senior specialist
and managing things rather than people.
Unfortunately it is commonplace for the most skilled specialist to be
promoted to head of department even though they may have little talent for
managing staff, or for that matter managing operations generally. The result
“
”

Dept
Manager
Section
Manager
Section
Manager
Deputy Dept
Manager
Director
16
of this is often that the task of managing the staff is delegated to an assistant
or deputy to the manager, creating the undesirable ‘one-to-one’ and the
unnecessary extra layer referred to earlier.
The structure illustrated above almost certainly has too many layers and, as
a result, arguably too many staff. It is possible to challenge such a structure
by seizing the opportunity to reduce the layers, justifying this by insisting
on best practice in the design of some individual jobs. The structure could
be redesigned as shown below.
Specialist Department
Manager
Director
Team
Leader A
Team
Member
Team
Member
Team
Member
Team
Leader B
Team
Member
Team
Member
Team
Member
Departmental
Managers
have a span of
8 or 10
Other
Departmental
Managers
Is there an
opportunity
to reduce
the staff?
New job for
specialist
Director has
a span of 4

17
Layers and grades – a warning
It is never wise to assume that jobs in the same layer (counting down from
the top) need to be assigned to the same rank, grade, pay, or other indicator
of seniority. This has become a bad habit in some organisations, particularly
in the public sector. It has the effect of forcing a large increase in the number
of layers, thus causing a population explosion with the attendant increases
in cost. Some interest groups, not excluding managers themselves, can
see this as an opportunity to increase the numbers employed. Of course,
productivity dies.
No institution can possibly survive if it needs geniuses
or supermen to manage it. It must be organised in such
a way as to be able to get along under a leadership
composed of average human beings.
Peter Drucker
2.4 Levels of work
Jaques’ model
Although the 8 x 5 benchmark is a useful check, there is another tool to help
control and optimise the number of layers. The levels of work method is an
attempt to limit the number of layers by forcing them to correspond to a
generically defined level or rank of job type.
The concept was developed by Elliott Jaques, a Canadian organisational
psychologist, and is based on time span analysis. He classifies managerial
work into seven categories or levels based on the length of time which each
type of job requires in order for it to achieve its aims and fulfil its purpose.
In this way he creates associations between type of job, the weight of
accountability and the organisational hierarchy.
Once a job is positioned at one of Jaques’ levels, it is possible to fashion a
correspondence with a layer in the structural chart.
“
”

18
Using the levels of work method to optimise the number of layers
This is the sequence of steps to apply the levels of work principle to the task
of reorganising an existing structure, or to creating a structure from scratch:
1 Pinpoint current problems: duplication, too many levels, too many layers,
too much management, excessive cost.
2 Examine the characteristics and requirements for every managerial job.
3 Define all jobs by title, job code and function, and map them to the
appropriate levels of work.
4 Assign locally appropriate definitions to all the levels needed, noting that
not all of Jaques’ levels are relevant to or necessary for every organization.
5 Consult the job grade structure and amend it to reduce or, less likely,
increase the number of grades if that makes sense.
6 Force a reduction in the number of managerial jobs if possible.
7 Complete the new scheme for job grades, layers, reporting structure and
employee numbers.
8 If necessary, create senior non-managerial streams to accommodate high
grade non-managerial staff.
9 Draw out and sense-check a new structure chart, auditing the number of
layers and widths of spans and calculating the average spans.
One of Jaques’ levels can accommodate more than one layer in an
organisational structure if this is required to organise the work or manage
processes. Start by defaulting to a model in which there is just one level to
one layer before resorting to two (rarely more) layers.
Corporate leadership
Scan the world and craft a sustainable company – longer term
CEO and Executive Board(s)
Strategic management/Strategic advisory
Plot direction for major/multiple business units and functions
Create the climate for successful change
Major/multiple Business Unit and Function heads
Business unit management/Expert advisory
Manage/develop business unit or function
Reinforce the climate for successful change
Smaller/single business unit
and function head
Senior level professional staff
General management/Senior advisory
Manage/develop business unit or function
Reinforce the climate for successful change
General Manager
Mid-level professional staff
Management/Advisory
Get it right first time
Manage/advise on people, technology,
and market systems and practices for
best results
First level of true ‘management’
Qualified professional staff
Lead team/Produce
Create ‘winning’ teams
Short-term problem solving
Team leaders
Senior clerical, junior
technical/professional staff
Operate/Produce
Create/deliver products/
services to specification
Shop floor operator
First level administration staff
First level customer facing staff
VII
VI
V
IV
III
II
I

19
In the same way, it is not necessary that every level should contain a layer.
Not filling all levels with jobs is an effective way of designing a leaner
structure, provided you can be sure that the management of processes or
the decision-making capacity of the organisation is not impaired.
If you want to change the culture, you will
have to start by changing the organisation.
Mary Douglas
More on levels and layers
With levels, it is the principle that is important. The mechanics are secondary.
The exercise is all about achieving ‘best fit.’ A series of iterative steps will
almost always be needed to find the equilibrium.
It is not always necessary to stick with Jaques’ time-span definitions. A
custom model may work better, provided it is given a clear and robust logic
that works in the specific context. That could be based on the local grade
structure if the warnings given earlier are heeded. However, Jaques’ model
has a proven record and should not lightly be set aside.
Theoretically, a pyramid of only six layers with an average span of seven is
possible in an organisation of more than 117,649 people – enough capacity
for most organisations.
“ ”
Layer
0 17
1 77
2 497
3 3437
4 2,4017
5 16,8077
6 117,6497
Span Staff

20
2.5 Job design
In this book we will not dwell long on this huge subject. From the perspective
of organisational design, we are mainly concerned with jobs which derive
from, and are directly relevant to, the processes for whose performance the
employees and their manager are accountable. This sounds obvious. But
often, much ‘work’ goes on that does little to fulfil a department’s purposes.
To avoid too much unproductive work, arrange for the tasks and decisions
embedded in jobs to be entirely functional and directed unerringly towards
how outputs such as product or service effectiveness can be optimised. Only
then should inputs be defined, such as time, skills, tools and materials; in
other words, use the design process to create an effective test of whether
value is being added.
The art and science of designing jobs is an effort to balance the behaviour
and performance that managers want with the welfare of the jobholder. In
relation to the latter, motivation starts with skilfully designed content and
structure of jobs.
Don’t chase the latest management fads.
The situation dictates which approach best
accomplishes the team’s mission.
Colin Powell
2.6 Line and function
It was common in the past to refer to line and staff structures. Although the
term is less heard today, it remains the clearest terminology with which to
refer to the juxtapositions of conventional, ‘family tree’ charts.
On any chart some people are said to be in line jobs (‘in the line’) within the
layered hierarchy through which is threaded the vertical chain of command.
Others are said to be in a ‘staff’ relationship to line jobs because they work in
what we now call the functional departments, and so have a more ‘horizontal’
relationship to the rest of the organisation. Specialist jobs mentioned earlier
are also staff jobs.
“
”

B FC D E
A
These days there is much analysis of what is meant by ‘functional’ and ‘line’,
the relationships they ought to have to each other, and the ratio between
them or the balance they represent. So what is a function, and how does it
relate to other parts of an organisation?
Operations and support
A job or the work it performs in any organisation is generally one of two
main types – operational or in support of operations. The taxonomy works
like this.
In the structure above, A is CEO or COO or General Manager of the
organisation and B, C and D are the line reports for operational departments
such as sales, manufacturing and engineering. These jobs manage activities
which add value for customers through marketing, selling and delivering
the products or services for which the organisation exists.
E and F are also direct line reports of A, but their jobs are functionally to
manage activities that support B, C and D. Thus E could be the Chief Financial
Officer and F the Head of Human Resources (HR). Information technology
(IT) is another support function and others include legal, health, safety and
environment (HSE), estates, and security.
But some activities may be less easily classified into support or operation
because context makes a difference. For example, distribution as an activity
is usually operational if its performance is critical to adding value to the
21
B FC D E
A
SupportOperations

22
customer in a direct way, but it may, in some organisations where that is not
the case, be defined as ‘merely’ functional, particularly if its expense is a
small proportion of total cost.
Whether engineering is or is not a support function also depends on
context. It may merely support the organisation in a way that a maintenance
department would, but if what the company sells includes the value added
by engineering design, then it is almost certainly not a support function but
an operation.
Functionally supportive relationships
The importance of the taxonomy here is to help decide how best to
organise work, draw charts that best illustrate what we mean, and to aid
understanding. In the chart sketched below, D is a divisional sales manager
with three territorial sales teams led by G, H and I. D also has specialist
support from J, an IT manager responsible for sales systems, X responsible
for the HR support, and Y responsible for financial reporting and analysis.
Day to day J, X and Y all report to D and the solid line drawn between each
of them and D illustrates that.
But X and Y also have functional relationships with E and F, often called
‘dotted line’ relationships. This means that, although their day to day line
reporting is to D, for certain aspects of their job they are responsible to E
and F. This is usually illustrated on charts with a dotted line.
B FC D E
G J
H X
I Y
A

23
For Y, Finance Director F will probably provide direction on accounting
rules and standards, reporting formats and timetables, budgeting, and
other generic and recurring finance functions. F would have had a major
influence in selecting Y for the job and would be the judge of Y’s professional
competence.
The conventional definition of management is getting
work done through people, but real management is
developing people through work.
Agha Hasan Abedi
A point of interest is the degree to which X and Y are expected to follow the
lead or requirements of their functional heads F and E where this may not be
to the comfort of, or consistent with, the requirements of D. That questions
the strength or weakness of the dotted line. The policy must be set by A
and it must be made clear to D, E and F. D must also ensure that X and Y
understand the boundaries and behaviours that define their jobs.
Note that J has, in this case, no functional head at the level of E and F so all
IT policy is down to D, advised and supported by J. In the units run by B and
C one would expect there to be the same or similar relationships paralleling
those just described for D. But there are no rules that say the structure
beneath D should be symmetrical. Indeed it is important that the design
of each be configured according to their processes, scale and other features.
Functional governance – a variation on the ‘dotted line’
Functional governance is a more robust variation of the support relationship.
To the functional support already described, functional governance adds a
duty on Y and X to use a broad discretion in reporting to F and E by exception.
This process is commonly used to report actual or proposed departures by
D from any common set of understandings on good practice set by F or E,
and endorsed by A. The effect is to limit D’s unilateral power to engineer
opportunities for business policies and initiatives. By this means, D’s
decisions become subject to checks and balances from outside the division.
“
”

24
Thus the governance process (sometimes called the functional college) can
subject D’s decisions which, in the opinions of Y or X, need further review at
the next level up the functional tree, and if necessary beyond, where matters
can be discussed (one hopes dispassionately) between senior functional
and operational managers.
Finance functions are the most vital of all the potential functional colleges
because they control most of the information that an organisation produces
and on which its decisions are based. But HR is also becoming more essential
in this context because of the amount and range of compliance now required.
I try to buy stock in businesses that are so
wonderful that an idiot can run them.
Because sooner or later, one will.
Warren Buffett
Needless to say, general managers, such as managing directors of operating
subsidiaries, generally dislike functional colleges because covert decisions
and actions cannot easily be kept from their own line. They are more likely
to be relaxed about a more technical functional college such as IT.
The notion of functional governance is most useful where a general
management position, such as the Managing Director of a company in
a group (see the diagram opposite), has acquired a significant degree of
autonomy from the group CEO. The Finance Director, who line reports to
the Managing Director, is permitted an open line of reporting to the head
of function, the group CFO. That then legitimately bypasses the Managing
Director, releasing information and independent judgement by exception,
thus curbing the Managing Director’s decision making autonomy.
By the same token the Regional Accountant also has a direct line to the
Finance Director. The three finance jobs inhabit the finance function college.
This organisational device allows a high degree of devolution day-to-day
while at the same time operating a ‘stop loss’ or ‘braking mechanism’ to
arrest rogue decisions made lower down the corporate structure.
“
”

25
Group CFO
Finance
Director
Managing
Director
Group CEO
Operations
Director
Regional
Operations
Manager
Other
Regional
Accountant

29
3.1 The starting point for corporate organisational design (OD)
The design of a new structure is not a single task that can be performed in a
couple of hours. It involves a good deal more than sketching out a series of
charts until it ‘looks right’. Best practice entails a process with a sequence
of steps. It requires proper research to achieve a thorough understanding of
the business model in all its aspects, including all the key processes that add
value and give competitive advantage to the enterprise. A quick and easy
approach will be built on sand.
It is the business model which tells us how the company or institution earns
its bread. However, all too often there is no shared understanding of how this
works by those who labour at it. Even though all important decisions made
by managers are or should be elements of it, the properties of the model
itself may be articulated rarely or referred to only infrequently in discussions
and decision making forums, even among the most senior managers.
In designing how an organisation should be structured and function we
believe in the absolute necessity of using the model as the foundation stone.
The model is the reason the business exists and a failure to take account of
the breadth and depth of what the organisation is, and is trying to be, results
in structures that are not fully fit for present or future purpose. The business
model will be liberated by good organisational design which is rooted in
that model.
The sequence of steps for researching the model of the business and the
process of designing structure is illustrated below.
3 The business model
The business model
Mission,
policies
and
corporate
objectives
Business
strategies
Value
chain –the
design of
process
and jobs
The OD process
Configuration
and structure
Accountability and
measures of performance

30
3.2 Mission, policies and aims
The business model is defined first by the policies that originate from the
needs of stakeholders. These are owned collectively by the Board, not by its
executives. They define longer term fundamentals of the business and may
include statements of mission – what the organisation hopes to achieve or is
there to do or an expression of its values. Policies should also acknowledge
important aims such as the acceptable rate of return on shareholders’ funds
or targets for growth of sales or profit.
Policies are not immutable and will be reviewed or confirmed from time
to time, perhaps even annually. But frequent changes of policy suggest a
business not at ease with itself, even perhaps a headless chicken.
Typically, policies are about where to trade and what in, how to
fundor resource the business, where to locate it, the direction in
which to develop it, the questions of growth, acquisition, the place
of research and development (RD) and almost anything that
sets the course of business operations for the medium or even the
long term.
3.3 Strategy
Strategy is proposed by executives and ratified by the board. It must be a
derivative of policy. Its purpose is to describe the means by which executives
think policy can be realised and its priorities addressed. The board, for
example, may prefer to maximise short-term margins rather than long-
term business growth. It may wish to prioritise growth in new markets after
recognising incipient decline in current ones. And, were there a policy to
aim for, say, a 19% return on net assets, executives will determine which
strategy will best achieve all these objectives.
The business model
Mission,
policies
and
corporate
objectives
Business
strategies
Value
chain – the
design of
process
and jobs
The OD process
Configuration
and structure
Accountability and

Obvious examples of strategy formation are decisions about markets,
channels, selling methods, products, pricing and terms of trading, supply
chain, manufacture, quality and everything that underpins, in the private
sector at least, the search for competitive advantage.
In the public sector the search for cost-effective ways to fulfil policies
presents equally taxing strategic challenges, and represents equally
important influences on organisational design. Not-for-profit organisations
have equivalent outputs and benchmarks of success.
The distinction between policy and strategy can easily become blurred,
particularly if the board consists mainly of executives, and even more so
if executives are also the principal shareholders. And successful strategies
may morph into policy. The essential point is that policy is about what to do
(or try to do) and strategy is about how to do it.
3.4 Value chain - process and job design
Processes and methodologies and the design of managerial and staff jobs
are put in place because they are the practical means, tools and techniques
used to implement business strategy. But there must be logic behind the
sequence and configuration of processes. They are linked in what we usually
think of as a value chain.
Although strategy creates the competitive advantage or the basic means of
fulfilling objectives, processes actually add the value day by day. Strategy
may be described in generalisations. Descriptions of processes need to get
down to detail.
31
The business model
Mission,
policies
and
corporate
objectives
Business
strategies
Value
chain – the
design of
process
and jobs
The OD process
Configuration
and structure
Accountability and

32
When travelling on a journey, even if there
are only three of you, make one a leader.
Muhammad
We can describe them by writing operating procedures, drawing flow charts,
and by listing the content and responsibilities of individual jobs. Managerial
jobs demand a formal design as much as the organisational structure itself.
What actually needs to be written down might vary, but managers need
to know enough about what is expected to be able to perform confidently.
Others around them also need to know, so that they can work together
effectively.
Process and job design together are the business model’s largest and most
detailed body of knowledge, and provide the foundation for designing
structure.
3.5 Configuration of structure
Our concern now becomes the structural dimension of design – how
managers and staff report, and how they relate to each other within the
framework that we configure and call structure. Managerial structures
must be derived from the value chain, flowing logically from the physical,
administrative, and managerial processes of the business. We segment the
organisation, or chop it up into chunks like departments, which the logic
of the value chain provides us with. Corporate structures go further and
probably owe more to the policies and strategies of the model (rather than
processes), and should be derived directly from those.
The job of top managers is to get the structure right and to maintain, modify
and renew it. That of the middle and junior ranks of managers is to operate
The business model
Mission,
policies
and
corporate
objectives
Business
strategies
Value
chain – the
design of
process
and jobs
The OD process
Configuration
and structure
Accountability and
“ ”

33
faithfully to process and to structure. It is worth noting that operating to
process does not necessarily imply a rigid and bureaucratic style of conduct.
It simply means sticking to what the given job is and using common sense.
Design of structure should attempt to make, or to enable, managers to
be effective by emphasis on economy, through optimising the numbers
of managers and the number of their reporting staff, and on efficiency by
optimising the depth of managerial layers and width of managerial spans.
Section 2 describes the tools and techniques that can be used.
3.6 Accountability and measures of performance
Control is at the heart of any manager’s job. And it remains true that, if
managers are to be held to account, what is being managed by them must
be measured appropriately using, for example, key performance indicators
(KPIs).
So an organisational design requires that there be measures that quantify
or qualify the performances of the processes being managed, and the
corresponding individual performances of managers. If in any instance
this proves not to be possible, it is the managerial job or the process being
managed whose design might be searched for flaws.
The business model
Mission,
policies
and
corporate
objectives
Business
strategies
Value
chain – the
design of
process
and jobs
The OD process
Configuration
and structure
Accountability and
The business model
Mission,
policies
and
corporate
objectives
Business
strategies
Value
chain – the
design of
process
and jobs
The OD process
Configuration
and structure
Accountability and

34
3.7 Balance in all things
The aim of good design is to achieve a balanced and harmonious structure
that will give the organisation a structure that is robust and flexible enough
to help it achieve its potential.
The relationship between the business model and how the organisation
should be put together may sound obvious. But our experience has been
that companies are frequently keen to look for solutions to problems
by changing the structure of the organisation without first asking if the
problems originated in fundamentals such as strategy, or even in confused
corporate objectives.
For example, a large retail organisation operating in many countries
used franchises in some territories and in others operated its own stores.
Following its acquisition by a group known for its leading global brand, it
began to question how its retail operations should be organised in pursuit
of improved performance. It was advised that no changes to organisational
structure should be attempted until it had resolved matters at the heart of
its business model.
You can’t build a strong corporation with a lot of
committees and a board that has to be consulted every turn.
You have to be able to make decisions on your own.
Rupert Murdoch
Some of the specific issues identified included whether the strategy of the
business should be dictated by brand promotion or by expert retailing?
Retailing strategy emphasises pragmatic local decisions using processes that
aim to maximise margins in stores. Brand-led strategy requires a consistent
global (or global region) approach to marketing that requires different
types of management process that usually constrain how store managers,
particularly franchisees, operate. Without a policy decision to govern which
strategy should be followed, it is useless and dangerous to consider the
issue of whether organisational design should favour the franchise or store
ownership structures.
“
”

In fact it was also right to ask if the problem might be even deeper, requiring
policy formation at the level of the acquirer group. Was it clear about what
returns it needed from its investment and when? And what were, or are, its
expectations from the acquisition?
It is common for organisations to change policies without reappraising
strategy, to change strategy without recognising new requirements
for processes or for the design of managers’ jobs or changed resource
requirements. Equally common are situations where policy and strategy
are not even articulated or understood. In which case, processes cannot be
underwritten by logic and in turn organisational design itself is merely the
result of fragmented decisions arbitrarily taken.
Management problems are not respecters of
the company organisation, nor of the talents
of the people appointed to solve them.
Anthony Stafford Beer
Structural configurations will either liberate or limit the effectiveness of
the managerial process and affect the overall health of the organisation
accordingly. How its structure is configured is a major influence on how it
can and should work, and on how it is, and is intended to be, controlled by
its managers.
35
A congruent approach to
organisational design
Structural
configuration
Operating
processes
Performance measures/
accountabilities
Business
model and its
strategies
“
”

36
The quality to be sought most in organisational design is that of balance
– a balance between well-defined processes and lean, elegantly configured
structuresgivingeffectivejuxtapositionsofmanagerialjobs.Theharmonious
concept is illustrated in the diagram on the previous page, remembering
that process (how managers manage) is derived from the strategy and all the
other properties of the business model.

39
4.1 Follow the business model
We come now to discuss the design of organisation at the macro level – how
businesses and business units, as opposed to departments, are put together
so that they work well.
As the previous section indicates, designing a corporate, organisational
structure requires an intimate knowledge of the enterprise and every aspect
of its business model. There should be a recognisable strategy covering the
more common points of reference listed below and this should, through the
design of its processes, influence the organisation’s structural form.
n Markets, customer types and channels to market.
n Products – groups, range, pricing method and categories.
n Locations – geography and logistics.
n Production facilities for manufacturing (such as factories), or service
organisation (such as franchising).
n Business support functions such as finance and HR departments.
n Supplychainsourcesandlogistics(suchasphysicaldistribution,inventory
management, sub-contract fulfilment, et cetera).
n Commercial strategy (such as pricing, contracts, et cetera).
n Inter-trading terms – trading within the wider corporate structure.
n Financial reporting and profit centres – how business unit performance is
measured and at what points in the structure.
n Shared service centres, outsourced transactions and administration –
purchases from within or from outside the wider organisation.
4.2 Integrated and devolved structures
Corporate structures fall broadly into two types – the integrated and the
devolved. Both provide a macro view of organisation, and sometimes require
complex organisational design at the level of the whole enterprise.
Any design that is devolved will introduce the idea of autonomy, putting a
significantly higher degree of power and self-sufficiency directly into the
hands of many managers. Accountability for profit itself is usually not far
away.
4 Corporate structures

40
A conventionally integrated company, on the other hand, is itself a single
business unit and often a single profit centre, even if it is a very large entity.
Some integrated organisations are monolithic, an example of which, the
multi-dimensional matrix, is described at Section 6.4.
The only things that evolve by
themselves in an organisation are disorder,
friction, and malperformance.
Peter Drucker
4.3 Integrated organisations
All companies default to a conventionally integrated structure until their size,
or their internal or customer-facing complexity, pushes them in the direction
of the more devolved features found in the profit centre and business unit
types of structure.
In an integrated, centralised organisation the managers of operations and
support functions all report to a point of control at the top of a unified
hierarchy. Usually, this is the single position where accountability for profit
is not shared with any other executive post.
Authority is delegated vertically into functional and operational silos, layer
by layer within the hierarchy. Power is not dispersed laterally into profit
centres or business units as in a devolved model. The entire organisation is
the only true profit centre.
In integrated organisational models other notions of accountability have to
be designed into the information system – those of profitability of product,
market, channel, et cetera. Some managers will be held to account for these.
But unilateral accountability is often difficult to isolate, and profitability is
not as categorical a definition as profit. These complexities are met and dealt
with in Section 6 on matrix structures.
“
”

41
The structure above is organised as a series of specific operational units
(dark) alongside support departments (light). The operational units are
normally parts of one uninterrupted value chain running end-to-end
through the whole enterprise, giving it its integrated character. There is little
or no organisational segmentation around geography, markets, customers,
products, and so on. Sales, manufacturing and support functions are
managed together, usually under a management team of departmental or
functional heads and their chief executive.
This integrated structure places limits on the degree to which managers
can make independent decisions. They must make only those relevant to
their function, since all departments need to plan and act in concert with
a single end in view. The mechanism for this is the integrated budget. The
chief executive or an equivalent is alone responsible for the overall budget
and accountable for the realisation of profit.
But monolithic examples of integrated structures are often inherently
unhealthy in their design. In larger companies this can include behaviours
by top ranking executives which have not been seen since the days of
medieval baronies.
As an example, a great engineering organisation once had a structure in
which the chief executive’s main role was as peacemaker between functions
that behaved more like hostile factions. The heads of two product groups
competed for resource from other functions formed around procurement,
manufacturing and engineering, each of which regarded themselves as
profit centres even though they sat on the same value chain, just as they
would have done in any other integrated organisation. Fortunately, a
change of leadership resulted in a change of business model that enabled
General Manager
Human
ResourcesFinanceBuyingSales Manufacturing MarketingEngineering

42
the structure to morph naturally into a highly successful devolved model.
Managerial dysfunction and flawed structural design often go hand in hand.
Usually, the solution also calls into question the business model.
Leaders create an environment in which everyone
has the opportunity to do work which matches
his potential capability and for which an equitable
differential reward is provided.
Elliott Jaques
The question remains: why would an organisation not want to adopt a
devolved, business unit structure? The answer is that there are examples of
healthy integrated structures both large and small. Integrated structures are
often less expensive to run. Devolved structures may encourage duplication
of tasks, processes and functions as they become replicated in many places.
This can ramp up cost and create a need for administrative overlays to
maintain control, compliance or transparency.
Then again, an organisation that is without much diversity of product or
market is quite likely to have no need to devolve responsibilities to a greater
number of managerial positions. Increasing their number itself increases
complexity and problems of communication and flexibility unless there is a
true need for devolution.
4.4 Devolved structures
Complexity and scale
As a generalisation, devolved structures offer more help in managing
complexity than the integrated model. The need to segment organisations in
order to firmly fix accountability, to encourage a self-reliant and enterprising
team of senior managers, and to control a multi-faceted strategy brings
with it challenges which sooner or later force organisations to devolve their
structures to some degree.
In his classic book, The Practice of Management, Peter Drucker describes
how corporations evolved in the first half of the 20th century. He noted
that as soon as companies began to grow they needed to grapple with the
“
”

43
necessity of devolution. He observed how over time, they were forced by
circumstance to change from the integrated form to another, which he called
the federal.
The terms federal and decentralised are no longer common currency in
organisational design, having largely been replaced by ‘devolved.’ But
the idea and application behind all these words is the way in which the
governance and structure of all larger corporations would have started to
evolve at some time in their histories.
Devolution was once considered innovative. Before it joined the devolution
revolution, the Ford Motor Company grew as a fully integrated, functional
monolith. Control went directly up a single hierarchy towards one man,
notwithstanding the huge scale of Ford. In contrast, General Motors began to
create subsidiary managerial platforms in which power, with accountability
for local results, was pushed outwards and downwards throughout the
organisation in a way we recognise as devolved.
For the first time, the results achieved by specific units of the wider
corporation could be ring-fenced. Results could be reported and
differentiated according to the products being made, or by a level in the
supply chain, such as a parts manufacturer, or by a geographical market.
It was soon appreciated that this empowered managers, induced better
performance from them, and became a powerful enabler of growth. Ford,
General Electric and all the other giants of mass manufacturing followed
suit, creating the birth of consumerism so eloquently described by Drucker.
The speed of the boss is the speed of the team.
Lee Iacocca
When to devolve
Today the choice is usually not whether to devolve in the federal mode, but
at what point on the path of growth and complexity it should be done. And
when it is being done, what logic or principle should be followed. Complexity,
as well as size, matters in organisational design although the two are often
bound together as a single challenge.
“ ”

44
Today, businesses seem to become more complex earlier and earlier in their
development. This may have something to do with technologies opening
more channels through which customers can be reached, and globalisation
and international development opening up many markets simultaneously.
It is also easier to manage longer and more fragmented supply chains by
offshoring and outsourcing. Today it is much more likely that managers will
be comfortable with this ‘change is normal’ approach to life.
Size may be the final determining factor. A manager cannot lead if his or
her reach is so extended that control is loosened and effectiveness reduced.
The point at which this becomes a game-changer on structure varies greatly
within the industrial spectrum.
From these examples it is clear that devolution into business units follows,
first and foremost, the logic provided by the organisation’s external
interfaces for both sales and supply. Accordingly, we shall now focus on the
principles upon which effective devolution rests, and in Section 6 on the
form of devolution known as the matrix.
Autonomy
The distinguishing characteristic of a devolved unit of organisation is the
decision-making autonomy it hands over to its managers. All business unit
managers have some degree of autonomous authority, a measure of freedom
to decide and act without referral upwards. Typically this freedom may
allow discretion on levels of expenditure, on allocation of resources, even
the development and pursuit of strategies.
In an echo of Jaques’ concept of ‘time-span of discretion,’ there is a close
connection between the degree of autonomy and the interval of time between
which a business unit manager is called to account. This suggests that time
is a factor in defining autonomy as well as in measuring accountability.
In practice, the degree of autonomy residing in organisational structures
varies widely. For example, at an extreme end of the spectrum a group of
companies might use the principle to invest broad discretionary power in
a business unit manager by providing only its capital, its mission and an
expectation of trend in profit performance or growth, reviewed after, say,
three years. The rest is up to the executive officer in command.

However that would be too loose a mandate for most main boards to swallow.
They would be uncomfortable about losing so much control; and are likely
to be more prescriptive, requiring at least annual or perhaps quarterly
reportingwithpossiblechallengesandinterventions,monthlyprofitandloss
reporting to budget, and caps on capital expenditure and rafts of business
case justifications for initiatives and innovations. Banks and investors may
also limit autonomy by imposing regimes of financial governance such as
covenants.
It is part of the task of a designer or auditor of organisational design to
measure the degree of autonomy a manager of a devolved unit has been
granted. Pinning this down helps define a key attribute of structure which
should make an appearance somewhere in the description of the business
model. An appropriate degree of autonomy is one that sits comfortably with:
n the mission and objectives associated with the business model
n the risk assessments done on the relevant strategies
n the overall scale and complexity of the enterprise.
Dreams have their place in managerial activity,
but they need to be kept severely under control.
Lord Weinstock
Power of devolution
Essentially, autonomy describes and measures how a corporate organisation
embraces the principle of devolution – a letting go of power formally
recognised through the design of organisational structure.
The power of devolution lies in giving practical effect to a principle which
can have a huge impact on the performance of managers. It is to push any
given degree of managerial accountability down to the lowest competent
level in the organisation. In a large well known group of companies that level
is defined not by whether the individual job holder is up to the task (that is
a given because of excellent selection processes) but by a search for what
appropriate accountability can be embedded in each job in the hierarchy.
45
“ ”

46
There is no suggestion that this ought to determine the number of levels
or layers appropriate to an organisation’s structure. But there would be a
tangible connection between any Jaques-type ‘level of work’ definition
and the level or layer to which decisions and accountability for results are
devolved. Devolved organisations, of course, spread accountability widely.
4.5 Profit centres
Creating profit centres
Profit centres are archetypal business units where the devolved principle
is demonstrably clear, and autonomous accountability easily measurable.
So it is common for enterprises to choose to configure their organisation
around this idea. It can be an easy, almost compelling hypothesis to test at
an early stage of a restructuring project. Recognising and locating the profit
centres from revenue streams can be the key design task from which much
else follows.
Profit centres, like all business units, often become a key feature of the
business model and, in some cases, are even enshrined in the corporate
policy or strategy. Once put in place they make performance and the origin
of profit more transparent. Profit centres can vary in size from the building
blocks of very large corporations to simple product or customer segments of
a business that has outgrown its former integrated structure.
A profit centre must have most of the following characteristics:
n control over the origin of its revenue – by market, channel, territory,
product group or any equivalent measure
n control over the managerial and structural entity to which the revenue is
credited – a whole company or a division or discrete unit of it
n a discrete body of cost that can be more or less realistically associated with
the corresponding revenue
n the difference between the revenue and the costs being the centre’s profit
or contribution to overhead and profit
n the ability to measure all of the above, together with an adequate system
of financial and operational information that supports decision making
and performance measurement
n a recognisably discrete body of resources – managers, the staff, assets,
tools, et cetera, and the ownership of processes
n a sub-set of the business model and the corporate policy and strategy to go
with it.

47
The organisational principle being followed is that the work, enterprise
and resources that a profit centre will liberate can be hard-wired into its
culture and system of performance reporting, thereby creating managerial
accountability at its most direct, measurable and transparent.
Measuring profit centres
Specifically, a profit centre is a unit of an organisation whose manager has
accountability for the revenues attributable to it and the corresponding costs.
In a corporate structure shown below where B, C and D are profit centres
within company A, all of A’s revenue is being generated by the managers
of B, C and D, before it is consolidated in A’s profit and loss (PL) account.
There is a range of corporate functions bearing the usual variety of overhead
costs at E.
There are costs directly attributable to B, C and D, so that a bottom line of
‘profit’ can be determined for each of them. This may include direct costs
alone (generating a bottom line we can call gross margin), or also include
some operating costs not directly incurred by individual products but
indirectly incurred – such as their cost of distribution (generating a net
gross margin or contribution).
The cost of sales, conventionally made up of variable costs, is the obvious
first charge against the revenue of a profit centre. It becomes more complex
when we have to decide which semi-variable and fixed costs to charge against
the revenue of the profit centre. It is perhaps obvious that the profit centre
payroll and the people managed in the profit centre have to be co-terminus.
Within profit centres B, C and D, there may be overheads that bear no relation
to the profit centre’s products, outputs or revenues. But they might be an
appropriately apportioned or allocated charge against the profit centre.
Some overhead costs may simply be shared between profit centres, including
some that occur at the corporate centre far from the authority of the profit
centre manager. A view must be taken whether to allocate or apportion these
B C D E
A

back to the profit centres. The pros and cons of this could be debated at
length and there is no absolutely right or wrong methodology.
Fortunately, for the purposes of organisational design we need not be
too dogmatic about what costs to include. It is more important to have a
thoroughgoing awareness of what is or is not included, and above all to
ensure consistency across all the profit centres in any organisation.
The relative performances of B, C and D are easy to calculate at gross margin
but at gross profit and net profit, the results are affected by the treatment
of shared cost. There is, for example, a case for charging the cost of E direct
to PL.
Typically, profit centre managers ought to have authority over the main
influences on profit – prices and sales volumes – and the main costs. These
are the discretionary elements of a trading strategy which must operate
within the company’s policies on markets, capital, financial returns (return
on sales and/or return on capital employed) and matters of corporate
compliance.
4.6 Support functions
The scale of the profit centre in relation to its parent organisation often
makes a difference to the entrepreneurial latitude allowed to the profit centre
manager, and the rigidity of the requirement to comply with directives from
the centre. Some core or operational functions such as sales, supply chain,
and manufacture or distribution enable the profit centre business model to
operate. Other functions such as finance and HR support these activities
and thus provide the internal engine that allows the organisation itself to
function. These fixed or semi-variable overheads may or may not be profit
centre functions.
48
B C D A
Revenue 100 120 80 300
Cost of sales 80 90 60 230
Gross margin (GM) 20 30 20 70
Profit centre costs 5 5 5 15
Gross profit (GP) 15 25 15 55
Share of E’s overhead cost 4 5 3 12
Contribution: B+C+D = A 11 20 12 43

49
Support functions fall into four types, two weak and advisory, the other two
stronger and more authoritative.
Shared service support
The simplest of the four types of support are the transactional and generic
services common in most organisations – financial accounting, HR, IT, legal
services,estates,HSEandsoon.Conventionally,theyweredepartmentalised
but are increasingly given quasi-autonomous status as ‘shared service’ units.
They normally report into the organisational centre and support the entire
enterprise, but they can be wholly or partly positioned within operating
functions or business units, which is to say they are devolved. Sometimes
they exist both centrally and in a devolved part of the structure. However,
too much unnecessary duplication of the same or similar processes at the
centre is a common and expensive mistake.
Project or coordinating units
These usually consist of multi-disciplinary teams set up to pull together
core operating, functional work on specific projects, usually in support of
initiatives or innovation related to introducing new technology, products,
markets or assets. Such units are frequently temporary and their outputs are
advisory rather than authoritative.
Business units Business units Business units Business units
Shared services Policy
Projects Operational core
CEO

50
Policy units
Support also comes in the form of functions that dispense governance and
policy. In very large companies these may be head office organisations like
finance, HR, corporate communications and legal. They can be positioned
in the person or office of the most senior executives in the organisation.
They are powerful, but the power is often confined to professional, technical
and behavioural standards for processes and managers, rather than
to operational and executive management. In large organisations they
sometimes have powers to instruct or recommend the horizontal integration
of functions performed in devolved business units. An example of this would
be where a policy function mandated the coordination of sourcing supplies
for several brands or products.
In time, every post tends to be occupied by an
employee who is incompetent to carry out his duties.
Work is accomplished by those employees who have
not yet reached their level of incompetence.
Lawrence J Peter
Quasi-core functions
A fourth type is less administrative in character and more direct and
influential in supporting executive operations. Examples include quality
assurance, engineering, logistics and even marketing. The classifying of
these as support can be debatable; the processes used in carrying them out
need careful study to decide if they add value directly or simply support the
functions that do. So they will either occupy a place in the structure at the
corporate or business unit centre, or as a mainstream operating function.
“
”

5.1 Corporate parenting
If policy, mission or objectives have been clearly articulated, the role of the
corporate parent in owning profit centres, companies, divisions, and other
devolved entities that have been given degrees of autonomy, ought to be
obvious. There should be more than a hint as to why devolution is the policy,
and what purposes the devolved business unit structure is meant to serve.
This is best practice because a corporation, as it grows, often finds it
increasingly difficult to demonstrate the value added by the centre. It is
sometimes alleged that organisational centres (head offices) destroy rather
than create value, and that if the corporation is a group of companies it may
be worth less than the sum of its parts. Thus it may come under pressure to
justify its existence, or at any rate its size, by demonstrating to investors that
the centre is a powerhouse of governance and policy that forces financial
performance and is astute strategically.
To avoid this trap, the design and composition of the centre should, ideally,
be put through the same investigation and tests on structure as any of its
constituent businesses or business units.
There can be great reluctance to do this. The central organisation owns its
own autonomy, and usually chooses to avoid the stress brought on by any
challenge to its form or performance. However, avoiding the issue altogether
invites conclusions about ‘them’ at the centre and ‘us’ in the working
businesses which may reverberate unhealthily.
Collinson Grant has worked with a number of organisations to define
the function of the corporate centre and to reshape the relationship with
operating divisions. In these examples, the centre can be seen as:
n a dispassionate investor/shareholder – offering minimal strategy and few
managers or operational capabilities, but simply demanding a return on
its investment and compliance with a number of audit requirements
53
5 The corporate centre

54
n a parent of a number of loosely integrated and ‘lightly’ managed business
units – applying a fundamental strategic direction, allocating support,
and providing common ‘governance’ processes and controls, but leaving
many important and all operational decisions to local, accountable
managers
n a parent of tightly managed business units – applying numerous controls
and making sufficient interventions to give local managers a well-defined
set of guidelines within which to operate.
When managers take the first view – that the group or ‘PLC’ is an investor –
they rarely intend that as a criticism. Rather, it reflects a desirable state
of affairs. They can run their operating units as completely independent
businesses, with their own cultures and identities, and have the autonomy
and freedom to develop them as they see fit.
However, when managers take the third view – that their operating units are
tightly controlled – they often do intend that as a criticism. They reckon they
have too little discretion in forging their own destinies. This view could be
shaped by a range of factors: the history of the business within and outside
the group; the personal relationship with the relevant people at group; the
extent to which the activities of one operating unit are tied up with those of
others; or a perception of unnecessary intervention from the centre.
There is generally a recognition that large and diverse groups do need to set
some common disciplines and controls in evaluating the risks of projects
or acquisitions. ‘Codes of practice’ can also be extended into ‘soft’ areas
of managerial practice and behaviour. Such a mechanism balances purely
financial measures of performance. And compliance (or otherwise) with the
standards gives group another indicator of the health of each of its operating
businesses.
5.2 Structure and operations of the centre
Amajorchallengeforanyambitiousbusinessistoensurethatitissufficiently
robust to support growth and then sustain performance. This naturally
encompasses the function of the centre. Pertinent questions are:
n Would the development of the function and responsibilities of the centre
help to improve control? Or would it simply add bureaucracy?

55
n Does the group rely too heavily on the personal intervention and
commitment of a few vital managers, stretching them to the limit?
n Does a ‘hands off’ approach work for established operating units? Would
it be suitable or grossly inadequate for newly acquired or fledgling
businesses?
n How effective are the structure and the supporting systems and processes
at recognising and responding to early warning signals?
n Are the basic assumptions – particularly about the competences needed
and the market focus – apt?
n Is the configuration of operating units appropriate? Could they be
redefined and reduced in number to create discrete strategic business
units, each with a unique service capability and/or market focus?
5.3 A model for the function of the group centre
What elements of strategy, policy, standards, processes and systems should
be imposed from the centre? How should they be put into operation? How
should their effectiveness be measured? A framework to express this is
shown in the table below.
As a starting point for illustration, the HR strategy and the core processes
that support it could be mapped into this framework. An assessment of what
needs to go where would have to consider:
n the risks and implications of diversity versus commonality
n the nature of the competences required
n where specific capabilities should sit within the overall organisation.
This approach can be applied to any of the functions and processes within
a business. In aggregate, it would define and size the centre relative to the
organisation as a whole.
Functional strategy/process Strategy and policy – led by
Centre Local operating
company
n Operational
execution
n Led by
Centre
Local operating
company

59
6.1 Three degrees of complexity
The matrix has evolved as a solution to accommodating accountabilities
within complex devolved structures. It was originally popularised by the
white goods industry in the United States about forty years ago. Initial
enthusiasm for it had, after a decade of application, given way to more mixed
opinions. But it was the first conscious attempt to define a way of dealing
with corporate organisational challenges at their most complex – situations
where authority over people is split because managers or employees report
and are accountable to two or more managers. They may be required to take
instruction from, and therefore to have performance judged by, two bosses
(sometimes more) residing in different parts of the organisation.
Behind this is the problem that as organisations become larger, spread
their activities over more markets, products, and sources of supply, they
respond by becoming increasingly devolved but laterally wired together.
As this feature develops, it becomes harder to govern any one activity from
just a single point of management. Structurally, the matrix is neither wholly
devolved nor wholly integrated.
The key to being a good manager is keeping the people
who hate me away from those who are still undecided.
Casey Stengel
The big question that haunts all matrix organisations is whether conflicting
instructions will descend on managers from different directions. If those
instructions are about different aspects of their work and responsibilities,
they may be mutually compatible. If they are about the same aspects, they
may conflict as one authority within the structure struggles to impose itself
on the other without a clear-cut right to do so.
Organisations usually face this crunch when they become sufficiently
large and devolved that they sell multiple products in multiple markets
(particularly in different countries) and/or through different channels and/
or fed by alternative sources of supply.
6 Matrix structures
“ ”

60
The matrix structure has many variants and exists on a continuum in which
it becomes weaker or stronger towards the extremes. However examination
is helped if we can recognise three main categories. The first we can call the
loose (or weak) matrix because it is the least organisationally challenging.
The second we can call the strong matrix. Finally, there is the classic matrix,
an application of structural design where complexity is at its greatest.
6.2 Loose matrix structures
The loose principle
The box diagram below illustrates the principle of loose matrices where
problems of conflict are usually avoided because each authority exercises
it in a different sphere of responsibility for process. This can be illustrated
as follows.
We call the managers on the A, B, C, and D axis executive and those on the
1, 2, and 3 axis functional. Functional managers control the training, skills,
standards and HR aspects of work within spheres such as engineering,
design, product and marketing centres of expertise, finance, and support
functions generally.
Executive responsibilities are for project or operational activity and therefore
have an executive character. Executive managers may second or adopt
managers and staff who are based in the functional departments which are
their ‘professional home’ (such as finance or HR) and redeploy them to the
executive tasks for which they, as executives, are responsible. The executive
or operations departments, such as production, sales or projects, may be
organised as devolved business units based on geography, markets, type of
customer, types of product or operational execution.
Executive
Functional A B C D
1
2
3

Business
unit 1
Business
unit 2
Business
unit 3
Business
unit 4
Corporate
projects
Shared services Policy support
CEO
Business
unit 1
Business
unit 2
Finance
support
Business
unit 3
Business
unit 4
Corporate
projects
CEO
61
Thepeoplepositionedlikethis(inthecellsofthematrix)looktotheexecutive
manager for day-to-day operational direction and to the functional manager
for methods, standards, and other matters of compliance. Managers on
both axes have formal authority within a defined line and staff structure,
the key point being that in this weak matrix their powers do not conflict
because they are each managing different aspects of the employee’s work
and performance.
The diagram indicates four types of function we can associate with a weak
matrix:
Policy support function matrix
This construct was discussed at length in Section 2 where we drew a dotted
line between a finance support executive, such as a management accountant,
devolved to a business unit and a functional boss, such as a finance director,
who lives in policy support usually found at the level of the company or unit
board.

62
This type of loose or weak matrix works easily and is often a brilliantly
effective way of strengthening the capability of business units by devolving
to business operations the skills which have been honed in the specialist
functions under the functional head. This can all be achieved without the
functional head giving up control over the quality, standards, methods
and compliance of which it is the originator and guardian. On charts of
departmental organisations we saw this illustrated by a dotted line.
In the case of the finance function, these are the accounting practices,
reporting timetables, career paths, et cetera, where compliance with a
centralised view of life is accepted as normal. The requirements of the
executive manager on the other hand are likely to be the day-to-day
execution of the operational work which, in the case of the finance executive,
might include management accounting analyses, financial and commercial
research, the gathering of data and period reporting.
Almost all quality improvement comes via
simplification of design, manufacturing...
layout, processes, and procedures.
Tom Peters
Project function matrix
This construct can be organised in two ways. Individuals, teams and
managers may be based in a central project support pool and be assigned to
projects based in business units or corporately; or individuals can be drafted
from business units into project teams.
For the duration of the project, the day-to-day reporting line is to the head
of the business unit or corporate project leader, whichever is appropriate in
the context. Projects may be long or short and be continually or periodically
regenerated.
There will be a line of reporting (call it dotted) to the head of the central
professional resource located in Projects, such as a Chief Engineer, Chief
Research Scientist or Chief Designer. Career and professional matters remain
the province of the professional ‘father’ in Projects.
“
”

A benefit of the matrix structure is that people having similar sets of skills
are pooled together, giving them a sense of being at home professionally and
promoting the collegiate environment so important in getting professional
skills to flourish. At the same time, the diversity of experience acquired
in project-type secondments promotes experience exchange within the
professional group.
Project support matrices in some organisations may simply be a way of
organising off-line projects. For others it may be central to the business
model when, for example, very large amounts of value are added by research,
design, and development or when the essence of the business is to perform
projects or contracts for customers. In such cases the resource allocation
process is always operated from the professional centre. This can sometimes
become the subject of conflict where resources are scarce for specialist skills,
experience or talent. Generally however, the solution works very well and for
larger companies that struggle with using project resources flexibly across
many contracts, no better solution is likely to be at hand.
Shared services support matrix
Shared service units, the most common of which are the accounting
and human resources functions, can perform the transactional services
supporting an organisation’s business units more economically, efficiently
and effectively than when each discretely manages its own. But this is a very
weak form of matrix. Although the head of the service has two bosses, one of
them, the business unit head, is really more like an internal customer, and
their relationship, typically, is governed by service level agreements.
63
Division A Division B Division C Division D
Project 1 Project 2
CEO

64
Some shared service units combine with, or are an element within, a policy
support function (see page 61) in that their scope may extend beyond the
transactional and into advisory support and surveillance of compliance.
Core functions support matrix
In devolved structures, business units may be given a focus on customer-
facing activities like selling, marketing, and country management, to the
exclusion of other core activities like manufacture, supply or logistics.
Those could, like the customer-facing activities, be devolved. Instead,
these processes are organised as a unified central resource. In this type of
structure, they are cast in the role of functions supporting business units.
This matrix structure is usually very loose. Nevertheless conflict may arise
if managers in, say, logistics find themselves sandwiched between policy
supportdirectiveswithwhichtheyshouldcomply(forexample,aprohibition
on the expensive use of air freight) and the requirements placed on them by
business units using their weight by behaving as internal customers (and
who might insist on an air freight delivery).
Every company has two organisational structures:
the formal one is written on the charts; the other
is the everyday relationship of the men and women
in the organisation.
Harold S Geneen
6.3 Strong matrix structures
Matrix structures for support functions and projects are usually
straightforward to organise when the weak matrix structure is used. Then,
although managers and staff may have two bosses, the lines of authority can
be easily defined. Generally behaviour is intrinsically weighted in favour of
compliance with a minimum of conflict.
None of the above applies to the various forms of strong matrix where
one authority directly and uncomfortably overlays another. Here power
and political issues can be daily challenges. Conflict may be inherent and
it becomes more difficult to draft charts illustrating how the organisation
“
”

65
works. The strong matrix becomes not so much a solution as a compromise,
springing from a recognition that there is no better alternative.
The most common case is the problem of needing to organise managerial
activity by both product and market, and to hold the managers of both
equally to account. It is at its most acute where markets are associated with
countries that need dedicated (often resident) managers; where there are
markets associated with sales channels that require specialists; or there are
products with multiple branded lines.
Before the best compromise structure has been put in place, the problem
may appear to look like this:
Selling more than one product through more than one customer into more
than one country is an organisational problem that almost certainly cannot
be resolved by structural means alone. Nevertheless, an attempt must
be made to construct a balance of power between managers of product
operations (supply, manufacture, commercial, sales and marketing) and
managers of market operations (in-country, in-house, in-territory, and via
channel).
Loose/tight matrix
It may be acceptable to recognise that responsibility for the product is
subordinate to that for the market. The managers of the market may be
given a strong remit to determine, by acceptance or veto, the properties of
the product, its price structure, its methods of distribution, et cetera.
Market 1 Market 2 Market 3 Market 4
Support Support
Support Support
CEO
Product 1 Product 2 Product 3

66
In effect, the product managers become a support function as shown below,
where activity is mainly being driven by the market managers. However,
each market manager may see the requirements of each market as different.
This may create difficulty and expense for product managers who, in theory,
could be producing a different version of each of three products for each of
four markets, making twelve variations in all.
This can be mitigated by a variant of the structure above, illustrated in the
next diagram opposite. It creates a role for an additional senior manager to
pull together the management of all markets. This post would be filled by, for
example, a group or global sales or marketing manager, or chief operating
officer. This solution gives some stability to the relationship between product
managers and market managers, and will help to organise a more rational
strategy for products than might otherwise be the case.
The business unit matrix
The loose/tight structure may not be ideal if it would underplay the
legitimate contribution of product specialists with their intimate knowledge
of the properties and potential of their products. It may take a manager who
can be more dispassionate about how a product should be presented and
taken to market.
So, in the variation illustrated above, business units are formed to cover the
multiple products and markets. But this is at the expense of fragmenting
the managerial task over the four markets. This may be destabilising if the
markets overlap and the same customers inhabit two or more markets. It
also creates an extra layer of management cost.
Market 1 Market 2 Market 3 Market 4
Support Support
Product 1 Product 2
Product 3
CEO

67
A more rational structure would show the above situation in a variation
illustrated below. But its feasibility might depend on whether the
organisation can report profit centre numbers with enough information
about costs and revenues.
Neither this nor many of the other matrix structures can be efficiently
managed and controlled without systems for accurate period reporting and
performance measurement of both process and management.
Business unit 1
Product
Market
Function
Product
Market
Function
Product
Market
Product
Market
Business unit 2 Business unit 3 Business unit 4
Support Support
CEO
Market 1
Product
profit centre
Product
profit centre
Market 2 Market 3 Market 4
Support Support
CEO
Product
profit centre
Product
profit centre
Product
profit centre
Product
profit centre
Product
profit centre
Product
profit centre
Product
profit centre
Product
profit centre
Product
profit centre
Product
profit centre

68
In any great organisation it is far, far safer to be
wrong with the majority than to be right alone.
John Kenneth Galbraith
6.4 The classic matrix
Product and market matrix
All the examples of matrix structures so far, both of the loose or weak type,
and the strong, have been designed to mitigate the potential for conflict that
can arise when people have two bosses – an organisational structure where
they have more than one line of direct reporting.
The classic matrix meets head on the challenge which this presents. In
doing so, like any dog with two masters, the manager or staff member in this
position yields less to authority and more to their own sense of judgement
and, hopefully, of responsibility. Clear line reporting breaks down.
In the illustration the market/product profit centre managers shown in each
cell are in this position. They are accountable for the sales revenue and
the margin yielded from each product by each market in the matrix. Note
that in the example it cannot be assumed that all markets sell all products.
Note also that an individual salesperson or manager may be responsible
Functional Product
business unit 1
Product
business unit 2
Product
business unit 3
Market
division
1
Market/product
profit centre
manager
Market/product
profit centre
manager
Market/product
profit centre
manager
Market
division
2
Market/product
profit centre
manager
Market/product
profit centre
manager
Market
division
3
Market/product
profit centre
manager
Market/product
profit centre
manager
Market/product
profit centre
manager
Market
division
4
Market/product
profit centre
manager
“ ”

for the activity and results of more than one product (cell in the matrix) in
any market. In the chart, the nine cells might be covered by, say, four profit
centre managers.
The essence of this structure is that accountability has passed to the
managers of the profit centres who report to their market and product bosses
with a greater degree of operational discretion than in any of the other
structures we have considered. Their lines of accountability to both bosses
are put more into question and they have a stronger hand to play than in the
other forms of matrix outlined earlier.
For here it is the matrix itself that sets the behaviours. Control and
accountability upward through the organisation is sacrificed to an enlarged
degree of devolution. And, by way of a footnote, the ability to draw a
conventional line and staff chart for this classic matrix structure has reached
the limits of possibility.
There are many variants on this theme when it comes to the detail of who
takes responsibility for decisions which include prices, products, range,
promotional initiatives (their devising, tuning and implementation),
channels and selling methods, and much more. There is no single preferred
approach to putting a structural framework around these. The least to be
hoped for is that the three parties always involved in the classic matrix
work harmoniously and to the same agenda. At best a strong and bright
talent, preferably without excessive ego, emerges in one of the positions in
the matrix, and is supported by the other two. Here behaviour rather than
structure brings success or failure.
The multi-dimensional matrix
In large, multi-market, multi-product, global organisations there are yet
more possibilities for added complexity. The chart below is an organisational
diagram of a well-known global company that seeks to organise by:
n support function – finance, HR, legal, et cetera
n two major groupings of product and two major types of service
n four types of customers defined by industry
n four global regional markets
n four types of customers within multiple countries
n groups of marketing intermediaries defined by channel.
69

70
The product and services dimension on the left hand side operates in a less
heavyweight, more background mode than the market dimension on the
right, which is strongly focused on customers.
All the labelled boxes are managed by an individual senior executive with
conventional line and staff reporting above and below. Clearly, completely
discrete accountability for revenues, margins, costs and profit is impossible.
And strategies for products and markets cannot be the province of any single
executive. Line reporting overlays of product, market, channel, et cetera,
are necessary and critically important in understanding trading events and
performance throughout the business.
The role of the CEO becomes more crucial by several orders of magnitude
in larger companies with these complexities. It is a paradox that such
organisations may need to fall back from more and more devolution towards
a greater degree of centralisation just to hold products and markets together
Major product
group
Major product
group
Global
services
Head of
markets
Support function Support function
Support function Support function
CEO
Region 4Region 3Region 2Region 1Service BService A
Customer group 1
Customer group 2
Customer group 3
Customer group 4
Customer group 1
Customer group 2
Customer group 3
Customer group 4
Customer group 1
Customer group 2
Customer group 3
Customer group 4
Customer intermediaries

in a coordinated way. The CEO may talk about executive performance to a
conference of executives rather than to individuals in turn.
However, the logic of this business model is that the monolithic structure
brings a competitive advantage of its own. The matrix transforms from a
devolved structure into an integrated one, squaring the circle and bringing
us back to the integrated model we first examined.
The secret of all victory lies in the
organisation of the non-obvious.
Marcus Aurelius
Far from being undesirable, with maximum devolution seen as the ideal,
this monolithic matrix may become the best option in preventing the empire
succumbing to disintegration and sprawl. The history and size of the global
empire, in this case, has created a business model whose competitive
advantage is its scale and hegemony.
In this organisational environment, solutions to accountability and control
may involve introducing structural features such as multi-disciplinary
committees led by the chief executive; or a chief operating officer, positioned
at the node where policy is set, who can review strategy and process with
precision and regularity. This is also the ‘land of the project’ where most
competitive initiatives are conceived, planned and implemented by multi-
disciplinary teams of executives seconded from around the organisation.
6.5 The matrix environment
To adopt or not to adopt a matrix structure
The matrix design problem is at its most acute in deciding who should
have accountability for taking products to market. As we have seen in
the strong and the classic matrix models, managers with territorial
responsibility (say regional or country managers) and managers with
product responsibility compete for influence over the same piece of business.
Although responsibilities for roles and tasks can be differentiated and the
exercise of authority coordinated, it is in measurement and the fulfilling
of accountability, the ‘who got the sale’ issue, where most difficulties tend
to lie.
71
“ ”

72
But matrix structures are often also employed when there is no particular
structural solution being sought. Some CEOs seem strangely drawn to them
for reasons that have nothing to do with necessity but more because they
seek to avoid conflict with subordinates in handing out jobs. They tend to
be indecisive about whom best to hold to account for results and the matrix
structure fudges this problem nicely. The structure is adopted by choice, as
a convenience, not because the context requires it.
However to put a more positive spin on this, matrix structures are rightly
associated with that style of management which likes to emphasise
cooperative values, teamwork and harmony, and collective accountability.
Nevertheless, we are not fans of the matrix, and to unnecessarily embrace a
complex structure when it has the effect of complicating measurement and
accountability cannot be recommended. If certain behaviours are thought
desirable (like teamwork and cooperation) it is better to induce them through
training and incentives, or by the way in which people are managed.
The faux matrix
It is common for managers to refer to a structure as a matrix when it is
actually nothing of the kind, but merely drafted to appear so. Not all charts
drawn in the shape of a matrix operate as one in managerial terms. The
process of managing, and the line reporting structure may, in practice, be
one which is quite conventional and owes nothing to a matrix except how it
has been represented on paper.
Todistinguishbetweenarealandfauxmatrixitisnecessarytofocusattention
on how the process is actually performed or would be best performed – in
terms of who does, or should do, what and how. In Section 3 we suggested
that when creating a design, the first task is to study processes and the jobs
they require before going on to define accountability and how it is best
measured. The organisational structure can be designed around them.
Note that decision taking is a real process. Decisions, like the processes in
which they are embedded, need owners (those responsible for taking them).
Once the processes for making decisions are recognised, accountability for
them will lead towards the right organisational structure, metrics and KPIs
which will (if there is balance in the organisational structure) confirm it.

73
In drawing the structure, give proper weight to all these points. For example,
the chart below appears to be a drawing of a matrix.
But it is probably not telling us anything other than the chart below, which is
the straightforward story of an organisation of four business unit divisions
and three of the usual centralised support functions.
In practice there is unlikely to be a process in which the heads of the three
functions line report to any, let alone all, of the divisional business unit
heads as the matrix version of the chart appears to suggest.
Division A
Finance
HR
Division B Division C Division D
CEO
IT
CEO
Division A
Finance
HR
Division B Division C Division D
CEO
IT

74
Again, if there are three production plants each supplying some or all of
the four trading divisions (chart below) then the relationship may be one
in which each plant manager reports to four divisional heads who instruct
their subordinates about requirements for supply. But it is more likely that
the divisions act simply as internal customers and the plant managers line
report to the CEO, even though the chart is not drawn that way.
The correct way to draw the most likely relationship is shown below so that
the line reporting is clear and layers and spans are made transparent.
Drawing matrix diagrams like that below is fine. It indicates usefully which
plants supply which divisions – an aspect of the organisational story. But this
is an organisational explanation, an annotation of the true organisational
chart that will or should properly show line and staff reporting.
The grid diagram below probably also tells the truth about the operating
relationships where PM is a product manager. But this is not a line and
staff chart, merely an explanatory memorandum in support of a process
description.
Division A Division B Division C Division D Plant 1 Plant 2 Plant 3
CEO
Sales and
Marketing 1
Sales and
Marketing 2
Sales and
Marketing 3
Engineering PM PM PM
Manufacturing PM PM PM
Commercial PM PM PM
Plant 1
Plant 2
Plant 3

7.1 The property of control
Control needs to be an intrinsic element of organisational structure. A
structure is only in balance and fit for purpose when it is anchoring the
controls that sustain every healthy business. This is largely achieved by clear
accountabilities and well designed, transparent business reporting. These
themes have already been discussed but even so are worth a final look.
The essence of control is accountability. This is present when a structure has
in its chart, or within its narrative description:
n clarity on who reports to whom, for every managerial node in the structure
n clarity on which managers own (are accountable for) which processes
n measures of process economy, efficiency and effectiveness which use data
on financial and resource inputs, the ratios for transformation of inputs
into outputs, and the outcomes achieved against expectation or target
n KPIs (key performance indicators) to measure the fulfilment of
accountabilities. KPIs can be stand-alone salient figures, but more
frequently are expressed as ratios or the relationships between two or
more bits of data owned by a manager (such as ‘average units produced
per hour’)
n a logical hierarchy of measures that cross refer to, help quantify, or
reconcile with the financial reporting of the business, and are free from
ambiguity.
7.2 Accountability and responsibility
Because both of these terms have resonance in designing managerial
structures, it is useful to avoid using them interchangeably. Responsibility
is a general term that relates to a job holder’s duties in an operational or
functional field of activity – the scoping of the job. Thus one can be held
responsible for health and safety, for the credit control department, for eight
people, or for teamwork.
Accountability introduces the concept of accounting for performance and
results – measurement, appraisal, and judgement about degrees of quantity
and quality for all the things for which there is responsibility.
77
7 Control and structure

78
Thus a manager with responsibility for health and safety might hold
accountability for there being no accidents. The manager responsible for
credit control could be accountable for limiting debtors over sixty days to
2% of the total. Typically, accountability is associated with the KPIs of the
department or, for the financial results, against the budget for the division.
It is not possible to describe a job definitively without adequate reference
to both of these terms, but best practice has it that without accountability,
control of business management will be weak.
In a well-designed structure, accountability and measurement will be
clear. They will be robustly supported by forms of management accounting,
particularly those that unitise costs and margins, use standards for both of
these, and then employ them as the foundation of the system of budgetary
control. A structure cannot reach its optimum design unless underpinned
by these properties. For this reason business reporting is a key component
of organisational design.
Reduce the layers of management.
They put distance between the top of
an organisation and the customers.
Donald Rumsfeld
7.3 Business reporting
The design of the monthly management information pack and its associated
analyses are rarely thought of as an aspect of designing organisational
structures. We take the opposite view. Organisational structure ought not
to be reviewed without including detailed reference to the financial and
operational numbers that quantify the scale and shape of an organisation,
and point to how it is to be controlled.
The numbers for which managers are held to account are an essential
element in defining jobs and placing them in context among all the others
in the organisational hierarchy. Whether we are measuring sales turnover,
payroll cost, supply chain, inventory values or whatever, the more senior the
job the more money is at stake.
“
”

So, the most important principle to be observed is that of ensuring the best
possible congruency between the structure of organisation and the design
of business reports. No managerial node should be outside the scope of
business reporting, and most if not all should be recognisable from a cross
reference to the business reporting suite.
It ought to be possible to make a good fist at drawing a structural chart from
a sufficiently detailed set of management accounts, and conversely, to base
the design of a reporting suite by working back from a chart.
79

81
8. Shared services and outsourcing

Cost can be saved by bringing routine transactions into one ‘shared service
centre’ offering economies of scale and efficiency. But there are pitfalls. Not
all organisations benefit. Checks should be made to determine whether
the right conditions exist at the outset. Similarly, outsourcing can be
advantageous, but also carries certain dangers.
8.1 Shared service centres
Shared services generate real savings by reducing costs, applying consistent
standards and focusing on value-adding activities. This is not about
‘centralisation’. All parts of the business need to trust the shared service
centre (SSC) and to delegate the handling of the selected activities to the
unit without trying to ‘dabble’ or ‘interfere’ in day-to-day routines. The SSC
should take on tasks that allow it to exploit economies of scale and provide
the critical mass necessary for the cost-effective implementation of improved
systems. But it can be remote from its internal customers and not exposed to
day-to-day problems in the field. And every new structure within a business
creates its own reason for being, tensions and rigidities – and attracts its
own particular costs. There can be other pitfalls:
n consolidating effort in one place tends to diminish rather than improve
the flexibility to reduce costs quickly in difficult times
n a ‘centrist’ approach can discourage innovation and stifle the essential
initiative to respond promptly to change
n servicelevelagreementsarenotinfallible;performancehastobemeasured
and managed robustly
n costs can escalate rapidly when new tasks are taken on board without a
thorough review of how else they could be done.
The investment required to create truly common IT systems and ways of
working can be underestimated. The transition can become protracted and
expensive. Duplicate people and equipment may be required for training
and activities at the new centre long before savings can be made elsewhere.
83
8 Shared services and outsourcing

84
8.2 Do the right conditions exist?
A well designed and run SSC that supports the business effectively can
benefit cost, service and operations. But not all businesses lend themselves
to this approach and some questions need to be answered before an
expensive project or investment is launched:
n Is there really a case for combining activities in an SSC?
n Are there high-volume, common, repetitive activities that will give
economies of scale?
n Do activities have links to the front line that cannot or should not be
broken?
n What personnel will still need to be retained in busines units (BUs) to
handle local administration and queries? (In small units, the saving could
be minimal.)
n Will combining make it easier or more difficult to automate?
n What investment is required?
n Are the processes robust or do they need to be upgraded first?
n Could many of the savings be achieved by other/simpler means?
n How much disruption and cost will the transition cause?
n How long will the transition take, allowing for setting up and training?
n What risks could the business face from severe disruption?
Assess the business case rigorously
The characteristics of the business and the investment required do not
always provide a ready business case for creating an SSC. The nature of the
transactions does not always make it easy for a remote team to handle them –
particularly where local links to the customer are important to avoid disputes.
And a common failing is to underestimate how many employees have to be
retained by local units to handle queries and other administration when
some duties are taken over by an SSC. There are cases where organisations
go ahead and invest to create an SSC without accurately assessing the costs
and benefits, only to discover that the expected savings cannot actually be
achieved.
In the right environment, savings in the region of 20% should be achieved by
the introduction of shared services for transactional financial services and
internal services such as payroll. In other cases the analysis might question
the very need for the function at all – or at least suggest that it need only be
retained on a much smaller scale. In HR, for example, there is often scope

85
for considerable cutting back. For other functions such as training, legal
services and property, the benefit will show in improved service.
Initial cost savings can be quickly eroded without rigorous control and
clarity in accountability. To maximise effectiveness, managers should seek:
n to install accurate measurement of results and related rewards
n to analyse carefully the organisation’s current costs and services
n to define the core processes with adequate flowcharts and supporting
narrative
n to benchmark practices internally and against other companies
n to use measures that are simple, tangible and acceptable to everyone
n to focus on ‘leading indicators’ of business drivers to forecast results
n to establish processes for continuous review and improvement.
It is also vital to assess carefully the impact of proposed changes on the
relationships with customers and agreed service levels. Do not proceed
without a clear plan that demonstrates the return on investment (with a
healthy contingency for the cost of IT).
Shared service centre: evolution and maturity
SSCs allow an evolutionary approach to raising performance and reducing costs
Multiple
systems
Non-standard,
complex
processes
Multiple
locations
Headcount
100%
Remove
complexity
Non-standard,
simple
processes
Multiple
locations
Multiple
systems
Headcount
80%
Multiple
systems
+
standardise
processes
Standard,
simple
processes
Multiple
locations
Headcount
60%
Common
systems
+
implement
common
system
Standard,
simple
processes
Multiple
locations
Headcount
50%
Common
systems
+
establish new
organisation
+
centralise
transaction
processing
Standard,
simple
processes
Single/few
locations
Headcount
40%
e.enabled
system
+
Implement
e-enabled
common
system
+
virtual
transaction
processing
Standard,
simple
processes
Virtual
locations
Headcount
25%

86
So the design of an overall service level agreement (SLA) for the SSC must
reflect all the requirements of the enterprise. Some of the most important
might be quite difficult to define and measure. There is no value in having
a robust credit control department that minimises the debtor days, if it
alienates all your customers so they stop buying. There has to be a balance
between the hard measures and the softer needs of the organisation.
Do not underestimate the risks
There have been some very high profile cases in which the risks to the
business were not understood or managed effectively. Some years ago,
one of the UK’s largest engineering and design consultancies set up a
shared service centre to handle, among other things, its sales ledger.
The transition encountered severe problems and for a time the business
was unable to invoice all of its customers properly. The consequences
for cash flow were so drastic that what was normally a very successful
corporation was at risk of failing. Thankfully it recovered quite quickly,
but its experience is a warning for all.
8.3 Outsourcing
Outsourcing and sub-contracting have become something of a panacea.
‘Our own business does not do it well, but there must be someone out there
who can do it better!’ And that may be true. But it does not follow that
outsourcing is the best solution.
Potential
pitfalls
to guard
against
Sub-optimal
performance
if left
unchecked
Failure to control
SSC drives costs
BUs lack the power
to challenge SSC
SLAs met but
delivery still poor
Measures not
aligned to outcomes
SSC unilaterally assumes
new functions
Inadequate
benchmarking of costs
SSC workers are
remote from users
Poor awareness
of BU difficulties

When an activity is ‘non-core’ there may be a case for considering jettisoning
it completely and asking another organisation to provide that service under
the terms of a robust contractual agreement. Well defined back-office tasks
can be obvious candidates, but other support jobs might be suitable. Even
outsourcing a more central process may offer real benefits. Other providers
may be able to offer and maintain a critical mass in expertise and the lower
labour costs that come from remote location. Many tasks, administrative
and otherwise, have electronic inputs and outputs, so work can easily be
’exported’ from country to country.
But not all outsourcing ventures by large banks, for example, have been
successful. In call centres, product information, local knowledge and
even language have become major considerations. Services for directory
enquiries and train timetables are among several that have been prone to
cultural misunderstanding.
There has to be an absolute distinction between the essential competences
that must always be kept within the business and the less critical activities
that could be handled by a third party. But the boundary is not always clear
and could shift. Is it wise to allow even the most mundane of contacts with
a customer to be handled by someone else? And do service level agreements
that tend to measure mainly hard statistics about performance allow you
sufficient confidence that transactions are being managed as you would
wish? Even more than with an internal shared service centre, it is important
to measure the softer aspects to confirm that the customer’s experience of
the service is satisfactory.
The desire to concentrate on a specific market, product, or geographic area
ebbs and flows with the perceptions of opportunity, competition and risk.
Given the time it takes to move from concentration to diversification (or
back), the timing of the decision is vital to success.
Outsourcing is a decision that can take years to reverse – it’s strategic!
When a business jettisons one of its core competences, it may become
just a secondary player competing in a different market. While that may
be the intention, it can have unexpected consequences that turn out to be
irreversible.
87

89
9. How lean structures keep
down cost

9.1 Good design is naturally lean
By adhering to the guidelines on good design described in this book the
structures created will tend towards being leaner than if a more instinctive
and unsystematic approach is taken. The following are examples drawn
from earlier in this book:
n Minimise layers and make structures as flat as possible.
n Optimise spans and apply the 8 x 5 test.
n Avoid one-to-one line reporting.
n Distinguish between operational managers and specialists who may be
important but are not true managers (because they do not or should not
have subordinates).
n Consciously seek to limit the number of managers.
n Look for misshapen, inelegant, unbalanced and overblown structures that
suggest it would be possible to have fewer layers.
n Cross-reference layers and grades but do not allow grades or levels to
determine the number of layers.
n Take job design seriously and, if the business is large enough, use a
design specialist to combine multiple tasks within discrete jobs and thus
minimise hidden unproductive capacity.
n Achieve the best balance between functional support jobs and operational
jobs, avoiding duplication and considering shared services.
n Challenge the need for functional overhead everywhere, including the
corporate centre.
n Check that strategy is derived from a fully articulated business policy so
that both give rise only to effective processes.
n Base organisational structure on processes of proven relevance to the
business model (that they add value).
n Check that accountability can be measured at every managerial node.
n Check for congruence between the structures of organisation and
management’s financial and operational reports.
n Employ matrix structures only when they accurately reflect how the
organisation has to work.
91
9 How lean structures keep down cost

92
9.2 Mapping the functional alignment of the population
One of the techniques for examining the leanness of a structure is to locate
and quantify population by its functions and professional taxonomies
throughout the indirect staff of an organisation, which will typically account
for at least 60% of the total headcount even in manufacturing companies.
The larger and more complex their structures, the more they tend towards
matrix characteristics and have multiple levels of devolved management, the
more they are likely to conceal the real number and locations of occupations
of the same type.
For example, in some organisations jobs in finance, HR, logistics,
engineering, and other generic occupations can be found scattered across
a wide range of territorial or business unit locations. And resource groups
such as secretaries and clerical support grow and take root in an unplanned
and unjustified way. There is a simple analytical technique that can drill
down into the organisational data and map, quantify and profile these
groups or functions. Other generic types of jobs such as direct and indirect
blue collar production, the head office staff and so on, can all be included
to give a comprehensive view of the type illustrated in the following table.
The data above is taken from an actual case and shows, for example, that in
this integrated group employing just over 3,000, nearly 5% have financial
jobs that are threaded through just about every division. In fact all the
customer-facing divisions carry out their own invoicing, a feature never
previously noticed let alone questioned by senior finance managers at head
office because they never thought of it as relevant to their own function. The
view given by the chart data suggests another option.
Because conventional business reporting rarely provides any analysis
from this perspective, when such data is exposed for the first time it can
reveal a worrying profile of jobs scattered randomly without process logic
or functional theme. This can be the key to opportunities for making the
organisation leaner by realigning the structure so that the same or similar
jobs are given more effective managerial cover and perhaps assembled into
more robustly designed and effective processes.

93
FunctionGroupHQCustomer-facingdivisions(CFDs)OperationsTotal%
Head
Office
Service
Division
Export
Division
Customer
Division
Energy
Division
Industrial
Division
CFD
total
Manufact’g
Division
Group
Purchasing
Group
RD
Operations
total
Group
total
Functional
profile
Finance3241612174998127191494.9
HR593738244520194431474.9
IT73421321213311.0
FacilitiesManagement336325110.4
HSE21213911100.3
Legal511060.2
Communications71251422314331.1
Supportfunctionsub-totals1101244222995202413047538712.8
ExecutiveDirectors660.2
Secretarial/Clerical5020319456316824144422608.6
ProjectManagers6227161287410515953.1
BusinessManagers33812336921912334126
BusinessImprovement1452102153210321.1
Resourcegroupsub-totals632910139491373555834910151917.2
Logistics110731312201281185417230110.0
Purchasing9111415448496561522458.1
Commercial114118240250.8
Quality21582154010515571.9
CustomerServices214115320321.1
Sales2211130150.5
CorporateBusinessDevelopment7347325190.6
Aftermarket920110110.4
Marketing527070.2
Engineering0539131352122302122731956.4
Operatingfunctionsub-totals222818837441714682571382241790730.0
Indirectstafftotals1956933398122403102535620235593181360.0
Directemployees(bluecollar)367524521283158411540896121140.0
Totalgroupemployment195105408122174531134011972147514893024100.0

94
We found in another, even larger devolved organisation, the group human
resources director had no ownership of several hundred HR professionals
scattered about the whole enterprise, and was ignorant of the total number.
Given that each level in the hierarchy of HR managers had no interest in
reducing their own resources, it was entirely fair to conclude that the HR
overhead was not subject to any control over total headcount or cost. The
same was true for most other functions and resource groups. Functional
alignment even of core operating activities revealed some surprises, and
scope for challenging the way things were.
This analysis paints a picture of where each type of employee is located, how
many there are relative to other groups, and how rich or poor the various
managed units are in their different types of staff. In addition, the size of the
organisational centre and what it consists of is also made plain.
Any assessment of the facts revealed by functional analysis has to be
based on judgments. But anomalies, inconsistencies and even previously
unnoticed and possibly scandalous offences against common sense will
come to light. The technique can produce uncomfortable but valuable
reading that, with the right action, can reduce cost when the design of an
organisational structure is redrawn.
A case study in making an organisation leaner by using functional alignment
as a tool of analysis
After a lean project in the company whose functionally aligned population
is illustrated above, the case for structural change in each function
was examined. After looking at the possibilities for improved processes,
including introducing shared services, the following restructuring plan was
developed.

95
FunctionHead
Office
Prof’l
advisory
Trans-
actional
Services
Division
Export
Division
Customer
Division
Energy
Division
Industrial
Division
CFD
Total
Manufact’g
Division
Group
Purchasing
Group
RD
Operations
Total
Group
Total
Notes
Finance2012323558183977110Managementaccountingonlyindivisions
HR1770551515107HRconcentratedinSSC
IT1025775547IncreaseinITresourcetosupportSSC
FacilitiesManagement272211Someoutsourcing
HSE16229
Legal156TakeoutofGHQintoSSC
Communications12622413Centraliseforbetteruniformityandconsistency
Supportfunctionsub-totals224914637783055313130322%reductioninsupportfunctions
ExecutiveDirectors66
Secretarial/Clerical632510156305011126127172Secretarialblitzed
ProjectManagers12271612874151590
BusinessManagers238123369231334126Leftintactduringorganisationalchange
BusinessImprovement1255531
Resourcegroupsub-totals12550158034341142777748142518%reductioninresourcegroups
Logistics1495015659090205NewSSCdepartmentadoptsbetterprocesses
Purchasing155540408181191NewSSCdepartmentadoptsbetterprocesses
Commercial2141182426
Quality2215821540121256
CustomerServices214101103737Casetostrengthenaccepted
Sales222291717Casetostrengthenaccepted
CorporateBusDevelopment77Deemedonlyacorporatefunction
Aftermarket921111Maybestrengthenedlater
Marketing5277
Engineering539131352122422264186Retainscarceskillsasapolicy
Operatingfunctionsub-totals425104914836191513632252224774323%reductioninoperatingfunctions
Indirectstafftotals387930027235776129569533326259147120%reductioninindirectstaff
Directemployees(bluecollar)36752452128315841408811196Virtuallynochangeasdirectsoutofscope
Totalgroupemployment387930063310101113423101011746612402667

LeanGHQ
focusedon
governance

Two-levelshared
servicecentre

Sectorbusinesses’
overheadsreduced
andmovedtoSSC

Group
Purchasing
eliminated

RDoverhead
embeddedin
Manufacturing
Division

96
The effect of its implementation is illustrated in the two charts below. The
measured changes between the base deployment and number of people and
that projected was as follows.
Total Total Saving Saving Profile Profile
Function Old
population
New
population
no.
increase/
decrease
%
increase/
decrease
Old
%
New
%
Finance 149 110 (39) (26) 4.9 4.1
Human Resources 147 107 (40) (27) 4.9 4.0
IT 31 47 16 52 1.0 1.8
Facilities
Management
11 11 0.4 0.4
HSE 10 9 (1) (10) 0.3 0.3
Legal 6 6 0.2 0.2
Communications 33 13 (20) (61) 1.1 0.5
Support function
sub-totals
387 303 (84) (22) 12.8 11.4
Executive Directors 6 6 0.2 0.2
Secretarial/Clerical 260 172 (88) (34) 8.6 6.5
Project Managers 95 90 (5) (5) 3.1 3.4
Business Managers 126 126 4.2 4.7
Business
Improvement
32 31 (1) (3) 1.1 1.2
Resource group
sub-totals
519 425 94 42 17.2 16.0
Logistics 301 205 (96) (32) 10.0 7.7
Purchasing 245 191 (54) (22) 8.1 7.2
Commercial 25 26 1 4 0.8 1.0
Quality 57 56 (1) (2) 1.9 2.1
Customer Services 32 37 5 16 1.1 1.4
Sales 15 17 2 13 1.5 0.6
Corporate
and Business
Development
19 7 (12) (63) 0.6 0.3
Aftermarket 11 11 0.4 0.4
Marketing 7 7 0.2 0.3
Engineering 195 186 (9) (5) 6.5 7.0
Operatingfunction
sub-totals
907 743 (164) (18) 30.1 27.9
Indirect staff totals 1807 1465 (342) (19) 59.9 55.1
Direct employees
(blue collar)
1211 1196 (15) (1) 40.1 44.9
Total group
employment
3018 2661 (357) (12) 100.0 100.0

The proportion of indirect staff – 2, 3, and 4 in the pie charts – fell from 60%
of the whole to 55%, and there was an enormous change in the proportion of
indirect overhead. Overall the effect was as follows:
Even on unchanged turnover, the effect of fitting the organisation to a new
structural disposition was dramatic.
9.3 Core and Support
Another useful analytical tool of organisation is to apply the taxonomy of
Core, Improvement and Support (CIS) to the numbers in employment and
their cost. This can lead to similar results as functional alignment, and the
two tools can easily be applied in the same lean project.
97
Old profile
■ 1 ■ 2 ■ 3 ■ 4
New profile
■ 1 ■ 2 ■ 3 ■ 4
13%
17%
30%
40%
11%
16%
28%
45%
Old structure New structure Change
Sales £50m £50m nil
Sales per employee £16.6K £18.8K +13%
Employees 3018 2661 -11.8%
Indirect staff 60% 55% -5%
Profit £2.2K x 2661 employees = £8.5m +17.1%

98
The nomenclature used earlier in this book – Operations and Support are
synonymous with Core and Support – is normally applied to an analysis of
the total cost of employing people.
In general, Core operations add value to what is supplied to, and paid for
by, customers. This may also be loosely associated with the direct, often
variable, cost of personnel.
Support functions default to being the fixed or semi-variable, indirect or
overhead cost of people. ‘Improvement’, a sub-set of Support, is functional
support activity associated with planned change, the maintenance of, or
improvement to, quality.
The CIS model looks towards an idealised balance between these three
categories of resource. All can be quantified by measuring time spent on
specific activities defined as one of the three categories. The balance of
employees’ time spent between Core operations, Improvement activity and
functional Support can be measured using the Collinson Grant methodology
known as Process Activity Analysis (PAA).
Core 51%
Support 43% Improvement 6%
B FC D E
A
SupportCore

99
The ideal proportions are usually thought to be in the region of 75% (Core),
5% (Improvement), and 20% (Support), although many organisations
struggle to achieve this because they find it difficult to manage overhead.
The result of the analysis provides a strong indication of where resourcing
priorities lie, and whether a change in the population’s structure is desirable.
Even more important than reducing the Support cost is to spot and resist an
adverse trend towards an ever larger Support population relative to the Core.
9.4 Duplication
The biggest scope for organisational change to produce a leaner structure
can often be found by identifying where work of the same or similar sort is
being done in more than one place. This gives sharp pointers to units that
can be eliminated, or combined and reduced.
Functional alignment and CIS often provide pointers towards scope for
change, but careful study of organisational charts illustrating structural
juxtapositions, narratives on job design, and almost any other data
on population profiles can reveal opportunities for amalgamation and
consolidation, merger, elimination, and reduction of activity and population
generally.

101
10. The inter-trading organisational
model

10.1 The inter-trading challenge
When there is inter-trading of goods or services between the subsidiary
companies or profit centre business units of a common parent organisation,
managers in the subordinate businesses and in the common parent may
both need to answer a series of questions:
n What are the best organisational corporate structures and configurations
for maximising the parent organisation’s profit?
n How would that affect profit and loss reporting, and how would the
financial returns of each of the inter-trading businesses be measured
objectively?
n What principles should be adopted and transactional methods used to
report inter-trading business?
n How can commercial and planning disputes between inter-traders be
avoided?
n How should inter-trading prices and terms be set?
n What information systems and support for decisions are needed to
understand and control inter-trading transactions?
Allofthesequestionsarecentraltothecorporatebusinessmodel(seeSection
3). We have already stressed the importance of choosing an organisational
design that supports a trading strategy rather than allowing inter-trading
to develop by custom and practice and thereby possibly distorting profit.
In short the answers to the questions should determine how the structure
reflects the value chain, the trading strategy and the general model.
Nevertheless, designing and managing an organisation that inter-trades
raises significant challenges in which we have taken more than a passing
interest.
The problem with managing either a business or
a prison by periodic rather than continuous inspection
is that the ‘variables’ are likely to be seriously out
of control before the discrepancy is noted.
Anthony Stafford Beer
103
10 The inter-trading organisational model
“
”

104
10.2 The inter-trading business model
Inter-trading becomes necessary when the value chain of a company’s
products or services passes through two or more of its subordinate units,
both accountable for profit and loss. This makes the profit centres into
buyers and sellers of partly finished goods within an enclosed trading
universe, system of exchange or value chain. Despite being locked together
in common ownership, each needs to price and pay for goods and services
within conventional systems of accounting, so that each can make margin
on the transaction and report profit. So, how to do it?
Taking a simple example, an oil company has four subsidiaries: to conduct
exploration; to pump oil from its wells; to refine the product; and to market
the oil. All have accountability for their own profit and loss. Each business
has to quantify the increments in value that occur inside the profit centre as
products move through successive stages of the parent organisation’s chain
of value. The results they report will depend, to some degree, on how the
terms for inter-trading are designed, essentially the price of the trade.
Rank is a great beautifier.
Edward Bulwer-Lytton
Inter-trading prices are used to assign value to the product as the first
business trades with the second (and in a long chain such as oil) the second
with the third and so on. In a fully managed system of inter-trading, these
prices would be fixed by a method determined or approved by the parent
company. For reasons that will become clear, it is barely an option for
the prices to default to a free market price as would be the case between
unrelated willing buyers and sellers not under common ownership.
Business models that rely on inter-trading can be qualified in four ways:
n transactions take place between sibling organisations under the authority
of a single parent or common owner
n the parent organisation is minded to oblige the subsidiary businesses to
sell some or all of their goods and services to each other – it mandates the
inter-trade
“ ”

105
n the sibling organisations add value to the products or services traded
between them, causing them to share the same value chain
n the sibling organisations are accountable for, and report through, their
own, separate, profit and loss statements – they are discrete profit centres.
Normally all four conditions are present when businesses inter-trade.
10.3 Resource allocation and the value chain
In conventional capitalism, resources are allocated in response to supply
and demand, and to cost and price. Prices tend to find their own level when
they are struck: when willing buyers and willing sellers act on the relative
strengths of supply and demand. This is so fundamental to industry and
commerce as we know it that it is easy to forget that other options for setting
prices may be needed in some circumstances.
In an inter-trading regime, the business model dictates that value will be a
function of policy, not of free market economics. Though it cannot be based
on subjective need, inter-trading is nevertheless a process that requires
regulation and intervention. Units bound together by common ownership
and a common value chain cannot create price as the free market does.
In these circumstances, the utmost care is needed to obtain a true and fair
view of the reality behind how incremental value is measured and added
into profit.
The concept of the value chain helps to point up the stages through which
goods or services pass as value is added by each profit centre in the chain.
Each value chain has a unique configuration shaped by the processes and
the positions they hold in the sequence of links. Value chains vary in length
(number of business units in the chain) and size (value added) of the links.
Value chains may be designed deliberately to adopt an effective organisation,
or may be a simple evolution. It is best if an evolved configuration and
methodology for valuing outputs are replaced by a design that fairly and
logically determines where profit and margin should be taken.
Although financial accountants have various ways of dealing with inter-
trading bookkeeping, the information of interest to managers is what
supports such decisions as, in the example of the oil company:

106
n How profitable is drilling?
n What are the margins on refining?
n What is the return on capital employed in marketing end products through
retail filling stations?
The answers are valuable, not because they will be right or wrong, but
because they will, potentially, inform logical, profit-seeking decisions on
investment, cost control, volumes and so on.
In addition, decision-supporting information must help the company:
n to retain control over its business model, particularly when inter-trading
assumes some complexity, and there is danger that transparency may be
lost
n to set transfer prices so that it can take profit in the amount and at the
point in the value chain that best serve the company’s interests
n to promote efficiency and effectiveness in managing the processes in all
the company’s inter-trading business units.
10.4 Making inter-trading work
Inter-tradingmaybeattheheartofabusinessmodelorperipheraltoit.Which
of these depends on what proportion of revenue or cost is represented by the
value of inter-trading transfers. Anything above, say, 5% could materially
influence the reported financial results of the inter-trading parties. So about
this point there begins to be a need for a considered policy on inter-trading,
and a regular system that reports and accounts for inter-trading transactions.
Of these, the most important is the inter-trading price itself.
The common parent of inter-trading businesses must exercise authority to
fix the method by which the inter-trading price is set: its sibling business
units are impotent to do so. The business model that binds them into mutual
trade removes their option to negotiate and agree on a market price that
reflects the strengths of their respective positions. And because the goods
and services that they are directed to trade are in the same value chain, the
supplier is obliged to plan capacity and to offer uninterrupted supply of a
quantity exactly equal to the user’s expressed demand.

107
This business model has the effect of imprisoning both parties in artificially
weak positions because they are prevented from negotiating freely or to any
real purpose. Neither is owner of the business model. Each party’s cards
are known to the other in a way not usual in conventional, third-party deals.
It may be that an inter-trading price that would suit both parties, and that
recognises some mutual ‘goodwill’, might resolve the problem of setting an
agreed price. However, that could not be relied on. Goodwill can be in short
supply in ‘corporate families’ as in others. Sibling organisations, abandoned
by the common owner to ‘slog it out’ through bullying or bluff and counter
bluff, have sometimes resorted to expensive internecine strife; or to cartel-
type agreements to cover up inefficiencies that act against the interests of
their common owner.
More importantly, a transfer price that suited both parties might not report
profit in the way that suited the other important party, the common owner
of both businesses, which must understand and correctly record the wealth
creation taking place in the value chain under its ownership. It must know
how its wealth is created; where its profits come from; and how effective its
business model is.
Management is efficiency in climbing the ladder
of success; leadership determines whether the
ladder is leaning against the right wall.
Stephen Covey
There are a number of ways for the owner of the business to set inter-trading
prices. But to set them with confidence, it is even more necessary than usual
to know the costs. A known relationship with cost supports the decisions
managers will have to make. For example, is the inter-trading price to be
more than, the same as, or less than the understood cost? And if more than
cost, how much more?
Usually costs are researched from internal data in order to test the margin
available at a given price. This will be followed by a series of commercial
judgements and financial plans, after which sales and marketing activity
can take place. A lot will depend on whether the same or similar goods are
inter-traded and sold in the open market concurrently.
“
”

108
Many inter-trading prices are set to a cost-plus formula of a type common
in the public sector, or in the private sector when a contract is expected
to run for a similarly long and uncertain period. An example might be a
groundworks contractor in a construction business which is a sibling to the
design and build business in the same group. In other cases cost plus means,
say, £1 each plus 15% to give a 15% return on sales, or £1 each plus 20p
to yield a 15% return on capital employed, calculated from the company’s
budget.
The policy applied to setting inter-trading prices might properly be
influenced by the lower commercial risk that is associated with inter-
trading. Conventional calculations of return assume that high risk of the
loss of capital, or of an under-utilisation of capacity because of unstable
demand, justify an aspiration to commensurately high returns. Inter-trade
is not exposed on these fronts.
10.1 Transfer costing
Banking profit early in the value chain
Where output passes through several subsidiaries before being sold to an
external customer, the owner of the business model can choose to turn cost
into profit and loss by adding most margin in one or more profit centres that
come early in the value chain. Margins and profit in the first profit centre are
then fatter than those budgeted or realised from the second or from others
further down the chain.
Adding margin early defines the business model. It is justified as an effective
means of squeezing the last drop of profit out of increasingly challenging
inter-trading prices on subsequent trades, particularly at the last one to a
third-party customer. These low inter-trading prices and relatively high
targets for margin and profit can be the instrument for forcing cost out.
Sometimes the strategy is articulated in the claim that it ‘prevents sales
teams from giving away margins’ to external customers. That is to say,
when salesmen have no margin to give away, they cannot be seduced by
customers into agreeing to low tariffs or spot prices, ‘low ball tenders’, or
other stratagems to shift volume or achieve total revenue targets in hard
markets.

In extreme cases, any potential margin from the open market may already
have been exhausted before the last amount of value has been added. The
problem that this may present is that the margin to the external customer
becomes harder to know, and the customer’s motivation for buying harder
to interpret. This is particularly so if the value chain is long and the
relationship between sibling parties at the high and low ends of the chain is
remote. The control over the profitability of products may not be centred in a
recognisable place. Furthermore, the sales personnel and profit-responsible
managers who are selling to ‘real’ customers may become demoralised when
little or no profit can be made and attributed to them. This can adversely
affect the sales and marketing effort.
Another concern is that the high margin yields won early in the passage of
the product down the value chain may lead to complacency in managers
there, disguising inefficiencies which, if they could be seen for what they are
and tackled, would create higher profits for the parent.
Taking margin at the external point of sale
The opposite strategy is also commonplace. Little (sometimes no) margin
is embedded in inter-trading prices. All or most is reserved for the final,
external sale. The logic of this is that the only real margin is that realised
from prices in a free and open market. To know and report this, and to base
all analysis of the profitability of products on it, is to achieve the truest
measure of profitability.
There can then be a focus on getting the maximum sales effort in one place.
To avoid the ‘low balling’ problem, it is usual to apply controls on the sale
prices, if necessary to a design approved by the owner of the business model,
and possibly even under its direct authority.
It is common for companies to suffer systemic bidding failure if sibling
businesses need to collaborate to win bids. Each business often blames the
otherfordestroyingcompetitiveness,usuallywithallegationsofgreedaimed
at those early in the chain. This problem is usually solved by instituting
a ‘group’ procedure for approving tenders in which there is an oversight at
the level of group (parent organisation) of any bid price or tariff to which two
or more profit centre business units are contributing.
109

110
Equitable returns
As value is inter-traded between the businesses they own, some parent
companies consider that each subsidiary that adds value should be allowed
to earn a fair return on the capital used or the revenue generated. The return
that is then targeted is a calculation reflected in the approved inter-trading
prices.
There are always as many views about what represents a fair return as there
are parties with an interest in it, even after any agreement on the best type
of return to be targeted, whether it be on sales, capital employed, net assets
or fixed cost.
And there are further challenges to be faced when, for example, a ‘fair’ figure
of, say 5% on sales, is thought to be a fair average but it is felt ‘more fair’ – or
more challenging – for some of the relevant business units than for others.
The technology, innovation, or quality required might not necessarily be
paralleled by the costs of production. Some business units in the chain may
think it fairer to be asked to add value at 5% ROS, while others would be
happier with, say, 10%. All such judgements can feel subjective.
This problem is compounded if reasonable returns are agreed on but the
sibling organisation that sells to the external customer is having trouble
hitting its targeted return because the market has stopped paying the prices
necessary for it to do so. In that scenario, those subsidiaries which are only
inter-trading will win out even though they may have the most scope for
reducing costs. The sibling business which is the final adder of value and
has the vital task of selling is overexposed to failure. This is not a recipe for
harmony up and down the value chain.
If the value chain is short, a fast and flexible response to external market
problems may be possible. If the value chain is long, that may, because of
the number of links in it, be more difficult. The difficulty may be affected by
the extent to which players in the value chain have an investment in it. Is the
product line central to their sales effort or merely peripheral? If they have
large external markets for the partly processed product, or for unrelated
products, transfer prices are less crucial to financial outcomes and easier
to set without fuss. The parent organisation remains the natural arbiter in
all of this.

What should be avoided is an arbitrary agreement for a percentage return –
a sort of ‘ex gratia gift of margin or profit’. In a case in point, a manufacturer
used a sibling organisation as a depot for low-cost stocking before onward
transit to a third sibling for final manufacture and external sale. In an
informal arrangement, the depot service was ‘awarded’ a nominal 5% of
the value of goods crossing its threshold, thereby ill-advisedly distorting the
measurement of relationships between costs and revenue.
Synthetic market valuation
It is not unusual for a product to be sold in the open market and also
transferred downstream to a sibling business for further processing before
sale. In such cases the open market price provides a sound basis for setting
the inter-trading price.
Even if the inter-trading sibling company does not sell externally, it may
still be possible to research price equivalents when there is a third party
organisation that does. The inter-trading value can then be synthesised from
market research.
The problem with this attractive option is that open market prices may vary
with location, product, customer, rebates earned, or otherwise. Nevertheless
a realistic price can often be synthesised. Data are simply projected or
extrapolated (incorporating ‘what ifs’ and ‘if onlys’) to arrive at convincing
inter-trading prices.
10.2 International inter-trading
As we have seen, inter-trading takes place exclusively between separate
business units that are accountable for profit and loss where they exist
within a single parent company or group. But these parts of the group,
though linked by the same value chain, may be located in more than one
country. In one respect, the management accounting for profit and loss in
each domain can just ignore this. The effect of currency movements can, of
course, be shown as a variance. Exposure can be excluded from or included
in the inter-trading price according to the use to which information on, say,
margin is to be put. Thus the system of accounting for profit in one place
or another, in order to measure added value consistently, can be entrusted
entirely to the finance department.
111

112
A greater challenge, however, may be the need to employ different transfer
prices to fix profit in each country by a method, or at an amount, that
satisfies different tax regimes. Some tax authorities are said to be suspicious
of any parallel transfer prices used for management information or cost
accounting. They may spot that these, were they applied to tax, would
improve its yield. So it may be prudent to keep well away from the official
audit trail any transfer prices that are primarily used as internal measures
and controls over performance.
10.3 Terms of inter-trading
Formal terms of trading such as credit given or taken, conditions of sale,
warranty and so forth are sometimes necessary, sometimes not. Sibling
companies may comply with a consistently applied group practice in these
matters or may not. They may be self-accounting or served by a shared
service centre. The inter-trading parties may not only be separate profit and
loss accounting centres but have separate banking arrangements and their
own KPIs and budgets controlling cash flow. Terms of trading will therefore
cover transactions for which invoices will be rendered against terms for
credit given and taken, perhaps even with penalties for non-compliance
with the agreement or contract.
Terms of trading also need to have provision for how inter-trading prices can
be changed and, particularly, how and when adjustment may be made for
cost inflation, such as in the cost of commodities. Getting a price increase
(or resisting it) in a transfer costing relationship can be one of the more
frustrating and time-consuming tasks undertaken by a business executive.
None of this matters too much, provided that there are rules in place for
behaviour and process.
When things go wrong, for example if inter-traded goods or services are
alleged to be unsatisfactory, it is usual to have ready a code of practice that
determines the provision for restitution, if any. It is important to have put
in place a robust process to handle this well before it is actually needed. As
with the inter-trading price itself, buyers and sellers are bound together by
the ultimate owner of the business to whom both belong.
Neither can walk away, even though the dissatisfaction might in other
circumstances make this a natural course of action. With inter-trading,
agreement on restitution in the wake of a complaint can be especially

113
difficult to reach, while recourse to law is impracticable and unlikely, even
though this has been known in cases where the parent was weak and there
was a policy vacuum.
A longer list of things that can go wrong, and for which some prior agreement
on corrective action or compensation may be needed, includes:
n providing sub-standard quality, or failing to meet specification, and any
associated costs of returning or disposing of the goods
n poor performance on delivery, possibly leading to consequential loss of
performance in downstream processes, including the costs of overtime
required to make up for delays and the penalties imposed by external
customers
n unscheduled deliveries causing unplanned stocking, with the associated
costs, which may include writing off the value of work-in-progress
n costs associated with cancelled orders, for whatever reason or cause
attributed.
There may be other problems to be covered that are specific to the inter-
trading context. These need to be anticipated before actual conflict arises,
bearing in mind that settling these or any problems between inter-trading
partiesismoredifficultthaninconventionalcustomer/supplierrelationships.
Hazards of this sort are present in all inter-trading arrangements. The parent
company ought to devise, or agree and authorise, some principles that
provide a consistent line in restitution. It might be that the answer to some
problems is not to be found in restitution at all. Not all parent organisations
would, for example, wish time and money to be used pursuing disputes,
and sending money backwards as well as forwards down the value chain.
For example, inter-trading suppliers’ performance can be monitored, and
action taken by the parent to resolve cases of under-performance through
managerial rather than financial solutions.
Subsidiary company and divisional accounts usually include inter-trading
revenues and costs as an essential part of the published management
accounts. Consolidated group accounts, however, will invariably treat inter-
trading revenue transactions as eliminations, avoiding double-counting
their income.

114
When any organisational entity expands
beyond 21 members, the real power will be
in some smaller body.
Cyril Northcote Parkinson
10.4 Strategy and structure
In a business with a conventional, unitary organisational structure, all
value-adding activity, by definition, takes place within the single profit
centre for which its chief executive and managerial team take responsibility.
There is, of course, a value chain, but it will pass through the departmental
and functional areas that make up the structure. Such a model offers no
inter-trading challenges.
But that structure, a single profit centre, can be changed without the need for
new,incorporatedsubsidiaries.Profitcentrescanbecreatednotwithstanding
and might be justified on the grounds that they make it easier to measure
managers’ performance and accountabilities, and to exercise control. Each
profit centre may be a self-contained, independently managed division of
the larger business. This may enable the parent company:
n to place a limit on managerial spans or unit size
n to distinguish particular activities or skills in order to sharpen focus
n to recognise the effects of location, site, sunk and new investment, and
so on.
The centre, the common owner or parent, will have particular regard for
the importance of attributing profit to, controlling the cost of, and pursuing
sales from the devolved platform it has created. Whether its devolved units
are incorporated operating subsidiaries or not, they become inter-trading
siblings and subject therefore to the need to comply with some system or
rules on price, terms, behaviour and so on.
Entities created by devolving a business into profit centres may seek the
benefits of synergy in adding more value together than would be possible
for each alone. This possibility may be discovered by serendipity, or may be
deliberately conceived at the corporate centre. In this context the synergy is
almost always associated with an over-riding objective owned by the group
parent or common owner rather than by the individual inter-trading units.
“
”

The oil company is a good example of this. Its vertically integrated
configuration is its main competitive advantage over organisations that have
more fragmented or looser structures. The synergy that comes with inter-
trading helps to exploit to maximum effect the profit to be made from the
scarce raw material. Keeping control of every point along the value chain
maximises the profit to be wrung out of ownership. Only inter-trading can
achieve this. Furthermore, this business model creates barriers to market
entry by competitors.
The use of a vertically integrated model, with multiple business units strung
along the value chain, does not prevent those units from trading their
outputs with third parties. A vital decision for those managing the whole
organisation is how much to release into the market at each intermediate
point and how much to retain until the process at the end of the chain –
retail sales – is complete.
As we know, large quantities of crude oil end up as petrochemical products
because supply and demand and pricing are used to optimise profit.
Making sales at multiple points along the chain generates cash flow for the
investment that sustains the chain. This decision also faces the company
discussed in the next section.
10.5 Reviewing the value chain
It is just as important to analyse and understand the inter-trading data of
a business as its conventional external transactions. However, this is often
ignored. There is a tendency to assume that inter-trading and intra-group
sales are a lower ‘grade’ of sale for which the only reporting necessity is a
figure needed for group eliminations. But more analysis is safer. The impact
of the inter-trading configuration on the business model ought to be clear
and the value chain mapped and fully understood.
In the real example shown below, the parent company operates six
incorporated manufacturing profit centres, (DON, WQ, BAC, SSS, BTS and
PTB). Each one is based on different processes, operating for historic reasons
on different sites. Each reports to the parent through its own profit and loss
account. All the businesses make external sales as well as inter-trading
with peer profit centres supplying downstream processes. The profit centres
also inter-trade with three other home and overseas divisions engaged in
distribution (CUK, CFE and BNZ).
115

116
The data on sales, margin and margin percentage in the following three
tables were collected by special analysis from the accounting record. The
data in the first table were used to create the value chain map shown opposite.
Sales inter-trade value
£K SJGroupinter-tradesales(extractonly) External
salestotal
Inter-
trade%
Salesby/to DON SSS PTB WQ CUK
DON 147 221 13491 151 28055 33
SSS 61 153 11170 1
PTB 114 106 31958 1
WQ 255 4,257 20843 18
CUK 18460 –
Margin inter-trade value
£K SJ Group inter-trade margins External
margin total
Sales by/to DON SSS PTB WQ CUK
DON 59 99 6476 38 15150
SSS 37 77 3798
PTB 89 58 17577
WQ 122 1831 14590
CUK 4615
Margin inter-trade %
% SJ Group inter-trade margin % External
sales%
Inter-
trade%
Salesby/to DON SSS PTB WQ CUK
DON 40 45 48 25 54 48
SSS 61 50 34 53
PTB 78 55 55 67
WQ 48 43 70 46
CUK 25 –

The map shows a value chain of 23 inter-trading flows between nine profit
centres. In some flows, such as DON and WQ, inter-trading transactions take
place in both directions, making a complex picture, imperfectly understood
by managers.
Equivalent maps were also drawn for margin and margin percentage. Others
can be drawn for different product types, and for volume and added value.
117
20843
WQ
DON SSS
BAC BTS
PTB
1883
BNZ
16086
CFE
5129
CUK 18460
83
114
153
82
74
66
28055
11170
13491
255
392
172
31958
221
1399
827
4257
111
151
172
673
61
147
8669
106
Flows on the internal value chain £50K
Key:
Distributors
External sales
Manufacturers
£K internal
transfers on the
value chain
55
172

The data in the third table show margin percentages achieved by the sellers
on peer-to-peer inter-trading, and the weighted average margin percentages
for all the inter-trades. No consistent pattern of margin percentage can be
discerned in the inter-trades. So the coherent regime for pricing that one
would have hoped for is lacking.
The conclusion to be drawn is that the profit centres inter-trade without the
strategic direction from the centre that would give a defining characteristic to
the business model. The profit and loss accounts of the divisions lack some
meaning because the inter-trading revenue is calculated from unstructured
prices, yielding percentage margins from which no business information or
standard for control may usefully be derived.
The full consequences of such an absence of analysis, and loss of control,
can result in misinformation on a wider canvas. First, there is the apparent
inability to understand the significance of profit and loss and its elements:
cost, margin, and average selling price.
Second, it is not clear how the internal value chain can best be configured
and optimised or, therefore, how the company’s business model should best
function to promote its efficiency and effectiveness. These are just some of
the questions that could be asked:
n Are all the subsidiary profit centres necessary?
n Is there a case for rationalising the managerial structure, and the scope
and scale of accountability for profit?
n Need all the existing business units report profit and loss?
n Could some processes be contracted out to add more value?
It is common for the flows of inter-trading not to be understood by senior
managers for whom the properties and health of the business model should
be the primary concern. For there to be managerial control over a business,
inter-trading data need to be reported for:
n inter-trading sales
n margin and margin percentage on inter-traded sales
n volumes of sales
n average selling prices (ASPs)
n all the above broken down, if necessary, by product or process
n comparison between inter-trading data and the equivalent for external
sales.
118

119
10.6 Contracting out
A parent organisation that is in control of its business model, the
configuration of its value chain, and its inter-trading terms is in a position
to consider whether there would be any advantage in contracting out the
processes of its business units. In this way it might be able to control the
value chain and the inter-trading pricing structures without having to own
all the assets employed.
The oil industry again provides an example. In fact, oil companies often
contract out major processes such as drilling. For contracts to work properly,
it is essential that the inter-trading prices should add value and that risk
should be taken fully into account.
In extreme cases of contracting out, inter-trading offers examples of virtual
models in which the owner of the business model will itself own very few of
the assets employed in the value chain. Despite this, it may be able to retain
control over the configuration of the value chain and the inter-trading prices
charged, and to seize the lion’s share of profit.
In a real example, a company requiring shipping containers for its leasing
business contracted out their manufacture. It exercised control over the
whole manufacturing chain of value by negotiating the inter-trading prices
for the raw materials and the value added by the first and second-tier
supplierswiththefinalbuilder.Theoptiontoputallitseffortintonegotiating
the lowest possible purchase price of fully built containers was replaced by
a series of separate contracts with each supplier in the chain. Each contract
set an inter-trading price, a part of which was remitted directly to it as the
owner of business model. It thus banked margin at each stage as the product
moved down the value chain as well as from the final process of leasing,
which it operated through processes and resources wholly its own.
A particularly complex example of such a virtual value chain owned by a
single owner in control of the business model is to be found in the insurance
industry. An insurance company may own the model through which it
conducts its business while sharing the value chain with other participants
with whom it contracts. Thus an insured customer buys a policy from a sales
and marketing agent such as a bank or a broker for whom the commission
is a negotiated inter-trade price at the front end of the value chain. The
underwriter’s fee is another inter-trade price. In the event of a claim, the

120
claims handler and (if appropriate) the specialist medical or other service
provider are remunerated through further inter-trade prices pre-arranged
with the insurance company.
The whole value chain is owned and commercially orchestrated by the
insurance company, whose business model determines and controls the
operational relationships of all parties. The inter-trading prices all along
the complex value chain are devised and owned by the company to which
all other players are symbiotically linked. At one time insurance companies
themselves performed these processes on what was then a wholly internal
value chain in a single profit centre model. Now, although contracting out
these processes, they continue to own the network of what are in effect a
series of virtual inter-trades off a tariff that remains their own.

121
11. The transfer costing
organisational model

11.1 Approach
Transfer costing is Collinson Grant’s term of choice for a specific model
of organisational structure and control. Having a much wider industrial
application than inter-trading, it combines a particular structural template
with a regime for management accounting and business reporting that is
supported by a robust philosophy of accountability and control developed
by us in conjunction with a major industrial company and client of long
standing.
The result sought is an organisation that maintains tight control over its
costs and margins in pursuit of sustainable profit. This model comes into its
own in stable industries where improvements in profitability can be made
by operating disciplined management of margins and costs. It should not
be confused with international transfer price accounting between differing
national taxation regimes.
11.2 A tough regime for managers
Most managers would be offended were it suggested they do not aim to
pursue profit. However, the reality is that most businesses are not run by
senior managers exclusively dedicating all their actions and decisions to
that end. However, successfully implemented, transfer costing allows no
place for managers to hide their short or long-term performance from the
scrutiny of those to whom they report.
The unforgiving nature of the transfer costing regime pushes pursuit of profit
harder than any other we know. Not all organisations are comfortable with
the amount of transparency to which business reporting expose managers’
actions and decisions. But where radical, sustained improvement is desired
such as post-acquisition, during a turnaround or a change of management,
the results can be startling.
123
11 The transfer costing organisational model

124
11.3 Organisational design
For transfer costing to work successfully, it must be possible for the
foundationsofanorganisationtoberestructuredaroundthreediscretesenior
managerial remits involving three very different types of accountability.
The business is structured as a single profit centre with two generically
defined divisions. Transfer costing is often referred to as the operator/trader
concept because the majority of managers reside in one or other of the two
structures of the organisation.
n The Operations division is a production and/or a supply function with
both fixed and variable direct costs where value is added by processing
or procuring products, and the main financial concerns of managers are
costs and volume of throughput.
n TheTradingorCommercialdivisionisasalesfunctionwithresponsibilityfor
pricing, product mix and inventory, and selling (and possibly distribution)
costs, and the main activities are marketing, sales, commercial and others
that face the customer, and where the main financial concerns of managers
are prices, margins and volume of sales.
n There is also a corporate centre housing overhead functions and costs
which are largely unrelated to the day to day managing of the operations
and trading divisions but are to do with general management, direction,
strategy and financial control of the company.
n The Operations and Trading divisions may be sub-divided into smaller
production or sales centres by scale (size and number of processes),
diversity (types of product, process, markets, channels, territories, sites)
or length of value chain (number of discrete processes), so that the design
can accommodate an organisation of any size.
n Sub-divisions of Trading are often known as customer-facing business
units (CFBUs) – an important label since much of the value of the transfer
costingstructureandaccountingistoemphasisethecriticalaccountability
of sales and marketing managers for looking to the market to develop
margins on their sales and to leave operating cost accountability to
operations managers.
The structure shown here is more radical than it first appears. The divisional
Director of Operations is responsible for cost, not profit, as are all his
subordinates. And the divisional Director of Trading is responsible for all
the margins made on sales to external customers, except for any effect (for
better or worse) that is caused by variation in cost.

125
This concentration of power can be difficult for some managerial teams
to swallow. However, where context dictates there might be more than
one Operations and/or Trading division, which can ease the problem of
implementing changes in managerial hierarchies.
11.4 Accounting for transfer costing
Production or supply by Operations is to the order of the Trading division.
Work in progress moves within the Operations division at standardised costs
until the outputs (or procured finished goods) are complete. Then, their
value at the final standardised cost (the transfer cost) is charged to Trading’s
inventory or sales.
Thus:
Transfer cost per unit produced x no. of units = transfer value.
Operations is credited with the value of output at the transfer cost as income
on its internal account. The difference between that total transfer value and
the total cost of Operations’ expenditure is a (cost of sales) variance which
serves as the performance criteria for the Operations division. Budgets may
be set for break-even or a favourable variance if the assumption is that costs
will be reduced year-on-year.
Sales units
Divisional Director Divisional Director Corporate Centre
CEO
MarketingSalesUnit Managers Commercial
Finance HR IT

126
Trading division is credited with the value of sales to customers. From this
income to its internal account are debited the transfer costs credited to
Operations, the balance being gross margin. Gross margin totals are debited
with the actual costs of selling and marketing to record net margin.
The simplified version of performance reporting illustrated above can be
elaborated into a full set of management accounts. These will draw out more
detail for multiple production and selling centres or units, full cost and
sales analysis which can be overlaid on the managerial structure of layers
and spans, and detailed analysis of variances by volume, mix, price and
expenditure.
11.5 Making transfer costing work
This is not the place for detailed instruction into the mechanics of a system
of transfer costing or, for example, templates for business and performance
reporting. But in order to understand the importance of considering transfer
costing as an organisational design option, the following is germane.
n There should be an intense focus on systems of management that support
executive decisions and promote action to create profit. This bias to action
has two clear, enduring priorities – to optimise margins and to control
costs.
n Margin and cost are both controlled, but measured separately so that
change affecting the one can be known, free from the knock-on effects
of the other. For example, actual costs of production, however they vary,
Thus: Actual Budget Variance
Transfer value credited (transfers to Trading) 95 100 (5)
Actual direct cost of production and supply 80 78 (2)
Actual Operations Division indirect costs 25 22 (3)
Surplus or (deficit) on Operations (10) 0 (10)
Thus: Actual Budget Variance
Actual sales 150 160 (10)
Transfer value debited 95 100 5
Gross margin 55 60 (5)
Gross Margin % 37 38 (1)
Selling costs and Trading overhead 10 11 (1)
Net margin 45 49 (4)

stand isolated from any effect on margins which other managers are
accountable for.
n Sales employees concentrate upon selling prices to enhance margins,
and are allowed to ignore production or supply costs which are the sole
accountability of operations managers.
n The system of information and control that fixes accountability for
financial results must exactly overlay the organisational structure. Thus,
information reporting, measurement and accountability are layered so
as to mirror the managerial chain of command. Operational controls are
important but are connected to and absorbed into reports for financial
control.
n Management and cost accounting is an integral part of day-to-day
management for everyone. Management accounting does not just keep
the score – it supports decisions using variance analysis.
n Internal markets can obscure trading reality – so the value chain must be
transparent and the movement of value reported with rigorous consistency.
Cost is considered to be the only rational basis for valuing work-in-progress
passing down the internal value chain. Margins artificially attributed by
policy or negotiation to production or distribution business units are not
used here.
n Business units and sites that produce product cannot generate margin or
profit, only cost; added value is considered to yield a margin only at the
point of sale to a third party.
n Units of production or supply are valued using fixed-cost standards
for each operational stage. Partly finished product and stage costs are
transferred between points, into inventory, and at full-cost value when the
final product is sold by trading.
n Profit and loss can only be measured for a full-service business unit. It
comprises Operations’ variances carried forward to profit and loss,
Trading’s net margins also carried forward, and the cost of the central
overhead, perhaps allocated back to activities and included in the cost of
Operations.
n The values budgeted for costs and margins are derived from budgeted
unit costs and prices at budgeted volumes: and these are always clearly
reported against actuals.
n Note that Operations adds value but cannot add contribution or profit;
and Trading adds contribution; while profit is only registered in the profit
and loss account – the statement that consolidates all revenues and all
costs, including those at the corporate centre.
127

128
11.6 Transfer costs and transfer value
Setting transfer costs
The transfer costs from which the transfer costing system takes its name
are the units of value or currency that are applied to production or supply
outputs. The transfer cost at which each product is valued will be based
on the organisation’s knowledge of its costs. From this an expected or
standardised cost is developed by means of any appropriate tool of cost-
finding analysis.
Transfer costs are not standard costs in the sense that they are routinely
updated. Once set, they are moved infrequently in order to benchmark
costs conservatively, and prevent the effect of cost increases migrating into
the margins for which Trading managers have accountability. That would
weaken their challenge, which is to build margins from their sales and
marketing strategy and effectiveness.
Calculating transfer value
In a simple illustration, a manufacturing division which we call Operations
makes three products – A, B and C – bearing transfer prices set at £50, £70
and £20 respectively.
In a period, output is A = 10 units, B = 20 units, and C = 25 units so, by
multiplication, £2,400 of transfer cost value has been created that can be
credited to Operations and the debit transferred to the Trading account in
exchange for the goods.
If Operations’ cost of production is more than £2,400, it records a deficit on
its account. If less cost has been incurred, it records a surplus. Either result
shows as a positive or negative charge against profit and loss.
In the same way, Trading, which has ‘paid’ £2,400, will make a positive
margin if it sells for more than that, or a loss if not.
Ideally, costs will remain close to what is expected and selling prices will be
such that margins can feed profit and loss to give adequate returns on sales
or capital.

129
11.7 Accountability and control
Transfer costing allows managers to be held to account: in Operations
for their costs against a target of break-even, and in Trading for margins
achieved against target or budget.
The margins are measured from the cost base on which transfer prices are set
and not from the actual cost in each accounting period. So margin is neither
enhanced nor eroded by the skill or fortune with which manufacturing costs
(materials, conversion and manufacturing overhead) are managed.
Arranging the structure of the organisation around these two distinct types
of operational accountability, while employing the simple accounting
devices of transfer costs and the accounting mechanics of transfer costing,
ensures measurement and control over the two most critical influences on
profit in the types of businesses where the system is most effective.
For a complete philosophical embrace and the best results, all strategy,
role definition, information management, reporting, and related aspects
of business management must be organised around both Operating and
Trading dimensions as clearly distinct aspects of decision making and
accountability (though not each to the exclusion of the other, since they are
joined at the hip in the common pursuit of optimum profit).
Although it is accepted that some costs may need time if they are to be
challenged and reduced by managers, the accounting system does not
provide for this – it reports the nature and size of a cost problem to managers
and quantifies the challenge accordingly. There is no soft option just as there
is no blame. Thus:
n When volume falls, the challenge is to make all costs fall in direct
proportion.
n When volume increases, the challenge is to stop costs from rising in
proportion.
Measures of performance should shine like beacons – they are the most
visible signs of what managers must regard as important. When linked to
financial incentives, the light shines so much more brightly.

130
11.8 A transfer costing universe
The canvas over which the simple principle of transfer costing operates
becomes the organisation’s transfer costing universe. Within it the
organisational design, and the mechanics of management accounting,
costing and business reporting are all subordinated to the policies and
protocols that the regime imposes.
There is a body of best practice that can be adopted when transfer costing
is used in this way as a tool of managerial control. It can be applied in a
balanced way throughout all parts of the universe, and is based on our
experience of inducing compliant managerial behaviour throughout the
structure.
In terms of both the organisation and its systems, the scope of a transfer
costing universe must be defined by the network of potential flows of
transferred value. The network illustrated on page 117 is a good example of
this.
Those parts of the organisation participating in the universe must be
bound to a single point of governance. This has to be capable of initiating
and owning the method of valuing transfer prices, the definitions of
accountability to which they give rise, and the best practice policies to be
applied, for example, on the treatment of variances.
This bars any participating part of the organisation from defining any other
method of valuation, or from setting or agreeing on any other trading term,
either unilaterally or bilaterally. One consequence of this is that transfer
prices cannot be devised until the policies – the rules of trading to be held
in common – are articulated, understood and verified for compliance. So,
it is highly desirable that a suitable part of the organisation should have
stewardship of the rules and of the concept and substance of the best
practice.
11.9 Role of the centre
The organisational structure associated with transfer costing helps to
emphasise the difference between the operational roles of directors
and managers in Operations and Trading and the general and financial
management in the corporate centre, which manages the overhead functions
that serve the whole organisation.

While the former turn the handle on the business machine throughout
the month, the role of the latter is governance, direction, policy, objective-
setting, authorisation of resources, and critique of performance and strategy.
Conventionally configured businesses often fail to observe this distinction,
with the result that everyone from top to bottom in the organisation is mired
in the swamp of routine tasks and day-to-day management.
Under transfer costing, the centre works like the bridge of the ship – making
decisions about destinations, the rate of passage, the hazards to be navigated,
the priorities that need to be observed and so on.
Only small central teams are consistent with this role – transfer costing
structures are highly devolved, and managers are focused on the priorities
and accountabilities made very clear by the centre.
Central functions like HR and IT, legal and estates, and even quality
assurance, may be wholly or partly devolved to Operations and Trading.
Where functions are devolved the costs lies with those divisions, and in the
case of Operations are included in the unit transfer costs.
11.10 Transfer costing – a case study
A leading quoted company produced roadstone and construction aggregates.
It owned and operated a large number of quarries extracting and grading
crushed rock, and pits extracting, screening and grading sand and gravel.
In addition to the extractive operations, some primary processed minerals
underwent secondary processing to make asphalt for road surfacing and
ready-mixed concrete for general construction.
Sales were made to customers directly from the primary processing sites, and
material was also transferred to secondary processing plants from where the
added-value products were sold on to third parties. Asphalt plants, usually
located at primary production sites in largely rural locations, were run by
quarry site managers responsible for the whole site’s profit and loss. Ready
mixed concrete operations came under a separate body of managers in a
separate profit centre, at numerous and mainly urban locations.
131

132
This traditional structure was replaced by a transfer costing regime
with its different conventions for organisational structure and controls.
Implementation of transfer costing was varied for quarries and the ready
mixed concrete activities, to reflect the greater complexity of the latter. The
organisational structure to support the transfer costing regime is shown
below in a fully developed form that includes regional devolution of both
Operations and Trading, and devolved central functions for finance, HR,
and so on.
The benefits of the new transfer costing structure became apparent in the
worst sector recession yet experienced by any of the company’s managers.
Despite sales falling by more than 30%, this Collinson Grant client was the
only operator in the sector to trade profitably, because trading managers
refused to destroy margin by ‘buying turnover’, and operating managers
worked hard to control the production mix and reduce costs.
A simplified version of the line and staff structure from this case study is
illustrated below. The abbreviated diagram scopes an organisation of four
large regions, a workforce of 5,000 and a turnover of £1.5bn.
CFO Regional Directors x 5
CEO
Financial
Controllers x 4
Consolidation
Accountants x 2
Regional HR
Director
Financial
Controller
Operations
Financial
Controller
Trading
Technical
Manager
Marketing
Manager
Margin Centre
Managers x 8
Production Centre
Managers x 15
Regional
Operations
Director
Regional
Trading
Director
Regional CFO Head of IT
(shared)
Purchasing
Cost Accountant
Invoicing and Sales Ledger
Credit Control
■ Corporate Centre ■ Regional Centre ■ Operations Centre ■ Trading Centre

133
11.11 Managing cost reduction
Managers both in the centre and in Trading are likely to expect Operations
managers to win improvements in productivity to lower the cost base. It
is an objective of transfer costing to put downward pressure on the costs
of operators. This can be achieved in a number of ways, one of which is
to respond to the achievement of surpluses by budgeting reduced transfer
costs in some or all subsequent budget years – to move the goalposts, if
possible time after time. Another approach is to ratchet back the transfer
costs in expectation of, or as a directive to achieve, a lower cost base.
A plan for cost reduction over a number of years ahead is essentially an
element of the strategy for the business as a whole. Trading will always make
clear what it sees as the imperatives, but best practice also sets targets for
margins on sales and for profit. While cost reduction is delivered through the
downward pressure of the ratchet, Trading is expected to increase margins
over and above any saving achieved.
In the long term, overheads such as direct costs are not treated as ‘fixed’.
Once accountability for costs is established, a manager’s job is to tackle both
fixed and variable costs as variable downward. This concept is essential if
volumes are falling, and should be desired when volumes increase.
While the desire of Trading for reduced transfer prices can be safely
assumed, it is the balanced view from the centre that should command the
policy going forward. That is to say: how much reduction and how soon?
In the integrated company model, Operations is accountable to the centre
for meeting the challenge, and responsible to Trading for its delivery to the
agreed time, specification and cost. The centre owns the transfer costs and
the finance department is the best steward of them because the values they
represent calibrate the performance of the Operations and Trading divisions.
In the end, all business operations can be
reduced to three words: people, product and
profits. Unless you’ve got a good team, you
can’t do much with the other two.
Lee Iacocca
“
”

134
Given their customary views, it is common for Trading managers to think
about the possibilities of increasing margins by sourcing externally, or
making changes to the specification or mix of products being marketed.
Of course they cannot make radical change to supply chain policy
unilaterally, but will make representations. The centre has the veto having
the responsibility for calculating the effect on corporate profitability of any
change or restructuring.

135
12. Organisation of branch
networks

137
12.1 Models
Many companies reach their customers and provide their services through
a network of local or regional branches. Collinson Grant has considerable
experience in improving performance by getting organisational practice in
branches to align with a corporate business model, designing footprints,
lines and substance of reporting, and specifications for the exercise of
authority.
Even in days when online trading is growing rapidly, a branch network has
many advantages, which is why some models run both bricks and mortar
and e-trading side-by-side.
Bricks and mortar advantages include: access and proximity to customers;
knowledge of local customs and buying habits; dispersed logistics
(sometimes a negative); closer relationships with local suppliers and so on.
But such an organisational model demands clarity on a number of important
decisions that could either be made locally or centrally. Typically the branch
manager reports to a central head office. The managers there, as well as in
the branches, all need to know where authority and responsibility lies for
the enterprise generally, what authority is delegated and who is accountable
for what. The organisational design needs thinking through at a high level to
arrive at an effective model for the corporate business and for each business
unit. The business model must have a design in which it is clear where
decisions are best taken so confusion cannot reign nor mistakes affect profit.
Since branches are more or less businesses in microcosm, the organisational
design must consider accountability and decision taking in all the key
areas that can impact on profit: pricing; terms and conditions for trading;
procurement and inventory; trading policies; payroll and operating costs;
managing employees; and processes and systems.
Broadly, there are two types of organisational model for managing a
branch-networked business. The decision on which of them to embrace is
determined by the choice of corporate business model and of contextual
matters like the scale of operations.
12 Organisation of branch networks

138
Our view is that most businesses ought probably to default to the first, which
we call the ‘loose/tight’ model with its related protocols and structures,
unless there is positive economic advantage in the other type, which we call
the ‘centric’ model.
The main reason for our preference for loose/tight is that it emphasises
wide accountability for profit by managers carefully selected for the
entrepreneurial qualities they possess. In short, get the right person and
give them as much responsibility as they can reasonably and responsibly
handle.
The loose/tight model
Much of what is written here can be taken as a guideline for what we call
the ‘loose/tight’ model for control – a framework particularly suitable for
managing branches. This sets out the wide discretion allowed to branch
managers within a tightly targeted and measured accountability to the
centre. A regime for loose/tight control will have most or all of the following
characteristics:
n a significant degree of devolved decision taking for local managers
involving:
➤ light touch supervision from above
➤ longish timescales between being called to account
➤ a limited amount of day-to-day support from the organisation’s centre
n strong and challenging financial and operational objectives set by the
centre for the branch and its manager
n a high-level financial measure of performance such as profit or return on
capital.
Managers of a local branch should have:
n some authority over prices and terms for customers, which products and
services to offer, and the sourcing of products and the purchase prices
paid
n an influence on the approved levels of inventory, and the stock turns
n substantial freedom to pursue business wherever it may be found
n substantial authority over staffing levels, operating costs and processes
and procedures
n freedom to recruit employees and to discharge or discipline them (within
a code of good practice)
n an opportunity to win capital allocations for good business cases.

139
To make this work a company has:
n to put considerable effort into and be systematic about recruiting branch
managers so that a much higher proportion of appointments are successful
than is the norm
n to put considerable time and effort into thorough supervision of financial
and operational plans, given the relative infrequency with which they will
be tested
n to have a clear but firm procedure for holding managers to account
n to have clear and relatively simple measures for assessing performance
n to set challenges based on business results not compliance with process
n to set challenges that stretch managers, and annually review where the
goalposts are
n to give serious consideration to paying meaningful incentives.
The centric model
The preferred loose/tight model for managing branches has a rival that is
a consequence of modern data processing, retrieval and storage, and rapid
communications. By running mathematical algorithms at the centre to
analyse real-time sales and inventory, supermarkets and other retail chains
can replace a range of local managerial functions.
These techniques enable small groups of people based centrally to determine
and control supply in a way that would have been ill-advised in an earlier
age. The model is suitable for fast-moving consumer goods (FMCG), many
retail businesses and some others like them where the scale of operations
supports investment in a powerful, networked infrastructure; and the
sophistication of computerised modelling functions mean that they must
be in the hands of specialists at the centre. This may also be true for some
distributors of industrial goods or where most customers are trade buyers.
The scope of this centric model affects product selection, procurement,
pricing, promotions, and all functions facing both customers and suppliers –
everything in fact except those used to manage the local costs and resources
for which the branch manager in the centre of the value chain could still
retain responsibility.
Other factors tend to support the centric model. The regulatory environment
governing employment and work is undermining the autonomy that can be

140
granted to the first line of managers. The risk of decision taking needs to be
spread across more than one level of manager, and to involve specialists in
supporting roles which are located, inevitably, at the centre. Responsibility
for business support costs such as property and utilities are often more
easily controlled from the centre, and even locally consumed goods and
services may be centrally procured in bulk or outsourced through centrally
placed contracts.
12.2 Authority
The essence of the organisational design and managing of branch networks
is how authority and decision taking is exercised and where control is vested.
The branch manager’s authority is embedded in the design of the
organisational model. This in turns defines the accountability held. The
scope of branch authority can be broad, and includes many dimensions of
which the following examples are the most important.
12.3 Authority to set prices
List prices
If the list prices available to branch customers are mandated by the centre,
and it denies pricing discretion to branch managers, it is able:
n to be certain that the margins it seeks will (certain other influences apart)
be realised
n to discourage customers from shopping around from branch to branch to
search for or negotiate a better price
n to dissuade managers from chasing turnover at the expense of margin and
profit.
Central authority over prices tends to imply ‘national’ list prices but even
where the centre allowed them to vary by region, the above three points
would more or less still hold good.
When branches have authority over their own list prices:
n they can respond flexibly to local competition, for example by making
smaller, more frequent adjustments to price than can be done from the
centre

141
n they find it easier to take market share from competitors (when that is the
policy).
It is possible to give branches some degree of authority to set list prices so
as to create local competitive advantage, with the revenue and margins
achieved as a result being captured in the budget. After the budget is
approved, the centre could exercise control over the degree or frequency
with which prices might then be modified and the budgetary assumptions
changed. But essentially the prices set by the branch manager would be
those paid by the customers.
Spot prices
Spot prices are off-list prices that can be used to favour some customers for
some products, or all customers for some products. Spot prices are likely
to be used alongside the list prices, to which all prices default if neither
customer nor product qualifies for spot.
By their nature spot prices are without longevity, and may be associated with
a single promotion. They can also be invoked for a one-off sale to a single
customer. For example, a trade customer quoting for a large job may wish to
quote competitively, and ask the branch manager for a spot reduction on the
price of a vital material so that both branch and customer may profit from
their respective sales.
The centre may retain authority to authorise spot pricing, so whenever
branch managers believes there is a case they would seek authorisation from
the centre. This gives the centre an opportunity to recognise and consider
the effect of a lower price for that sale on the planned margins and profit.
But an approval process by which the centre exercises total control over all
types of spot pricing might be so laborious as to make control an impractical,
blunt instrument.
Branch managers may be given authority to agree or veto sales at spot prices
proposed by the centre. Here there can be both risk and opportunity in equal
measure, price having the greatest and most direct effect on profit of all the
factors of influence. There is a need for keen judgement on the amount of
margin to yield, to which customers and on which products. The effects can
be good and bad at the same time.

142
Spot pricing is a tool for the branch to get or keep sales when it needs to act
decisively to exert competitive pressure or defend market share.
However, used too frequently and freely the branch’s customers may learn
from experience about the branch manager’s selling behaviour and may,
in effect, be given an unintended lesson in how to negotiate prices down.
Branch managers who are known to be spot price ‘pushovers’ may find
turning the situation around and preventing the permanent erosion of
margin and profit to be all but impossible.
The centre could ration spot opportunities by giving a standing authorisation
for a limited number or value of spot concessions, thus forcing branches to
prioritise opportunities and use the facility selectively. In this way the centre
could retain a measure of control over margins and the branch profit plan
by writing into its budget the effect on sales and margin of the spot pricing
it has authorised in advance.
Furthermore such reduced control over margins can be mitigated if the
centre gives branch managers an incentive to meet targets for profits and
capital employed that cannot be achieved if spot pricing is used too liberally.
Discounting, rebating and giving away concessionary value
Discounts are reductions applied to the list prices of selected products, or
to groups of them. Their availability is advertised to all customers or, in
some cases, to pre-qualifying customers (perhaps those who have earned
vouchers) or selected purchases, such as ‘buy one, get one free’ (BOGOF).
Rebates are reductions applied not to the product but to a customer’s
spending, often on all of it but sometimes excluding certain types of
products or expenditure. Exclusions normally include sales where another
offer such as BOGOF already applies. Reductions are usually conditional on
the customer’s exceeding certain volumes or values of expenditure on any
particular occasion or over a certain period.
Concessionary value is a non-cash benefit. An example is free delivery for
major customers or for qualifying purchases where expenditure is more
than a certain value.

143
Authority to modify prices in these ways can be kept at the centre in the
same way, and for the same reasons, as for list prices and spot pricing.
Discount offers can be planned as schemes by managers at the centre, by
general sales managers or marketers, and implemented over the heads of
branch managers. Branches may have a choice to opt in or opt out, or it
might be compulsory.
The options for branches to promote the sales of some products to all
customers, or of all products to some customers, follow the same logic and
have the same effects as those described for spot pricing.
Where this authority exists at branch level, it is the primary tool available to
the manager to increase sales and market share, and broaden the customer
base.
Meetings are a symptom of bad organisation.
The fewer meetings the better.
Peter Drucker
It is common to find situations where absolute authority does not and
need not exist at either branch or centre. A broad platform for promotion
is created by the branch, submitted to, and authorised by the centre. This
will include competitive pricing operated with freedom by the branch but
within the envelope of an approved plan, the financial effects of which may
have been bought off at the centre through the budget approval process and
monitored through the period accounts.
“ ”

144
12.4 Authority over credit and customers
Terms for trading and credit
Terms and conditions are normally set at the centre because the specialised
legal knowledge required is only likely to be available there. The centre’s
control of the ‘Ts and Cs’:
n ensures that the content is professionally written, and so legally watertight
and tested
n prevents both branches and their customers from wittingly or unwittingly
undermining the company’s rights, the most important of which are to be
paid, and to reject liability where appropriate
n ensures universal and consistent application of Ts and Cs throughout the
branch network, so preventing erosion of the company’s policies, and
making compliance easier to monitor
n makestheplanningofcashflowmorepredictablebyinstitutinguniversally
applied standard credit terms (for example 60 days), and thus improving
liquidity.
Setting terms and conditions at branches:
n avoids excessively legalistic terms and conditions set by lawyers at the
centre that may be too restrictive or complex for some customers to
swallow, with the possible result that the branch becomes an unattractive
trading partner
n may be informed by a clearer insight into what terms are appropriate for
the local trade
n may provide a better informed appreciation of the real risks, including
credit risks, from which the company needs to be protected.
It is always open to the branch to seek approval from the centre for unique
terms for special customers.
Granting credit
The process of granting credit to a customer may be formal or informal,
lengthy or short, and done at the branch or at the centre. Companies are
inherently vulnerable to mistakes of policy on this subject.

145
Accountability for granting credit is usually vested in a person such as the
branch manager or in a protocol designed and approved at a high (therefore
central) location and which could be operated either from the branch or at
the centre.
The centre is better able:
n to design a secure and consistent process of credit checking
n to operate any formal process, with economies of scale giving a lower cost
per application
n to cover more professionally the legal aspects designed into a system and
its operation, such as the process for collection of overdue debts
n to operate credit checks dispassionately and thus objectively
n to put customers on ‘stop’ dispassionately and thus objectively.
The benefits of which are:
n better quality assumptions when planning cash flow and fewer bad debts
n centrallycoordinatedinformationaboutthetotalcredittakenbycustomers
buying from a variety of branches.
The end result should be fewer defaults and firmer control over cash.
At the branch:
n cheaper, quicker decisions that are less encumbered by bureaucracy may
stimulate sales
n applying local knowledge can actually lead to more informed decisions
about credit worthiness and therefore reduce risk
n customers may be classified, with more precision than is possible at the
centre, as ‘no credit’, ‘21 days’, ‘45 days’, et cetera. Although standard
terms are 30 days, the branch slots each customer into a more diverse but
structured classification.
Customer catchment areas for branch trading
This is usually a problem only where customers are trade and account
customers.

146
If the centre designs the geographical perimeter that prescribes the trading
area allocated to the branch, then:
n the centre is well placed to take a view of the optimum density of the
branch network, including the number of branches and the estimated
turnover and customer numbers for each
n this creates an orderly network in which the number of branch units is
not excessive in relation to sales and margin, and it is possible to keep
to a minimum the need for branches to make incursions outside their
allotments
n branch managers can be directed to focus marketing and sales on their
given territory and not dissipate effort on peripheral business
n customers are given no encouragement to shop around between branches
and to undermine branch margins by seeking ever better deals.
It is also feasible to take a more organic approach to the formation of
branches that fits with the natural behaviours of branch managers and
customers.
Natural catchments are always, to some extent, subject to the way customers
gravitate to branches – something that cannot always be anticipated and
planned, and which does not necessarily reflect the obvious geography.
Too tight a definition of boundaries may be pointless – the exact location of
branch premises and access to them is likely to cause customers to ignore
the designated boundary between branch sales territories, whoever they are
set by. Branch managers are rarely in favour of creating a line except where
it might stop customers deserting them for a more successful branch.
Customers cannot be refused if they beat a path to the door of a branch
outside their ‘area’, because the alternative is that they will buy from a
competitor – the customer has the right to choose. And branch managers
with a talent for competitive selling should be allowed to exercise it freely in
a spirit of encouraging enterprise.

147
12.5 Authority over procurement
Selection of the product range offered at the branch
A company’s product range defines a large part of its business model. Most
companies are associated with and create a brand around the customers’
perception of what they sell or do.
This is an issue that concerns the most senior managers and is therefore
nearlyalwaysassociatedwiththecentre.Butthecentrecannotbeomniscient.
It is the branches that operate on the front line, where there is much to learn
about customers’ wants from the customers themselves.
Restricting branches to ordering and stocking from a product range
mandated from the centre:
n ensures that the benefits of comprehensive, expert and probably expensive
research reach every part of the branch network
n can cross-reference procurement and sales across the widest canvas
possible
n uses to the maximum the buying power of the whole organisation
n creates a strong bond between the brand and the range by offering
consistency in customers’ expectations and experiences
n allows branches’ performances to be compared like-for-like (because
there is less variation in the mix of products)
n permits sales performance to inform the selection of product ranges for
the whole branch network.
Problems can arise if the branch has complete authority to stock the range of
products, because the unique character of the range could cause:
n confusion among customers if a branch’s mix of products fails to reflect
the company’s main strengths; or if the differences in inventory between
branches is too great
n a gap to appear between the branch’s offering and the company’s brand
that may dilute customers’ perceptions of what is on offer and damage
sales
n the value of the company’s brand ultimately to disappear if the branch
does not behave as a part of a network
n dangers of abuses involving branch managers, suppliers and customers,
that the centre may struggle to spot, correct or control.

148
Conversely, if the branch has no say in the choice of products, important
potential benefits are lost. Branches:
n may have a better ear for what (their) customers want than the centre
n can move more quickly than the centre to meet new demands
n can be a source of information (particularly if many branches speak with
a single voice) about the need to shift the core product range in a new
direction as customers’ tastes change.
Branch managers always inherit a core range, either formally and precisely
defined, or through their understanding about the ‘business the company is
in’. So it would be unusual for them to want to move away from the perceived
core.
However, if the centre has no capacity for accepting feedback from branches,
it cuts off the best channel of information about customers’ wants and
may lose competitive advantage. So the branch may be entrusted with the
authority to select non-core product ranges to search for a more exact match
between its inventory and what the local market wants.
Selection of suppliers
Selecting the product range is a different task to selecting the suppliers
of the products. A given range may present many options for choosing
suppliers, while at the same time some products may only be available from
a single source.
If the centre selects all suppliers, several of the same effects are felt as when
the centre selects the products. For example, it:
n ensures that the benefits of comprehensive, expert and probably expensive
research on value for money and strategic procurement reach every part of
the branch network
n can integrate procurement with sales on the widest canvas possible
n maximises the buying power of the whole organisation
n allows branches’ performances to be compared like-for-like (because
products are bought at the same price for all branches and the mix of
products varies less).

149
In addition:
n it is possible for the company to exercise absolute control over prices and
terms by using a single negotiating platform
n earnings from purchase rebates and discounts can be retained at the centre
and not subjugated to the result of mismatched sales or procurement by
branches
n the effectiveness of the centre’s procurement team can be measured in the
purchase rebates and discounts obtained
n the centre has the resources to investigate the origins of supplies and can
check that their provenance complies with regulations on quality, safety,
environment, et cetera
n the company can use objective criteria to rationalise the number of
suppliers and the supply chain.
For some products, the branch may have a more intimate knowledge
of sources of supply, particularly if these are local. Other advantages of
procurement by the branch are that:
n using local suppliers selected by the branch may mean lower transport
costs
n procurement by the branch may be informal and done with little expense
by the branch manager or the staff within their normal working hours.
However:
n low-volume purchases at a branch will usually be at a higher unit price
n the informal process for procurement at the branch may be unsafe, or
lacking advantage in a number of ways – shorter credit terms, less
watertight guarantees, poorer security of supply, et cetera.
The centre ought to be able to distinguish and measure the overall benefit
between choosing suppliers for core and non-core products by branches
and the centre itself.
Negotiating discounts on purchasing, rebates, concessionary value, terms,
conditions and credit from suppliers
The centre represents the purchasing power of the company and this should
be brought to bear when selecting products and suppliers.

150
A branch, however, can take a more forensic view of opportunities for value,
being more selective in claiming concessions from some of its suppliers on
some of its higher selling products.
The authority to negotiate need not be the exclusive property of the centre
or the branch. But for the branches to continue to negotiate after the centre
has already won some concessions would probably require sanction at the
centre. Without this, branches may negotiate in vain. The centre can put
suppliers on notice that branches may come and ‘ask for more’.
12.6 Authority over inventory policy
Deciding the volume and value of stock to hold
In managing inventory the organisation attempts to optimise two conflicting
variables:
n the availability of stock in relation to demand
n thevalueofcapitalinvestedininventorycomparedwithdesirablefinancial
and operational targets for business performance, such as stock-turn.
At one time it was a major function of the branch to reorder and so maintain
the availability of products without which sales would atrophy and profits
evaporate. That became largely unnecessary as soon as technology allowed
branches to be restocked on a push system in which information about sales
is transmitted daily (or with even greater frequency) from tills directly to
computer systems operated at the centre.
Re-ordering can be triggered automatically by systems from either the
branches or the centre. The other option is manual or near manual re-
ordering done locally, in which case managing of demand signals and
managing working capital are functionally separate, laborious and iterative.
When the centre has control over inventory:
n well-designed systems can optimise inventory by making use of all
material data such as seasonality, flexibility to respond to weather,
transport logistics, suppliers’ lead times, et cetera (as in the chain store
retail sector)

151
n sophisticated programs can balance re-ordering parameters and the ‘open
to buy’ control of working capital in a way that traditional branch-based
manual or computer-assisted, semi-automatic methods cannot)
n stock turn can closely reflect the short-term patterns of sales
n cash flow can exactly reflect sales
n involuntary stock-outs become rare, being always automatically controlled
by the system to stay within working capital limits, or confined to endemic
risks associated with suppliers, logistics and other external contingencies.
Central control over re-ordering without computerised system support (as in
the past) is problematic and to be avoided.
What happens at the branch depends very much on the business model and
operational context. Factors influencing inventory include:
n the number of stock-keeping units (SKUs) to be reordered, and the degree
to which simple kanban systems can be applied, (that is to say the size of
the manual task of inventory control)
n the suppliers’ lead times (re-ordering windows may be critical)
n theoverheadcostofpayingemployeesdedicatedtoinventorymanagement
n the capital available for investment in system support
n whether local knowledge on patterns of demand supports reordering
decisions better than knowledge at the centre.
Writing down stock
The write-down of stock has a big effect on key financial measures of
performance.
Writing down stock (or avoiding doing so) is a major source of error and
malpractice by profit centre (branch) managers.
Independent judgement by disinterested professionals can be critical to
arriving at the true and fair view, and this is more likely to be found at the
centre,wherearm’s-lengthjudgementsarealsolikelytobemoreconservative
and therefore safer.
However, there is a significant cost if all branches are audited for stock
write-down, at arm’s length, every year. Local knowledge of demand may
give effect to better judgements about slow moving stock. And risks can be

152
mitigated with incentives or penalties to encourage branches to make true
and fair local decisions on write-down.
12.7 Control of resources, costs and operations
Budgeting fixed or variable costs such as permitted numbers of staff
The number of staff is usually a significant, locally variable operating cost
and a driver of other direct and indirect costs.
The centre has options in which it can either seek control over:
n each element of branch cost such as payroll, or
n only over the total branch cost.
In either case (elements of cost, or the total cost), these can be controlled, by
the centre or locally, as a fixed sum or as a sum flexed against an appropriate
variable. For example, the size of the staff may be controlled as a variable
of productivity – sales per employee, transactions per hour worked, or £
margin per £ payroll.
Alternatively, the centre may confine its authority to approving budgeted
headcount and/or fixed payroll because it requires certainty about the cost
as well as having a solid basis for comparison of the key operating costs by
branch and around the network.
A monthly reporting process may operate from the centre that will show the
actual and budgeted manpower and cost associated with it.
However, the centre may require no control over costs at all, confining its
control to high-level financial measures, such as profit and return on sales
or capital employed, and possibly some indication of trading health such as
new customers won.
On the other hand, the centre can be over-prescriptive, removing scope for
branch initiatives to respond flexibly to changing circumstances.
Staff, payroll, and productivity can be matters solely for the discretion of
branch managers who need to exercise their own local controls in order to
achieve the financial performance required from them.

153
Thus reporting from the branch could be skeletal, confined say, to indicatives
of financial health – sales, margin and profit – with the analysis of numbers
‘in-between’ being assumed.
This saves cost and time for reporting and debate, but the risk is increased of
misrepresentation and disguised underperformance that it would be better
to detect as early as possible.
Setting rates of pay and assigning jobs to a structure of grades
The centre’s HR professionals – a substantial overhead cost for many
companies – are qualified to create robust structures for grades and pay
rates based on job content and evaluated differentials.
The centre has the knowledge and opportunity to research and set rates of
pay nationally or regionally, but is unlikely to be able to discover for itself
the local market rates in each branch area.
The advantages of doing this at the centre are that:
n uniform company structures inform branch managers how to organise the
staff, functions and tasks
n common pay structures avoid glaring differentials, which can cause
problems when employees move between branches.
But branches should have some freedom of manoeuvre, because:
n it is at the branch that local labour markets are best understood. Branch
managers can decide to recruit into grades and pay with freedom, in
accordance with their own direct responsibility for results
n branch managers are best placed to judge the most economic balance
between the efforts made by the staff and the rewards paid to them,
usually without the company running an excessive risk.
Hiring, firing and disciplining employees
The competence of the centre’s HR professionals can be relied on to make
good final decisions on all matters of managing personnel, so the company’s
processes should be constructed around that function. Conversely, branch
managers usually lack HR skills and legal knowledge, exposing the company
to greater, avoidable risk.

154
Putting the centre in charge of HR:
n prevents branch managers from making serious errors in law and practice
n may undermine the authority of the branch manager, if too many decisions
about local matters are made remotely.
Branch managers understand the local context and can make practical,
expedient HR decisions accordingly. Branch managers often have only a
few employees reporting to them, have probably hired them and know them
personally, and if they have devolved accountability for performance, HR
decisions are usually handled sensibly.
If the branch has responsibility for HR:
n decisions are often made more quickly, which is usually best for everyone
n poor decisions by too many branch managers may give the company a
bad name as an employer and also prove costly in employment tribunal
awards and severance payments.
In practice, a scale of gravity determines how and where disciplinary cases
should be dealt with. The centre would need to ensure that managers receive
appropriate training to equip them for whatever decisions they are expected
to take.
If you don’t like change, you will
like irrelevance even less.
General Eric Shinseki
Design or specification of operating practice
The centre may choose to design and define processes, procedures and
practices that are mandatory at the branch. These fall into two categories.
First, there are those whose purpose is to regulate the connections between
the centre and the branch, such as the regular reporting flows in which the
form and content are prescribed to make assimilation, consolidation and
inter-branch comparison easier.
“ ”

155
Second, there are those practices done wholly within the branch, such as
recording staff sickness, or disposal of waste materials and countless others.
For example, the centre may prescribe that all employees clock in. It may
direct that waste be recycled in a certain way. It may wish to direct how an
insurance claim is made in order to disseminate throughout the network the
most efficient or foolproof method. Or it may choose to leave these things to
the discretion of the branch.
The centre may issue Standard Practice Instructions (SPIs) to branch
managers, and require the whole network to comply as directed. Prescribing
process and practice at branches from the centre comes at a cost, because
it consumes time and effort in both places: enforcement, if taken seriously,
also has a cost.
Butoftenthecostislessthantheconsequencesofdoingnothing,particularly
for:
n financial recording for audit and business reporting
n regulated activities carrying a risk that the centre needs to mitigate
n activity in which errors, omissions and poor practice at a local level can
result in corporate liability or dispute.
Branches are an integral part of the company in a way that an operating
subsidiary is not. For this reason, branch managers should be expected to
comply with SPIs, or the equivalent, and accept the support they afford in
the interests of efficiency and effectiveness both at the centre and at the
branch.
However, the centre has the option to give the lightest possible touch to how
branches conduct themselves, particularly when branches are measured
against high-level financial objectives.
If confidence is high, it pre-supposes that risk is being managed and that the
costs associated with managing from the centre can be reduced.
12.8 Authority to allocate and invest capital
Access to capital is always conditional, and it is the centre that defines
the conditions. Access can be planned, provided for in the budgeted year,
arranged in an emergency, or made available for an opportunity.

156
Capital would normally be allocated on application supported by a business
case prepared by the branch. The centre is able to make judgements about
need or opportunity from a comparative standpoint and can allocate scarce
capital around its branch network according to the best returns promised.
Conventional control of this type can only be operated from the centre.
Branches hardly ever hold capital sums: it is the role of the centre to raise
it and make it available where it will productively feed the business. But
profitable branches may generate free cash and can be incentivised by the
right to apply for and use some of it.
The case for capital expenditure is often associated with local enterprise, so
an excessively tight central control over capital could stunt initiative. Capital
requirements at the local level are often for relatively small sums needed
quickly, so a lengthy vetting process at the centre may snuff out opportunity.

159
The days are long gone when the only copy of the company’s organisation
chart was pinned up on a large sheet behind folding doors on the wall of the
Personnel Director’s office. Yes – it happened, and not only in fiction. Now,
in some organisations, modern practice is for out-of-date charts on scraps
of A4 paper to gather dust in the bottom drawers of departmental managers’
desks. Is this progress? It could not happen in your organisation, of course!
13 A word in conclusion

163
Collinson Grant is a management consultancy. We help firms all over
Europe and worldwide to restructure, merge acquisitions, cut costs,
boost performance and profit, and manage people. This builds long-term
relationships. We have kept some clients for over 40 years.
Our emphasis is on results and value for money. We expect to give a first
class return on the investment in us. So we do not recommend action unless
we are sure that the outcome will be worth it. We are not afraid to give bad
news, or to champion ideas that may not be welcome.
14.1 Skills – the sort of work we do
Most of our work is on three themes – organisation, costs, and people. We
use this simple framework to manage complex assignments – often with
an international dimension – and to support managers on smaller, more
focused projects. We help them:
n to restructure and integrate – following acquisitions or to improve profits
n to rationalise the supply chain – we examine every process and interface
to improve efficiency and service
n to set up financial and managerial controls – we create robust systems to
improve decision-making and reduce risks
n to introduce Lean manufacturing and refine business processes – we
analyse and improve how work is done, and use new ways to create
change and make it stick
n to cut costs – we make systematic analyses of overheads, direct costs
and the profitability of customers and products; this helps managers to
understand complexity, and to take firm steps to reduce it
n to manage people – we draw up pay schemes and put them into effect,
guide managers on employee relations and employment law, get better
performance from people, and manage redundancy.
Our consultants have held senior executive line positions with responsibility
for profit and loss. We work in many different sectors for large private and
public companies and also in the public sector.
14 Collinson Grant

164
In projects involving organisational change, we have worked with managers:
n to design and implement new organisational structures
n to review layers and spans of control to find opportunities to reduce costs
n to integrate managerial teams with new accountabilities and controls
n to improve profit and turn around failing companies
n to recruit new staff and/or manage redundancies
n to outsource functions and set up shared service centres
n to serve as interim directors.
Our approach and consulting style is grounded in pragmatism, urgency,
value for money and objectivity. Programmes are customised to the
requirement. Documentation is what is necessary to the case.

167
Our readers continue to enjoy the quotations which pepper our publications.
Some of those quoted are more familiar than others, so here are some brief
introductions.
Page reference
2 William Edwards Deming (1900-1993) was an American
statistician, professor, author, lecturer and consultant
whose work in Japan helped that country’s manufacturers
to conquer world markets.
9 Russell Lincoln Ackoff (1919-2009) was an organisational
theorist and a pioneer of systems thinking. As Professor of
Management Science at the Wharton School, University
of Pennsylvania, he made a speciality of pointing out how
managers’ real behaviour differed from that prescribed by
management scientists.
10 Geoffrey M Bellman (born 1938) was a business executive
with several Fortune 500 companies before becoming a
consultant. His books include Getting Things Done When
You Are Not in Charge.
15 John Adair (born 1934) was an officer in the Scots Guards,
an adjutant with a Bedouin regiment in the Arab Legion,
and a deckhand on a fishing trawler before going on to
write 40 books. More than a million managers have taken
part in the action-centred leadership programmes he
pioneered.
17, 40, 143 Peter Drucker (1909-2005) was an Austrian-born
American consultant, educator, and author. He invented
the concept known as management by objectives and
was one of the first to spot trends such as privatisation,
the power of marketing, the information society, and
‘knowledge workers’.
19 Mary Douglas (1921-2007) was a British social
anthropologist who investigated symbolism, culture, and
the role of the individual in society.
15 Notes on quotations

168
20 Colin Powell (born 1937) was chairman of the US Joint
Chiefs of Staff from 1989-93 and Secretary of State from
2001-05.
23 Agha Hasan Abedi (1922-1995) was the Pakistani founder
of the Bank of Credit and Commerce International, which
collapsed in 1991 after becoming implicated in money
laundering.
24 Warren Buffet (born 1930) controls Nebraska-based
Berkshire Hathaway. His value-investing philosophy
earned him the nickname ‘Sage of Omaha’. Although a
billionaire, he is renowned for personal frugality and
plans to give away 99% of his wealth.
32 Muhammad (c570-632) was orphaned by the age of six
and worked as a shepherd and caravan leader until he
received visions calling on him to destroy polytheistic
idols around al ka’ba (the cube), a shrine in Mecca. His
followers took control of the Arabian peninsula and Islam
was born.
34 Rupert Murdoch (born 1931) inherited the Adelaide
News from his father and acquired other newspapers
in Australia, New Zealand and the UK before building
a global media empire. At the peak of his power, Forbes
magazine named him the 24th most powerful person in
the world.
35, 103 Anthony Stafford Beer (1926-2002) was a professor
at the Manchester Business School. An early advocate
of computers in business management, he led
groundbreaking research into cybernetics before
renouncing material possessions and moving to mid-
Wales during the 1970s.
42 Elliott Jaques (1917-2003) was a Canadian psychoanalyst
and organizational psychologist, who invented the
concept of the ‘time-span of discretion’, based on the idea
that the higher a person sits in a hierarchy, the longer he
can work unsupervised.
43, 133 Lee Iacocca (born 1924) made his name as a salesman but
also played a part in designing the Ford Mustang and the
Lincoln Continental. After the Ford Motor Company fired
him in 1978, he got his revenge by leading the revival of
the Chrysler Corporation.

169
45 Lord Weinstock (1924-2002) built the General Electric
Company into one of Britain’s leading industrial
conglomerates. He hoarded £3bn worth of cash but his
successors spent it badly, destroying the share price.
Weinstock’s personal stake, once worth £480m, was
reduced to £2m.
50 Lawrence J Peter (1919-1988) co-authored The Peter
Principle, a humorous book about the way hierarchies
work. Subtitled ‘Why things always go wrong’, the thrust
of its argument is that employees promoted on merit
eventually ‘rise to the level of their incompetence’.
59 Charles ‘Casey’ Stengel (1890-1975), was an American
baseball outfielder and manager. His nickname reflected
his Kansas City origins but sportswriters also dubbed him
‘The Old Perfessor’ because of his wit and quotability.
62 Tom Peters (born 1942) was a management consultant
with McKinsey before becoming an author and
motivational speaker. His first book, In Search of
Excellence, preached the gospel of empowerment at all
levels of an organisation.
64 Dorset-born Harold Geneen (1910-1997) emigrated to the
United States with his parents and studied accounting. He
was president and CEO of International Telephone and
Telegraph during the 1960s, when sales increased from
$765m to $17bn.
68 John Kenneth Galbraith, (1908-2006), was a Canadian
economist and author whose support for the ideas of
John Maynard Keynes was evident in books such as The
Affluent Society. He was an adviser to President John F
Kennedy, who appointed him US Ambassador to India.
71 Marcus Aurelius (121-180), was a Roman emperor and
stoic philosopher who defeated the Parthians and the
Germanic tribes.
78 Donald Rumsfeld (born 1932) holds the records for being
the youngest and the oldest person to have served as US
Secretary of Defence, holding the post under presidents
Gerald Ford and George W. Bush. In between, he was
president of G D Searle Company, a pharmaceutical firm.

170
104 Edward Bulwer-Lytton (1803-1873), was an MP, Secretary
of State for the Colonies, and sat in the House of Lords.
He also found time to be a poet, playwright and novelist,
coining the phrase ‘the great unwashed’.
107 Stephen Covey (1932-2012) wrote The Seven Habits of
Highly Effective People, a self-help book which argued the
case for character and principles and sold more than 25
million copies. A devout Mormon, he was sent to England
as a missionary in his youth.
114 Cyril Northcote Parkinson (1909-1993) was a naval
historian and author best known for Parkinson’s Law,
which argued that work, especially in a government
bureaucracy, ‘expands to fill the time available for its
completion’.
154 General Eric Shinseki (1999-2003) was wounded in
Vietnam, served as Chief of Staff of the US Army from 1999-
2003, and was appointed Secretary of Veterans’ Affairs in
2009.

Complexity, Direct costs, Employee relations,
Employment law, Implementing change, Integrating organisations, Lean,
Managerial controls, Organisational design, Overheads,
Performance management, Pricing, Process improvement, Procurement,
Reward, Supply chain, Transitional management, Value chain analysis,
Workforce planning
collinsongrant.com
Costs
United
Kingdom
People
Performance
Mainland
Europe
Organisation
Restructuring
United States
of America
Productivity

Managing organisational design

More Related Content

What's hot (19)

Viewers also liked (12)

Similar to Managing organisational design (20)

More from Cheshire East Council (17)

Recently uploaded (20)

Managing organisational design